Understanding and Treating Panic Disorder: Cognitive-Behavioural Approaches [1 ed.] 0471490679, 9780471490678

Citation preview

Understanding and Treating PanicDisorder Cognitive-BehaviouraI Approaches Steven Taylor

(£)WILEY

1

UNDERSTANDING AND TREATING PANIC DISORDER

PORTSMOUTl-1 & S.E. Hh~NTS.

PSYCttOLOGY SERVICE

The Wiley Series in

CLINICAL PSYCHOLOGY Steven Taylor Alan Cartt Max Birchwood, David Fowler, and Chris Jackson (Editors) Dominic H. Lam, Steven H. Jones, Peter Hayward, and Jenifer A. Bright

Understanding and Treating Panic Disorder: Cognitive-Behavioural Approaches Family Therapy: Concepts Process and Practice Early Intervention in Psychosis: A Guide to Concepts, Evidence and Interventions Cognitive Therapy for Bipolar Disorder: A Therapist's Guide to Concepts, Methods and Practice

Titles published under the series editorship of:

J. Mark G. Williams

School of Psychology, University of Wales, Bangor, UK

Peter Salmon

Psychology of Medicine and Surgery: A Guide for Psychologists, Counsellors, Nurses and Doctors

William Yule (Editor)

Post-Traumatic Stress Disorders: Concepts and Therapy Treating Complex Cases: The Cognitive Behavioural Therapy Approach

Nicholas Tarrier, Adrian Wells, and Gillian Haddock (Editors) Michael Bruch and Frank W. Bond (Editors) Martin Herbert, Eric Emerson, Chris Hatton, Jo Bromley, and Amanda Caine (Editors)

J.Mark G. Williams, Fraser N. Watts, Colin MacLeod, and Andrew Mathews Phil Mallon

Beyond Diagnosis: Case Formulation Approaches in CBT Clinical Child Psychology (second edition) Clinical Psychology and People with Intellectual Disabilities

Cognitive Psychology and Emotional Disorders (second edition)

Multiple Selves, Multiple Voices: Working with Trauma, Violation and Dissociation A list of earlier titles in the series follows the index.

UNDERSTANDING AND . TREATING PANIC DISORDER Cognitive-Behavioural

Approaches

Steven Taylor Department of Psychiatry, University of British Columbia, Canada

JOHN WILEY & SONS, LTD Chichester · New York · Weinheim • Brisbane • Singapore • Toronto

Copyright©

2000 by John Wiley & Sons Ltd, Baffins Lane, Chichester, West Sussex PO19 lUD, England National 01243 779777 International (+44) 1243 779777 e-mail (for orders and customer service enquiries): [email protected]. uk Visit our Home Page on http:/ /www.wiley.co.uk or http:/ /www.wiley.com

All Rights Reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, electronic, mechanical, photocopying, recording, scanning or otherwise, except under the terms of the Copyright, Designs and Patents Act 1988 or under the terms of a licence issued by the Copyright Licensing Agency, 90 Tottenham Court Road, London, UK WlP 9HE, without the permission in writing of the Publisher. Other Wiley Editorial Offices

John Wiley & Sons, Inc., 605 Third Avenue, New York, NY 10158-0012, USA WILEY-VCH GmbH, Pappelallee 3, D-69469 Weinheim, Germany Jacaranda Wiley Ltd, 33 Park Road, Milton, Queenslad 4064, Australia John Wiley & Sons (Asia) Pte Ltd, 2 Clementi Loop #02-01, Jin Xing Distripark, Singapore 129809 John Wiley & Sons (Canada) Ltd, 22 Worcester Road, Rexdale, Ontario M9W lLl, Canada

British Library Cataloguing in Publication Data

A catalogue record for this book is available from the British Library ISBN 0-471-98704-2 (cased) ISBN 0-471-49067-9 (paper) Typeset in 10/12pt Palatino by Best-set Typesetter Ltd., Hong Kong Printed and bound in Great Britain by Bookcraft (Bath) Ltd, Midsomer Norton, Somerset This book is printed on acid-free paper responsibly manufactured from sustainable forestry, in which at least trees are planted for each one used for paper production.

For Jessica May Quinn

CONTENTS About the Author Foreword Preface

PART I Chapter 1 Chapter 2 Chapter 3 Chapter 4 Chapter 5 Chapter 6 Chapter 7 Chapter 8 PART II

Chapter 9

lX

Richard J.McNally

. . . . . . . ........

.

Xl Xlll

THEORETICAL FOUNDATIONS AND EMPIRICAL FINDINGS Panic attacks, panic disorder, and agoraphobia ................... Comorbidity: panic disorder in context . . . . . . . . . . . . . . . . . . . . . . . . Cognitive models . . . . . . . . . . . . . . . . . Feared sensations: what are their causes? . . . . . . . . . . . . . . . . . . . . . . . . Meta-analyses of treatment outcome . . . . . . . . . . . . . . . . . . . . . . . Treatment outcome: a closer look . . . Combining psychological treatments with drug therapies ............ Predicting treatment outcome . . . . . .

1 20 37 58 81 111 155 178

CLINICAL PROTOCOLS AND PROCEDURES

Assessment . . . . . . . . . . . . . . . . . . . . . . .

217

Chapter 10

Developing a case formulation

261

Chapter 11

Cognitive-behaviour therapy: an overview ...................

285

Cognitive interventions

308

Chapter 12

............

VIII

CONTENTS

Chapter 13 Chapter 14 Chapter 15

Chapter 16 Chapter 17 References Author Index Subject Index

Interoceptive and situational exposure . . . . . . . . . . . . . . . . . . . . . . .

339

Adjunctive cognitive-behavioural interventions ..................

.

366

Strategies for improving treatment adherence and preventing relapse .......................

.

388

protocols for ............ .

401

Cognitive-behavioural special populations

Other psychological treatments

....

.

427

443 497 507

ABOUT THE AUTHOR Steven Taylor trained as a clinical psychologist at the University of Melbourne, Australia (M.Sc., 1987) and at the University of British Columbia, Canada (Ph.D., 1991). He is currently an Associate Professor in the Department of Psychiatry at the University of British Columbia. Since 1992 he has been Associate Editor of Behaviour Research and Therapy. He has published over 100 journal articles and book chapters, and recently published an edited volume, titled Anxiety sensitivity: Theory, research, and treatment of the fear of anxiety (1999, Mahwah, NJ: Erlbaum). Dr Taylor received early career awards from the Canadian Psychological Association, the Association for Advancement of Behavior Therapy, and the Anxiety Disorders Association of America. He is actively involved in clinical teaching and supervision, and maintains a private practice in Vancouver, Canada. His research interests include cognitive and cognitive-behavioural treatments and mechanisms of anxiety disorders, particularly panic disorder, posttraumatic stress disorder, and obsessive-compulsive disorder.

FOREWORD Great strides have been made in the cognitive-behavioral treatment of panic disorder during the 1990s. During the previous two decades, clinicians targeted agoraphobia avoidance, and progress was modest. But when panic-or more specifically, fears and misconceptions about panicrelated bodily sensations-became the main target of interventions, major advances began to occur. Indeed, contemporary cognitive-behavior therapy (CBT) is at least as effective as the pharmacologic interventions once thought to be uniquely potent against panic. Despite these achievements, challenges remain. Third parties, exemplified by managed care companies, have been increasingly reluctant to reimburse for treatments devoid of empirical support. On the one hand, this would seem to be good news for CBT clinicians who have long been attentive to the evidential basis for their practice. On the other hand, CBT clinicians lack the institutional resources possessed by the pharmaceutical industry that enable it to disseminate information about its effective interventions. How, then, should clinicians spread the news about CBT? One approach is for experts to write excellent, accessible texts that bring mental health professionals up to speed with regard to empirically supported, cost-effective interventions. This is precisely what Steven Taylor has done in this masterful new book on panic disorder. The merits of his book are many. Taylor is a renowned and prolific clinical researcher in the anxiety disorders field who has written the clearest, most detailed account available on contemporary CBT methods for treating this syndrome. Taylor's book differs from other panic disorder books in several ways. Explicating methods pioneered by David M. Clark and David H. Barlow, Taylor provides a thorough, step-by-step account of how to conduct cognitive therapy, interoceptive exposure, and exteroceptive exposure. Unlike other practice-oriented books, Taylor has chapters on enhancing treatment adherence, incorporating adjunctive interventions, and customizing validated methods within an idiographic, case formulation framework. But perhaps the most distinctive feature is Taylor's inclusion of chapters that cover the research literature that provides the evidential

xii

FOREWORD

basis for the techniques described therein. Thus, his book is a hybrid volume: both scholarly monograph and highly practical treatment manual. In summary, this excellent book should be read by all mental health professionals who treat panic disorder. It provides the tools one needs to help these patients in the fastest, most effective way possible, and it documents the science that undergirds the methods. Richard

J. McNally,

Ph.D. Harvard University Cambridge, Massachusetts December 1999

PREFACE Panic disorder, with or without agoraphobia, is a common and often debilitating condition that typically begins in adolescence or early adulthood. It tends to wax and wane in severity, particularly in response to life stressors, and can be chronic if untreated. According to a recent statement on practice guidelines, the American Psychiatric Association (1998) concluded that cognitive-behaviour therapy (CBT) has 'considerable evidence of efficacy in the treatment of panic disorder' (p. 2). Thus, it is not surprising that the various forms of CBT are gaining widespread acceptance among mental health professionals. In recent years a number of excellent books have been published on the nature and treatment of panic disorder. These include books dealing primarily with theoretical issues (e.g., McNally, 1994), therapist manuals that describe the basics of CBT (e.g., Craske et al., 1994), and self-help guides (e.g., R. R. Wilson, 1996). What is lacking is a comprehensive clinician's guide; one that integrates theory, empirical findings, and treatment guidelines, to provide a framework for understanding and treating both routine and complex cases of panic disorder. The present volume aims to provide such a guide. The book is divided into two sections. The first covers the theoretical foundations of cognitive-behavioural treatments for panic disorder (with or without agoraphobia), and the relevant empirical findings. We begin with a review of the descriptive psychopathology and natural history of panic disorder, followed by a review of the major cognitive theories. Treatment outcome studies of behavioural and cognitive-behavioural therapies are then reviewed. For the purpose of comparison, we also will review the empirical literature on other treatments for panic disorder, such as pharmacotherapies, with the aim of helping the clinician decide whether a CBT protocol or some other treatment is most appropriate. This section of the book also reviews the research on integrating CBT with pharmacotherapies. The section concludes by reviewing the literature on predicting important clinical outcomes, such as treatment dropout, response, and relapse. The reviews in these chapters provide empirically based guide-

XIV

PREFACE

lines for selecting the optimal treatment. Thus, the chapters are relevant to both clinical researchers and clinical practitioners. The second section of the book builds on the first. It describes the clinical protocols and procedures for cognitive-behavioural assessment and treatment. Basic procedures and variations of CBT are described, along with reviews of other psychological treatments. This section also describes a case formulation approach to treatment. That is, a process for developing an understanding of the patient's problems, including a description of the predisposing, precipitating, and perpetuating factors. The formulation guides the clinician in selecting procedures from empirically validated treatment protocols, and also helps determine the timing of these interventions. The book describes treatment protocols for uncomplicated cases as well as guidelines and strategies for dealing with more difficult cases. The latter include cases of panic disorder that have failed to respond to conventional CBT approaches, and cases in which panic disorder is comorbid with other disorders (e.g., comorbid posttraumatic stress disorder). Protocols are also described for implementing CBT in specific settings (e.g., emergency rooms, rural settings), specific populations (e.g., children, adolescents, the elderly), and particular cultural milieux (i.e., culture-specific aspects of treatment). This book is intended for a wide audience of mental health professionals, including psychologists, psychiatrists, social workers, counselors, psychotherapists, and students in the mental health professions (e.g., clinical psychology graduate students and psychiatry residents). The book is not intended as a stand-alone guide; it should be used in conjunction with appropriate clinical supervision. As such, this volume is intended to aid training in CBT. Throughout this volume three conventions will be used to simplify and streamline the text. First, the term panic disorder will be used to refer to panic disorder with or without agoraphobia, even when reviewing the older studies that described their patients as agoraphobics. The latter designation was used in studies published prior to DSM-III (American Psychiatric Association, 1980). In those studies, most if not all patients described as having agoraphobia would meet latter-day criteria for panic disorder with agoraphobia. Studies of patients with agoraphobia without panic will be explicitly identified as such. The second convention adopted in this volume concerns duplicate publications and papers using overlapping samples. In reviewing the literature on a given topic (e.g., the literature on predictors of treatment outcome),

2

PREFACExv

papers will not be cited if they used samples that overlapped with those of other studies. To include duplicate studies would give a misleading impression of the degree of support for a given conclusion. The third convention is that this book will draw the distinction between early and contemporary forms of CBT for panic disorder. The former will be called first generation CBT or CBTl. This typically consisted of situational (in vivo) exposure and some form of cognitive restructuring, such as self-instruction training, rational-emotive therapy, or Beck's (1976) cognitive therapy. These cognitive methods were devised prior to the development of contemporary cognitive models of panic (e.g., Clark, 1986), and so were unlikely to emphasize the vicious cycle of panic (see Chapter 3). Cognitive restructuring in CBTl tended to focus on agoraphobia, and often taught patients to use coping statements during exposure exercises (e.g., statements such as 'I can cope with my anxiety'). In contrast to CBTl, second generation CBT (CBT2) refers to the treatment developed by Clark and colleagues (e.g., Clark, 1989; Clark & Salkovskis, 1987), Beck (1988), and Barlow and colleagues (e.g., Barlow & Cerny, 1988; Craske et al., 1994). CBT2 focuses largely on panic attacks. This book is largely about CBT2. The patient is presented with a cognitive model of panic attacks (e.g., Clark, 1986), and cognitive exercises are used largely to correct catastrophic beliefs about arousal-related body sensations. Interoceptive and situational exposure exercises are used to challenge these misinterpretations. Breathing retraining also is used, particularly in cases where panic attacks are associated with hyperventilation. Many people contributed, directly or indirectly, to this book. I would like to express my gratitude to the many teachers and colleagues who have stimulated my thinking on the nature and treatment of panic disorder. Thanks also to Ingrid C. Fedoroff Ph.D., and Adrian Wells Ph.D. for their perceptive comments on previous drafts of this book, and to Jonathan Fleming M.D., who provided useful advice on medical contraindications to interoceptive exposure exercises. Thanks also to Michael Coombs, senior editor at John Wiley & Sons Ltd, for his patience and sound advice. Finally, I would like to thank my patients, who over the years have taught me a great deal about panic disorder. Case examples described in this book have been modified to protect patient privacy and confidentiality. Following the guidelines set out by Clifft (1986), the cases illustrations are either disguised to eliminate identifying information, or represent composite descriptions, combining material from several patients.

Part I

THEORETICAL FOUNDATIONS AND EMPIRICAL FINDINGS

Chapter 1

PANIC ATTACKS, PANIC DISORDER, AND AGORAPHOBIA 'When I was a kid I was stung by a wasp while playing in the back yard. My face puffed up like a balloon and my eyelids swelled until I couldn't see. My throat closed up till I could hardly breathe. My mother rushed me to the hospital where they gave me adrenaline. The doctor told me I was lucky to be alive. Ever since I've worried about my health, and particularly about allergic reactions. This morning the air was so humid and smoggy that I had trouble catching my breath. I started to worry that I might be allergic to smog. My mouth went dry and I felt a lump in my throat. That really scared me. I started to breathe faster so I would get enough air. But things only got worse. I felt dizzy, my face went numb, and my heart started pounding. My chest was so tight that I could hardly catch my breath. I was paralyzed with fear and sure I was going to die. Frantically, I grabbed my cell phone and called for an ambulance. By the time it arrived I felt better. It wasn't an allergic reaction. Just my nerves. Next time I might not be so lucky.'

This description provided by a panic sufferer clearly shows that panic attacks can be terrifying experiences. Panic disorder 1 is a common and debilitating condition. It often co-occurs with other psychiatric disorders, as in the above-mentioned example, which was a case of panic disorder and hypochondriasis. The purpose of this book is to describe the nature of panic disorder, and to describe specific treatment strategies. We will review treatments of uncomplicated panic disorder, as well as strategies for treating comorbid cases. The latter are common in clinical practice and present special challenges to the clinician. The present chapter begins by providing some background into the nature, costs, and course of panic disorder.

1 Throughout

this volume, the term panic disorder will be used to refer to panic disorder with or without agoraphobia.

4

PANIC ATTACKS, PANIC DISORDER,AND AGORAPHOBIA

THE COSTS OF PANIC DISORDER Social and Economic Costs In the United States, anxiety disorders account for an estimated 32% of the total economic costs of psychiatric disorders, exceeding the costs of schizophrenia (21%) and mood disorders (22%) (Rice & Miller, 1993). These costs include lost productivity, social welfare expenditures, and direct health-care expenses. Similar findings are likely to be found in Britain, Australia, and other Western countries. Compared to other psychiatric disorders, panic disorder is a leading cause for seeking emergency department consultations (Weissman, 1991). It is also a leading cause for seeking mental health services, surpassing schizophrenia and mood disorders (Boyd, 1986). Panic disorder exceeds the costs associated with many other anxiety disorders such as social phobia, generalized anxiety disorder, and obsessive-compulsive disorder (Greenberg et al., 1999; Kennedy & Schwab, 1997; Rees et al., 1998). Boyd (1986) observed that 'anxiety disorders are often referred to as mild psychiatric disorders ... but the high rates of treatment for panic disorder suggest that it might be better viewed as a severe psychiatric disorder' (p. 1573). Indeed, panic disorder is associated with impaired occupational and social functioning and poor overall quality of life (Edlund & Swann, 1987; Katerndahl & Realini, 1997; Leon et al., 1995; Markowitz et al., 1989). .. The high costs are partly because panic disordered patients-particularly those in the early stages of the disorder-are most likely to present to their primary care physician or hospital emergency departments, thinking they are in imminent danger of dying or 'going crazy' (Katemdahl & Realini, 1995). In these settings patients may undergo a series of extensive and expensive medical tests before panic disorder is finally diagnosed. Ruling out serious organic conditions is good clinical practice, but substantially contributes to the costs that panic disorder places on health-care systems. This is especially true when panic disorder is mistaken for some other disorder. Unfortunately, this sometimes happens.

Physical Costs People with panic disorder, tion, report poorer physical der is associated with high norms adjusted for age and

compared to people in the general populahealth (Markowitz et al., 1989). Panic disorcholesterol levels, compared to population gender (Hayward et al., 1989). There is sug-

PANIC DISORDER 5

gestive evidence that panic disorder is associated with heightened risk of morbidity and mortality due to cardiovascular and cerebrovascular diseases (Coryell et al., 1982, 1986; Fleet & Beitman, 1998; Martin et al., 1985; Weissman et al., 1990). The cause(s) of these associations have yet to be established. They may be a result of maladaptive attempts by panic sufferers to dampen their hyperarousal, such as by smoking, consuming excessive amounts of alcohol, or overeating. Maladaptive coping may be compounded by a sedentary lifestyle, which is common in people whose lives are severely restricted by panic attacks and agoraphobia.

PANIC DISORDER Diagnosis, Prevalence, and Course According to DSM-IV (American Psychiatric Association, 1994), panic disorder is defined by recurrent, unexpected panic attacks, followed by at least a month of either (1) persistent concern about having more attacks, (2) worry about the possible implications or consequences of the attacks, or (3) a marked change in behaviour as a result of the attacks (e.g., avoiding situations associated with attacks, such as quitting a stressful job). Panic disorder has a bimodal age of onset, typically developing either between ages 15 and 19 or 25 and 30 years (Ballenger & Fyer, 1996). It has a lifetime prevalence of 1.5-3.5%, making it one of the most common psychiatric disorders (Eaton et al., 1991; Kessler et al., 1994). Clinical features of panic disorder are similar across genders (Oei et al., 1990), although the disorder is diagnosed more than twice as often in women as men (Kessler et al., 1994; Weissman et al., 1997). The disorder can be chronic if untreated (Breier et al., 1986; Uhde et al., 1985). The prevalence, course, gender distribution, and age of onset of panic disorder appear to be generally consistent throughout the world (Weissman et al., 1997).

Panic Attacks Defining Features Panic attacks are discrete episodes of intense fear or discomfort, accompanied by four or more of the 13 symptoms shown in Table 1.1. Panic attacks tend to occur suddenly, reaching peak intensity within l0min (American Psychiatric Association, 1994), and typically lasting for about 20min (Barlow & Craske, 1988). Attacks with fewer than four symptoms are called limited symptom attacks.

6

PANIC ATTACKS, PANIC DISORDER,AND AGORAPHOBIA

To illustrate the frequency of symptoms during panic attacks, data from 62 panic disordered patients are shown in Table 1.1 (Rapee et al., 1990). Panics were assessed by two weeks of prospective monitoring using a panic attack diary, and also retrospectively assessed by asking patients to recall symptoms of their typical attacks. Although the findings differ to some extent with the different methods of assessment, both methods indicate that the most common symptoms were palpitations, dizziness, fear of losing control or going crazy, and trembling. Similar findings have been reported in other studies (e.g., Barlow & Craske, 1988). The symptoms listed in Table 1.1 provide only a hint of the cognitive events that take place during panic attacks. During the attacks, catastrophic thoughts fill the person's mind, unchallenged and uncontrolled. The person has difficulty thinking clearly, including difficulty in questioning whether the symptoms are really dangerous (Beck, 1988). As we

Table 1.1 Prevalence of symptoms during panic attacks

Palpitations Fear ?f losing control or gomg crazy Feeling dizzy, unsteady, lightheaded, or faint Trembling or shaking Sweating Shortness of breath or smothering sensations Chest pain or discomfort Derealization or depersonalization Chills or hot flushes Fear of dying Paresthesias (numbness or tingling sensations) Nausea or abdominal distress Choking

Prospective monitoring (2 weeks): % of patients reporting symptoms during any of their attacks

Retrospective report: % of patients reporting symptoms during their typical attacks

87 82*

82 69

71

90

71 64

62

74 68 66

56 49

50 58

49 42

58 69 47

33 18

50 36

44

From Journal of Anxiety Disorders, 4, Rapee, R. M., Craske, M. G., & Barlow, D. H., Subjectdescribed features of panic attacks using self-monitoring, 171-181, Copyright (1990), with permission from Elsevier Science. *For the prospective monitoring, fear of losing control was assessed separately from fear of going crazy. A total of 82% reported fear of losing control, and 33% reported fear of going crazy.

PANIC DISORDER 7

will see in later chapters, a thorough cognitive assessment is important for understanding panic disorder and for planning treatment.

Symptom Definitions and Descriptions Some symptoms-such as hot flushes or chills-are easily described and understood by clinicians and patients alike. Other symptoms are either more difficult to describe or are described in many different ways. To avoid confusion, the following are definitions and descriptions of these symptoms. When patients describe fear of losing control they often mean that they are frightened of going crazy during a panic attack. This is why fear of going crazy and fear of losing control are grouped together as a single symptom in DSM-IV. However, these fears are not always synonymous. Fear of losing control can refer to a number of other feared consequences, including fear of doing something embarrassing such as trembling in public, or fear of fleeing in a dramatic fashion when panic strikes (e.g., dropping one's groceries and bolting from a supermarket). Fear of losing control can also refer to fear of striking out and attacking someone. Fear of harming others may occur during a panic attack if the sufferer's path to safety is obstructed when they are attempting to escape a panicprovoking situation, such as a crowded department store.

Palpitations refer to an uncomfortable

or abnormal awareness of heart beats. Symptoms include heavy beating of the heart, fluttering in the chest, skipped beats, rapid heart beating, irregular or 'extra' heartbeats, or feeling one's pulse pounding in the neck (Fiengo & Michelson, 1996). Choking sensations include a sense of throat tightness or constriction. Another choking sensation is known as globus (Kellner, 1991), which is a sensation of having a ball or lump in one's throat or a feeling that something is stuck in the throat.

Chest pain during panic attacks most commonly occurs in the left inframammary region. It also may occur on the right or left side in other locations, such as the pectoral or central regions. The pain may be sharp or stabbing in quality. Occasionally, it is constricting and cramplike, and may be difficult to distinguish from angina (Bass, 1990).

Paresthesiasrefer to numbing, prickling, or tingling sensations that typically occur in the hands, but also occur in the lips and face, and sometimes in the feet.

Dizziness refers to the illusion that the person or the person's environment is spinning (giddiness or vertigo). Patients sometimes confuse dizziness with depersonalizationand derealization.As defined in DSM-IV, deperson-

8

PANIC ATTACKS, PANIC DISORDER,AND AGORAPHOBIA

alization is an alteration in the perception or experience of the self, characterized by feeling detached from one's mental processes or body. This is sometimes described as feeling like an outside observer of oneself, or like being in a dream. Derealization is an altered perception or experience of the world; things seem strange or unreal, and people seem unfamiliar or mechanical. The sufferer may describe feeling dazed, detached, and confused, with emotions muted in intensity. Objects may appear distorted in colour or size, or otherwise unfamiliar. Sounds may appear muffled or •appear to come from a distance. Colours appear dull or washed out, and body sensations, thoughts, and the external world may seem to be somehow fading away (Fewtrell & O'Connor, 1995). Does DSM-IV Describe All the Major Panic Symptoms? Case reports of panic attacks typically describe combinations of the various symptoms listed in Table 1.1. Apart from Sigmund Freud, few authors have mentioned other panic symptoms. Freud (1894/1949) described many of the DSM-IV panic symptoms in his description of 'anxiety attacks'. He observed that along with feelings of anxiety there may be an accompanying disturbance of any one or more of the bodily functions, such as respiration, heart's action, vasomotor innervation, or glandular activity. The patient lays stress on one or other of these symptoms, and complains of 'heart spasms', 'difficulty breathing', 'drenching sweats', 'ravenous hunger' and the like. (p. 80)

'Ravenous hunger' is rarely seen in contemporary descriptions of panic attacks (Cox & Taylor, 1999). Yet, Freud (1894/1949) referred to this sensation several times in his descriptions of panic. He also observed that hunger was often associated with giddiness. This raises the question of whether attacks characterized by ravenous hunger were true panics, or whether they were actually hypoglycemic episodes. DSM-IV appears to cover the most common symptoms of panic attacks, although the cognitive symptoms (fears of dying, going crazy, or losing control) are only summary descriptions of the many forms that these fears can take. How Do Episodes of Panic End? Episodes of panic come to end for a variety of reasons, including escape or avoidance behaviours, such as taking PRN (as needed) medications, distracting oneself, doing breathing exercises, or fleeing a feared situation (Radomsky et al., 1998). Panic attacks also come to an end because of factors associated with physical exhaustion; it becomes increasingly difficult for the body to maintain an extremely high level of arousal for an extended period of time. Some patients describe a refractory period

PANIC DISORDER 9

following their attacks, during which time panics are unlikely to occur. Retrospective self-reports from panic patients suggest that the refractory period can last from minutes to weeks (Radomsky et al., 1998).

DSM-IV Panic Typology Unexpected Panic Attacks DSM-IV specifies three different types of panic attacks. This distinction is based on the relationship between the probability of panic and the presence or absence of cues (triggers) for the attacks. Unexpected (uncued) panics are attacks that the sufferer perceives as occurring spontaneously or 'out of the blue'. These characterize panic disorder, although they occasionally occur in other disorders (Barlow et al., 1985), and occasionally occur in people without any psychiatric disorder (Norton et al., 1992). People who occasionally experience unexpected panic attacks are at risk for developing full-blown panic disorder (von Korff & Eaton, 1989). One form of unexpected panic are nocturnal panic attacks, in which the person awakens from sleep in a state of panic. These attacks typically occur during non-REM sleep and are not usually precipitated by dreams (Craske & Barlow, 1989). They occur in about 69% of people with panic disorder (Mellman & Uhde, 1989), and are phenomenologically very similar to daytime panics (Craske & Rowe, 1997). People with panic disorder typically report that most of their panic attacks occur during waking hours. However, for some people the attacks are mostly nocturnal, as the following case shows. This case also illustrates some of the subtle avoidance associated with panic disorder.

Mrs V. described a long-standing history of unexpected panic attacks. These occurred either as she was falling asleep or while asleep. The attacks were not associated with nightmares or other dreams, and an evaluation at a sleep clinic failed to reveal an organic cause. Often the attacks would wrench her out of a deep sleep. Symptoms included intense palpitations, dyspnea, dizziness, and fear that she was dying. During some attacks she was unable to move her arms or legs for a few moments (isolated sleep paralysis). This was especially frightening. Mrs V. rarely had daytime panic attacks. During the day she attempted to avoid all sources of stress, and also avoided activities and situations that produced body sensations. This included the avoidance of rich foods, sexual intercourse, and all forms of physical exertion. Mrs V. even tried not to walk too fast, fearing that tachycardia could be bad for her heart.

10

PANIC ATTACKS, PANIC DISORDER,AND AGORAPHOBIA

Situationally Bound Panic Attacks The second type of panic attack consists of situationally bound (cued) panics, which are defined as attacks that are almost invariably triggered by specific situations. For example, a person with panic disorder might always panic in crowded elevators. Situationally bound panics also occur in other disorders, particularly other anxiety disorders. People with specific phobia or social phobia, for example, might panic if suddenly exposed to an intensely feared stimulus.

Situationally Predisposed Panic Attacks These are often but not invariably triggered by a given situation. Being in a supermarket line-up, for example, may increase the probability that a panic-disordered person will panic. Situationally predisposed attacks occur in panic disorder and occasionally in other disorders. The probability of panic is influenced by a variety of factors, including crowding, heat, humidity, and other contextual variables. A detailed assessment, as described in later chapters, is often needed to identify all the relevant panic cues.

Strengths and Limitations of the DSM-IV Typology The DSM-IV distinction among unexpected, situationally bound, and situationally predisposed panic attacks serves as a useful reminder that panics can differ in the extent that they are perceived (by the patient) as being cued by specific stimuli. The panic typology also can be useful in making differential diagnoses (see below). Yet, it has important limitations, especially for understanding and treating panic disorder. The main shortcoming is that the DSM-IV typology describes only some of the parameters of panic; the typology simply describes how panic attacks may vary in terms of the relationship between cues (identified by the patient) and the probability of panic. Panic attacks-regardless of whether they are unexpected, situationally bound, or situationally predisposed-can vary in many other ways, such as duration, frequency, number of panic symptoms occurring during the attack, intensity of symptoms, types of catastrophic thoughts, and ways the attacks end (e.g., whether they are terminated by performing particular escape behaviours). Later chapters will show how these variables are important iri understanding the factors that exacerbate or ameliorate panic attacks.

PANICDISORDER 1 1

The Protean

Nature

of Panic Disorder

The frequency and severity of a person's panic attacks typically fluctuate over time and circumstance. The attacks are typically frequent during episodes of stress, although this is not invariably the case. Sometimes the attacks are initially unexpected and then become more predictable, and sometimes less frequent as the person learns to identify and avoid panicevoking stimuli (American Psychiatric Association, 1998). Some sufferers report that their panics vary simply in terms of intensity and predictability, while other panic patients describe having many different 'types' of panic, characterized by distinctly different symptom profiles. One day a person might have an attack characterized by intense palpitations and fear of dying. The next day the attacks might be characterized by intense depersonalization and fear of insanity. People with panic disorder often describe having a mix of full-blown and limited symptom attacks. Examples of the latter include discrete episodes of dizziness or episodes of intense nausea, often accompanied by fears of impending catastrophe (e.g., fear of physical collapse). A question of theoretical and clinical interest is why full-blown attacks are experienced on some occasions and limited-symptom attacks on others. Does this reflect differences in the nature or strength of catastrophic beliefs, or does it reflect other factors, such as escape or distraction from the triggering stimuli? As these questions imply, to understand panic disorder it is important to assess limited symptom panic attacks, as well as episodes where the patient 'comes close' to panicking.

Precipitating Factors Stressful Life Events Initial panic attacks typically occur in 'agoraphobic' situations, such as driving alone or traveling on a city bus, and often in the context of some form of stressor (Breier et al., 1986; Faravelli et al., 1992; Lelliott et al., 1989; Shulman et al., 1994). Commonly reported stressors include separation, loss, or illness of a significant other, being the victim of sexual assault or other forms of interpersonal violence, and financial or occupational stressors (Faravelli, 1985; Faravelli et al., 1985; Finlay-Jones & Brown, 1981; Katon, 1984; Kendler et al., 1992; Michelson et al., 1998; Murrey et al., 1993; Pribor & Dinwiddie, 1992; Saunders et al., 1992; Stein et al., 1989; Tweed et al., 1989; Uhde et al., 1985; Walker et al., 1992).

12

PANIC ATTACKS, PANIC DISORDER,AND AGORAPHOBIA

Although recent studies provide useful information on the types of stressors associated with the onset of panic disorder, the role of stressful life events has been noted for over a century. Freud (1895a/1949), for example, described the role of stressors in the development and exacerbation of panic disorder: 'A man of forty-five ... first had an anxiety attack ... on receiving the news of the death of his aged father: from that time onwards a complete and typical anxiety-neurosis with agoraphobia developed; further, a young man who fell a victim to the same neurosis on account of disagreements between his young wife and his mother, and developed agoraphobia afresh after every domestic quarrel; a student who was rather an idler and had his first anxiety-attacks during a period of hard cramming under the spur of paternal displeasure.' (pp. 111-112)

Other disorders-such as other anxiety disorders, mood disorders, and personality disorders-also have been linked to stressful life events (American Psychiatric Association, 1994; Murrey et al., 1993; Pribor & Dinwiddie, 1992; Zlotnick et al., 1996). This suggests that stressful life events play a nonspecific role in the development of psychopathology; stressors do not invariably lead to panic disorder.

Other Precipitants In about 30% of patients, the initial (and typically unexpected) panic attack occurs while the person is intoxicated with, or in withdrawal from, a psychoactive substance such as marijuana, cocaine, or anesthetic (Aronson & Craig, 1986; Ballenger & Fyer, 1996; Last et al., 1984). Preliminary research suggests that seasonality also plays a role in panic onset. Lelliott et al. (1989) found that of 57 patients with panic disorder, more had their first panic in late spring and summer than in fall and winter, and in warm weather than cold. Precipitants of Panic: What are the Common Elements? An important question is why various factors-such as weather conditions, psychoactive drugs, and stressful life events-increase the risk for developing panic attacks and panic disorder. Why are these factors associated with panic attacks in some people but not others? And why do panic attacks persist after stressful events have passed, or after psychoactive substances have long been metabolized?

l i

l

PANIC DISORDER l 3

Various vulnerability factors have been postulated. According to Freud (1895a/1949), stressors were provoking or contributory factors that are neither necessary or sufficient, but can interact with specificpredisposing factors to produce specific forms of psychopathology. Similar views predominate today, although the hypothesized specific factors are far different from Freud's proposition that libidinal energy is transformed into anxiety (and panic) because of faulty sexual practices or lack of appropriate sexual outlet. Stressors, agoraphobic situations, and hot or humid weather conditions all can cause arousal-related body sensations, which may be catastrophically misinterpreted, thereby producing panic attacks. Drugs that sometimes precipitate panic disorder also cause a variety of body sensations. Marijuana, for example, increases heart rate (Beaconsfield et al., 1972). Panic attacks may occur because the person catastrophically misinterprets these sensations. These cognitive factors are discussed in Chapter 3.

Precipitants and Age of Onset Why does panic disorder typically develop between the ages of 15 and 30? Major life changes typically occur during this period, such as leaving the security of the parental home, starting a new job, starting a family, and other changes in role functioning. Similarly, experimental use of psychotropic drugs is most likely to occur between adolescence and early adulthood. Thus, panic disorder may develop in adolescence or early adulthood because the precipitating events are most likely to be experienced during this period.

Differential Diagnosis Panic disorder is not diagnosed if the attacks are the direct physiologic result of acute intoxication or withdrawal from a substance (e.g., intoxication of caffeine, amphetamines, or cocaine; withdrawal from alcohol, barbiturates, or benzodiazepines ). Panic disorder also is not diagnosed if panic attacks are entirely due to a general medical condition (e.g., entirely due to hyperthyroidism, hyperparathyroidism, pheochromocytoma, vestibular dysfunctions, seizure disorders, or cardiac conditions such as supraventricular tachycardia) (American Psychiatric Association, 1994). Panics arising from these sources may be best treated with approaches other than those described in this book. Accordingly, differential diagnosis plays an important role for planning appropriate treatment.

14

PANIC ATTACKS, PANIC DISORDER,AND AGORAPHOBIA

Panic disorder is diagnosed if the attacks continue even when the precipitant is no longer present (e.g., when the patient is no longer intoxicated with, or in withdrawal from, a psychoactive substance) (American Psychiatric Association, 1994). Panic disorder is also diagnosed if the attacks cannot be entirely explained by the effects of a drug or a general medical condition. Drugs or medical conditions can be precipitants or exacerbators of panic disorder, as illustrated by the following case. Mr E. was a middle-aged diabetic man who had experienced a number of hypoglycemic episodes in which he lost consciousness. As a result of these experiences he acquired an intense fear of having further hypoglycemic episodes, and developed recurrent panic attacks and widespread agoraphobic avoidance. His diabetes appeared to contribute to his panic attacks in that his hypoglycemic episodes led him to acquire an intense fear of arousalrelated sensations, including fears of dizziness, faintness, sweating, and palpitations. Whenever he experienced these sensations-regardless of whether or not they were due to hypoglycemia-he catastrophically misinterpreted them as indications of impending syncope, and therefore panicked.

Panic disorder is not diagnosed if the attacks are better accounted for by another disorder, such as another anxiety disorder. To illustrate, a person might seem to have recurrent, unexpected panic attacks. But on further inquiry it might be found that the person has a specific phobia and is victim to recurrent, unexpected exposures to the phobic stimulus (e.g., a person with spider phobia who has repeated, unexpected exposure to large spiders). Here, the attacks only appear unexpected; the attacks are actually cued panics triggered by unexpected exposure to the phobic object.

AGORAPHOBIA

t

( t

r

T ti

h p a

3

C

Description

and Diagnosis

tE 0

In 1871, Carl F. 0. Westphal coined the term agoraphobia. His description bears much resemblance to the syndrome as currently defined:

tJ

e 0

For a few years now patients have come to me with the peculiar complaint that it is not possible for them to walk across open spaces and through certain streets and that, due to the fear of such paths, they are troubled in their freedom and movement .... They gave me hints, or spoke candidly,

I1 C

ti

AGORAPHOBIA l 5 that they might be laughed at or considered to be insane due to the peculiarity of the matter. This fear of walking though spaces, e.g., streets, described the major phenomenon to such an extent, that I composed the term Agoraphobia, fear of spaces. Though, in addition, the fear was related to certain other situations, and therefore the selected term ... is not entirely exhaustive .... Without any exceptions, all patients mentioned that they absolutely do not know the reasons for this fear. It comes by itself; a sudden occurrence, strange thing, that appears when attempting to cross a square, or while even thinking about it. ... [The patient's] perception (observation) of a monstrous width of a square and also the thought of it, makes him believe that something will happen to him while being in a state of fear and confusion .... The condition can be lessened or forced to disappear through an escort, especially while engaging in a conversation; at the sight of a vehicle going in the same direction, or seeing an open door in one of the houses located on abandoned streets, and so forth. Furthermore, the pleasurable stimulation of alcohol makes it easier to overcome the painful condition. (Westphal, 1871, translated by Knapp & Schumacher, 1988, pp. 59-74)

Westphal noted that his agoraphobic patients experienced unexpected panic attacks and situational panics, along with anticipatory anxiety and 'anxiety about their anxiety' (Kuch & Swinson, 1992). Westphal did not make the connection emphasized today, between panic attacks and agoraphobia (Boyd & Crump, 1991). This is a curious omission, given the prominence of panic in his case descriptions. It was not until Freud that the link between panic attacks and agoraphobia became apparent. Freud (1895b/1949) observed that 'in the case of agoraphobia, etc., we often find the recollection of a state of panic;and what the patient actually fears is a repetition of such an attack under those specific conditions in which he believes he cannot escape it' (p. 136, emphasis in original). Today, agoraphobia is defined as 'anxiety about being in places or situations from which escape might be difficult (or embarrassing) or in which help may not be available in the event of having a panic attack ... or panic-like symptoms (e.g., fear of having a sudden attack of dizziness or a sudden attack of diarrhea)' (American Psychiatric Association, 1994, p. 396). Agoraphobia usually develops as a consequence of full or subclinical panic disorder (Ballenger & Fyer, 1996). People with agoraphobia tend to fear and avoid a wide range of situations, including being alone outside the home, being at home alone, crowds, bridges, elevators, and traveling by car, bus, train, or airplane. Often the person is better able to endure these situations when with a trusted companion such as a parent or spouse. In clinical settings, over 95% of people with agoraphobia also have a current or past history of panic disorder (American Psychiatric Association, 1994). Accordingly, the focus of this book is on panic disorder-with

16

PANIC ATTACKS, PANIC DISORDER,AND AGORAPHOBIA

or without agoraphobia-rather than on agoraphobia without panic. Those rare cases of agoraphobia without a history of panic disorder respond to treatments similar to those used for panic disorder with agoraphobia.

Varieties of Agoraphobia Dimensions Factor analytic studies reveal that agoraphobia is composed of several distinct but correlated factors (dimensions), which are hierarchically organized. Starting at the top of the hierarchy, agoraphobia forms a factor distinct from other phobias, such as social phobia, animal phobia, and blood/injury /illness phobia (Arrindell et al., 1991a, 1991b). In turn, the agoraphobia factor is composed of several correlated lower-order factors (subfactors), including fear of public places, fear of open spaces, and claustrophobia (Arrindell et al., 1995; Cox et al., 1993; Hamann & Mavissakalian, 1988; Johnston et al., 1984; Kwon et al., 1990). In turn, each of these factors are composed of distinct but correlated sets of factors. For example, claustrophobia is composed of fear of physical restriction and fear of suffocation (Rachman & Taylor, 1993). Thus, agoraphobia consists •of a multi-layered hierarchical arrangement of factors. Factor analytic studies indicate that if each factor represents a discrete set of causal mechanisms (Cattell, 1978), then agoraphobia arises from a mix of mechanisms that range from specific (e.g., those pertaining to fear of suffocation) to general (e.g., those influencing all agoraphobic fears: see Taylor, 1998, for details). Each factor may need to be targeted in order to produce comprehensive and enduring reductions in agoraphobia. The causes of these factors remain unknown. Each factor may arise from maladaptive beliefs about particular stimuli.

I

r s e

t e a

p 0

cl

1'

:c

Subtle Agoraphobia If agoraphobia is defined by the fear of situations associated with panic or panic-like experiences, then fears of a variety of other stimuli can be regarded as forms of agoraphobia. People with panic disorder often fear and avoid activities, substances, and situations that evoke arousalrelated sensations. These include physical exercise, emotionally evocative movies, humid or stuffy environments, sexual intercourse, drinking coffee, and so on (see Barlow & Craske, 1994, for a comprehensive list). Fear and avoidance of these stimuli are not adequately measured by exist-

A

a, ht

cl ol

st of of

t

AGORAPHOBIA 17

ing agoraphobia scales, and during clinical patients often neglect to report these fears. sent a form of subtle agoraphobia. Little is or factorial structure (dimensions) of subtle

interviews panic disordered Accordingly, the fears repreknown about the prevalence agoraphobia.

Clinical Course Not all people with panic disorder develop agoraphobia. For those who do become agoraphobic, this usually occurs within the first year of the onset of recurrent panic attacks (American Psychiatric Association, 1994). Breier et al. (1986) found that patients who misperceive their first panic attack as a sign of catastrophe (e.g., heart attack, brain tumor, or impending insanity), compared to patients who recognized their first panic as simply an anxiety reaction, developed agoraphobia more rapidly. Although agoraphobia typically develops as a consequence of full-blown or subsyndromal panic disorder, exceptions have been documented. Agoraphobia sometimes develops after the occurrence of panic-like episodes, such as attacks of diarrhea in people with irritable bowel syndrome. As agoraphobia worsens, the sufferer may become increasingly dependent on significant others, demanding that they accompany them when they have to leave home or enter social situations. There may be considerable changes in the family system, with multiple family members being affected by the patient's increasing dependency and avoidance of agoraphobic situations. Significant others may be compelled to take over many of the patient's responsibilities, such as earning wages, shopping, and child rearing responsibilities such as attending school meetings (Katon, 1994).

Differential Diagnosis Agoraphobia can be distinguished from other disorders characterized by avoidance behaviour. This is done by assessing the person's focus of apprehension or reasons for avoiding (Craske, 1991). Agoraphobic avoidance is characterized by fear of having panic or panic-like attacks. Avoidance in other- disorders is associated with different concerns. This was demonstrated by McNally and Louro (1992), who examined the reasons for fear of flying in people with agoraphobia and in people with a specific phobia of flying. Both groups feared and avoided flying, but did so for different

18

PANIC ATTACKS, PANIC DISORDER,AND AGORAPHOBIA

reasons. Agoraphobics avoided flying for fear of having a panic attack, whereas specific phobics avoided flying for fear that the plane might crash. Although many agoraphobic situations are social situations (e.g., shopping malls), agoraphobia and social phobia can be distinguished by examining the foci of apprehension; people with agoraphobia are frightened mostly of panic (or panic-like) episodes, whereas people with social phobia are frightened mostly of ridicule or rejection by others. Sometimes the distinction between the two can be difficult, such as when the person is frightened of having a panic attack because it would be embarrassing. Here, the two disorders are distinguished by identifying the main source of apprehension. If the person is primarily concerned about panic attacks (including attacks in nonsocial situations), then avoidance behaviour would be conceptualized as agoraphobia. In some cases, however, both social phobia and agoraphobia would be diagnosed. For further discussion on the distinction between these disorders see Mannuzza et al. (1990). In cases of space phobia (Marks & Bebbington, 1976), the person presents with agoraphobic avoidance, but the focus of apprehension is typically on fear of falling in the absence of nearby visual support (i.e., absence of a nearby object that they can fix their gaze upon, such as a wall). People with space phobia have neurological soft signs and appear to suffer from disturbed integration of vestibulo-ocular reflexes as a result of diverse neurologic lesions (Marks, 1987). Sufferers may be unable to travel across rooms without crawling on their hands and knees. Unlike agoraphobia, space phobia does not respond well to situational exposure therapy, and is not characterized by the fear of panic or panic-like episodes (Marks, 1987). To complicate this differential diagnosis, it is not uncommon for people with panic disorder and agoraphobia to have mild vestibular abnormalities (see Chapter 4). These can exacerbate agoraphobic avoidance, but are not sufficient to account for agoraphobia because similar abnormalities are commonly found in the general population Qacob et al., 1996b). In contrast to the treatment refractory nature of space phobia, it appears that milder vestibular abnormalities can be treated with exposure therapy, particularly exposure to movements or stimuli that induce dizziness (Yardley, 1994). With regard to differential diagnosis from general medical conditions, DSM-IV states that if an associated medical condition is present (e.g., a cardiac condition), agoraphobia is diagnosed when the fear of being incapacitated or embarrassed by the development of symptoms (e.g., fainting) is clearly in excess of that usually associated with the medical condition (American Psychiatric Association, 1994).

SUMMARY AND CONCLUSIONS 19

SUMMARY AND CONCLUSIONS Panic disorder (with or without agoraphobia) is common, costly, and often debilitating. Agoraphobia typically develops as a consequence of recurrent panic attacks or panic-like symptoms. Panic attacks and agoraphobia are heterogeneous, and may arise from multiple pathogenic processes. The heterogeneity has important implications for treatment; it underscores the importance of developing an individualized understanding of the patient's problems in order to plan treatment.

COMORBIDITY: PANIC DISORDER IN CONTEXT Comorbidity refers to the co-occurrence of two or more disorders. These may occur either at the same time (concurrent comorbidity) or at any time in a person's life but not necessarily at the same time (lifetime comorbidity). There are two main reasons for devoting an entire chapter to comorbidity. First, panic disorder (with or without agoraphobia) is commonly comorbid with a variety of disorders. Understanding the causes of comorbidity may shed light on the causes of panic disorder. Second, comorbidity can pose special challenges for treating panic disorder, sometimes requiring the clinician to modify existing treatment protocols. This chapter will review the most common forms of comorbidity, and consider the implications for understanding panic disorder.

COMORBIDITY MODELS To put the discussion in perspective, we will first consider the major models of why disorders co-occur at a rate significantly higher than what would be expected on the basis of chance. These models have been described by several writers (e.g., Docherty et al., 1986; Frances et al., 1988; Starcevic, 1992; Tyrer, 1988) and can be summarized as follows. •

•

•

Predisposition model. This model states that particular disorders, such as separation anxiety disorder or dependent personality disorder, are predisposing factors for the development of another disorder such as panic disorder with agoraphobia. Complication model. According to this model, some disorders are consequences of other disorders. For example, depression or avoidant personality disorder may arise as a consequence of panic disorder. A version of this model states that panic disorder may amplify or exacerbate previously existing personality characteristics, rather than creating a personality disorder de nova (Starcevic, 1992). Common diathesis model. This model proposes that panic disorder and some other disorders arise from a common diathesis, such as elevated

COMORBIDITY BETWEEN PANIC DISORDER 21

•

•

neuroticism. An extreme form of this model states that many disorders are simply different manifestations of the same disorder. Another variant of this model is a hierarchical model proposing that psychopathology arises from common and specific factors (Taylor, 1998; Zinbarg & Barlow, 1996). Some factors are common to a range of disorders, while others are specific to panic disorder. Forme fruste model. According to this model, some disorders (e.g., generalized anxiety disorder) are formes frustes of panic disorder or agoraphobia. That is, incomplete, modified, or chronic subsyndromal expressions. For all practical purposes it is very difficult to distinguish this model from the common diathesis model. Interaction model. Two or more co-occurring disorders, according to this model, interact or mutually exacerbate one another, so that the disorders become more severe when they occur together than they would be if they occurred separately. For example, borderline personality disorder is characterized by the tendency to produce interpersonal crises, which may exacerbate panic disorder. In turn, panic attacks and agoraphobia may exacerbate features of borderline personality disorder, such as fear of abandonment.

Some of these models conceptually overlap, and more than one model may apply to a given set of co-occurring disorders. The following sections will review the research on the comorbidity between panic disorder (with or without agoraphobia) and other disorders, and examine how the comorbidity models can account for the co-occurrences.

COMORBIDITY BETWEEN PANIC DISORDER AND OTHER ANXIETY PROBLEMS

Childhood Psychopathology Behavioural Inhibition

Behavioural inhibition is characterized by fear of novel stimuli, particularly social stimuli. Severe behavioural inhibition has been identified among children of panic disordered patients, and has been linked to childhood anxiety disorders (Biederman et al., 1990; Rosenbaum et al., 1988). It appears that behavioural inhibition is a nonspecific risk factor for anxiety disorders in general (Turner et al., 1987, 1996). Accordingly, severe behavioural inhibition may be a predisposing factor for a variety of disorders, or perhaps arises from a diathesis that is common to panic disorder (with agoraphobia) and other disorders.

22

COMORBIDITY: PANIC DISORDER IN CONTEXT

Separation Anxiety Disorder This disorder is similar in some ways to severe behavioural inhibition. Separation anxiety disorder is characterized by excessive anxiety about separation from the home or from those to whom the person is attached (American Psychiatric Association, 1994). Klein (1964) found that half of 32 panic disordered adults had an early history of severe separation anxiety. As children, many were unable to attend school, and avoided summer camp. Adult patients with other disorders (e.g., schizophrenia, mood disorders, or personality disorders) tended not to have this childhood problem. Klein proposed a predisposition model to account for the findings. That is, agoraphobia and panic attacks in adulthood were said to develop from childhood separation anxiety. Panic disorder with agoraphobia was thought to be triggered in adulthood by loss-related events such as bereavement or separation from a loved one. Klein proposed that some panic attacks were false alarms triggered by actual or threatened separation from a significant other. This observation is consistent with the fact that people with severe agoraphobia are unable to venture far from home without being accompanied by a spouse, parent, or other 'safe' person. However, not all patients with panic disorder and agoraphobia show this pattern, and panic attacks are not always triggered in adulthood by actual or threatened separations from significant others. This led Klein to hypothesize two types of panic disorder, differing in terms of whether or not there was a history of severe separation anxiety during childhood. More recent studies generally support the view that a history of severe childhood separation anxiety is more likely to be found in adults with panic disorder, compared to adults with other disorders and to adult normal controls (Battaglia et al., 1995; Silove et al., 1993, 1995, 1996). A history of severe childhood separation anxiety is associated with an earlier age of onset of agoraphobia (Berg et al., 1972; Breier et al., 1986; Klein, 1964). Unfortunately, little is known about the mechanism underlying separation anxiety. Similarly, it is unclear what might cause separation anxiety to develop into panic disorder with agoraphobia. A possible cognitive mechanism is discussed in Chapter 3.

Anxiety Disorders in Adulthood According to epidemiologic surveys, lifetime comorbidity between panic disorder (with or without agoraphobia) and other anxiety disorders is common. The most commonly comorbid disorders include social phobia (occurring in 15-30% of people diagnosed with panic disorder), obses-

MOOD DISORDERS 23

sive-compulsive disorder (8-10%), specific phobia (10-20%), and generalized anxiety disorder (25%) (American Psychiatric Association, 1994). Comorbidity tends to be even higher among people seeking treatment for panic disorder. To illustrate, Breier et al. (1986) found that 48 (80%) of their 60 panic patients had concurrent comorbid generalized anxiety disorder. A common diathesis model provides the most parsimonious account of the high comorbidity between panic disorder and other anxiety disorders. However, it is possible that other comorbidity models account for some patterns of comorbidity. For example, generalized anxiety disorder might sometimes arise as a consequence of panic disorder. That is, the occurrence of panic attacks may lead to worry about one's health and safety. Worry then may become a learned response for avoiding danger in general. Thus, worry may spread to other domains (e.g., finances, social and occupational functioning, wellbeing of others), and so generalized anxiety disorder may develop.

MOOD DISORDERS Major Depression Among people who develop panic disorder, 50-65% will develop major depression at some point in their lives. In two-thirds of these cases depression develops concurrent with, or after the onset of, panic disorder (American Psychiatric Association, 1994; Kessler et al., 1998). The risk of developing major depression is more closely linked to the severity of agoraphobia than to the severity or frequency of panic attacks (Ball et al., 1994; Thompson et al., 1989). It may be that this association is partly due to confounded diagnostic criteria. Depressed people may refrain from entering agoraphobic situations because they lack the energy to go out, or because they pessimistically predict that they would not enjoy outings to movie theatres, shopping malls, or other public places. Sometimes this may be misdiagnosed as agoraphobia, thereby leading to the overestimation of the relationship between agoraphobia and major depression. Although this could account for the strong association between major depression and agoraphobia, it does not account for depression in panickers who are not agoraphobic. People with panic disorder often become demoralized about their panic disorder, despairing that terrifying panic attacks will always be a part of their lives. Demoralization may spiral into major depression. This is a complication model of comorbidity, where depression is a complication or consequence of panic disorder. This model accounts for depression arising during or after the development

24

COMORBIDITY: PANIC DISORDER IN CONTEXT

of panic disorder, but does not explain why depression sometimes occurs first. Other comorbidity models may be needed to fully account for the relationship between panic disorder and depression. Patients with coexisting panic disorder and major depression tend to have more severe depression and panic disorder than people who have only one of these disorders (Andrade et al., 1994; Breier et al., 1986; Grunhaus et al., 1994; McLean et al., 1998). This is consistent with an interaction model, where panic disorder and mood disorders exacerbate one another. Another possibility is that these disorders arise from a common diathesis.

Suicidal Ideation and Behaviour Some studies have found that panic disorder is a risk factor for suicidal ideation and suicide attempts (e.g., Weissman et al., 1989). Other studies failed to replicate this finding (e.g., Beck et al., 1991; Rudd et al., 1993). Recent findings suggest that suicidal ideation and attempts are better predicted by disorders comorbid with panic disorder (e.g., major depression, substance abuse, or schizophrenia) than by panic disorder per se (Hornig & McNally, 1995). According to a recent consensus statement from experts on panic disorder, 'it remains unclear whether uncomplicated panic disorder, especially panic disorder without agoraphobia, is associated with a high risk of suicide attempts ... and whether suicide attempts by individuals with panic disorder are actually related to or caused by panic symptoms, as opposed to being a result of comorbid conditions' (American Psychiatric Association, 1998, p. 22). It may be that panic disorder sometimes leads the person to become depressed and suicidal. Thus, suicidal ideation and behaviour may sometimes be a complication of panic disorder. However, other comorbidity models may equally apply, such as a common diathesis model. Clinicians should be alert to the possibility of suicide risk in panic disordered patients. Indeed, it has been recently recommended that all patients presenting with panic attacks should be asked about suicidal ideation and past suicide attempts, and about conditions likely to increase suicide risk, such as depression and substance abuse (American Psychiatric Association, 1998). Even though panic attacks are characterized by fear of dying, the clinician should not assume that this precludes suicidal ideation. Patients may experience suicidal thoughts during panic attacks and in between attacks. The following case illustrates one way that fear of dying and suicidal ideation may coexist.

HYPOCHONDRIASIS 25

Mr K. was a 49 year-old laboratory technician with panic disorder and generalized anxiety disorder. His panic attacks were often triggered by chest tightness, accompanied by shooting pain in his chest and arms. He had been extensively examined by various medical specialists, but no organic abnormality was found. Yet his chest pain persisted and so he repeatedly requested further investigations, including potentially risky procedures such as cardiac catheterization. Mr K.'s cardiologist refused all requests for further testing, telling Mr K. that his pains and panics were simply due to stress. Whenever Mr K. experienced chest pain he would go to his computer and perform searches of the MEDLINE database for possible causes. Although his searches were intended to provide reassuring information, they often led him to learn of rare and deadly cardiac conditions, which further added to his anxiety. Mr K.'s chest pain and panic attacks were so frightening that he seriously contemplated suicide as a means of escaping his pain and suffering. He illicitly obtained syringes, saline, and a lethal amount of potassium, which he carried in his briefcase. During his panic attacks Mr K. feared that he would have a heart attack and die. Yet, during the attacks he also had thoughts of killing himself. He envisioned a lethal injection as quicker and less painful than a fatal heart attack.

HYPOCHONDRIASIS Hypochondriasis is defined by a preoccupation with fears of having, or the idea that one has, a serious disease. These fears and beliefs arise from misinterpretations of one or more body signs or symptoms (American Psychiatric Association, 1994). Among panic disordered patients presenting to general medical outpatient clinics, comorbid hypochondriasis is concurrently diagnosed in 16-25% of patients (Barsky et al., 1994; Noyes et al., 1994). In an outpatient anxiety disorders clinic, Furer et al. (1997) found that 10 (48%) of their 21 panic disordered patients also met DSMIV criteria for hypochondriasis. Hypochondriacal concerns can arise before, during, or after the onset of panic disorder (Furer et al., 1997). Panic disorder and hypochondriasis may tend to co-occur because both arise from a general tendency to catastrophically misinterpret body sensations. In other words, there may be a cognitive diathesis common to both disorders (see Chapter 3).

26

COMORBIDITY: PANIC DISORDER IN CONTEXT

PAIN-RELATED DISORDERS

Chronic Chest Pain It is not uncommon for people with panic disorder to present to primary care physicians with recurrent chest pain as their main complaint. Although their pain is typically a symptom of panic disorder, they may deny having panic attacks. Instead, they describe their symptoms in terms of unexpected bouts of chest pain (Katon, 1994). The other panic symptoms (e.g., anxiety, impending doom) are perceived by the patient as normal reactions to a life-threatening disease. This can lead the clinician to overlook panic disorder.

Other Types of Chronic Pain Compared to the general population, people with panic disorder have a higher prevalence of migraine headaches (Breslau & Davis, 1993; Stewart et al., 1989). Breslau and Davis, for example, found that people with a history of migraines, compared to those without, were almost 13 times more likely to develop panic disorder during a 14 month follow-up period. Other types of recurrent pain can co-occur with panic disorder. The pain can occur in between panic attacks and may increase in intensity during attacks. Kuch et al. (1991) assessed a consecutive series of 141 panic disordered patients presenting to an anxiety disorders outpatient clinic. Although none presented with pain as their chief complaint, 38% reported chronic pain; i.e., persistent pain for longer than 6 months, which interfered with vocational and leisure activities. A total of 8% reported daily use of analgesics. Although some patients reported chest pain, pain was most commonly located in the head, shoulders, and lower back. Panic disordered patients with chronic pain, compared to those without pain, had greater anxiety and depression, and more severe agoraphobic avoidance. Recent studies provide some clues as to why panic disorder and chronic pain tend to co-occur. Results of structural equation modeling supported the view that anxiety sensitivity-the fear of arousal-related sensationsproduces fear and avoidance of conditions associated with these sensations, such as pain. In turn, fear and avoidance of pain leads to greater pain severity and disability, by such mechanisms as analgesic abuse and muscle deconditioning (Asmundson & Taylor, 1996). Similarly, anxiety

SUBSTANCEUSE DISORDERS 27

sensitivity is a risk factor for panic attacks (see Chapter 3). Thus, a common diathesis model may account for the comorbidity, where anxiety sensitivity contributes to both panic disorder and chronic pain.

IRRITABLE BOWEL SYNDROME Irritable bowel syndrome is a chronic gastrointestinal disorder characterized by at least three months of abdominal pain and disturbed defecation (bloating, diarrhea, constipation, passage of mucus), in the absence of identifiable gastrointestinal pathology (Lydiard et al., 1994). It occurs in 8-17% of the adult US population (Drossman et al., 1982; Manning et al., 1978; Thompson & Heaton, 1980) and is likely to have similar rates of occurrence in other Western countries. People with panic disorder commonly have gastrointestinal symptoms (Katon, 1984). A recent epidemiologic survey found that symptoms of irritable bowel syndrome were more than four times more common in people with a history of panic disorder, compared to people with histories of other psychiatric disorders or no psychiatric disorder (Lydiard et al., 1994). In samples of patients presenting for treatment of panic disorder, the prevalence of concurrent irritable bowel syndrome ranges from 17-41% (Lydiard et al., 1991; Noyes et al., 1990a). It is not known why irritable bowel syndrome commonly occurs with panic disorder. It may be that both disorders share a common diathesis.

SUBSTANCE USE DISORDERS

Substances as Precipitants of Panic Panic disorder can be precipitated by the use of psychotropic drugs such as marijuana or cocaine (Ballenger & Fyer, 1996; Shulman et al., 1994). People who chronically use these substances may be at greatest risk for developing panic disorder. In a study of 95 cases of cocaine-induced panic it was found that panic attacks typically developed after years of cocaine use (Louie et al., 1996). Long-term abuse increases the likelihood that the person will experience intense intoxication- or withdrawal-related body sensations. These may be catastrophically misinterpreted, thereby resulting in panic attacks. In other words, panic disorder may sometimes arise as a complication of drug intoxication and withdrawal, particularly in people with elevated anxiety sensitivity.

28

COMORBIDITY: PANIC DISORDER IN CONTEXT

Alcohol Abuse and Dependence Elevated Co-Occurrence with Panic Disorder Alcohol misuse appears to play an especially important role in the precipitation, maintenance, and exacerbation of panic disorder. An epidemiological survey revealed that among people with panic disorder, 40% of men and 13% of women had alcohol abuse or dependence in the past six months (Leon et al., 1995). These rates were substantially higher than those for people with obsessive-compulsive or phobic disorders (men: 14-27%, women: 4-7%) and higher than those in people with no anxiety disorder (men: 7%, women: 1%). For substances other than alcohol, panickers had low rates of substance abuse or dependence (men: 3%, women: 5%). These rates were not higher than those with obsessive-compulsive or phobic disorders (men: 6-8%, women: 2-5%), but were somewhat higher than in people with no anxiety disorder (men: 2%, women: 1%; Leon et al., 1995). Panickers with alcohol abuse or dependency appear to have a more severe panic disorder than those who do not abuse or become dependent on alcohol or other sedative/hypnotics (Cox et al., 1990).

Self-Medication of Anxiety and Panic People with panic disorder may abuse alcohol (and other sedating drugs) to dampen their anxiety and panic symptoms. This is a complication model of comorbidity, where the person uses alcohol or other sedating substances to 'medicate' their panic disorder (Quitkin et al., 1972). This model is consistent with the finding that panic disorder typically precedes alcoholism in people with both disorders (Mullaney & Trippett, 1979; Stockwell et al., 1984). It is also consistent with patients' reports that they use alcohol to reduce anxiety and panic (Bibb & Chambless, 1986; Norton et al., 1989). Laboratory research also shows that panic disordered patients drink alcohol as a means of dampening anxiety and panic sensations (Kushner et al., 1996). There may be subgroups of panic patients who are at especially high risk for developing alcohol problems (Kushner et al., 1996). Individual differences in sensitivity to the anxiolytic properties of alcohol (Levenson et al., 1987), beliefs about alcohol use as a coping method, and lack of availability of alternative coping responses (Cooper et al., 1992) all may play a role. Gender-related socialization also may be important, where it is more socially acceptable for men to use alcohol as a coping method (Cox & Taylor, 1999). The use of alcohol as self-medication is illustrated in the following description, which was taken from the diary of Edvard Munch, the

SUBSTANCEUSE DISORDERS 29

Norwegian neo-impressionist among other problems.

painter, who suffered from panic attacks,

'I felt these frightening pains near my heart. I was frightened. In the morning, felt dizzy when I got up. Quickly something to calm me. Rang for service. Port wine, a half bottle-that helped. Felt better. Then some coffee and a bit of bread. Renewed anxieties. Out into the street, to the first restaurant, one drink, two drinks-that helped. Out into the street. Everything will be alright . . . . I drank and drank yesterday like all days, and went to bed in a stupor'. (Munch, 1909, cited in Heller, 1984, p. 190)

Munch's account also illustrates some of the common triggers to panic, such as postural hypotension (dizziness caused by standing up too quickly) and caffeine-induced anxiety.

Self-Medication of Withdrawal Symptoms

Acute withdrawal from alcohol produces increased activity of the central nervous system, which leads to arousal-related sensations such as tachycardia, tremulousness, sweating, nausea, and agitation (George et al., 1988). Alcohol withdrawal 3 SDs) compared to the other ESs. Once computed, mean ESs can be compared by performing statistical tests or comparing confidence intervals. When the mean ESs are based on a small number of trials then investigators sometimes interpret the results simply by visually inspecting the pattern of means. Treatment outcome moderators or predictors can be identified by correlating particular variables with the ESs obtained in each trial or study. For example, the percentage of people with major depression in each study can be correlated with the ESs to determine whether pretreatment depression predicts treatment response. For further details the reader should consult one of the many texts and articles on meta-analysis (e.g., Hunter & Schmidt, 1990; Rosenthal, 1991; Wolfe, 1986). Meta-analysis, compared to the conventional (narrative) review method, has the advantage of being less prone to reviewer biases (Light & Pillemer, 1984). However, meta-analysis is not without its limitations. The main concern is that the quality of the studies that go into the metaanalysis influence the validity of the conclusions that are drawn. However, the same is true of the conventional (narrative) review method. Moreover, meta-analytic methods can control for various aspects of the studies, such as methodological quality. Unfortunately, not all metaanalyses control for these factors.

META-ANALYSES OF PANIC DISORDER In reviewing the often complex patterns of meta-analytic findings we will focus on results that compared behavioural and cognitive-behavioural therapies to the other main therapies used in panic disorder. For some treatments (e.g., psychodynamic therapy) there were too few trials to be examined meta-analytically.

Meta-Analyses of Behaviour Therapy The earliest meta-analyses focused on the degree of clinically significant change produced by behaviour therapy, which consisted primarily of situational exposure to agoraphobic situations. Panic attacks were not the

88

META-ANALYSES OF TREATMENT OUTCOME

focus of treatment. Indeed, many of the studies included in the metaanalyses were conducted before panic disorder was officially introduced into the diagnostic nomenclature (i.e., DSM-III). Trull et al. (1988)

These authors conducted a meta-analysis of studies using the Fear Questionnaire (Marks & Mathews, 1979) to assess agoraphobic avoidance. The main findings of this study, along with its methodological strengths and weaknesses, are summarized in Table 5.1. Treatments consisted of unspecified behavioural therapies, administered for a mean of 13 h. For each study, mean scores were calculated before and after treatment and these were compared to mean scores from data from two control groups (college students and community controls) collected by Trull and colleagues. Using community norms, the average patient began treatment at the 97.3 percentile, and progressed to the 68.0 percentile at posttreatment, and the 65.5 at followup (median followup duration= 18 weeks). Findings were less impressive when compared to the student sample. Using student norms, the average patient began at the 99.9 percentile, and progressed to the 98.7 percentile at posttreatment, and 98.2 at followup. Trull et al. argued that community norms were more appropriate because, compared to the students, the community sample was demographically more similar to the patient samples. If this assumption is accepted then the results suggest that behaviour therapies produced clinically significant changes, with gains maintained at followup. Unfortunately, it is not known whether the followup results reflect the effects of additional treatment sought during the followup period. This is a problem of all metaanalyses reviewed in this chapter; none were able to determine whether additional treatment was sought during the followup period because this information is rarely reported in treatment outcome studies. Trull et al. found that treatments using situational exposure were associated with the greatest improvement compared to the other (unspecified) behavioural interventions. Patients treated by doctoral-level therapists (presumably psychologists or psychiatrists) improved more than patients treated by nondoctoral therapists (e.g., pre-doctoral graduate students). Group therapies produced better outcomes than individual (one-to-one) therapies. Studies of CBT2 appearing after Trull et al.'s meta-analysis suggest that group and individual treatments are equally effective (Lidren et al., 1994; Neron et al., 1995). Unfortunately, Trull et al. did not look at dropout rates or at the rates of improvement due to waiting list or placebo. However, it seems unlikely that patients in these conditions would move into the normal range of

META-ANALYSES OF PANIC DISORDER 89

functioning. Trull et al. 's study has the advantage of comparing studies on a common outcome scale, thereby avoiding the problem of different studies using measures that differ in terms of reliability, validity, and sensitivity to treatment effects. A disadvantage is that few outcome variables were assessed (Table 5.1). In principle, Trull et al.'s use of normal controls is a strength because it provides information on whether treatment caused patients to move into the normal range of functioning. An important concern, however, is whether these controls are representative of the population at large. Student controls are unlikely to be representative. Community controls were selected from a single location (Lexington, Kentucky) and it is not known whether this sample was representative of the US general population. A further problem was that the community samples completed a telephone-administered version of the Fear Questionnaire, whereas patients and student controls completed a pencil-and-paper version. It is not known whether these differences produced different results. Jacobson et al. (1988)

Another meta-analysis of behaviour therapy was published by Jacobson et al. (1988). The results, summarized in Table 5.1, were derived from the raw data of 11 previously published studies. Treatments consisted of situational exposure or imaginal exposure, sometimes combined with other interventions (e.g., cognitive restructuring, diazepam, eclectic psychotherapy). Within-trial effect sizes were averaged across measures. Most outcome measures assessed the severity of agoraphobia or related phobias. Averaging across studies, Jacobson et al. classified 58% of patients completing treatment as clinically improved, and 27% as ending treatment with little or no agoraphobia. Most patients maintained their gains between posttreatment and followup (mean followup duration = 6 months). A small proportion (5%) of patients were classified as having deteriorated during this interval. Recovery rates were greater for ratings of specific fears (e.g., the patient's worst fear) than for broader measures (e.g., global severity of agoraphobia). Involvement of the patient's spouse in therapy (e.g., as a coach or cotherapist) appeared to increase treatment efficacy. The use of a systematic program of home-based exposure exercises also seemed to improve efficacy. The results of this study suggest that a heterogeneous group of behavioural therapies was moderately effective in treating agoraphobia. Dropout rates were not examined. The percentage of treatment responders is likely to be lower if dropouts were taken into consideration. Another concern is whether the results are representative of other treat-

'-0 0

s: m

Table 5.1 Meta-analyses of behaviour therapy: Trull et al. (1988) and Jacobson et al. (1988)

:i>1 Author(s) & year

Period

Trull et al. (1988)

1975-87

# studies included

Main findings

19

Agoraphobia: * Behaviour therapy produced clinically significant reductions * Gains maintained at followup (median followup duration: 18 weeks) * Situational exposure was the most effective behavioural treatment * Group treatment was more effective than individual treatment * Doctoral therapists more effective than nondoctoral therapists

Methodological strengths and limitations

z

~

Depression: * Behaviour therapy produced clinically significant reductions

Strengths: * For a given outcome variable, all effect sizes were computed from a single scale * Compared whether treatment moved patients into the normal range of functioning Limitations: * Little information on the types of behavioural therapies included * Did not examine other treatments * Limited assessment of outcome variables (e.g., did not assess attrition) * Apparently did not assess for effect size outliers * Did not weight effect sizes by sample size * Representativeness of control groups was not established * Did not report whether different behavioural therapies differed in terms of treatment duration

continued

ti;:'"'

i,

~

(./)

m (./)

0-n -l

:;o

m

~

s: m z-l

0 C

r1 0

s: m

Table 5.1 (continued) Author(s) & year

Period

Jacobson et al. (1988)

1976-86

# studies

Main findings

Methodological strengths and limitations

included

11

Agoraphobia: * 58% of treatment completers were clinically improved * 27% of patients had little or no agoraphobia at the end of treatment * Gains generally maintained at followup (mean followup duration: 6 months) * Treatments more effective when spouses were involved and when structured home-based practice exercises were used

Strengths: * Examined clinical significance of treatment effects by classifying patients as clinically improved or recovered * Obtained raw data so that classifications could be calculated for each patient * Weighted effect sizes by sample size Limitations: * Did not examine other treatments * Apparently did not assess for effect size outliers * Representativeness of data sets was not established * Did not report whether different behavioural therapies differed in terms of treatment duration * Limited assessment of outcome variables (e.g., did not assess attrition)

s: m ~ )>

z)>

~ en m

en

0.,,

:;; z

()

Cl

en

0;;

z)>

~ (/)

m

(/)

0"Tl -i

;,:,

m

~

s: m

z-i 0 C -i

n 0

s: m

META-ANALYSES OF PANIC DISORDER 97

Between-trial effect sizes were computed, with medications compared to pill placebo. On several outcome measures (panic attacks, agoraphobia, and global improvement), antidepressants and benzodiazepines were more effective than placebo and equally effective to one another (Table 5.4). However, antidepressant trials tended to be longer than benzodiazepine trials (means = 16 and 7 weeks, respectively), which may have biased the results. It is possible that benzodiazepines may have been superior if administered for as long as antidepressants. Moreover, some types of benzodiazepines may have been more effective than others. High-potency benzodiazepines (e.g., alprazolam, clonazepam, and lorazepam) were developed as treatments of panic because the lowpotency benzodiazepines (e.g., diazepam) were thought to be ineffective (Gorman, 1994). Accordingly, it would have been useful if Wilkinson et al. had examined the efficacy of specific medications. Boyer (1995)

In a later study, Boyer (1995) conducted a meta-analysis of placebocontrolled, double-blind studies of SSRis, imipramine, and alprazolam. SSRls included paroxetine, fluvoxamine, zimeldine, and clomipramine. Boyer found that SSRis were significantly superior to both imipramine and alprazolam (Table 5.4). For imipramine and alprazolam, higher dose was associated with larger effect size. It was not possible to examine the dose-response relationship for all SSRis because different drugs were used. However, for the five fluvoxamine studies the relationship between dose and effect size was not significant. The superiority of the SSRis remained significant, but was less pronounced, when they were compared to studies that used higher doses of imipramine (>150mg/day) or alprazolam (>4mg/day). Thus, it is unlikely that the superiority of SSRls was due to under-dosing of imipramine or alprazolam. All drug trials tended to be brief (median= 8 weeks) and so it is not known whether the pattern of results would be maintained for longer duration of treatments, or when patients were withdrawn from their medications. Comment

Meta-analysis of drug studies suggests that antidepressants and benzodiazepines are more effective than placebo, and that SSRls may be the most effective drugs of all. Attrition tended to be lower for SSRI than nonSSRI antidepressants, which further underscores the potential value of SSRls.

'-0

CXl

~ m

Table 5.4 Meta-analyses of pharmacotherapies: Wilkinson et al. (1991) and Boyer (1995) Author(s) & year

Period

# studies

Main findings

Methodological strengths and limitations

► )>

~

~

included

(/l

m

Wilkinson et al. (1991)

1964---90

19

Attrition: Globalclinicalratings: Panicattacks: Agoraphobia:

PP>AD> BDZ (AD, BDZ) > PP (AD, BDZ) > PP (AD, BDZ) > PP

Strengths: * Examined median effect sizes because means were skewed * Effect sizes appear to have been weighted by sample size * Homogeneous group of studies: included only double-blind placebo-controlled studies Limitations: * Examined drug treatments only * Did not control for differences in treatment duration (AD mean = 16 weeks, BDZ mean= 7 weeks) * Did not examine subclasses of drugs (e.g., low vs. high potency BDZs) * Did not report followup data * Dropout rates were examined 'qualitatively' rather than metaanalytically. No data on mean dropout rates were reported

(/l

0 "Tl

;;! m

~

~

m

z-l 0 C

r1

~m

~~.::, 1

•------•--·•

Boyer (1995)

1982-93

27

%

panic-free:

.a

•-•--••~•-v-

•-•-•

(SSRis, imipramine, alprazolam) > placebo SSRis > (imipramine, alprazolam) SSRis > (high-dose imipramine, high-dose . alprazolam) . . 1m1pramme= alprazolam

-v-.-.------•--•-~-

---------------------

Strengths: * Examined subgroups of pharmacotherapies * Examined effects of high-dose treatments * Homogeneous group of studies: included only double-blind placebocontrolled studies Limitations: * Examined drug treatments only * Did not examine attrition * Did not report whether outlying effect sizes were omitted or whether effect sizes were weighted by sample size * Did not assess whether treatments differed in duration * Did not compare treatments in terms of their efficacy for agoraphobia, anxiety, or depression * Followup results not reported

AD Antidepressant medications (monoamine oxidase inhibitors or tricyc!ic antidepressants) BDZ Benzodiazepines (low or high potency) PP Pill placebo SSRI Selective serotonin reuptake inhibitors (e.g., fluvoxamine) For attrition, x > y indicates that treatment x was associated with a higher dropout rate. For other outcome variables, x > y indicates that treatment x was more effective than treatment y.

s:. m 5>' )>

z)>

'::::

(/)

m (/)

0 ,,

:{;

z

(")

CJ (/)

0 ;;o CJ m

;;o -0 -0

100

META-ANALYSES OF TREATMENT OUTCOME

Meta-Analyses of Behavioural, Cognitive-Behavioural, and Drug Treatments Over the past decade several large meta-analyses have been published comparing drugs treatments to behavioural and cognitive-behavioural therapies. The findings of these studies, along with a summary of their major strengths and weaknesses, are summarized in Tables 5.5-5.9. For our purposes, the studies all have an important weakness, which is not listed in the tables because it applies to all of the studies. None of the studies evaluated CBTl separately from CBT2. Given that CBT2, unlike CBTl, was specifically designed to eliminate panic attacks, a comparison between the two would have been informative, particularly in light of Chambless and Gillis' (1993, 1994) finding that CBT2 was among the most effective treatments (Table 5.3). There are two other methodological weaknesses that are common to some of the meta-analyses in the tables. The first is that some combined all antidepressant medications into a single 'antidepressant' category, which included SSRis (e.g., fluvoxamine) and tricyclics (e.g., imipramine). Recall that Boyer (1995) found that SSRis were more effective than imipramine and alprazolam (Table 5.4). An important question is whether SSRis are more effective than behavioural and cognitive-behavioural therapies. A second weakness concerns the treatment followup results from all the meta-analyses reporting those data. Followup results are difficult to interpret because it is not known whether patients received additional treatment during the followup interval. Additional treatment could account for any apparent 'maintenance' of treatment gains from posttreatment to followup. Despite their limitations, the following meta-analyses have been influential in shaping the views of researchers and practitioners, and so they merit review. In the following sections we will review these metaanalyses, and attempt to determine which conclusions can be methodologically justified. Mattick et al. (1990) In one of the earliest attempts to determine the comparative efficacy of drug and psychological treatments, Mattick and colleagues computed within-trial effect sizes for several outcome variables (Table 5.5). A concern with meta-analyses is that they typically include low- and highquality research. Narrative reviews often exclude or downplay the importance of low-quality studies. Mattick et al. addressed this issue by including only data from studies using accepted treatment procedures.

Table5.5 Meta-analysis of drug and psychological treatments: Mattick et al. (1990) Period

# studies

Main findings

Methodological strengths and limitations

Panicattacks:(Alprazolam =imipramine) > PP Imipramine + situational exposure = imipramine = situational exposure CBT (alprazolam = PP) CBT y indicates that treatment x was more effective than treatment y.

0 ;u 0 m ;u

~

0 ~

0

N

;:

Table 5.6 Meta-analysis of drug and psychological treatments: Cox et al. (1992a)

Period

# studies

Main findings

m

Methodological strengths and limitations

included 1980-90

34

Attrition: Globalseverity:

Imipramine (29%) > situational exposure (15%) > alprazolam (10%) Imipramine = alprazolam =situational exposure Mean effect sizes for all three conditions >0

Panicattacks:

Agoraphobia:

Anxiety:

Imipramine = alprazolam = situational exposure Only alprazolam mean effect size was > 0 Imipramine = alprazolam = situational exposure Mean effect sizes for alprazolam and situational exposure > 0 Imipramine = alprazolam = situational exposure Mean effect sizes for all three conditions >0

Depression:

Imipramine = alprazolam = situational exposure Mean effect sizes for imipramine & situational exposure > 0

For attrition, x > y indicates that treatment x was associated with a higher dropout rate. For other outcome variables, x > y indicates that treatment x was more effective than treatment y.

-~;,:;::,.,.;,,;:

Strength: • Examined specific treatments rather than groups of treatments; e.g., imipramine instead of antidepressants as a group. Limitations: * Effect sizes were not adjusted for outlying values. • Effect sizes were not weighted by sample size. • Examined only three treatments. • Did not compare treatments to control conditions such as pill placebo or waiting list. * Did not examine followup data. * Did not report whether treatment groups differed in terms of treatment duration.

j;;

z:i> )> ~ m

(/J

(/J

0 "TI -I

;u

m

~

;:

m

z-I 0

C -I

n 0 ;:

m

Table 5.7 Meta-analysis of drug and psychological treatments: Clum et al. (1993) Period

# studies

Main findings

Methodological strengths and limitations

included 1964--90

29

Attrition:

PP > drug treatments AD>BDZH Psychological coping treatments > (PP, PsyP, WLC)

Panicattacks:

(AD, BDZH, psychological coping, drug + psychological coping) > (PP, PsyP) Agoraphobia: (AD, BDZH, BT,psychological coping, drug + psychological coping) > (PP, PsyP) Anxiety: (AD, BT,psychological coping, drug + psychological coping) > (PP, PsyP) BDZH = PP, PsyP Depression: (AD, drug + psychological coping) > (PP, PsyP) BDZH = PP, PsyP

Strengths: * Only controlled studies were included. This could be taken as a methodological strength, although it resulted in a small number of studies. Moreover, uncontrolled studies can be used if they yield the same magnitude of effects as controlled studies. Limitations: * Effect sizes were not adjusted for outlying values. * Effect sizes were not weighted by sample size. * For many analyses, confounded type of treatment with type of control group. Results for unconfounded comparisons were based on very small numbers of studies, thereby precluding statistical tests. * Did not control for treatment duration. * Did not examine followup data.

AD Antidepressant medications (selective serotorun reuptake inhibitors, monoamine oxidase inhibitors, or tricyclic antidepressants) BDZH High-potency benzodiazepines (e.g., alprazolam, clonazepam, lorazepam) BT Behaviour therapy (typically situational exposure) PP Pill placebo PsyP Psychological placebo (e.g., attention placebo) WLC Waiting list control + Indicates that one treatment was combined with another; e.g., imipramine combined with situational exposure For attrition, x > y indicates that treatment x was associated with a higher dropout rate. For other outcome variables, x > y indicates that treatment x was more effective than treatment y.

s: m ► )>

z)>

~ (/) m (/)

0 ,, ~

z ()

'O (/)

0 ;;o 'O

m

;;o 0

w

Table 5.8 Meta-analysis of drug and psychological treatments: Gould et al. (1995) Period

# studies

Main findings

Methodological strengths and limitations

included

0

.I>,.

1974-94

43

Attrition: Globalseverity:

Panicattacks: Costof treatment($):

AD (25%) > BDZ (13%) (Drugs (20%) =drugs+ CBT (22%)) > (CBT(6%)= controls (7%)) (CBT alone, drugs alone, drugs + CBT) > controls CBT most effective, especially CR + interoceptive exposure CBT ::o:CBT + drugs ;:o:drugs Situational exposure + imipramine = imipramine AD=BDZ CBT gains maintained at followups of at least 12 months CBT > drugs > controls Imipramine + situational exposure= imipramine Imipramine and group CBT had the lowest cost

Strengths: Examined effect sizes at treatment followup Compared treatments in terms of costs * Examined effect sizes for global response to treatment * *

Limitations: * Confounded type of treatment with type of *

* *

* *

*

AD BDZ CBT Controls +

Antidepressant medications (selective serotonin reuptake inhibitors, monoamine oxidase inhibitors, or tricyclic antidepressants) Benzodiazepines (low or high potency) Cognitive-behaviour therapy (CBTl or CBT2) Included waiting list, pill placebo, and psychological placebo Indicates that one treatment was combined with another; e.g., imipramine combined with situational exposure

For attrition,

Ill'''''

'Pc,,,..;o-t\,,,e-.,.

control group Did not report whether outlying effect sizes were omitted or whether effect sizes were weighted by sample size Did not assess whether treatments differed in terms of duration Did not compare treatments in terms of their efficacy for agoraphobia, anxiety, or depression Did not separately analyze high- and lowpotency BDZs 'CBT' grouping was very broad, including various behavioural and cognitivebehavioural therapies, bibliotherapy, paradoxical intention, and other interventions For several comparisons, differences between effect sizes were not statistically examined. Conclusions were based on visual inspection of the data

x > y indicates

00 ~ .. ~ ..,.._,,.,..,a.l,-,1-~·,.,,

that treatment .,._,.,

i-.d.,o,:,--

x was associated ,.,...._.. + ,._-

.. ~-.,..~·.,.

with a higher ....,. __

...,...,...,..

dI"?pout rate.

--~o-i...--

,-.1,_.,.

-...,.

.. .,..._...,

Table 5.9 Meta-analysis of drue: anc:l_Qsy:choloirkaltn~atmPnts: v;in Rr1lknm

pf

.......

nl /1QQ7,

s m

~ )>

z)>

~ Cf> m Cf>

0 ,, --j

:;o

m

~

sm z --j

0

C

(i

@ m

---------------------------~

.....

&

For

---

-·•

..

.n""' ...., .... --

other

outcome

11.....'-U ... au;:.:,ulcu ---

variables,

ut::auut=uL.x. was assoc1atea w1tn a higher dropout rate.

x > y indicates

that

treatment

x "'as

more

effective

than

treat::rnent

y.

Table 5.9 Meta-analysis of drug and psychological treatments: van Balkom et al. (1997)

Period

1964--95_

# studies included 106

Main findings

Attrition:

AD = AD + BT = BDZH = BT = PPM + BT = PPM = PP + BT = PP, PsyP, WLC (mean dropout rate across all studies= 20%) Panic attacks: (AD, AD+ BT, BDZH, PPM) > (PP, PsyP, WLC) (BT,PP + BT, PPM + BT) = (PP, PsyP, WLC) AD = AD + BT = BDZH = BT = PPM + BT =PPM= PP+ BT Agoraphobia: (AD, AD + BT, BDZH, BT, PPM + BT, PPM, PP + BT) > (PP, PsyP, WLC) (AD+ BT)> (AD, BDZH, BT, PPM+ BT, PPM) AD +BT= PP + BT AD = BDZH = PPM + BT = PPM = PP+ BT (AD, AD + BT, BDZH, BT, PPM + BT, PPM) Anxiety: > (PP, Psy P, WLC) PP+ BT = (PP, PsyP, WLC) AD + BT > (BT,PPM + BT} AD= AD + BT = BDZH = PPM = PP + BT (AD, AD + BT, PPM) > (PP, PsyP, WLC) Depression: BDZH = PP + BT = (PP, PsyP, WLC} AD + BT > (BT,PPM + BT) AD = AD + BT = BDZH = PPM = PP + BT

Methodological strengths and limitations

Strengths: * Excluded outlying effect sizes from analyses * Rated the quality of each study and assessed effects of quality on treatment outcome Limitations: * Effect sizes not weighted by sample size * Did not look at whether controlled trials had different effect sizes from uncontrolled trials, or whether treatment conditions differed in the % of controlled trials * Did not examine separate sub-classes of treatments (e.g., selective serotonin reuptake inhibitors, CBT2) * Did not assess whether treatments differed in terms of the number or duration of sessions * Unduly broad classification of treatments as 'psychological panic management'

s: m j;;

}>

z)>

~ (/) m (/)

0 ""T1

":i;.

z n

CJ

Antidepressant medications (selective serotonin reuptake inhibitors, monoamine oxidase inhibitors, or tricyclic antidepressants) BDZH High-potency benzodiazepines (e.g., alprazolam, clonazepam, lorazepam) Behaviour therapy (typically situational exposure) BT pp Pill placebo Psychological panic management (e.g., paradoxical intention, bibliotherapy, CRl, CR2, progressive muscle relaxation, breathing retraining, PPM imaginal exposure). Situational exposure was not included in PPM. PsyP Psychological placebo (e.g., attention placebo) WLC Waiting list control + Indicates that one treatment was combined with another; e.g., imipramine combined with situational exposure For attrition, x > y indicates that treatment x was associated with a higher dropout rate. For other outcome variables, x > y indicates that treatment x was more effective than treatment y.

AD

vi

0 ;u

CJ

m ;u 0

u,

l 06

META-ANALYSES OF TREATMENT OUTCOME

The authors also conducted regression analyses to identify relations between study features and effect sizes for agoraphobia. These effect sizes were uncorrelated with treatment duration, proportion of treatment dropouts, and type of outcome measure. On all outcome measures, imipramine and alprazolam were equally effective, and both more effective than pill placebo. Monoamine oxidase inhibitors and diazepam were no better than pill placebo. The most effective of the behavioural treatments were the ones that used situational exposure. On measures of panic attacks, combined imipramine and situational exposure was no better than imipramine alone or situational exposure alone. However, on measures of agoraphobia, anxiety, and depression, the combination treatment was most effective. CBT (consisting largely of CBTl trials) tended to be more effective than situational exposure and more effective than imipramine on most outcome measures except agoraphobia. Here, CBT was less effective than situational exposure. Effects of situational exposure (with or without imipramine) and CBT appeared to be maintained at followup (mean followup durations, situational exposure: 16 months, CBT: 8 months). Cox et al. (1992a)

These investigators focused on what were the main treatments at the time: imipramine, alprazolam, and situational exposure. CBTs were classified as situational exposure if they emphasized this form of exposure and did not focus on interoceptive exposure. Dropouts were most frequent for imipramine, followed by situational exposure and alprazolam (Table 5.6). For most treatments and outcome variables, the within-trial effect sizes were significantly greater than zero. This finding is difficult to interpret because waiting list controls and placebo conditions also have effect sizes greater than zero (Mattick et al., 1990). There were no differences between imipramine, alprazolam, and situational exposure on any of the outcome measures. However, like all meta-analyses of panic disorder, the number of trials tended to be small (across treatments, there was a mean of five trials for each outcome measure). There may have been insufficient statistical power to detect differences between treatments. Clum et al. (1993)

This was a meta-analysis of controlled outcome studies. Thus, betweentrial effect sizes were computed for a range of outcome variables (Table 5.7). Treatments included high-potency benzodiazepines, antidepressants, other pharmacotherapies, behaviour therapy (situational exposure

META-ANALYSES OF PANICDISORDER l 07

and/ or imaginal exposure), 'psychological coping treatments', and combined pharmacological and psychological treatments. 'Psychological coping treatments' was a heterogeneous category consisting of numerous techniques and combinations of techniques (e.g., cognitive restructuring, relaxation training, paradoxical intention, and bibliotherapy, sometimes in conjunction with exposure therapies). It was not reported whether some of these 'coping treatments' were more effective than others. Attrition was greater for pill placebo than for drug treatments. Within the drug treatments, there were more dropouts for antidepressants than for high-potency benzodiazepines. Psychological treatments had more dropouts than controls. However, these analyses were based on a very small number of studies and so may be unreliable. On a range of outcome variables, most treatments were effective relative to controls. Behaviour therapy and psychological coping treatments were among the most effective, whereas on some measures high-potency benzodiazepines were no better than placebo. Pretreatment severity of agoraphobia was unrelated to treatment outcome. Gould et al. (1995)

These authors found that CBTs also had the lowest attrition rates (Table 5.8). In terms of global outcome, CBTs were most effective, followed by the combination of CBT and pharmacotherapy, and then followed by pharmacotherapy alone. Among the CBTs, treatments combining cognitive restructuring with interoceptive exposure were the most effective. Imipramine and group CBT were the least expensive treatments. Pretreatment severity of agoraphobia was unrelated to global outcome. At followups of 6 months or more, CBTs were the most successful at maintaining gains. For imipramine and alprazolam, gains tended to be lost over time when patients were tapered off their medications. Gains were maintained for combined situational exposure and pharmacotherapyi.e., treatments in which patients received these interventions during the acute phase and then treatment was discontinued. Unfortunately, it is not clear whether patients sought additional treatment during the followup interval, and so the followup results are difficult to interpret. van Balkom et al. (1997)

In the mos.t recent and largest meta-analysis to date, van Balkom et al. (1997) examined the efficacy of drug and psychological treatments for panic disorder. This was an extension of an earlier meta-analysis (van Balkom et al., 1995) and so only the more recent study will be reported

108

META-ANALYSES OF TREATMENT OUTCOME

here. Excluded were studies of beta-blockers, buspirone, psychoanalysis, and other treatments because of insufficient numbers of relevant trials. To simplify analyses, the authors created a single class of control conditions, consisting of a combination of pill placebo, psychological placebo, and waiting-list controls. They had planned to analyze completer and intentto-treat data, but found that the latter was not reported in 92% of studies. Accordingly, like the other meta-analyses, they examined only treatment completers. van Balkom and colleagues classified the following techniques under the category of 'psychological panic management' (PPM): bibliotherapy, paradoxical intention, progressive muscle relaxation, and treatments consisting of all or some of procedures used in CBTl or CBT2 apart from situational exposure (van Balkom et al., 1997; Anton J. L. M. van Balkom, personal communications, 22 September and 10 August 1998). Thus, PPM included breathing retraining alone, imaginal exposure alone, but not treatments consisting largely of situational exposure. Some treatments classified as PPM were developed before the second generation cognitive-behavioural therapies became available, and were not designed to eliminate unexpected panic attacks (e.g., Emmelkamp et al., 1978, 1986; Emmelkamp & Mersch, 1982; Emmelkamp & Wessels, 1975). Oddly enough, one treatment was simply the imaginal equivalent of situational exposure to agoraphobic situations (Emmelkamp & Wessels, 1975). It is unclear why these treatments were classified as 'psychological panic management.' Indeed, an impetus for developing second generation CBT was to overcome the limitations of the older therapies (Barlow, 1994). van Balkom et al. noted that the effect sizes varied widely for PPM. This is not surprising, given the heterogeneity of treatments included in this category. The authors found that the following did not predict treatment outcome: duration of disorder, percentage of dropouts, percentage of females, age at start of treatment, percentage of patients with panic disorder without agoraphobia, and overall quality of study. Other findings are summarized in Table 5.9. Analysis of posttreatment effect sizes led van Balkom et al. to conclude that overall, antidepressants combined with situational exposure was the most potent short-term treatment for panic disorder with or without agoraphobia. As with all the meta-analyses reviewed in this chapter, van Balkom et al.'s (1997) study suffered from a number of shortcomings that undermine the strength of their conclusions. Among the most important are (1) it is not clear whether treatments differed in terms of duration, (2) CBT2 was not specifically examined, and (3) the category called 'psychological panic management' was so heterogeneous that the effect sizes for this category

SUMMARYAND CONCLUSIONS 109

are difficult to interpret. Given the superiority of SSRis over other antidepressants (Boyer, 1995), it would have been useful to see a comparison between SSRis and CBT2. van Balkom and colleagues also conducted a meta-analysis of the studies reporting followup data (mean followup interval= 1.2 years) (Bakker et al., 1998). Treatment gains appeared to be maintained. However, it was not known whether the results were influenced by patients seeking further treatment between the end of the target treatment and the end of the followup period. Followup data were available for only 15% of patients included in the meta-analysis of post-treatment effect sizes, and for many followup studies it was not clear whether patients initially treated with pharmacotherapy were still on medication. Accordingly it is difficult to make much of the followup data. Comment The meta-analyses summarized in Tables 5.5-5.9 differ in a number ways, including the studies included, meta-analytic methods, grouping of treatment conditions, and to some extent, in the conclusions that were drawn. Each of the meta-analyses has its weaknesses, thereby rendering all conclusions tentative. The studies suggest that attrition from behavioural and cognitive-behavioural therapies tends to be equal to, or lower than, attrition from pharmacotherapies. Antidepressants, high-potency benzodiazepines, situational exposure, and CBTs appear to be equally effective to one another, and more effective than control conditions. However, important differences in treatment efficacy may exist within each class of treatment (e.g., within the class of CBTs, CBT2 may be more effective than CBT1). Such differences were not adequately examined in most metaanalyses. No reliable predictors or moderators of treatment outcome were identified, although meta-analysis is a crude method for investigating these variables. It may be asked whether psychological and drug studies enlisted patients of comparable severity. There is no reason to expect any differences. Similar inclusion/ exclusion criteria were used, along with similar recruitment procedures. Pretreatment scores on various clinical measures suggest that the patients enrolled in psychological treatments were as severe as those enrolled in drug studies.

SUMMARY AND CONCLUSIONS Meta-analysis provides a broad perspective on the efficacy of treatments. Individual differences are obscured or examined in a very limited way.

110 META-ANALYSES OF TREATMENT OUTCOME

Part of the problem is that treatment outcome trials often do not report sufficient information to perform a good meta-analysis. In general, the meta-analyses suggest that most of the major treatments for panic disorder were reasonably well-tolerated, and at least moderately effective. Most treatments appear to influence most of the major symptom domains. Tricyclic antidepressants (e.g., imipramine) and high-potency benzodiazepines (e.g., alprazolam) tend to be equally effective, although antidepressants tend to have higher attrition. Compared to tricyclics, SSRis tend to have fewer side effects and therefore may be associated with lower attrition. SSRis tend to be more effective than non-SSRI antidepressants and high-potency benzodiazepines. Situational exposure consistently emerged as an effective, well-tolerated treatment. An important question is whether there are ways of improving the efficacy of situational exposure. The use of structured exposure exercises, and addition of some forms of interpersonal skills training appears to be useful. Several studies investigated the effects of supplementing situational exposure with cognitive restructuring. The results offered mixed support for this combination, possibly because in some studies the cognitive restructuring was not based on the cognitive models described in Chapter 3. In other words, the focus was not on correcting catastrophic misinterpretations of body sensations. As we will see in Chapter 6, cognitive methods focusing on these misinterpretations appear be more effective than generic or non-specific cognitive restructuring. Few meta-analyses specifically examined CBT2, as distinct from CBTl. Although CBT2 is effective and well-tolerated compared to control conditions, there is insufficient evidence from the meta-analyses to determine whether it is more effective than other treatments. An important question is whether CBT2 is more effective than SSRis, which appear to be the most effective anti-panic drugs. Another important question is whether treatment efficacy is improved by combining CBT2 with medications such as SSRis. In the following two chapters we will explore these unanswered questions by taking a closer look at the individual treatment-outcome studies, including studies that have been completed after the metaanalyses were published.

Chapter 6

TREATMENT OUTCOME: A CLOSER LOOK In the previous chapter we saw that meta-analyses indicate that cognitive-behaviour therapies (CBTs) are among the most effective treatments, but possibly no better than many other interventions. The metaanalyses should be interpreted cautiously because they had numerous limitations, such as the tendency to lump first- and second-generation CBTs together. The purpose of the present chapter is to examine treatment outcome in more detail. We will address the following issues concerning the efficacy and modes of action of treatments for panic disorder, with particular attention to second-generation CBT (CBT2). •

•

•

•

•

Changes in vulnerability factors. Does CBT2 lead to changes in variables that theoretically cause panic attacks, such as elevated anxiety sensitivity? Do other treatments produce these changes? If so, how might this occur? What are the most important ingredients? CBT2 contains a number of different treatment components. Which components or combinations of components tend to produce the best treatment outcome? Comparative efficacy. How does CBT2 fare in studies designed to directly compare it with other therapies? In this chapter we will examine the comparative efficacy of monotherapies (e.g., CBT2 vs. imipramine). Combination therapies will be examined in the following chapter. Rates and patterns of change. How fast does CBT2 exert its therapeutic effects? What are the major patterns of treatment-related change? For example, is desynchrony a major pattern of outcome, where some symptoms (e.g., panic attacks) are reduced faster than others (e.g., agoraphobic avoidance)? Does CBT2 differ from other treatments in terms of rates and patterns of treatment-related change? Breadth of therapeutic effects. How broad are the treatment effects in terms of the range of clinical variables that are influenced? Are the therapeutic effects of CBT2 and other panic treatments limited to symptoms of panic disorder, or do these therapies influence a broad range of symptoms and domains of functioning?

112

•

TREATMENTOUTCOME: A CLOSER LOOK

Efficacy in clinical service settings. How effective is CBT2 when it is taken out of the research clinic and applied to unselected panic patients in routine clinical settings?

TREATMENT-RELATED CHANGES IN VULNERABILITY FACTORS CBT2 Reduces Anxiety Sensitivity Cognitive models of panic (Chapter 3) predict that panic attacks can be reduced in frequency and severity by either reducing the strength of belief in the dangerousness of arousal-related sensations (as indicated by reductions in anxiety sensitivity) or by teaching the patient ways of reducing or avoiding feared sensations. The latter can be achieved by breathing exercises to reduce hyperventilatory symptoms, and by relaxation techniques to reduce arousal. Reducing anxiety sensitivity is the preferred approach. To merely teach patients ways of dampening sensations is to provide them with a crutch for coping with panic, which does not necessarily help them learn that the sensations are harmless. From this discussion we can see that cognitive models predict that reduction in anxiety sensitivity is a sufficient but not necessary condition for reducing panic attacks. The outcome literature supports this prediction. Several studies have shown that reductions in anxiety sensitivity (and reductions in scores on measures of closely related constructs) are accompanied by reductions in panic frequency and severity (e.g., Bouchard et al., 1996; Clark et al., 1994, 1997c; Hoffart, 1995; Ost & Westling, 1995; Otto & Reilly-Harrington, 1999; Schmidt et al., 1997c). Telch (1996) obtained results from a statistical mediator analysis (Baron & Kenny, 1986) that also supported the hypothesis that reductions in anxiety sensitivity mediate reductions in panic frequency.

Other Treatments Also Reduce Anxiety Sensitivity Psychosocial Interventions CBT2 is not the only treatment capable of reducing anxiety sensitivity. Situational exposure, CBTl, and applied relaxation all reduce anxiety sensitivity (McNally & Foa, 1987; Ost & Westling, 1995; Sochting et al., 1998). Situational exposure and applied relaxation may reduce anxiety sensitivity because both methods encourage patients to enter feared situations,

TREATMENT-RELATED CHANGES IN VULNERABILITY FACTORS 113

which provides patients with the opportunity to disconfirm their beliefs about the catastrophic consequences of arousal-related sensations.

Pharmacotherapies Anti-panic medications such as imipramine and fluoxetine also tend to reduce scores on measures of anxiety sensitivity and related constructs, and these changes are greater than those occurring in waiting-list controls (Clark et al., 1997c; Otto & Reilly-Harrington, 1999). There are several ways that pharmacotherapies could reduce anxiety sensitivity. One possibility is that patients change their attributions when the prescribing clinician tells them to expect drug side effects. Fluoxetine side effects include sweating, chills, dry mouth, agitation, dizziness, and nausea. Imipramine side effects include tachycardia, dry mouth, blurred vision, dizziness, and vertigo. Alprazolam side effects include nausea and dizziness (Canadian Pharmaceutical Association, 1990). Panic disordered patients tend to catastrophically misinterpret sensations such as these. When placed on an anti-panic medication and told about the side effects, the patient may come to reinterpret many arousal-related sensations as medication side effects. Instead of interpreting the sensations as a sign of impending disaster, the patient may re-interpret them as signs that the medication is exerting its therapeutic effects. Thus, information about side effects may change the patient's beliefs about arousal-related sensations. In this way medications might reduce anxiety sensitivity and thereby reduce the frequency of panic attacks. This hypothesis may provide a partial explanation, but is unlikely to be the whole story. Not all arousal-related sensations can be interpreted as medication side effects, and not all patients are encouraged by their doctors to view these sensations as side effects. Therefore, it is necessary to consider other explanations of how anti-panic medications reduce anxiety sensitivity. One possibility derives from Lazarus' (1966, 1991) work on stress and coping. According to Lazarus, perceived threat is a product of two types of appraisal: primary appraisal ('Is is dangerous?') and secondary appraisal ('Can I cope?'). Anti-panic medications may change secondary appraisals without necessarily changing primary appraisals. Some patients may come to believe that their medications prevent the feared catastrophe from occurring; e.g., 'Palpitations are dangerous, but my Ativan prevents my heart from beating too fast.' Such patients report that they are less afraid of arousal-related sensations becausethey believe they have medication to ward off feared catastrophes. Thus, beliefs about medications may lead to reductions in anxiety sensitivity. As we will see in later chapters, these beliefs constitute a two-edged

114 TREATMENT OUTCOME:A CLOSERLOOK

sword; they may help patients in the short term, but also increase the risk for relapse once medications are discontinued.

Panic Provocation Challenges Cognitive models of panic predict that any treatment reducing anxiety sensitivity should also reduce the tendency for patients to panic when they are administered substances that induce arousal-related sensations, such as sodium lactate or air enriched with CO 2 . In support of this prediction, anti-panic drugs not only reduce anxiety sensitivity, but also reduce the panicogenic effects of sodium lactate and CO 2 (Fyer et al., 1983; Kelly et al., 1971; Liebowitz et al., 1984; Rifkin et al., 1981; Sanderson et al., 1994). Similarly, CBT2 reduces anxiety sensitivity and also reduces the effects of lactate and CO 2 (Brown et al., unpublished 1990, cited in Barlow, 1994; Clark et al., unpublished 1990, cited in McNally, 1994; Schmidt et al., 1997c; Shear et al., 1991b). The cognitive models of panic suggest that the treatment producing the greatest reductions in anxiety sensitivity should also have the strongest effect in reducing the panicogenic effects of lactate and CO 2 • CBT2 may be among the most powerful treatments in this regard, because it produces among the largest reductions in scores on anxiety sensitivity and related constructs (Clark et al., 1994, 1997c; Otto & Reilly-Harrington, 1999).

Comment A range of effective panic treatments can reduce anxiety sensitivity, including CBT2 and some pharmacotherapies. Some conjectures were offered as to how drugs produce this effect. As with anti-panic drugs, CBT2 reduces the panicogenic effects of sodium lactate and CO 2 • These findings are consistent with the cognitive models of panic; treatments that reduce the tendency to catastrophically misinterpret arousal-related sensations should also attenuate the panicogenic effects of substances that induce these sensations. It is not known whether panic treatments reduce other cognitive factors associated with panic disorder. One study found that CBT2 did not reduce interoceptive acuity (Antony et al., 1994). Although CBT2 may not alter the patient's ability to detect arousal-related sensations, it may reduce the tendency to attend to the sensations. In other words, it may reduce body vigilance (Schmidt et al., 1997b). Indeed, if patients no longer fear these sensations then they should no longer be vigilant for them.

CBT2:WHATARETHEMOST IMPORTANT INGREDIENTS?115

CBT2: WHAT ARE THE MOST IMPORTANT INGREDIENTS? CBT2 treatment packages include a number of components, such as psychoeducation (e.g., information about the nature and treatment of panic disorder), breathing retraining, cognitive restructuring, relaxation exercises, interoceptive exposure, and situational exposure. Exposure exercises are often framed as 'behavioural experiments' to test patients' beliefs about the consequences of arousal-related sensations. CBT2 packages also contain 'nonspecific' components, which are found in a wide range of therapies. These include factors facilitating the therapeutic alliance (e.g., the therapist's expressions of genuineness, accurate empathy, and positive regard toward the patient), and factors that mobilize hope and foster expectations of improvement (e.g., presenting patients with a convincing treatment rationale). Some of the meta-analyses in Chapter 5 suggest that CBT2 may be no more effective than a number of other treatments. Does that mean that the efficacy of CBT2 is entirely due to factors common to many panic treatments (nonspecific factors)? If not, then which components of CBT2 make the most important contributions to treatment efficacy? Can we streamline treatment by omitting some components? Which combinations of components tend to produce the most effective treatment? The following sections will examine the literature on these issues.

Nonspecific

Factors

Nonspecific factors contribute to the efficacy of psychological and pharmacological treatments for most if not all psychological disorders (Frank, 1973; Kirsch, 1999; Schaap et al., 1993). CBT2 focuses on changing variables thought to be central to panic disorder, such as anxiety sensitivity. This suggests that CBT2 consists of more than nonspecific factors. A handful of studies have tested this assumption. Nonprescriptive Treatment

Shear et al. (1994) compared 15 sessions of CBT2 to 15 sessions of 'nonprescriptive treatment' (NPT). The first 3 sessions of each treatment were very similar to one another, consisting of education about panic attacks. This included education about the cognitive model of panic, even in the NPT condition. Patients receiving NPT were informed about the role of misinterpretations of panic symptoms, and 'the informational material

l l6

TREATMENTOUTCOME: A CLOSER LOOK

included in ... NPT for panic [was] aimed at correcting misconceptions about panic, especially regarding fear of bo_dy sensations' (Shear et al., 1994, p. 400). NPT therapists also tried to help patients identify any obvious triggers of panics. The remaining 12 sessions of NPT consisted of reflective listening. NPT therapists did not give direct advice, other than telling patients that panics were not dangerous. The discussions during reflective listening sessions 'identified previously unrecognized feelings with associated somatic sensations and provided an explanation for the occurrence of sensations, previously seen [by the patient] as emerging "out of the blue'" (p. 400). NPT and CBT2 both reduced panic attacks and related symptoms, with no difference between treatments either at the end of treatment or at 6 month followup. The authors concluded that 'these findings raise questions about the specificity of cognitive behavioral treatment' (p. 395). Before accepting this conclusion it is important to consider the overlap among treatments and the manner in which CBT2 was implemented. CBT2 and NPT both contained psychoeducation aimed at correcting misconceptions about panic attacks, especially misconceptions about body sensations. These are interventions characteristic of CBT2. Thus, NPT and CBT2 were not only similar in terms of nonspecific factors; they also shared components specific to CBT2. The latter may have accounted for the lack of difference between the efficacies of NPT and CBT2. A further problem concerns the adequacy with which the treatments were conducted. CBT2 contains components other than those shared with NPT, such as cognitive restructuring exercises, interoceptive exposure, and situational exposure. In order to test whether these components are more effective than nonspecific factors, it is essential that the components be properly implemented. Shear et al. (1994) observed that 'CBT treatment adherence ratings were only moderately high, and this may mean that CBT was not delivered optimally' (p. 400). If CBT2 is not properly implemented, then it may be no more effective than a nonspecific intervention such as reflective listening. Failures to Replicate

In contrast to Shear et al.'s findings, several other studies have found that CBT2 is more effective than nonspecific interventions. Beck et al. (1992) compared 8 weeks of CBT2 to 8 weeks of brief supportive psychotherapy based on the principles of client-centered therapy. Supportive psychotherapy sessions involved a review of the patient's weekly panic attack records, along with the use of nondirective methods such as reflective lis-

CBT2: WHAT ARE THE MOST IMPORTANT INGREDIENTS? 117

tening. The principles of unconditional positive regard, genuineness, and accurate empathy were applied in order to establish a supportive, trusting, therapeutic relationship. CBT2, compared to supportive psychotherapy, produced significantly greater reductions in panic symptoms and general anxiety. After S weeks of treatment, 71% of CBT2 patients were panic-free compared to only 25% of patients treated with supportive psychotherapy. Craske et al. (1995) compared 4 sessions of abbreviated CBT2 to 4 sessions of nondirective supportive therapy. The latter was modeled on Shear et al.'s NPT and consisted of 1 session of education (identical to CBT2) and 3 sessions of reflective listening and nondirective discussion. CBT2 tended to be more effective than nondirective supportive therapy, although not all differences reached statistical significance. CBT2 led to significant reductions in worry about panic and phobic fear. Nondirective supportive therapy led to no significant changes on these measures. The proportion of patients who were panicking at pretreatment but panic-free after treatment was greater for CBT2 (54%) than for nondirective therapy (9%). Two other studies suggest that some components of CBT2 are more effective than nonspecific interventions. van den Hout et al. (1994) conducted a controlled cross-over study of cognitive restructuring and situational exposure. The first part of their study is relevant for our purposes. Here, panic patients were treated with either 4 sessions of cognitive restructuring (aimed at reducing the tendency to make catastrophic misinterpretations) or 4 sessions of 'association therapy'. In association therapy, The therapist displayed interest, respect and understanding of the patient's problems and encouraged the patients to discuss any current or past concern. No suggestion was made that matters could be seen from a different perspective or that behavior change could be beneficial. Association therapy was given to control for therapist attention and positive regard that is part of cognitive therapy. (van den Hout et al., 1994, p. 448)

Cognitive restructuring was superior to association therapy in reducing panic frequency. Association therapy had no beneficial effects at all. Neither treatment produced changes in agoraphobia or anxiety. Finally, a pilot study suggests that another CBT2 component-situational exposure-is more effective than a nonspecific intervention. McDonald et al. (1979) assigned panic disordered patients to 4 sessions of either exposure homework (i.e., planning and discussing situational exposure) or a control condition (discussion of marital, family, and social difficulties, and strategies for dealing with these). No exposure instructions were given in

11 8

TREATMENTOUTCOME: A CLOSER LOOK

the control condition. The exposure homework condition was significantly more effective at posttreatment and 1 month followup, as assessed by measures of agoraphobic fear and avoidance. To summarize, with the exception of Shear et al. (1994), treatment outcome studies indicate that CBT2 is more effective than nonspecific treatmentsthat is, more effective than treatments that are not designed to directly reduce anxiety sensitivity. Preliminary evidence also suggests that some of the components of CBT2--cognitive restructuring and situational exposure-are more effective than nonspecific interventions.

Targeting Key Cognitions: Focal versus Non-focal Interventions Another way of investigating the specificity of CBT2 is to examine the efficacy of focal versus non-focal forms of cognitive restructuring. Focal interventions are those that target cognitions said to be important in panic disorder (i.e., catastrophic beliefs about arousal-related sensations). Nonfocal interventions target other types of cognitions. Two studies have tested the prediction that focal interventions are more effective in reducing panic attacks. Salkovskis et al. (1991) reported a multiple baseline study of seven panic patients treated with either focal or non-focal cognitive restructuring. Other CBT2 methods were not used. Five patients were assessed during no-treatment baseline periods of various durations, followed by two sessions of cognitive restructuring aimed at reducing catastrophic misinterpretations (focal cognitive restructuring). The other two patients were also assessed during baseline periods, and then received two sessions nonfocal cognitive restructuring. The two patients then were assessed during another no-treatment baseline, and then received two sessions of focal cognitive restructuring. All patients were asked not to increase or decrease the extent to which they entered feared situations. Focal cognitive restructuring typically began by helping the patient understand panic attacks in terms of Clark's (1986) cognitive model. Evidence for and against catastrophic misinterpretations was reviewed and these cognitions were challenged. Evidence for and against alternative, noncatastrophic interpretations was also reviewed. Homework involved keeping a panic diary, in which patients wrote out rational responses to catastrophic misinterpretations. Non-focal cognitive restructuring was aimed at concerns in the patients' daily lives, excluding discussions of body sensations. Patients were told

CBT2:WHATARETHEMOST IMPORTANT INGREDIENTS? 119

that panic attacks are often the result of anxiety arising from life stressors, and so it was important to reduce stress as a way of reducing panic attacks. Therapists used a variety of cognitive techniques, including the methods for challenging beliefs associated with anxiety or depression (Beck & Emery, 1985; Beck et al., 1979). Homework consisted of using a thought record to write down negative thoughts (e.g., 'It's all my fault, I never do anything right') and then to challenge them. Panic attacks and strength of belief about the dangerousness of body sensations were reduced by focal cognitive restructuring but not by the non-focal intervention. Thus, panic attacks were reduced by cognitive procedures targeting panic-related beliefs, but not by procedures that targeted other maladaptive beliefs. Focal and non-focal interventions were also contrasted by Brown et al. (1997), using full rather than partial treatment packages. These investigators compared two treatments administered to 40 panic-disordered patients: CBT2 for panic disorder versus CBT for generalized anxiety disorder. In other words, one treatment was focal with respect to panic disorder whereas the other was non-focal. Each treatment lasted 12-18 sessions. Non-focal CBT 'focused primarily on the cognitions and beliefs relevant to interpersonal concerns involved in generalized anxiety' (Brown et al., 1997, p. 329), based on the assumption that 'patients will be less likely to experience panic attacks if their general level of anxiety is lowered through such methods as relaxation training and rational responding to anxious thoughts associated with daily living' (p. 332). An example of such thoughts includes worries about an upcoming visit to see relatives. Both treatments were associated with significant decreases in the tendency to make catastrophic misinterpretations of body sensations, and with significant decreases in panic frequency, anxiety, and agoraphobia. Gains were maintained at 1 year followup. No group differences were obtained at posttreatment or followup. Improvement in panic in both conditions was correlated with the reduction in scores on measures of panicrelated beliefs, which led the authors to conclude that their non-focal CBT may be an effective means of changing these beliefs. The validity of this conclusion depends on the assumption that therapists in non-focal CBT did not challenge catastrophic interpretations of body sensations. To assess therapist adherence to the protocol, two audiotapes were rated for each patient by a cognitive therapist who was blind to treatment assignment. Thus, for each patient only a fraction of treatment was reviewed (i.e., 2 of 12-18 sessions). Even with this limited sampling, protocol violations were detected for 8 of 19 patients treated with non-focal

i

120

TREATMENTOUTCOME: A CLOSER LOOK

CBT (i.e., 42% of patients). Brown et al. observed that 'such violations occurred wheri the therapist initiated discussion for the following types of catastrophic thoughts or beliefs: the fear of fainting, the fear of suffocating, the fear of going crazy or losing control, and the fear of vomiting' (p. 337). If more tapes had been rated then perhaps more protocol violations would have been detected. If therapists were challenging catastrophic interpretations during 'non-focal' CBT then it comes as no surprise that this treatment was effective in reducing panic disorder and the tendency to make catastrophic interpretations. It also comes as no surprise that the authors were unable to detect differences between this treatment and focal CBT. In summary, two studies have attempted to determine whether treatment is more effective when it focuses specifically on catastrophic interpretations of body sensations. Salkovskis et al.'s (1991) small study provides preliminary support. Brown et al.'s (1997) larger but methodologically flawed study found no advantage for focal treatment. The best available research (Salkovskis et al., 1991) provides tentative support for the importance of treatments that directly target catastrophic interpretations.

The Breathing Retraining Controversy Early Studies Early versions of CBT2 were based on the view that hyperventilation plays an important role in many panic attacks. Accordingly, treatment consisted of (1) brief voluntary hyperventilation to induce a mild panic attack, (2) explanation of the effects of overbreathing and education about how hyperventilation causes panic attacks, and (3) training in a breathing control technique so patients could control stress-induced hyperventilation (Clark et al., 1985). This treatment helped patients learn that arousal-related sensations are harmless (Salkovskis & Clark, 1986). Several studies have evaluated the efficacy of this treatment. Some studies found that as few as two or three sessions of breathing retraining was effective in reducing panic frequency, even when situational exposure exercises were not used (Clark et al., 1985; Rapee, 1985; Salkovskis et al., 1986a). Breathing retraining also raised abnormally low levels of blood pC0 2 to the normal range (Salkovskis et al., 1986a,b). Gains from these procedures were maintained at followup assessments ranging from 6 to 24 months (Clark et al., 1985; Salkovskis & Clark, 1986; Salkovskis et al., 1986a,b), although interpretation of these findings is complicated by

CBT2:WHATARETHEMOST IMPORTANT INGREDIENTS?121

the fact that some patients received further treatment during the followup interval. Some studies have evaluated whether breathing retraining improves the efficacy of CBT2 packages. Bonn et al. (1984) administered situational exposure with or without a breathing retraining intervention similar to that of Clark and Salkovskis (1987). Both treatments were effective in reducing panic, with no differences at posttreatment. At 6 month followup, however, the treatment containing breathing retraining was superior on a number of outcome measures, including breathing rate, panic frequency, and agoraphobia. The patients treated with exposure alone tended to deteriorate during followup, whereas those treated with exposure and breathing retraining maintained their gains or continued to improve. de Ruiter et al. (1989) compared three treatments: (1) situational exposure, (2) breathing retraining, and (3) situational exposure plus breathing retraining. In contrast to previous studies, none of these treatments significantly reduced panic frequency. The only trend was for breathing retraining without exposure to reduce panic frequency. Breathing retraining was no more effective than situational exposure in reducing breathing rate. Curiously, breathing retraining tended to decreaseinstead of increase end-tidal pC0 2 (i.e., the amount of CO 2 in the lungs after exhaling). This suggests that breathing retraining was improperly administered; patients were taking slower but deeper breaths (Ley, 1991, 1993). As a result, breathing retraining did not have its intended effect. As with many CBT2 studies, there were no data reported on whether the other treatments were properly implemented or whether patients were completing the homework exercises.

Is Breathing Retraining a Placebo?

I

'

With the exception of de Ruiter et al., other studies show that panic attacks can be reduced by as few as two or three sessions of breathing retraining. The patients in these studies were selected because their panic attacks appeared to be associated with hyperventilation. Does breathing retraining eliminate other types of panic attacks? Hibbert and Chan (1989) treated panic patients with two sessions of either breathing retraining or a control condition, followed in each case by three sessions of situational exposure or other anxiety management strategies (e.g., distraction, use of coping statements). The control condition consisted of discussions about the relationship between anxiety, stress, and personality, without reference to hyperventilation. The breathing retraining package contained

122

TREATMENTOUTCOME: A CLOSER LOOK

information about how breathing retraining would help patients control their panics, but did not contain the cognitive restructuring (reattribution) methods used in the studies by Clark and Salkovskis. Patients were divided into 'hyperventilators' and 'non-hyperventilators' on the basis of the similarity between panic symptoms and sensations experienced during 2min of voluntary hyperventilation. Breathing retraining was more effective than the control condition on measures of anxiety, but not on measures of panic or agoraphobic avoidance. There was no evidence that hyperventilators benefited more than nonhyperventilators. This led the authors to conclude that 'training in controlled breathing ... may have a non-specific effect in the treatment of patients with panic attacks' (p. 232). Garssen et al. (1993) expanded on this conclusion by proposing that breathing retraining is a rational placebo. That is, a method that works, even in the long-term, but not because of some specific effect on the tendency to hyperventilate. Garssen et al. proposed that breathing retraining works by 'the induction of a relaxation response, presenting a credible explanation for the threatening symptoms, giving a distracting task to practice when panic may occur, and promoting a feeling of control' (p. 141). Ley (1993) took issue with Garssen et al.'s claims, asserting that 'breathing retraining is not only effective on the basis of its stated rationale (viz., voluntary reduction in ventilation), but that reductions in anxiety oftentimes attributed to placebo effects may, in fact, be a consequence of adventitious reductions in ventilation' (p. 393). Ley criticized the arbitrary and unvalidated method used by Hibbert and Chan (1989) to classify patients as hyperventilators versus non-hyperventilators, and thereby cast doubt on the validity of the claim that breathing retraining was a nonspecific treatment. Ley concluded that 'while it cannot be proven that breathing retraining is not a "rational placebo", there is neither an empirical or logical basis for questioning the rationale' (p. 393, emphasis in original). More recently, Garssen et al. (1996) conducted an ambulatory monitoring study in which pC0 2 and panic attacks were recorded in 28 patients instructed to enter fear-evoking situations. Half the patients experienced one or more panics, yet a decrease in pC0 2 was observed for only one patient. These findings suggest that either hyperventilation is unimportant in causing panic, or that it plays a role in only a small proportion of cases. In view of the fact that breathing retraining is generally efficacious in reducing panic frequency, and that it normalizes pC0 2 levels (Salkovskis et al., 1986a,b), the results are consistent with the view that breathing retraining may have some placebo effect in addition to its specific effects in reducing hyperventilation.

CBT2:WHATARETHEMOST IMPORTANT INGREDIENTS? 123

No doubt there will be continued debate on the mechanisms of breathing retraining in the treatment of panic disorder. Even so, for some patients this intervention may sustain rather than reduce catastrophic beliefs (Salkovskis et al., 1996). This occurs when breathing exercises are misused as safety behaviours (see Chapter 14). Over the years, Clark and colleagues have decreased the amount of time devoted to breathing retraining, to the point that they see it as a very minor or unnecessary component of treatment (David M. Clark, personal communication, December 4, 1998). Barlow and colleagues continue to use this breathing retraining in their empirically supported CBT2 package because patients find it helpful; 'although we think breathing retraining makes the weakest contribution to changing patients' behavior, patients tend to attribute a lot of change to it' (Barlow, 1997, p. 36).

If breathing retraining is used, it is best implemented when three conditions are met: (1) the patient tends to hyperventilate, ei~her chronically or in response to stress, or experience recurrent chest pain due to chest breathing (see Chapter 16), (2) cognitive restructuring or exposure exercises have convinced the patient that the feared sensations have no harmful consequences, and (3) the patient wishes to be free of these harmless but unpleasant sensations.

Situational Exposure As in traditional behaviour therapy, CBT2 includes situational exposure exercises in which patients are asked to enter feared places or perform feared tasks. This may consist of therapist-assisted exposure, in which the therapist takes a patient or group of patients to a feared situation such as a shopping mall. Another form of situational exposure is programmed practice, where patients are encouraged to perform exposure exercises on their own as homework assignments (Mathews et al., 1981). The two forms of exposure have similar efficacy (Ghosh & Marks, 1987; Jannoun et al., 1980; Marks et al., 1983; Mavissakalian & Michelson, 1986a), and both are commonly used in behavioural and cognitive-behavioural therapies. Behaviour therapy for agoraphobia (e.g., Mathews et al., 1981) is based on the view that exposure for a sufficient period of time causes fear to habituate. In comparison, exposure exercises in CBT2 are designed as behavioural experiments. That is, methods for testing catastrophic versus noncatastrophic beliefs about what will happen when the person enters agoraphobic or other situations. Behavioural experiments emphasize the importance of experiencing feared body sensations, and provide a means of learning that the sensations are harmless. To illustrate, if a patient

124

TREATMENTOUTCOME: A CLOSER LOOK

believed that depersonalization leads to insanity, this person would be encouraged to enter a situation in which depersonalization has been experienced, such as a supermarket brightly lit with fluorescent lights, and remain there until sufficient evidence had been collected to test the catastrophic belief ('depersonalization will lead to insanity') and its noncatastrophic alternative ('depersonalization is a transient, harmless experience'). Exposure and Safety Behaviours

Behavioural experiments commonly involve the manipulation of safety behaviours, which are actions the patient uses to ward off feared consequences (Wells, 1997). For example, while in the supermarket a patient might intently scrutinize the labels on packages in an attempt to shake off feelings of depersonalization. Patients are asked to drop these safety behaviours in order (1) to test whether the behaviours reduce the feared sensations or make them worse, and (2) to conduct strong disconfirmations of catastrophic beliefs (e.g., allowing depersonalization to become very intense in order to test beliefs about its consequences). Even for severely agoraphobic patients, situational exposure in CBT2 is used primarily to disconfirm catastrophic beliefs, which in turn reduces agoraphobia. As with CBT2, therapists using traditional behaviour therapy also ask patients to refrain from safety behaviours (Mathews et al., 1981). An important question is whether treatment outcome is enhanced by eliminating these behaviours. In an experimental study involving a single session of exposure, Salkovskis et al. (1999) found the efficacy of exposure was improved when patients dropped their safety behaviours. Williams and colleagues (Williams & Zane, 1989; Zane & Williams, 1993) provided further data relevant to this issue. Drawing on self-efficacy theory (Bandura, 1986, 1988), Williams (1990) developed a treatment called guided mastery, which is a form of graded, situational exposure that emphasizes the identification and elimination of 'defensive maneuvers'. An example of the latter includes tightly gripping a shopping cart in order to avert the feared consequences of dizziness, such as physical collapse or loss of control. Defensive maneuvers are essentially the same as safety behaviours. Two studies have compared guided mastery to exposure-only. The latter was a form of situational exposure in which the safety behaviours were not deliberately eliminated. In the first study, Williams and Zane (1989) found that guided mastery was more effective than exposure-only in reducing agoraphobia and anxiety. The second study found the treatments were equally effective in reducing these symptoms, and both were

CBT2: WHAT ARE THE MOST IMPORTANT INGREDIENTS? 125

superior to a waiting-list control (Zane & Williams, 1993). Unfortunately, panic data were not reported in either study. To explain their failure to replicate the Williams and Zane's (1989) findings, Zane and Williams (1993) noted that in their 1993 study, patients in both the guided mastery and exposure-only conditions were informed about safety behaviours. This was done so that the authors could assess these behaviours in both treatment conditions. It may be that simply informing patients about these behaviours was sufficient to eliminate them. This might account for the lack of difference between Zane and Williams' (1993) guided mastery and exposure-only treatments. In their 1989 study, only guided mastery patients were told about safety behaviours. In summary, a common practice in CBT2 is to encourage patients to refrain from engaging in safety behaviours. Evidence suggests that this strategy improves treatment efficacy. Chapter 13 presents clinical guidelines on how to identify and eliminate safety behaviours.

Comparative Efficacy of CBT2 Treatment Packages CBT2 packages usually contain a number of different treatment components, such as cognitive restructuring, interoceptive exposure, situational exposure (behavioural experiments), breathing retraining, and various forms of relaxation training. In this section we will examine whether, for the average patient, some treatment components or combinations of components are more effective than others. Although many CBT2 components are effective, this does not necessarily mean that treatment is most effective when all components are used. For a given period of treatment (e.g., 12 weekly 1 hour sessions), the more CBT2 components one uses, the less time there is available for each component. Several dismantling studies have been conducted, comparing CBT2 packages consisting of various combinations of treatment components. Tables 6.1 and 6.2 summarize this extensive and complex literature. The tables are organized by selecting a criterion condition and reporting whether it exceeds the efficacy of comparison conditions. For each treatmentoutcome study, the intervention with the fewest treatment components is the criterion condition. This typically consists of a single component condition (e.g., cognitive restructuring alone), which is then contrasted with various comparison conditions. The latter typically consist of combinations of various CBT2 components (e.g., cognitive restructuring plus interoceptive and situational exposure).

l'v

°'

Table 6.1 Comparative efficacy of CBT2 components and combinations: posttreatment

Selected criterion conditions

Comparison conditions (x > y and y < x indicates x was more effective than y)

Outcome measure (summary)

I

Author(s)

-l

;u

m

~

s: m z

-l

WLC

AR

WLC y and y < x indicates x

() 0:,

--! N

:iE I

Composite index of improvement

~ )> ;;a

m --!

I

m

s 0 u,

--!

1

1

Hoffart (1998) (SEXPwas guided mastery) Hoffart (1998) (SEXPwas guided mastery) Ost et al. (1993) (CR similar but not equivalent to CR in CBT2) Telch et al. (1995)

s cl

0 ;;a ~

z--! z G) ;;a

m 0

m

z--! u,

""'

continuedoverleaf

IN

'°

Table 6.2

Selected criterion conditions

Commonly used package (CR + BR+ IEXP + SEXP)

Other

w

(continued) Comparison conditions (x > y and y < x indicates x was more effective than y)

0

Outcome measure (summary)

SEXP = CR + BR + IEXP = CR + BR + IEXP + SEXP SEXP= CR+ BR +PMR= CR + BR + PMR + SEXP

Global outcome (various measures) Agoraphobia

CR+ BR+ IEXP + SEXP = AR CR+ BR+ IEXP + SEXP > AR CR+ BR+ IEXP + SEXP;:: SEXP

Global outcome (various measures) Panic attacks Panic attacks and agoraphobia

Followup duration (years)

Author(s)

-f

:,0

m

~

s: m

1

Telch et al. (1995)

1.5

Rijken et al. (1992)

z-f 0 C -f

n 0

CR + BR + !EXP + SEXP > SEXP CR + BR + IEXP + SEXP = SEXP CR + BR + IEXP + SEXP = SEXP = CR + BR + IEXP CR + BR + IEXP + SEXP > AR +SEXP CR + BR + IEXP + SEXP > CR + BR+ IEXP

Composite index of improvement % achieving high end-state functioning Global outcome (various measures) Panic attacks

CR + IEXP + SEXP > CR + BR +SEXP

1

Ost & Westling (1995)

s: m )>-

1 1 1 1

Arntz & van den Hout (1996) Hoffart (1998) (SEXP was guided mastery) Telch et al. (1995)

1

Hoffart (1998) (SEXP was guided mastery) Telch et al. (1995)

0.75

Clark et al. (1994)

Composite index of improvement

1

Telch et al. (1995)

Panic attacks, agoraphobia, & global severity

0.5

Craske et al. (1997)

Key:AR= applied relaxation;BR= breathing retraining; CR= cognitiverestructuring; High end-state functioning= low overall severity of panic disorder symptoms; !EXP= interoceptiveexposure; PMR = progressive muscle relaxation;SEXP= situational exposure.

n

5

(/)

m :,0

5 0

A

CBT2:WHATARETHEMOST IMPORTANT INGREDIENTS?131

The bottom of each table also presents a commonly used CBT2 package as a criterion condition. This consists of cognitive restructuring combined with breathing retraining and situational and interoceptive exposure. This package was used in most studies comparing CBT2 to other treatments. As mentioned earlier, breathing retraining is now not typically used by Clark and colleagues but continues to be used by Barlow and colleagues. Table 6.1 shows that at posttreatment, many treatment components and combinations are superior to waiting list controls on a variety of outcome measures. The table also shows that the various components and combinations appear to be roughly equal to one another in terms of efficacy. There are few differences in efficacy, apart from a trend for the commonly used CBT2 package to be more effective than treatment packages consisting of either specific CBT2 interventions or combinations of interventions. Table 6.2 shows the comparative efficacy of components and combinations at followup, typically 1 year after the end of treatment. A problem in interpreting the results is that it is likely that some patients received some form of additional treatment during the followup period. Outcome studies usually do not mention whether or not this occurred. Additional treatment seeking can blur any differences between the treatments compared in Table 6.2. This is because the weakest of the treatments should be associated with the greatest treatment-seeking during the followup interval. Conversely, patients receiving the strongest treatment would be least likely to seek further treatment during followup. Thus, the differences between the weakest and strongest treatments may be attenuated by differences in the probability of seeking further treatment. Despite this potential problem, the table shows two trends worth noting. The first is that applied relaxation tended to be less effective than other components and combinations. The other is that the commonly used CBT2 package tended to be the most effective. An important question is whether these results hold up when additional treatment seeking is taken into consideration.

Comment Several CBT2 packages are effective in the treatment of panic disorder. The efficacy of CBT2 is not simply due to nonspecific factors, such as therapist warmth, genuineness, and accurate empathy. Cognitive restructuring tends to be more effective when catastrophic misinterpretations of body sensations are challenged, than when other sorts of cognitions are

132

TREATMENTOUTCOME: A CLOSER LOOK

targeted (e.g., worries about gaining approval from others). Situational exposure tends to be most effective when patients are encouraged to refrain from using safety behaviours. It may be that some patients benefit more from some treatment packages than others. However, for the average panic disordered patient, the commonly used CBT2 package tends to be the most effective. Interestingly, this package contains breathing retraining, which is considered by some clinicians to be a placebo.

CBT2 VERSUS OTHER TREATMENTS Remarkably few studies have compared the efficacy of CBT2 against other treatments such as pharmacotherapies. In this section we will review this small literature. In the following chapter we will examine the larger, comparatively more complex and controversial literature on combining drug treatments with behavioural and cognitive behavioural therapies.

Alprazolam Kiosko et al. (1990, 1995) compared alprazolam (gradually increased to a mean of 4.6mg/day), pill placebo, CBT2, and waiting list control. These investigators attempted to taper patients off alprazolam two weeks before the end of treatment. However, only 1 of 16 patients was able to discontinue; the remainder presumably either relapsed or were too frightened of the consequences of discontinuing their medication. The 15 patients were returned to their pre-taper dose of alprazolam. Once stabilized, they completed the posttreatment assessment. At posttreatment, CBT2 outperformed placebo and waiting list control on most measures, whereas alprazolam did not differ from waiting list control. The proportion of panic-free patients at posttreatment was greater for CBT2 (87%) than alprazolam (50%). There was a nonsignificant trend for CBT2 to outperform alprazolam in terms of posttreatment global severity.

Imipramine Two studies compared CBT2 with imipramine. Clark et al. (1994) found that imipramine (combined with self-exposure instructions) was less effective than CBT2 when outcome was assessed after three months of treatment. Imipramine was gradually increased to a mean dose of 233 mg/ day over 3 months, and then tapered over the next 3 months. Six

CBT2VERSUSOTHERTREATMENTS 133

months after entering the study, after imipramine patients had been tapered from the drug, the CBT2 and imipramine conditions did not differ in efficacy. However, CBT2 was superior when patients were re-assessed 9 months later. At that time, 90% of CBT2 patients were panic-free, compared to 55% of patients treated with imipramine. It is not known whether CBT2 would have been superior if imipramine patients had remained on the drug. CBT2 tended to be more effective than applied relaxation, and the latter was as effective as imipramine. Barlow (1998) recently reported the preliminary findings of a multicenter trial comparing CBT2, imipramine (dose not reported), and their combination. Results for the combination treatment will be reviewed in the following chapter. At posttreatment, CBT2 and imipramine worked approximately equally well, and both surpassed pill placebo. Six months after treatment discontinuation, however, imipramine was associated with significantly more deterioration than CBT2. Patients treated with CBT2 tended to maintain their gains. Barlow (1998) concluded that 'based on the follow-up data documenting the better durability of cognitive-behavioral therapy and the medical principle of utilizing the least intrusive treatment first, we recommend a treatment strategy of cognitivebehavioral therapy followed by pharmacological therapy in those patients who require it' (p. 82).

Fluvoxamine Two studies have compared CBT2 to fluvoxamine. Black et al. (1993) compared CBT2, pill placebo, and fluvoxamine (built up to a mean dose of 240mg/day). In contrast to other studies, CBT2 performed relatively poorly. It did not differ from placebo on most outcome measures. After 8 weeks of treatment, fluvoxamine produced greater reductions than CBT2 on measures of anxiety, depression, and global severity. Fluvoxamine and CBT2 did not differ in their effects in reducing panic attacks, with 73% versus 48% panic-free at posttreatment. Fluvoxamine was not withdrawn before the posttreatment assessment, and no followup data were reported. The number of patients who were panic-free at the end of CBT2 (48%) was much lower than that of many other CBT2 studies (M = 88%: Chambless & Gillis, 1994). There are a number of factors that could have accounted for the poor performance of CBT2 in this study. Black et al. used an abbreviated (8 session) CBT2 protocol, which may have undermined treatment efficacy. Their CBT2 was less effective than other brief CBT2 protocols (e.g., Clark et al., 1999). Black et al. used only a single therapist.

' 134

TREATMENTOUTCOME: A CLOSERLOOK

No data was reported on treatment fidelity, and so it is unclear whether their CBT2 was adequately implemented. Black et al. had a substantially higher proportion of dropouts from their CBT2 (36%) compared to other CBT2 studies (M = 8%: Chambless & Gillis, 1994). This suggests that Black et al.'s CBT2, compared to other CBT2 protocols, was less acceptable to patients. Sharp et al. (1996) compared a number of treatment conditions, including 12 weeks of CBT2, pill placebo, and fluvoxamine (100-150 mg/ day). Combinations of these conditions were also investigated, which are reviewed in Chapter 7. Patients were tapered off fluvoxamine or placebo prior to the posttreatment assessment. At posttreatment, CBT2 and fluvoxamine were both superior to pill placebo on a range of measures, including anxiety, agoraphobic avoidance, and % of patients who were panic free. There were no differences between CBT2 and fluvoxamine at either posttreatment or 6 month followup. At followup, 70% of CBT2 patients were panic free, compared to 69% of fluvoxamine patients. However, it appears that more fluvoxamine patients sought additional treatment during the followup interval, which may have spuriously inflated the efficacy of the fluvoxamine condition. There are other important limitations to this study. CBT2 was delivered by a single therapist, with a second therapist occasionally filling in. It is not clear whether the findings would generalize to other therapists. A further limitation was that interoceptive exposure was not used in their CBT2, nor was there any indication that breathing retraining was used. Thus, Sharp et al. (1996) did not employ the commonly used CBT2 package. As we saw in Table 6.1, this package tends to be more effective than partial packages. It is possible that CBT2 would have been superior to fluvoxamine if this package had been used.

Long-term Outcome Few studies have directly compared treatments against one another in terms of their long-term efficacy. That is, their efficacy one or more years after treatment. In followups ranging from 1 to 5 years, it has been found that panic patients can maintain their therapeutic gains from pharmacotherapies, particularly if they remain on their medications or if they receive other treatment during the followup interval (Andersch et al., 1997; Mavissakalian & Pere!, 1992; Nagy et al., 1989, 1993; Pollack et al., 1993). Relapse rates tend to be high when anti-panic medications are withdrawn, with 30-100% of patients relapsing within weeks or months after medications are discontinued (Dupont & Pecknold, 1985; Pyer et al., 1987;

I

t (

]

1

CBT2 VERSUSOTHERTREATMENTS l 35

Nagy et al., 1989; Noyes et al., 1989b, 1991; Pecknold et al., 1988; Rickels, 1990a; Sheehan, 1986). Particularly high relapse rates have been observed for high-potency benzodiazepines such as alprazolam (Pollack & Smaller, 1995). Relapse rates are 50% or higher even when benzodiazepines are tapered very slowly (Marks et al., 1993; Spiegel et al., 1994). Sometimes the panic attacks arising during drug withdrawal are more frequent or severe than the panics experienced before treatment (Noyes et al., 1991; Pecknold et al., 1988). This phenomenon is called rebound panic. It does not usually occur when behavioural or cognitive behavioural treatments are discontinued, perhaps because these treatments teach patients how to treat their panic attacks and associated symptoms. Several studies have shown that brief courses of behavioural therapies (e.g., 8-12 weeks of situational exposure) can reduce panic disorder, with gains generally maintained at followup intervals ranging from 3 to 9 years (Burns et al., 1986; Emmelkamp & Kuipers, 1979; Marks, 1971; McPherson et al., 1980; Munby & Johnston, 1980). Unfortunately, studies of these therapies typically reported outcome data only for agoraphobia and anxiety; effects of treatment on panic attacks are less often reported. Few patients are symptom-free after treatment with situational exposure. Some relapse or experience temporary exacerbations in panic disorder (Burns et al., 1986; McPherson et al., 1980; Munby & Johnston, 1980), often in reaction to stressful life events (Burns et al., 1986). Studies of CBTl and CBT2 suggest that for the average patient, gains are maintained or patients continue to improve at followups ranging from 1 to 6 years (Beck et al., 1992; Brown et al., 1995, 1997; Clark et al., 1985, 1997a; Cote et al., 1994a,b; Craske et al., 1991; Evans et al., 1991; Franklin, 1989; Hoffart, 1998; Hunt & Andrews, 1998; Sokol et al., 1989). For CBT2, the proportions of patients who are panic-free at followup ranged from 60 to 100% (Beck et al., 1992; Brown et al., 1995, 1997; Cote et al., 1994a,b; Craske et al., 1991; Hoffart, 1998). At followup, 48-57% are rated as having achieved high 'end-state functioning' (Brown et al., 1995; Craske et al., 1991; Hoffart, 1998). High end-state functioning has been defined as 'clinical cure, or functioning within a normative range' (Craske & Barlow, 1993, p. 19), as indicated by 'a low severity of overall panic disorder symptoms, including anticipatory anxiety, limited symptom episodes, and phobic symptoms in addition to panic attacks' (American Psychiatric Association, 1998, p. 9). There is insufficient information to determine how well CBTl performed in terms of these outcome variables. In summary, followup studies suggest that, unlike drug treatments, the gains from behavioural and cognitive-behavioural therapies are usually maintained long after treatment has ended. It may be, however, that some

136

TREATMENTOUTCOME: A CLOSER LOOK

patients seek additional treatment during the followup interval. In these cases it may be that the extra treatment contributed to good long-term outcome.

Fresh Symptom Emergence? The concept of the general neurotic syndrome (Chapter 2) suggests that if the general diathesis for neurotic disorders is not treated by panic therapy, then psychiatric problems other than panic disorder may arise during followup. Hafner (1976) conducted a 1 year followup of panic patients treated with 4 days of intensive situational exposure. During followup, most (67%) of his 39 patients were said to have 'fresh symptom emergence', defined as increases in scores on various psychopathology measures. Other studies suggest that Hafner's results are anomalous; four other studies of exposure therapy for panic disorder have found that very few patients develop new neurotic symptoms or disorders during the followup interval (Emmelkamp & Kuipers, 1979; Marks, 1971; McPherson et al., 1980; Munby & Johnston, 1980). In fact, out of the four studies only one patient developed a new disorder. That patient developed a new phobia, which responded to behaviour therapy (McPherson et al., 1980). The remaining studies reported no emergence of fresh symptoms or disorders. It is possible that exposure therapies-and possibly CBT2directly treat the general neurotic diathesis, thereby reducing the risk for new disorders.

Comment In summary, the small literature directly comparing CBT2 to pharmacotherapies suggests that the efficacy of CBT2 is equal to or greater than that of alprazolam and imipramine at posttreatment. It is unclear how CBT2 performs in comparison to fluvoxamine. Future research is needed to compare CBT to other pharmacotherapies, such as other selective serotonin reuptake inhibitors shown to be effective treatments of panic disorder (e.g., fluoxetine, sertraline, or paroxetine: American Psychiatric Association, 1998). Another important issue is whether CBT2 is effective for patients who have failed to respond to other treatments. Preliminary evidence suggests that CBT2 is effective in treating patients who have failed to respond to pharmacotherapies (Otto et al., 1999; Pollack et al., 1994a). Long-term followup studies suggest that CBTs and situational exposure are effective in the long term, and are likely to be more effective than

RATESAND PATTERNS OF CHANGE 137

short-term pharmacologic treatment. It is not known whether drug treatments would be as effective as exposure and CBTs if patients remained on their medications. Any conclusions about the long-term efficacy of panic treatments are necessarily tentative, because patients sometimes seek additional treatment during the followup interval. A further limitation of most outcome studies-including studies of CBT2 and other therapies-is that outcome was assessed cross-sectionally. That is, at followup patients were asked to report the number of panic attacks they experienced during, for example, the past 4 weeks. Such assessments overlook transient exacerbations in symptoms during the followup interval, and so can paint a overly favourable picture of the long-term performance of treatments. Brown and Barlow (1995) asked patients to report on the course of their symptoms 2 years after a trial of CBT2. Treatmentrelated gains were generally maintained, with 75% of patients being panic-free at followup. However, many patients described a fluctuating course of symptoms. About a third of patients who were panic-free at followup had panicked at some point over the previous year. A clinically important question is how we can help patients maintain or improve on their treatment-related gains. A related question is how we can best prevent relapse. Strategies for addressing these issues are examined in later chapters.

RATES AND PATTERNS OF CHANGE Do some treatments work faster than others? Rate of improvement is an important clinical variable because rapid improvement may enhance the patient's motivation to continue in treatment and thereby prevent premature termination (Penava et al., 1998). Patterns of change are also clinically important. Some treatments may produce desynchronous changes, where some symptoms change faster than others. Other treatments may produce synchronous reductions in symptoms. Knowledge of these patterns is important for selecting treatments.

Comparisons Among Treatments Several studies have examined treatment-related patterns of change. Intensive situational exposure (e.g., daily sessions of flooding) can markedly reduce agoraphobic avoidance within the first week of treatment (Teasdale et al., 1977). Compared to less frequent exposure regimens (e.g., weekly sessions), flooding reduces agoraphobic avoidance more

• 138

TREATMENTOUTCOME: A CLOSER LOOK

,,;

:;. ;('.

rapidly than it reduces anxiety experienced in agoraphobic situations (Foa et al., 1980). For less intensive exposure treatments, two studies reported that graded situational exposure reduced avoidance without altering panic frequency (Arnow et al., 1985; Klein et al., 1987). Fava et al. (1991) found that 12 weeks of graded situational exposure reduced panic attacks in 84% of patients, and that it produced greater and more rapid reductions in agoraphobic avoidance than in panic frequency. Marks et al. (1983) also found that situational exposure reduced avoidance by a greater amount than it reduced panic frequency. In contrast, Sochting et al. (1998) found that 8 weekly sessions of situational exposure reduced avoidance and panic frequency by roughly the same amount. Thus, there is mixed evidence for desynchrony, although most studies suggest that situational exposure produces greater and more rapid reductions in avoidance than in panic or anxiety.

.

T1

p Sl 1f

cl Sl

cl C

T A S)

d1

For CBT2, panic frequency drops markedly within the first 1--6 weekly sessions (Arntz & van den Hout, 1996; Clark et al., 1985; Craske et al., 1995; Laberge et al., 1993; Salkovskis et al., 1986a; Sokol et al., 1989). CBT2 tends to produce synchronous reductions in panic frequency and agoraphobic avoidance (Penava et al., 1998). In comparison, drug treatments tend to produce desynchronous changes, with faster reductions in panic frequency than in avoidance (Deltito et al., 1991; Liebowitz et al., 1986; Muskin & Pyer, 1981).

SE

ti, m p,

H aE h,

p,

It has been claimed that CBT2 is slower than alprazolam

to exert its therapeutic effects (Hegel et al., 1994), because alprazolam can sometimes reduce panic and anxiety within two or three days (Ravaris et al., 1991; Sheehan, 1982). The claimed superiority of alprazolam is overstated for two reasons. First, panic disorder can sometimes be successfully treated in one or two sessions of CBT2 (e.g., Clark, 1996). Second, the typical time course for CBT2 appears to be similar to that of alprazolam, with both reducing panic frequency within 1-6 weeks (Ballenger et al., 1988; Cross National Collaborative Panic Study, 1992; Pecknold et al., 1994; Penava et al., 1998).

0~

in

p,

m

p;

d fr eJ

T fr g el f, cl lE

CBT2 also appears to exert its therapeutic effects about as fast as other drugs. The anti-panic effects of imipramine become apparent after about 4-6 weeks (Cross National Collaborative Panic Study, 1992; Klein, 1980). Black et al. (1993) found that fluvoxamine produced improvement earlier than CBT2 on measures of anxiety. However, CBT2 and fluvoxamine did not differ at any point on measures of panic frequency or severity. Sharp et al. (1996) found that CBT2 and fluvoxamine worked at the same rate in reducing anxiety. It was not reported whether the treatments differed in speed of panic reduction.

vi

1 e ti ,J

j

11

I

RATESAND PATTERNS OF CHANGE 139

To summarize, drugs and CBT2 work as fast as one another in reducing panic attacks, and appear to be somewhat faster than situational exposure. Compared to drugs, exposure and CBT2 produce faster reductions in agoraphobic avoidance. Drugs tend to produce desynchronous changes, with faster reductions in panic than avoidance. Situational exposure tends to produce the opposite pattern. CBT2 tends to produce synchronous changes. For rapid reduction of panic attacks and agoraphobia, CBT2 appears superior to either drugs or situational exposure.

Treatment-related Increases in Panic Frequency Although CBT2 generally produces rapid, synchronous reductions in symptoms, some patients display a particular form of desynchrony during treatment. Clinically, it is not uncommon to see patients who have severe agoraphobia but infrequent panic attacks. As a program of situational exposure is commenced, these patients may report a transient increase in panic frequency, especially in the frequency of situational panic attacks, as they enter formerly avoided situations. How common are these transient exacerbations? de Beurs et al. (1993) assessed 28 patients who completed a 13 week course of CBT2. Patients had panic disorder with moderate-to-severeagoraphobia. One group of patients (n = 10) displayed a decreased in panic frequency and avoidance over the course of treatment. Another group (n = 7) displayed a decrease in agoraphobia but a transient increase in the frequency of situational panics. A third group (n = 11) showed an intermediate pattern. These findings suggest that transient panic exacerbation occurs in over a third of patients with moderate-to-severe agoraphobia. Interestingly, the groups did not differ in pretreatment severity of agoraphobia or panic frequency, thereby making it difficult to predict who would experience a transient exacerbation. This study has other important clinical implications. It suggests that panic frequency may not always be a good measure of how treatment is progressing. Treatment can be beneficial even though the patient is experiencing a temporary increase in panic. In fact, a transient exacerbation may facilitate treatment because it enables the therapist to better identify and challenge catastrophic beliefs. But transient exacerbations can sometimes lead patients to the mistaken conclusion that treatment is making them worse, and therefore increase the risk of patients dropping out of therapy. To avert this problem one should advise patients that they may experience a transient increase in panic, and that this 'side effect' is an indication that treatment is working.

140

TREATMENTOUTCOME: A CLOSER LOOK

Comment To summarize, CBT2 is among the best treatments for producing rapid, synchronous reductions in panic attacks and agoraphobic avoidance. This is important because rapid results are likely to keep patients from dropping out of treatment, and because reductions in all the major domains of panic disorder (e.g., panic attacks and agoraphobia) is an important goal of treatment. However, during treatment patients may sometimes experience a transient increase in symptoms.

BREADTH OF THERAPEUTIC EFFECTS As we have seen in this and the previous chapter, CBT2 and the other major panic treatments are effective in reducing panic attacks, agoraphobia, and related symptoms such as general anxiety and depression. In the following sections we will assess what other beneficial effects are associated with these therapies, and whether CBT2 offers any advantage in terms of its breadth of therapeutic effects.

Axis I Comorbidity Two studies have investigated whether treatments for panic disorder also reduce concurrent Axis I disorders that the patient might have at the time of seeking treatment for panic disorder. Brown et al. (1995) assessed changes in comorbidity after a course of CBT2. Comorbid disorders in their sample were most commonly generalized anxiety disorder (33% of sample), social phobia (14%), and mood disorders (major depression or dysthymia: 13%). Over the course of therapy, comorbidity decreased as panic disorder abated. However, the prevalence of comorbid disorders increased at 2 year followup, even though panic disorder continued to abate. The exception was comorbid generalized anxiety disorder, which reduced in prevalence at posttreatment and at followup. Tsao et al. (1998) also examined changes in comorbidity after panic patients had been treated with CBT2. As in Brown et al.'s study, the most commonly comorbid disorders were generalized anxiety disorder, social phobia, and mood disorders (major depression or dysthymia). At posttreatment, the severity of several comorbid disorders had declined (generalized anxiety disorder, social phobia, and posttraumatic stress disorder), and there were nonsignificant trends for declines in severity of depression and specific phobia. No followup data were reported and so it is not clear whether there were enduring reductions in comorbidity.

BREADTHOF THERAPEUTIC EFFECTS 141

If CBT2 for panic does reduce comorbid disorders such as generalized anxiety disorder, then there are a number of ways this could occur. It could be that CBT2 directly reduces (perhaps temporarily) some nonspecific psychopathologic factor. Another possibility is that the reduction in comorbidity is due to generalization of skills learned in panic treatment; skills learned to reduce one disorder are used to reduce another. A third possibility is that some comorbid disorders may be a consequence of panic disorder; once panic disorder is reduced, some comorbid disorders also abate (Brown et al., 1995). It is not clear whether CBT2 is unique in reducing comorbidity. As we will see later in this chapter, it seems likely that other panic treatments similarly reduce comorbidity. Unfortunately, neither the studies by Brown et al. or Tsao et al. included a no-treatment control group to compare the effects of CBT2 to the mere passage of time. It may be that both studies were simply detecting natural fluctuations in the severity of comorbid disorders. This possibility is consistent with the results of Brown et al., where most comorbid disorders tended to fluctuate in severity (i.e., decreased by posttreatment and then increased at followup). Thus, it remains to be demonstrated that CBT2 causes reductions in comorbid disorders. The same criticism applies to many of the studies on the breadth of therapeutic effects because most studies did not use a control or comparison group. A further concern is that the presence versus absence of comorbid disorders is a crude way of assessing the breadth of therapeutic effects. Dimensional measures of the severity of specific disorders or symptoms represent a more sensitive approach. Findings based on these measures are reviewed in the following sections.

Blood-Injury Phobia and Social Phobia These phobias are considered together because both are measured by the Fear Questionnaire (Marks & Mathews, 1979), one of the most widely used scales in treatment studies of panic disorder. Strictly speaking, the Fear Questionnaire measures severity of avoidance rather than phobia, although most studies refer to it as a measure of phobia because scores on the Blood-Injury and Social Phobia subscales have good validity in identifying people with specific phobias (see Chapter 9). One study of CBTl reported that the severity of blood-injury phobia increased with treatment (Barlow et al., 1983). This finding is anomalous because most studies found that various panic treatments-including situational exposure, CBTl, CBT2, progressive muscle relaxation, and applied relaxation-were associated with reductions in blood-injury

r 142

TREATMENTOUTCOME: A CLOSERLOOK

phobia and social phobia (Arrindell et al., 1986; Beck et al., 1994; Burns et al., 1986; Clark et al., 1994; Craske et al., 1995, 1997; Evans et al., 1991; Hafner, 1983; Kleiner et al., 1987; Michelson et al., 1988, 1990; Ost et al., 1993; Wade et al., 1998). Compared to waiting list controls, these reductions were significantly greater for CBT2 and applied relaxation (Beck et al., 1994; Clark et al., 1994) but not for progressive muscle relaxation (Beck et al., 1994). With regard to drug treatments for panic disordered patients, Marks et al. (1983) found that imipramine was no more effective than pill placebo in reducing social phobia or blood-injury phobia. Compared to other imipramine studies, the study by Marks et al. was unusual in that imipramine was no better than placebo on measures of panic attacks and agoraphobia. Other studies have found imipramine to be effective for panic disorder (see Chapter 5). Sheehan et al. (1980) reported that phenelzine and imipramine were more effective than placebo in treating panic disordered patients, and also more effective in reducing comorbid social phobia. This suggests that when these drugs are effective in treating panic disorder, then they also are effective in reducing certain other phobias. In summary, compared to no-treatment controls, CBT2 reduces blood-injury phobia and social phobia in patients whose primary problem (in terms of severity) is panic disorder. CBT2 is not alone in these effects; many other anti-panic treatments similarly reduce panic patients' scores on measures of blood-injury and social phobias. This may be because the treatments target a vulnerability factor common to many disorders.

E

E

j

Obsessions and Compulsions

j

Panic disorder and obsessive-compulsive disorder (OCD) sometimes cooccur (Chapter 2). Some pharmacotherapies are successful in treating both disorders (e.g., fluoxetine, fluvoxamine, and clomipramine). It may come as a surprise, however, that behavioural or cognitive-behavioural treatments for panic disorder should be effective in reducing obsessions and compulsions. These panic treatments do not focus on OC symptoms, yet a number of studies have found that panic treatments do reduce these symptoms.

t F

C

C

11

a

r r

t

Kolko (1984) reported a case study of panic disorder comorbid with OCD symptoms (e.g., compulsive cleaning and checking). Panic disorder was the focus of treatment. Paradoxical intention was used, in which the

C

F

i •I

.!

BREADTHOF THERAPEUTIC EFFECTS 143

patient was instructed to increase her anxiety and to attempt to elicit feared consequences during exposure to agoraphobic situations. As panic disorder abated, so did the obsessive-compulsive symptoms. Fava et al. (1988b) reported that patients with panic disorder had more obsessive-compulsive symptoms than normal controls, although no patient met diagnostic criteria for OCD. OC symptoms included those that were clearly unrelated to panic disorder, such as compulsive handwashing. After patients received 3-4 months of situational exposure to agoraphobic situations, panic disorder and OC symptoms both abated, to the point that patients no longer differed from controls in severity of OC symptoms. This occurred even though obsessions and compulsions were never the target of treatment. Michelson et al. (1990) similarly reported that CBT2 for panic disorder was associated with significant reductions in scores on measures of obsessions and compulsions. Neither study used a no-treatment control group to assess whether the change in OC symptoms was simply due to the passage of time.

If panic treatments reduce OC symptoms, then how might this occur? Fava et al. noted that OC symptoms tend to covary with the person's anxiety level, increasing during times of distress and decreasing when the person feels calmer. As treatment reduced the distress associated with panic disorder, the OC symptoms abated. Thus, the reduction in OC symptoms was seen as being part of some nonspecific effect of stress reduction. Although this may occur for mild OC symptoms, it seems likely that patients with panic disorder and full-blown OCD will need specific treatment for the latter; panic treatment probably would not be sufficient.

Alcohol Abuse According to clinical lore, substance use disorders should always be treated before attempting to treat a comorbid anxiety disorder; 'even persons who originally took alcohol or benzodiazepines for their anxiety disorder must cease their drug use before treatment for their anxiety disorder [would] be practical' (Andrews et al., 1994, p. 22). But some panickers use alcohol as a form of 'self medication' in order to dampen anxiety or panic (Quitkin et al., 1972). Once panic disorder is reduced, the motivation to consume alcohol or other anxiolytics should also be reduced. Consistent with this hypothesis, Lehman et al. (1998) studied three cases of comorbid panic disorder and alcohol abuse and found that CBT2 (for panic disorder) reduced alcohol consumption. Each of these patients developed drinking problems after they began having panic

L

144

TREATMENTOUTCOME: A CLOSER LOOK

attacks, and all believed that alcohol helped reduce their anxiety and panic. After 11 sessions of CBT2 (along with daily self-monitoring of alcohol use), panic disorder had remitted for two of three patients, and alcohol abuse had remitted for all three. At 6 month followup, one patient was free of both disorders, one had panic disorder, and the third had panic disorder and alcohol abuse. These findings are encouraging in that they suggest that CBT2 may produce clinically meaningful reductions in alcohol abuse. However, alcohol abuse, like other substance use disorders, has a high rate of relapse (Marlatt & Gordon, 1985). For patients with comorbid panic disorder and substance use disorders, it may be necessary to integrate specific treatments for substance abuse (e.g., Daley & Marlatt, 1997) into panic treatment protocols. This would be particularly important when patients abuse alcohol or other substances for reasons other than 'self-medication.' In these cases substance misuse is likely to persist even when panic disorder is reduced.

Personality Changes N euroticism

In Chapter 2 we reviewed the arguments and evidence for the claim that elevated neuroticism is a nonspecific factor contributing to many disorders, including panic disorder. This is consistent with the suggestive evidence that CBT2 and other panic treatments are broad in their therapeutic effects, reducing a variety of symptoms and disorders. A handful of studies have investigated whether panic treatments reduce neuroticism. Unfortunately, results were mixed and no study was controlled. Two studies reported null results. Mavissakalian (1985) reported a nonsignificant trend for neuroticism to decrease after treatment consisting of situational exposure for agoraphobia combined with either imipramine or pill placebo. The sample size of this study was small (analyses were conducted separately for 10 females and 10 males) and may not have had sufficient power to detect changes in neuroticism. In a 1 year followup of brief, intense CBT2, Evans et al. (1991) reported reductions in panic disorder but not neuroticism. Three other studies reported that neuroticism declined with panic treatment. Burns et al. (1986) found that situational exposure for agoraphobia reduced neuroticism at posttreatment, with gains maintained at 8 year followup. In a study of a treatment combining CBTl with breathing retraining, Andrews and Moran (1988) reported that panic disorder and

BREADTHOF THERAPEUTIC EFFECTS 145

neuroticism declined with treatment, as assessed at posttreatment and at 6 and 12 month followup. Hunt and Andrews (1998) found that CBT2 reduced panic disorder and neuroticism. Gains for the latter were apparent at 2 year followup but not at posttreatment. In sum, the findings are inconsistent as to whether behavioural and cognitive-behavioural treatments reduce neuroticism. The inconsistencies may be due to methodological factors, such as small sample sizes. Further research is needed before we can judge whether CBT2 or other panic treatments produce enduring reductions in neuroticism.

Personality Disorders

Several studies have claimed that panic treatments reduce personality disorders and traits. Noyes et al. (1991) assessed panic patients before treatment and 3 years later. Treatment consisted of 8~34 weeks of either diazepam, alprazolam, or placebo, each of which was followed by 'naturalistic' treatment. Changes in personality disorder from pretreatment to 3 year followup were assessed by a self-report scale, the Personality Diagnostic Questionnaire (PDQ: Hyler et al., 1983). Patients with the greatest reductions in panic disorder also had substantial reductions in avoidant, dependent, and histrionic traits. It is not clear whether these changes were due to panic treatment, because some patients may have received treatment for personality disorders as part of their 'naturalistic' treatment. Hoffart (1994) reported that combined behavioural-psychodynamic treatment reduced panic disorder and personality disorder. However, these effects are hardly surprising because part of the treatment focused on personality disorder (Hoffart & Hedley, 1997). Other studies apparently do not suffer these limitations, although they have problems of their own. In a double blind study of alprazolam versus placebo, Reich et al. (1987) assessed personality disorder with the PDQ. They found that patients recovering from their panic disorder (after treatment with either alprazolam or placebo) also improved in self-report measures of self-confidence and sociability, although their scores on these measures were still below those of normal controls. It is unfortunate that Reich and colleagues did not report results separately for each treatment. Given the entrenched, long-standing nature of personality disorder, it seems unlikely that pathological personality traits would be changed in any lasting manner by simply giving patients a pill placebo. In a study of situational exposure combined with one of three antidepressant medications (imipramine, trazodone, or desipramine), Mavis-

T

146

TREATMENTOUTCOME: A CLOSERLOOK

sakalian and Hamann (1987) reported that panic disorder generally declined with these treatments, as did scores on measures of personality disorder traits. The greatest reductions, as assessed by the PDQ, were in the following traits: intolerance of being alone, hypersensitivity to criticism, lack of self-confidence, and indirect resistance (passive aggression). Little change was observed in a number of other traits including dramatic behaviour, affective instability, social withdrawal, and desire for affection. Overall, scores on measures of most personality disorders declined with treatment, except for schizoid and antisocial personality disorders, which had low base rates to begin with. Interestingly, personality disorder scores declined regardless of whether or not panic disorder was reduced by treatment. Four studies have examined the effects of CBT2 on personality disorder. Rathus et al. (1995) reported that 12 weeks of CBT2 for panic disorder reduced personality disorder, as assessed by the second edition of the Millon Clinical Multiaxial Inventory (MCMI-II: Millon, 1987). Statistically significant reductions were obtained on the following scales: paranoid, schizotypal, borderline, histrionic, dependent, passive-aggressive, and aggressive/sadistic personality disorders. Unfortunately, this study was uncontrolled and contained no followup data. In a controlled study of personality change after panic treatment, Black et al. (1996) reported that scores on a revised PDQ declined after 8 weeks of treatment consisting of either fluvoxamine, CBT2, or pill placebo. CBT2, compared to fluvoxamine and placebo, was associated with the greatest reduction in personality disorder traits, specifically schizotypal, narcissistic, borderline, and obsessive-compulsive personality traits. Followup data were not reported. Hofmann et al. (1998b) reported that CBT2 and imipramine both resulted in global reductions in scores on a self-report measure of personality disorder-the Wisconsin Personality Disorders Inventory (Klein et al., 1993)-with no difference between treatments in their effects on panic or personality disorder. Although panic treatments may reduce personality disorder in the short term, in the longer term patients may tend to fall back into their longstanding, maladaptive personality patterns. Given this concern, a recent study by Hoffart and colleagues is of particular interest because a 1 year followup was conducted, using a structured interview to assess personality disorder. Hoffart (1995; Hoffart & Hedley, 1997) compared 6 weeks of CBT2 to 6 weeks of situational exposure (guided mastery). From pretreatment to 1 year followup, Hoffart found the number of avoidant and dependent personality traits decreased significantly for both treatments, with no difference between the two. Dramatic traits (i.e., histrionic, narcissistic, and borderline traits) did not change with treatment.

BREADTHOF THERAPEUTIC EFFECTS 147

To summarize, several studies have examined the effects of panic treatments on personality disorders. Most studies had methodological shortcomings, and so the results must be interpreted with caution. It appears that personality disturbance declines with many panic treatments, including CBT2. The changes in personality pathology appear to be largely nonspecific, with changes reported across a range of personality disorders. One of the more consistent changes is decreases in dependent personality disorder and traits. How do Panic Treatments Influence Personality Pathology?

There are several ways that CBT2 and panic therapies could lead to reductions on measures of personality pathology. Changes in some personality traits could be state effects (Reich et al., 1986, 1987). That is, distress (e.g., associated with panic disorder) may influence either the reporting of personality pathology on personality inventories (a reporting artifact) or may actually exacerbate personality pathology, leading disordered traits to become more extreme or more frequently expressed. When distress is alleviated (e.g., by reducing panic disorder) the endorsement or expression of personality pathology may also abate. Changes in some personality traits also could be an artifact of overlapping features. For example, 'intolerance of being alone' is a trait of borderline personality disorder and also a feature of agoraphobia. In the latter case the person is frightened of being without help in the event of a panic attack. Panic treatment can eliminate this trait, but this may simply reflect a lessening of agoraphobia rather than a change in borderline personality pathology. Vague items on personality questionnaires also may give a spurious impression of personality change. Noyes et al. (1991) noted that some items on personality disorder questionnaires assess panic-related symptoms; e.g., 'I find it easy to relax' is a reverse-scored item on the PDQ paranoid scale. After panic treatment, patients may agree with this statement simply because they are more relaxed, without necessarily having any changes in suspiciousness or other paranoid traits. Some personality disorder traits-particularly avoidant, dependent, and histrionic traits-may be consequences of panic disorder, and so changes in panic disorder may lead to changes in personality. According to Rathus et al. (1995), The possibility of state distortions notwithstanding, it is our contention that living with panic symptoms can have a markedly adverse effect on per-

148 TREATMENT OUTCOME:A CLOSERLOOK sonality (e.g., agoraphobic avoidance increases dependency, fear of panic attacks increases histrionic characteristics and avoidance). Therefore, when panic symptoms are alleviated, the related personality characteristics would also be expected to improve. (p. 166)

Mavissakalian and Hamann (1987) similarly suggested some personality disorder traits (e.g., lack of self-confidence) may be consequences of panic disorder, and therefore these traits should abate with panic treatment. Hoffart and Hedley (1997) suggested that severe panic attacks may activate beliefs that one is weak, helpless, and defective, and thereby lead to avoidant, dependent, or histrionic traits. With the reduction in these beliefs over a course of panic treatment, the pathological personality traits also abate. Cognitive restructuring for panic disorder also could improve personality functioning if patients learn to apply these therapy skills to restructure a range of distorted beliefs, including those associated with interpersonal functioning (Black et al., 1996). If CBT2 and other panic treatments can produce genuine changes in some personality traits, then what are the implications for clinical practice? Hoffart and Hedley (1997) found that pre- to posttreatment change in catastrophic beliefs-as assessed by the physical concerns subscale from the Agoraphobic Cognitions Questionnaire-significantly predicted reductions in avoidant and dependent traits from pretreatment to 1 year followup. Self-efficacy for coping with agoraphobia was a nonsignificant predictor. 'Overall, the results are consistent with the recommendations that, in cases of panic/ agoraphobia with comorbid Axis II disorder, one should start with the Axis I disorder and focus on catastrophic beliefs' (Hoffart & Hedley, 1997, p. 86).

This recommendation is likely to be more applicable to some personality disorders than others. Comorbid paranoid personality disorder, for example, is a risk factor for dropping out of panic treatment (see Chapter 8). Therefore, it may be better to focus on this personality disorder early in treatment. For other personality disorders, such as avoidant and dependent personalities, Hoffart and Hedley's recommendation could be followed. But even here there are two important caveats to consider. First, panic treatments (including CBT2) tend to have poorer outcomes for patients with a comorbid personality disorder (Chapter 8). Second, although panic treatments may reduce personality disorder, some patients are likely to have residual personality pathology at the end of treatment. That is, even during stable periods of remission of panic disorder with agoraphobia, these patients still tend to display 'a greater tendency than normals to see themselves as rather unassertive, indecisive, self-critical,

(

t r ~

1 l:

C

C

1

I\ fi

e 1

I\

tl

BREADTHOF THERAPEUTIC EFFECTS 149

and emotional individuals, who are easily frustrated and often feel rejected when criticized' (Mavissakalian & Hamann, 1992, p. 308). In summary, when panic disordered patients have comorbid personality disorder, it may be necessary to treat the personality pathology either before, during, or after panic treatment, depending on the nature of the personality disorder. Panic treatments such as CBT2 may reduce personality pathology, but these treatments may not be enough. In some cases it may be useful to combine CBT2 with treatments for personality pathology, such as Beck et al.'s (1990) cognitive therapy for personality disorders.

Assertiveness Lack of assertiveness can be regarded as a feature of an Axis I disorder (as in social phobia) or as a personality trait (as in avoidant or dependent personality disorders). Given its dual nature, we review it here in a section of its own. Situational exposure for agoraphobia failed to improve assertiveness in three studies (Emmelkamp, 1980; Emmelkamp et al., 1978; Emmelkamp & Mersch, 1982). In comparison, Michelson et al. (1985, 1988) found that assertiveness tended to improve after each of three treatments for panic disorder: situational exposure, paradoxical intention, and progressive muscle relaxation. Kleiner et al. (1987) found that CBTl for panic disorder improved assertiveness regardless of whether assertiveness training was included in the treatment package. CBTl (without assertiveness training) improved assertiveness in some studies (Emmelkamp & Mersch, 1982) but not in another study (Emmelkamp et al., 1978). To summarize, research suggests that situational exposure for agoraphobia generally fails to improve assertiveness. Other treatments for panic disorder, notably CBTl, tend to improve assertiveness. The effects of CBT2 on assertiveness remain to be investigated.

Marital Satisfaction Many studies have found that neither situational exposure nor CBTl for panic disorder produced changes in marital satisfaction (Arrindell et al., 1986; Bland & Hallam, 1981; Craske et al., 1989; Emmelkamp, 1980; Emmelkamp et al., 1992; Keijsers et al., 1994; Kleiner et al., 1987; Michelson et al., 1988, 1996; Monteiro et al., 1985). Other studies found that these treatments improved marital satisfaction even when the marriage was not a focus of treatment (Barlow et al., 1983; Cobb et al., 1984;

150

TREATMENTOUTCOME: A CLOSER LOOK

Emmelkamp et al., 1992; Himadi et al., 1986; Lange & van Dyck, 1992; Ost et al., 1984). Barlow et al. (1983) found that CBTl for panic disorder was associated with improved marital functioning only if two conditions were met: (1) the marriage was poor to begin with (as rated by both partners) and (2) when the spouse was involved in treatment. However, Cobb et al. (1984) failed to replicate this finding; the marriages of patients with marital difficulties at the beginning of treatment tended to improve, regardless of whether the spouse was involved. In sum, the average married panic disordered patient tends to either improve or show no change in marital satisfaction over the course of panic treatment. The studies reporting an improvement in satisfaction did not control for the mere passage of time. Therefore, it is not possible to determine whether marital satisfaction was improved by panic treatment or by some other factor. Improvement in agoraphobia as a result of situational exposure is occasionally associated with a deterioration in marital functioning (Hafner, 1977; Hand & Lamontagne, 1976; Milton & Hafner, 1979). This typically occurs in the context of longstanding marital or sexual difficulties, and can be effectively treated with marital therapy or sex therapy (Hand & Lamontagne, 1976). Hafner (1984) argued that marital functioning is most likely to be disrupted when treatment produces large, rapid reductions in panic disorder. This may occur during brief intensive treatments, such as Hafner's 4 day program in which patients received several hours of situational exposure each day. Deterioration in marital functioning after panic treatment does not always mean that the latter caused the deterioration. Deterioration may be part of a periodic cycle of improvement and worsening of marital functioning where deterioration sometimes coincides with the end of treatment. In other cases, however, panic disorder and marital functioning may be causally related, with changes in one problem producing changes in the other. For example, as panic disorder abates, patients become more mobile and less dependent on 'safe' others. Under these circumstances patients in unhappy marriages may feel more confident about leaving the relationship. A good functional analysis, as part of a case formulation, is important in understanding the relationships among the patient's problems. Further research is needed to examine the effects of panic treatments not only on marital functioning, but also on the quality of other dyadic relationships (e.g., among non-married heterosexual or same-sex couples) and on other types of interpersonal relationships. It is currently unclear

BREADTHOF THERAPEUTIC EFFECTS 151

whether the findings obtained for married couples can be generalized to other types of relationships.

Comment In Chapter 5 we saw that various panic treatments-including CBT2 and pharmacotherapies-are effective in reducing panic attacks, agoraphobia, anxiety, and depression. In this chapter we have seen that the effects of these treatments may be even broader. These treatments may reduce blood-injury phobia, social phobia, and personality pathology. Behavioural and cognitive behavioural treatments for panic disorder may also reduce obsessions and compulsions, and there is evidence that CBT2 can have a beneficial effect on alcohol abuse. CBT1 (and possibly CBT2) improves assertiveness, while situational exposure generally does not. CBT2 appears to have a greater impact in reducing personality disorder, compared to fluvoxamine and pill placebo. However, this was based on only one study (Black et al., 1996) and so the findings need to be replicated before we can have confidence in the superiority of CBT2. Situational exposure and CBT2 tend to either improve marital satisfaction or have no impact. Occasionally, marital deterioration occurs during these treatments, and so some form of couples therapy may be needed. In summary, although treatment may be focused on panic disorder, the beneficial effects may spread to other symptoms and problems. We have considered some of the possible explanations for these ripple effects, although much research needs to be done to better understand their causes. Although the findings are encouraging news for clinicians, they must be interpreted with caution. CBT2 and other panic treatments may influence a range of symptoms and clinical problems. However, many patients will require specific, additional treatments for comorbid problems. Personality pathology, for example, tends to abate over the course of panic treatment, but patients who have personality disorders when they begin panic treatment are likely to still have personality pathology (albeit of lesser severity) at the end of this treatment (Mavissakalian & Hamann, 1987). Panic disordered patients with severe dyadic (e.g., marital) problems may experience an improvement in dyadic functioning over the course of panic treatment, but are still likely to require specific treatment for relationship problems (Hand & Lamontagne, 1976). We are faced with the important problem of how best to treat patients with multiple clinical problems. Are these best treated by integrating panic treatments with other therapies, or is panic treatment undermined

152

TREATMENTOUTCOME: A CLOSERLOOK

by trying to treat too many problems at once? Sometimes sequential treatments are effective; e.g., treating panic disorder and then addressing other problems (McLean et al., 1998; Woody et al., 1999). But which sequence is optimal for which combinations of clinical problems? We will examine these important issues in later chapters.

EFFICACY IN CLINICAL SERVICE SETTINGS How effective is CBT2 when applied in routine clinical settings? That is, in settings .that may not necessarily have the inclusion and exclusion criteria used in choosing patients for treatment outcome studies. The latter often exclude panic patients for a variety of reasons, including age (e.g., older than 65), medications status (e.g., recent changes in the type or dose of psychotropic medication), severity of agoraphobia, and panic frequency (e.g., low frequency). Patients are also sometimes excluded if they have previously received a course of the therapy under investigation. Is CBT2 effective in clinics that do not have such stringent exclusionary criteria? Two studies have examined this issue. Wade et al. (1998) treated 110 panic patients in a community mental health center using Craske et al.'s (1994) CBT2 protocol. Patients were unselected with regard to age, medication use or changes, severity or frequency of panic attacks, or the presence of agoraphobia, comorbid Axis I disorders, personality disorder, medical problems, or treatment history. As in controlled trials, patients were excluded, however, if they had active alcohol or drug dependency, psychosis, or mental disorders caused by medical conditions. Many of Wade et al.'s patients had comorbid disorders, with the most common comorbid Axis I disorders being generalized anxiety disorder (21%) and major depression (20%). Treatment outcome was evaluated by assessing patients before and after treatment, and then comparing the results to those reported in two controlled trials of CBT2 (Barlow et al., 1989; Telch et al., 1993). Those trials excluded patients for various reasons, including age, unstable use of medication, severity of agoraphobia, and panic frequency. Wade et al. had far more dropouts than the controlled trials (26% vs. less than 7% for the controlled trials). For patients completing treatment, the outcome was similar to the results reported in the controlled trials. Wade and colleagues found that 87% of patients were panic-free at posttreatment, and significant reductions were observed on measures of anticipatory anxiety, agoraphobic avoidance, generalized anxiety, and depression.

SUMMARY AND CONCLUSIONS 153

Medication use also declined during treatment. The proportion of panicfree patients compared favourably to proportions reported by Barlow et al. (85%) and Telch et al. (85%). Thus, the findings suggest that CBT2 can be successfully used in community mental health centers and possibly in other clinical service settings. However, the high dropout rate raises concerns about the ability for patients to tolerate CBT2. In Chapter 8 we will review the predictors of treatment dropout, which can help identify ways of making treatment more acceptable. A limitation of Wade et al.'s research is that they did not assess patients at followup. Hunt and Andrews (1998) examined whether CBT2 delivered in a routine service setting is effective at 2 year followup. Eighty-four panic patients completed treatment in a hospital-based anxiety disorders clinic. A standardized treatment manual was used (Andrews et al., 1994) and the only criterion for inclusion was the diagnosis of panic disorder as the main clinical problem. The proportion of patients dropping out of panic treatment was not reported. Treatment produced significant reductions in symptoms at posttreatment, although the effects were not as large as those observed in controlled trials. Patients tended to continue to improve during the followup interval, although many received further treatment during this period.

In summary, CBT2 appears to be effective when used in routine clinical settings, but possibly not as effective as reported in controlled clinical trials. A clinically important question concerns the level of therapist training required to obtain satisfactory treatment outcomes in routine settings. Wade et al. (1998) and Hunt and Andrews (1998) used clinicians with postgraduate training in clinical psychology or psychiatry, who were trained in the use of CBT2. In our research (e.g., Taylor et al., 1996), we have found that BA-level research assistants can be. trained to successfully implement CBT2. These therapists received weekly supervision and were not required to treat complex cases of panic disorder, such as panic disorder with comorbid borderline personality disorder. Successful treatment of more complex cases may require more experienced therapists.

SUMMARY AND CONCLUSIONS In this chapter we have seen that CBT2 is one of the most effective treatments for panic disorder, with gains maintained at long-term followup. Unlike the results from meta-analyses reviewed in the previous chapter, a closer look at treatment outcome literature suggests that the most commonly used CBT2 package-consisting of cognitive restructuring, breathing retraining, and situational and interoceptive exposure-is likely to be

154

TREATMENTOUTCOME A CLOSERLOOK

more effective than its individual treatment components, particularly at long-term followup. It also appears that CBT2 is likely to be more effective than pharmacotherapies at long-term followup. CBT2 appears to work as fast or faster than other treatments, and has beneficial effects on symptoms other than those associated with panic disorder. Preliminary findings suggest that CBT2 may be effective in treating drug-refractory panic disorder. Despite these encouraging findings, CBT2 is not effective for everyone, and patients sometimes relapse. The optimal treatment package depends to some degree on the characteristics of the patient and their disorder. The efficacy of a second round of CBT2 for patients who have relapsed, or booster sessions for preventing relapse, has not been studied. Clinically, booster sessions seem helpful. Another important issue is whether treatment is most effective when CBT2 is combined with drugs. We turn to this in the following chapter.

COMBINING PSYCHOLOGICAL TREATMENTS WITH DRUG THERAPIES It is important for the behavioural or cognitive-behavioural therapist to have a good understanding of anti-panic medications and how they interact with psychological treatments. This is because it is common clinical practice to combine psychological and pharmacological treatments for panic disorder (Wolfe & Maser, 1994). Indeed, many clinicians believe the optimal treatment consists of drugs combined with some form of psychological therapy (Alexander, 1991; Fahy et al., 1992; Marriott et al., 1989). This view arose from the fact that even the most effective drugs or most effective psychological therapies do not eliminate panic disorder in all cases. Combination treatments may be a way to improve treatment outcome. However, some clinical investigators have cast doubt on this approach, claiming that drugs impair the efficacy of psychological therapies (Marks, 1987; Sanderson & Wetzler, 1993). This chapter will review the evidence bearing on this important controversy. The basics of drug treatment for panic disorder will not be covered; the reader is referred to texts or articles dealing specifically with this issue (e.g., American Psychiatric Association, 1998; Canadian Pharmaceutical Association, 1990; Wolfe & Maser, 1994). This chapter will focus on the effects of combining drugs with either situational exposure or second generation cognitive-behaviour therapy (CBT2). These psychological therapies are of interest because they are among the most powerful treatments for panic disorder. We will review the literature on combining these treatments with each of the following drugs: benzodiazepines (diazepam, alprazolam), imipramine, selective serotonin reuptake inhibitors (fluvoxamine, paroxetine), buspirone, and moclobemide. These drugs were selected for review because they have been the focus of research on combination therapies. This chapter will also review the literature on another important way of combining psychological treatments with drug therapies. Even when

156

COMBINING PSYCHOLOGICAL TREATMENTSWITH DRUG THERAPIES

panic disorder is successfully treated with drugs alone, patients are sometimes referred for behavioural or cognitive-behavioural therapy in order to help them discontinue their medications without relapsing. As we saw in Chapter 6, patients commonly relapse when anti-panic drugs are discontinued. We will examine the question of which psychological interventions are useful for helping panic patients successfully taper off their medications.

COMBINING PSYCHOLOGICAL TREATMENTS WITH BENZODIAZEPINES

Diazepam Benzodiazepines can be divided into two classes, low or high potency (defined on the basis of required dose). Diazepam is the most common low-potency benzodiazepine used in panic treatment studies. Three studies have examined the effects of adding this drug to psychological treatments of panic disorder. In an early study using a cross-over design, Johnston and Gath (1973) treated four patients with imaginal exposure combined with either diazepam (20mg) or pill placebo. The drug or placebo was taken just before each exposure session. On behavioural and physiological measures, treatment was more effective when exposure was combined with diazepam, regardless of whether patients believed they were taking diazepam or placebo. This study had several important limitations, including its small sample size and lack of followup data. Importantly, imaginal exposure is among the weaker behavioural treatments for panic disorder (Chapter 5), and is not used today as a stand-alone treatment for panic disorder. Therefore the findings have dubious relevance for contemporary treatments. In a larger double-blind study, Hafner and Marks (1976) treated 41 panic patients with situational exposure combined with either diazepam (1 mg/kg) or pill placebo. The drug or placebo were taken shortly before exposure. Diazepam did not reduce the proportion of treatment dropouts and did not improve treatment outcome immediately after treatment or 6 month followup. However, during the first half of the exposure sessions, patients on diazepam reported less discomfort and had fewer panic attacks than those on placebo. This suggested that diazepam made exposure a little easier to endure, but the drug effects were not sufficient to influence treatment outcome.

More recently, Wardle et al. (1994) treated 91 patients with situational exposure combined with either diazepam (5-15mg/day) or pill placebo.

COMBINING PSYCHOLOGICAL TREATMENTSWITH BENZODIAZEPINES 157

In the previous studies (Hafner & Marks, 1976; Johnston & Gath, 1973), the drug or placebo was administered shortly before exposure, and thereby used as an exposure aid. In comparison, Wardle and colleagues administered diazepam or placebo on a daily basis for 12 weeks and then gradually tapered patients off their pills. Situational exposure was conducted during weeks 5-12. There were no significant differences between the diazepam and placebo conditions on measures of panic attacks, anxiety, or agoraphobia shortly after treatment or at 1 year followup. Thus, diazepam neither helped nor hindered situational exposure. Treatment was effective regardless of whether patients received the drug. Interestingly, there was no evidence of relapse when diazepam was withdrawn. It may be that relapse is less likely when combined with situational exposure exercises. On the other hand, the dose of diazepam was very low, which might account for the low relapse rate and for the failure of diazepam to influence treatment outcome. In summary, diazepam-taken in low or moderate doses-may make situational exposure somewhat easier to endure, but not to the point that the drug influences the effects of situational exposure at posttreatment or 6-12 month followup. There is no reason to expect different results for other low-potency benzodiazepines such as clorazepate or chlordiazepoxide. It remains to be seen whether the efficacy of exposure therapy is altered when diazepam or related drugs are used in higher doses. However, as we will see in the following section, high doses of alprazolam (measured in diazepam equivalent doses) hamper the effects of psychological treatments. The same may be true for high doses of diazepam and its congeners.

Alprazolam The London/I'oronto Study

In panic treatment studies, alprazolam is the most widely studied highpotency benzodiazepine. In a study of 154 patients with panic disorder (with agoraphobia) treated in either London or Toronto, Marks et al. (1993a) compared the efficacy of four 8-week treatments: (1) alprazolam plus situational exposure, (2) alprazolam plus progressive muscle relaxation, (3) pill placebo plus situational exposure, and (4) pill placebo plus progressive muscle relaxation. Alprazolam was gradually increased to a mean dose of 5.8mg/day, which is equivalent to 58mg/day of diazepam. Marks et al.'s relaxation procedure was used as an attention placebo because previous research had shown it to have little effect in treating panic disorder. After eight weeks of treatment, the psychological inter-

158

COMBINING PSYCHOLOGICAL TREATMENTSWITH DRUG THERAPIES

ventions (exposure or relaxation) placebo was gradually withdrawn.

were discontinued

and the drug or

The four conditions did not differ in the proportion of treatment dropouts. After 8 weeks of treatment, exposure was more effective than relaxation, and the alprazolam was superior to placebo on some measures but not others. Figure 7.1 shows the results on a clinician-rated measure of global improvement. Improvement at the end of eight weeks is indicated by the total height of the bars; i.e., 'improved, did not relapse' combined with 'improved, relapsed' for each treatment condition. On this measure alprazolam plus exposure did not differ from placebo plus exposure. At 6 month followup, the effects of alprazolam (relative to pill placebo) disappeared on every measure, whereas the effects of exposure (relative to relaxation) persisted on most outcome measures. More than half of the alprazolam treated patients lost some or all of their treatment-related gains. In contrast, gains were lost by only 12% of placebo-treated patients. The proportion of patients who improved and did not relapse at 6 month followup was higher for placebo plus exposure than for alprazolam plus exposure (see Figure 7.1). In other words, alprazolam plus exposure, com-

II!!Improved, did not relapse

100

■

90

Improved, relapsed

80

-"'... Cl)

C

70

C

60

w 50 ;;

·;:

I-

40

0

30

~ a

20 10 0 AE

PE

AR

PR

Figure 7.1. Clinician-rated global improvement: proportion of patients who became much or very much improved and remained so at 6-month followup. AE = alprazolam plus situational exposure; PE = pill placebo plus situational exposure; AR= alprazolam plus relaxation tape; PR= pill placebo plus relaxation tape

Reproduced by permission of the Royal College of Psychiatrists from Marks, I. M., Swinson, R. P., Basoglu, M., Kuch, K., Noshirvani, H., O'Sullivan, G., Lelliott, P. T., Kirby, M., McNamee, G., Sengun, S., & Wickwire, K. (1993). Alprazolam and exposure alone and combined in panic disorder with agoraphobia: A controlled study in London and Toronto. British Journal of Psychiatry, 162, 776-787.

COMBINING PSYCHOLOGICAL TREATMENTSWITH BENZODIAZEPINES 159

pared to placebo plus exposure, impaired treatment outcome in the long term. Marks and colleagues drew the following conclusions. The high-dose alprazolam effect might continue as long as the drug is given but [that] seems redundant. It has only a non-significant additive effect which disappears on discontinuation and interferes with maintenance of gains ... Most patients accept exposure, which yields twice more therapeutic gains by eight weeks, and which is usually maintained thereafter, despite no further clinical contact. (Marks et al., 1993a, p. 784).

These findings created a good deal of controversy, particularly among advocates of pharmacotherapy. Methodological criticisms were made by Spiegel et al. (1993) and rebutted by Marks et al. (1993b). The rebuttal highlighted some of the obstacles in disseminating results that favour psychological treatments over drugs. Support by Upjohn, the drug company funding the research, stopped abruptly when the results become known. Thereafter, Upjohn's response was to invite professionals to critique the study they had nurtured so carefully before. The study is a classic demonstration of the hazards of research funded by industry .... One has to wonder what the response would have been had the London/Toronto results been the opposite to what we found, that is, had alprazolam turned out to be at least twice as effective as behaviour therapy from the end of treatment onwards. Would so many errors have been made [by those criticizing the study], so many 'ifs' and 'buts' been raised? It seems unlikely. Such results would have been emblazoned in international and APA [American Psychiatric Association] conferences. Every psychiatrist would have received glossy mailings with multicoloured graphs showing the superiority of drug over behaviour therapy. But alas! Behaviour therapy turns out to be at least twice as effective as alprazolam. (Marks et al., 1993b, pp. 792-793)

Other Studies

What are the effects of lower doses of alprazolam? Riley et al. (1995) found that at posttreatment, the efficacy of CBT2 plus alprazolam did not differ regardless of whether patients received low or moderate doses of alprazolam (means of 1.1 vs. 3.Smg/day, respectively). Echeburua et al. (1993) found that low dose alprazolam (1.Smg/day) impaired the effects of exposure. These authors compared four 8-week treatments: (1) situational exposure, (2) alprazolam, (3) situational exposure plus alprazolam, and (4) situational exposure plus pill placebo. The proportion of dropouts did not differ across treatments. Patients treated with exposure plus alprazolam improved as much as those treated by exposure plus placebo. Patients treated in the three exposure conditions tended to improve more than patients receiving alprazolam alone. At the end of 8 weeks all treatments

160

COMBINING PSYCHOLOGICALTREATMENTS WITH DRUG THERAPIES

were discontinued. At that point relapse tended to be higher for exposure plus alprazolam (34% of patients) compared to exposure only (18%), alprazolam only (15%), and exposure plus placebo (0%). Thus, there was a negative interaction between exposure and alprazolam; combined treatment was less effective than singular treatments. For patients who had not relapsed immediately after treatment, a 6 month followup revealed that the maintenance of gains for exposure plus alprazolam was no better than for exposure alone.

Comment Research suggests that adding alprazolam to either situational exposure or CBT2 does little to improve treatment outcome in the short term, and can impair treatment response in the longer term. It may be that these conclusions also apply to other high-potency benzodiazepines such as clonazepam and lorazepam, although this needs to be investigated. Is there a place for benzodiazepines in behavioural or cognitivebehavioural practice? The hazards may outweigh any benefits. Marks et al. (1993b) asked 'why put patients on alprazolam in the first place to create a subsequent problem of discontinuation that then has to be dealt with by exposure, when exposure therapy creates no such problem and is followed by greater and lasting improvement?' (p. 793). A UK ban on the alprazolam congener triazolam highlights the need for care in using benzodiazepines, even in combination with non-drug treatments (Marks et al., 1993a). A further concern is that benzodiazepines impair motor coordination and thereby increase the risk of road traffic accidents and similar mishaps (Barbone et al., 1998). It might be argued that benzodiazepines would be a useful addition if they worked faster than behavioural or cognitive-behavioural therapies. However, as we saw in Chapter 6, there is no convincing evidence that benzodiazepines work more rapidly than CBT2. Benzodiazepines could be used for patients who have failed to respond to CBT2. On the other hand, other anti-panic medications (e.g., SSRis) might be better suited for this task because they have fewer of the risks associated with benzodiazepines. Spiegel and Bruce (1997) suggested that benzodiazepines might be implemented as an adjunctive treatment if the patient is at imminent risk for dropping out of CBT2. However, neither diazepam or alprazolam are better than placebo in reducing attrition from exposure therapy (Marks et al., 1993a; Wardle et al., 1994). Spiegel and Bruce further speculated

COMBINING PSYCHOLOGICAL TREATMENTSWITH IMIPRAMINE

161

that benzodiazepines might increase patients' adherence to behavioural or cognitive-behavioural therapies by providing some initial relief from panic attacks and reducing the anxiety associated with exposure. However, if these drugs facilitated adherence to any significant degree, then surely this would make the combined treatment more effective than exposure or CBT2 alone. The fact that this has not been found suggests either that (1) benzodiazepines do not facilitate adherence to psychological therapies, or (2) if they do increase adherence, this increment is so small that it does not improve treatment outcome.

COMBINING PSYCHOLOGICAL TREATMENTS WITH IMIPRAMINE Short-term Efficacy Imipramine is the most commonly used tricyclic antidepressant in the treatment of panic disorder, and the only drug of this class to be examined in studies combining psychological and pharmacological therapies. After 12-26 weeks of treatment, several studies have found that imipramine combined with exposure tends to be more effective than imipramine alone, and also is more effective than exposure alone, and exposure combined with pill placebo (Agras et al., 1991 unpublished, cited in Telch & Lucas, 1994; Mavissakalian & Michelson, 1986a; Mavissakalian et al., 1983; Telch et al., 1985; Zitrin et al., 1980, 1983). The proportions of dropouts did not differ across treatment conditions in any of these studies. The greater efficacy of the combination treatments is not because each treatment influences different symptoms; both reduced a range of symptoms of panic disorder, thereby making the combination more effective than either alone (Mavissakalian, 1993; Telch & Lucas, 1994). These encouraging findings are tempered by other findings on short-term efficacy. Marks et al. (1983) found that combined imipramine and exposure was no more effective than exposure plus pill placebo. Barlow (1998) found that combined CBT2 and imipramine was no better than CBT2 plus pill placebo. Agras et al. (1991 unpublished) found that exposure plus imipramine, compared to exposure plus placebo, produced smaller changes in panic-related catastrophic beliefs. For patients in combination treatment studies, the target dose of imipramine was typically 150-200mg/day. Such doses, administered over at least six months, appear necessary for the combined imipramine and exposure to outperform exposure alone (Mavissakalian, 1993). In practice it can be difficult to attain this goal because of imipramine side effects

T

162

COMBINING PSYCHOLOGICALTREATMENTS WITH DRUGTHERAPIES

(e.g., dry mouth, blurred vision, agitation, insomnia), which are particularly prominent during the first two weeks of treatment (Pohl et al., 1988). Many patients are unable to tolerate these side effects. In one study 43% of patients were unable to attain this dose range because of side effects (Mavissakalian & Michelson, 1986a). Lower doses decrease the problem of side effects, but lengthen the treatment period (Alexander, 1991), which may increase the likelihood that the patient will drop out of treatment.

Efficacy at Followup In followups ranging from 2 to 5 years after treatment had ended, combined imipramine and exposure tends to be no more effective than either treatment alone (Cohen et al., 1984; Lelliott et al., 1987; Mavissakalian & Michelson, 1986b). To illustrate, Mavissakalian and Michelson (1986b) conducted a 2 year followup of the singular and combined efficacy of situational exposure and imipramine. The superiority of the combined treatment over exposure alone was found immediately after treatment, but not at followup. This was because of the loss of treatment-related gains when imipramine was withdrawn, along with the continued improvement of patients treated with exposure alone.

At 6 month followup, Barlow (1998) found that patients treated with imipramine plus CBT2 tended to relapse as their medication was withdrawn. Patients treated with CBT2 plus pill placebo tended to maintain their gains when the placebo was discontinued. Thus, there was a negative interaction between imipramine and CBT2; the addition of imipramine to CBT2 impaired the efficacy of the latter. Barlow concluded that 'there seems no advantage to combining drug and cognitivebehavioral therapy' (p. 82).

Comment Adding imipramine in the range of 150-200mg/day to either situational exposure or CBT2 may sometimes improve treatment outcome in the short term-providing patients are able to tolerate doses in this range. Any advantage of combined treatment tends to be lost at followup, and there is some evidence that imipramine may impair the efficacy of CBT2. Thus, the clinician should not assume that combination treatment is necessarily better. One should consider a trial of CBT2 alone before contemplating the addition of imipramine. Other tricyclic antidepressants (e.g.,

COMBINING PSYCHOLOGICAL TREATMENTSWITH SSRls 163

desipramine) have yet to be investigated in combination treatment studies. However, there is no reason to expect the results to differ from those obtained from studies of imipramine.

COMBINING PSYCHOLOGICAL TREATMENTS WITH SELECTIVE SEROTONIN REUPTAKE INHIBITORS (SSRis) Compared to tricyclic antidepressants and benzodiazepines, SSRis have fewer side effects and lower overdose potential. SSRis also do not induce the drug dependency effects associated with benzodiazepines. As we saw in Chapter 5, there also is evidence that SSRis are among the most powerful anti-panic medications. Accordingly, they are becoming increasingly popular in the treatment of panic disorder. A handful of studies have examined whether treatment outcome is improved by combining SSRis (fluvoxamine or paroxetine) with CBT2 or situational exposure.

Fluvoxamine de Beurs et al. (19956) contrasted four conditions, three of which were twopart interventions consisting of one treatment followed by another: (1) 6 weeks of fluvoxamine (up to 150mg/ day) followed by 6 weeks situational exposure, (2) 6 weeks of pill placebo followed by 6 weeks of situational exposure, (3) 6 weeks of a partial CBT2 package (hyperventilation exercises and breathing retraining) followed by 6 weeks of situational exposure, and (4) 12 weeks of situational exposure alone. The treatments did not differ in the proportions of dropouts. At the end of 12 weeks, agoraphobic avoidance was reduced in all four conditions, with the combination of fluvoxamine plus exposure tending to be most effective. The other three conditions did not differ from one another on this variable. Panic frequency was equally reduced in all treatments except for the partial CBT2/ exposure package, which had little effect. A difficulty in interpreting these findings is that patients in the fluvoxamine plus exposure condition received half as much exposure (6 weeks) as those in the exposure alone condition (12 weeks). It is not clear whether the combined treatment would have been superior in treating panic if the treatments had been matched for length of exposure therapy. No followup data were reported, and so it is not known whether the advantage of fluvoxamine plus exposure in reducing agoraphobia was maintained at followup.

164 COMBINING PSYCHOLOGICAL TREATMENTS WITH DRUGTHERAPIES

In another study, Sharp et al. (1996) compared the following 13 week treatments: (1) fluvoxamine, (2) pill placebo, (3) fluvoxamine plus CBT2, (4) pill placebo plus CBT2, and (5) CBT2 alone. The fluvoxamine dose was 100-150 mg/ day during treatment, and patients were tapered off the drug at the end of 13 weeks. The conditions did not differ in the proportions of dropouts. At posttreatment and 6 month followup, improvement in panic disorder was better in treatments containing CBT2. The largest and most consistent treatment gains were for patients treated with either CBT2 alone or CBT2 plus fluvoxamine, although none of the differences between the active treatments were statistically significant. All tended to be superior to placebo alone. At the end of treatment the percentage of panic-free patients was as follows: fluvoxamine (69%), fluvoxamine plus CBT2 (83%), pill placebo plus CBT2 (76%), and CBT2 alone (70%). The percentages of panic-free patients at 6 month followup were not reported. Fluvoxamine plus CBT2 tended to exert its therapeutic effects a little earlier than CBT2 alone; the combined treatment outperformed placebo at the week 4 assessment, whereas CBT2 alone did not outdo placebo until the week 6 assessment. This study had three important limitations. First, a partial rather than full CBT2 package was used; interoceptive exposure and breathing retraining were not used. The full package tends to be more effective than partial packages (Chapter 6). Second, one therapist delivered the CBT2, with another therapist occasionally filling in. It is not known whether the results from this therapist can be generalized to other therapists. Third, the treatment conditions differed in the proportions of patients seeking additional treatment during the followup period. Thus, the followup data must be interpreted with caution.

Paroxetine Oehrberg et al. (1995) found that paroxetine (up to 60mg/ day) plus CBT2 was more effective than CBT2 plus pill placebo in reducing panic and anxiety. The effects on agoraphobia was not reported. The treatments did not differ in the proportions of dropouts. It is not known whether gains were maintained once the drug was discontinued. As we have seen throughout this chapter, followup data suggests that various drugs impair the efficacy of psychological treatments. The same may be true of paroxetine. Compared to outcomes obtained in other studies, Oehrberg et al.'s CBT2 produced weak results. The proportion of patients who were panic-free by the end of treatment was less than 36% for CBT2 plus paroxetine, and

COMBINING PSYCHOLOGICAL TREATMENTSWITH OTHER DRUGS

165

less than 16% for CBT2 plus placebo. These are very poor outcomes. Studies of CBT2 typically report that a much higher proportion of patients become panic-free (M = 88%: Chambless & Gillis, 1994). This raises the question of whether Oehrberg et al.'s therapists were sufficiently proficient in CBT2.

Comment The studies on combined SSRis and psychological therapies suffer numerous methodological problems and therefore preclude any firm conclusions about the merits of combining SSRis (fluvoxamine or paroxetine) with either CBT2 or situational exposure. It is also unknown whether any beneficial effects of adding SSRis are maintained at long-term followup, when treatments are discontinued. The followup data for other anti-panic medications (e.g., imipramine, alprazolam) raise the concern that in the long run, combined SSRis and CBT2 may be less effective than CBT2 alone.

COMBINING PSYCHOLOGICAL TREATMENTS WITH OTHER DRUGS Buspirone Cottraux et al. (1995) compared CBT2 plus buspirone to CBT2 plus pill placebo. Patients received up to 60 mg/ day of buspirone, which was discontinued prior to assessing outcome. At posttreatment, adding buspirone (compared to adding placebo) improved outcome on measures of generalized anxiety and agoraphobia but not on measures of panic attacks or on the proportion of treatment dropouts. It is unclear whether the treatments were more or less effective than CBT2 alone. At the end of treatment the proportion of panic-free patients was 71% for CBT2 plus buspirone and 74% for CBT2 plus placebo. This is slightly lower than the proportion of patients who are panic-free after CBT2 alone (M = 88%: Chambless & Gillis, 1994). A year after the end of treatment, Cottraux et al. found no difference between their treatments on any measure. This was because gains were maintained for the CBT2 plus buspirone group, whereas the CBT2 plus placebo group continued to improve. Cottraux et al.'s findings suggest that buspirone may temporarily augment the effects of CBT2 on some measures (anxiety, agoraphobia) but not others (panic frequency). These results are consistent with the results

166

COMBINING PSYCHOLOGICAL TREATMENTSWITH DRUG THERAPIES

of placebo-controlled studies of buspirone alone. Research suggests that this drug is useful in treating generalized anxiety (Rickels, 1990b), but ineffective for treating panic disorder (Sheehan et al., 1990, 1993).

Moclobemide Loerch et al. (1999) compared eight weeks of (1) moclobemide, (2) CBT2 plus moclobemide, (3) pill placebo, and (4) CBT plus placebo. Moclobemide dose was up to 600 mg/ day and then tapered before the posttreatment assessment. Dropout rates were highest for moclobemide (44%), followed by CBT2 plus moclobemide (21%), placebo (18%), and CBT plus placebo (7%). At posttreatment, on a range of outcome measures moclobemide alone was no better than placebo alone. CBT2 plus moclobemide was no more effective than CBT plus placebo, and both were superior to the other conditions. At 6 month followup, CBT2 plus moclobemide tended to be superior to CBT plus placebo. However, the interpretation of followup data is complicated by additional treatment seeking during the followup period. It would have been informative to include a CBT2only condition to see whether this treatment outperformed CBT2 plus moclobemide and CBT plus placebo.

WHEN SHOULD COMBINED TREATMENTS BE USED? According to the practice guidelines published by the American Psychiatric Association (1998), 'combining medication and CBT can be especially useful for patients with severe agoraphobia and for those who show an incomplete response to either treatment alone' (p. 21). The research literature does not support this claim. Although CBT2 can help patients who fail to respond to pharmacotherapy (Otto et al., 1999; Pollack et al., 1994a), the available evidence provides no support for the conjecture that drugs plus CBT2 are superior to CBT2 alone for partial responders. Moreover, the addition of drugs to either CBT2 or situational exposure does not improve efficacy on measures of agoraphobia (Brown & Barlow, 1995; de Beurs et al., 1995b; Chambless & Gracely, 1988, study 2; Craske et al., 1991; Emmelkamp & van der Hout, 1983; Fava et al., 1995; Oei et al., 1997). For severe agoraphobia, CBT2 with an emphasis on situational exposure (behavioural experiments) is currently the treatment of choice. Partial responders may benefit from detailed assessment of why they failed to fully respond to previous behavioural or cognitive-behavioural therapies. The reasons for partial response would dictate the nature of treatment.

TREATMENT INTERACTIONS: MECHANISMS 167

In some cases an extended course of CBT2 or situational exposure can improve treatment outcome (Rapp et al., 1983).

TREATMENT INTERACTIONS: MECHANISMS State-dependent

Learning

Why do some drug treatments interfere with psychological treatments of panic disorder? Marks et al. (1993a) suggested that drugs (e.g., alprazolam) may interfere with psychological treatments because of statedependent learning. That is, the learning that takes place during treatment (e.g., learning that agoraphobic situations are not dangerous) may be available in the patient's working memory only when the retrieval conditions match the conditions under which the learning originally took place. Learning acquired under the effects of an anti-panic drug may not be retrieved when patients are in a drug-free state. The higher the drug dose the greater the discrepancy between drug and drug-free states. Therefore, if relapse after drug treatment is due to state-dependent learning, then the greater the dose of alprazolam the greater the risk of relapse once the drug is discontinued. Findings from the London/Toronto study failed to support the state dependency hypothesis; alprazolam dose did not predict outcome at any point (Basoglu et al., 1994a).

Attributions for Improvement Further analyses of the London/Toronto data revealed that the best predictor of relapse in both the alprazolam and placebo groups was the degree to which patients attributed their improvement to the pills rather than to their own efforts (Basoglu et al., 1994a). Patients who attributed improvement to their pills were less confident about coping and had more severe withdrawal symptoms during the taper period. Withdrawal symptoms may further lead patients to attribute treatment gains to their pills (instead of their own efforts) and also lead patients to expect to relapse once the medication is withdrawn (Basoglu et al., 1994a). Basoglu and colleagues further suggested that attributing improvement to effective selfinitiated action (i.e., self-directed exposure exercises), rather than to an external agent like a drug, reduces the chances of relapse. This is because attributing improvement to self-initiated action encourages the person to persist with exposure exercises even while experiencing drug withdrawal effects. People who attribute their gains to the drug are less likely to enter feared situations when they notice the waning effects of the drug. Con-

168

COMBINING PSYCHOLOGICAL TREATMENTSWITH DRUG THERAPIES

sistent with this hypothesis, correlational analyses indfrated that patients who made self-attributions for treatment gains made significantly more outings into agoraphobic situations during the tapering-off period than did those who attributed improvement to the pills (Basoglu et al., 1994a). Several studies have shown that poorer response to behavioural or cognitive-behavioural treatment is predicted by the duration of the patient's use of benzodiazepines. Those who have been long-time users, and are probably still taking benzodiazepines, tend to benefit the least from the psychological therapies, even when researchers control for the severity of panic disorder at the beginning of the behavioural or cognitivebehavioural treatments (de Beurs et al., 1995a; Sartory et al., 1989; van Balkom et al., 1996). Long-term users are likely to become psychologically and physiologically dependent on their drugs. Long-term drug use may undermine the patient's confidence in coping without drugs, and may increase the tendency to attribute any symptomatic improvement to effects of drugs rather than their own efforts. Similar mechanisms may account for the findings that other anti-panic drugs (e.g., imipramine) interfere with psychological therapies at followup. However, this remains to be tested. Even if attributions for improvement are relevant to understanding relapse from combined therapies, it remains unclear why combined treatments (e.g., imipramine plus exposure) tend to be no more effective than singular treatment (e.g., imipramine alone) during the acute phase of treatment. It may be that other mechanisms play a role in treatment interactions, such as the mechanisms outlined in the following section.

Incompatible Influences on Anxiety Sensitivity In Chapters 5 and 6 we saw that the major pharmacotherapies for panic disorder are effective in treating this disorder, and are capable of reducing anxiety sensitivity (fear of arousal-related sensations) and underlying arousal beliefs (i.e., beliefs about the harmfulness of arousal-related sensations). As predicted by cognitive models of panic, reductions in panic disorder are predicted by reductions in anxiety sensitivity. Drug treatments and psychological treatments both reduce anxiety sensitivity (Chapter 6). This raises the question of why drugs plus CBT2, or drugs plus situational exposure, aren't more effective than either treatment alone. And why do drugs sometimes interfere with exposure or CBT2? One possibility is that drug treatments hamper psychological treatments by suppressing feared sensations and thereby making it difficult to use CBT2 or exposure to reduce anxiety sensitivity. Recent research, however, provides little support for this conjecture (Riley et al., 1995). Another pos-

TREATMENT INTERACTIONS: MECHANISMS 169

sibility is that drugs and psychological treatments reduce anxiety sensitivity in different, incompatible ways. Each method of reducing anxiety sensitivity is effective on its own, but when combined they interfere with one another to such a degree that combined treatment is no more effective-and sometimes less effective-than the individual treatments. Psychological therapies (notably CBT2 and situational exposure) appear to reduce anxiety sensitivity by providing the patient with corrective information about the harmlessness of arousal-related sensations. In Chapter 6 we discussed some mechanisms whereby drugs could reduce anxiety sensitivity. One possibility is that by dampening arousal sensations, anti-panic medications enable patients to believe that their medications prevent the feared catastrophe from occurring. Patients may report that they are less afraid of arousal-related sensations because they have medication to ward off feared catastrophes. Also, some arousal sensations may be attributed to drug side effects, thereby producing a reinterpretation of sensations (e.g., 'These sensations are not the dangerous ones that lead me to panic, they are simply indications that the medication is doing its work'). Thus, beliefs about medications may lead to reductions in anxiety sensitivity. This is a contingent reduction in anxiety sensitivity; arousal-related sensations are no longer feared because patients believe they are under the protective effects of medication (e.g., 'Palpitations are no longer dangerous because I'm under the protective effects of medication'). Under these conditions, medications would not alter the patient's beliefs about the dangerousness of the sensations in the absence of medication. Drug treatments and psychological treatments may provide patients with conflicting messages. Psychological treatments teach patients that arousal-related sensations are harmless and that there is no need to use drugs to avoid or dampen the sensations. In contrast, drug treatments may lead patients to believe that arousal-related sensations need not be feared because medication will prevent the sensations from becoming dangerously intense. If drug and psychological treatments work in these conflicting ways, then both alone may reduce anxiety sensitivity, but their combination may be no more effective than either alone. Medications taken on a PRN (as needed) basis may be especially pernicious, because these drugs are often taken just before the patient enters a fear-evoking situation. Drugs taken at this time may be especially likely to prevent the patient from disconfirming catastrophic beliefs about arousal-related sensations (Brown & Barlow, 1995). The treatment-followup data are consistent with these hypotheses. Psychological treatments (e.g., CBT2) teach patients that arousal-related sensations are harmless. Armed with this knowledge, patients are less likely to relapse when they encounter stressors that induce arousal sensations.

170

COMBINING PSYCHOLOGICAL TREATMENTSWITH DRUG THERAPIES

In contrast, drug treatments teach patients that they are safe only so long as they are on their drugs. Thus, once medications are withdrawn, the fear of arousal-related sensations should return. As we saw in Chapter 6, relapse commonly occurs when drugs are discontinued.

Comment Much remains to be learned about why combined therapies typically perform no better than singular treatments in the short term, and are less effective in the longer term. We reviewed some hypotheses about the mechanisms underlying treatment interactions. Other mechanisms are possible, and further work needs to be done to better understand the way drugs and psychological treatments interact with one another. Such research may eventually shed light on ways of better integrating these therapies, if they are to be integrated at all.

MEDICATION DISCONTINUATION Patients typically desire to discontinue their anti-panic drugs so long as relapse does not occur. But relapse is common (Chapter 6) and so it is important to find ways of dealing with this problem. The fact that antipanic drugs can interfere with psychological treatments provides another reason for discontinuing these medications. The clinician faces two challenges when tapering panic patients off their medications. The first is to find ways of minimizing withdrawal symptoms. Although withdrawal symptoms have been reported for various anti-panic drugs, such symptoms are particularly prominent during benzodiazepine discontinuation, therefore making it difficult for patients to discontinue these medications. The second challenge is to prevent the reemergence of panic disorder. In the following sections we will review the strategies used to help patients discontinue benzodiazepines, and consider whether the interventions can be applied to withdrawal from other anti-panic drugs.

Benzodiazepine

Withdrawal

Withdrawal Symptoms Benzodiazepine discontinuation produces withdrawal symptoms even when the drugs are taken in medically recommended (therapeutic) doses. The symptoms are more likely to occur for benzodiazepines with short

MEDICATION DISCONTINUATION 171

rather than long elimination half-lives, such as alprazolam rather than diazepam (Tyrer et al., 1983). Withdrawal symptoms can occur after only four to six weeks of regular benzodiazepine use (Fontaine et al., 1984; Murphy et al., 1989), although about six months of use is required before most patients are likely to have withdrawal symptoms (Rickels et al., 1988). A variety of symptoms have been ascribed to benzodiazepine withdrawal (Ashton, 1984). These include anxiety, panic attacks, agoraphobia, depression, gastrointestinal symptoms (e.g., dysphagia, vomiting, diarrhea), perceptual distortions (e.g., illusions, increased noise sensitivity), motor disturbances (e.g., tremulousness, ataxia), and influenza-like symptoms (e.g., stuffy nose, aches and pains, postural dizziness). Most withdrawal symptoms are unpleasant but harmless, although seizures can occur if high doses of benzodiazepines are abruptly discontinued. When benzodiazepines are gradually tapered, withdrawal symptoms typically dissipate within a few weeks after the patient stops taking the drug, although some clinicians report that symptoms can persist for months (e.g., Ashton, 1984). The peak intensity of the withdrawal symptoms is unpredictable, sometimes occurring during and sometimes after drug taper (Petursson & Lader, 1981; Tyrer et al., 1983, 1990). How does one distinguish withdrawal symptoms from relapse of panic disorder? The two differ in terms of time course; after drug discontinuation, withdrawal effects typically abate with time whereas relapse tends to emerge gradually and slowly (Roy-Byrne et al., 1993). Withdrawal and relapse also differ to some extent in their symptoms, the former being associated with a broader range of complaints. The Benzodiazepine Withdrawal Screening Questionnaire (Tyrer et al., 1990) can provide useful information in assessing withdrawal symptoms. Managing Withdrawal Symptoms Gradual rather than rapid drug taper helps but does not necessarily prevent the occurrence of benzodiazepine withdrawal symptoms (Mellman & Uhde, 1986; Pecknold et al., 1988; Schweizer et al., 1990). Several cognitive-behavioural strategies have been developed to enhance the effects of gradual tapering (Higgitt et al., 1987; Nathan et al., 1986; Onyett & Turpin, 1988; Sanchez-Craig et al., 1987; Tyrer et al., 1985). These treatments were designed for people with benzodiazepine dependence in general, although they can be implemented with panic patients. The rationale for these interventions is that drug-taking is maintained by negative reinforcement (escape from distress), and so drug discontinuation should be easier if patients are taught ways of coping with distress.

r

172

COMBINING PSYCHOLOGICAL TREATMENTSWITH DRUG THERAPIES

A commonly used treatment is anxiety management training (Suinn & Richardson, 1971). This involves training in progressive muscle relaxation (see Chapter 14) followed by training in deliberately visualizing unpleasant events or states that the patient has experienced. Patients are then trained to terminate distressing scenes by visualizing scenes that evoke relaxation or a sense of competency. Patients are instructed to practice these methods as homework assignments. Anxiety management training and other cognitive-behavioural anxiety-management strategies have proved moderately effective in helping patients discontinue benzodiazepines, although patients often relapse. A recent development is complaints management training (Elsesser et al., 1996). This entails strategies used in anxiety management training but also includes other techniques intended to ease withdrawal symptoms and promote a sense of control over them. Treatment is symptom-focused. That is, the choice of interventions depends to some extent on which symptoms are most severe for a given patient. Techniques commonly used by Elsesser et al. (1996) are shown in Table 7.1. Some of these methods may be unfamiliar to the reader and therefore require explanation. Breathing exercises involve slow abdominal breathing with short pauses at the beginning and end of each respiratory cycle. The Valsalva manoeuvre involves deeply inhaling and then increasing the Table 7.1 Coping with benzodiazepine withdrawal monly used in complaints management training Withdrawal symptoms Anxiety or tension Restlessness or agitation Tachycardia Sleep disturbance or fatigue Depression or lack of energy Sweating or hot flushes Poor memory or concentration difficulties Tightness in chest and throat

symptoms: Methods comCoping methods

• • • • • • • • • • • •

Progressive muscle relaxation Anxiety management training Distraction Physical exercise Valsalva manoeuvre Progressive muscle relaxation Planning pleasant activities Cooling wrists Making lists and notes Slow abdominal breathing Progressive muscle relaxation Valsalva manoeuvre

From Elsesser, K., Sartory, G., & Maurer, J. (1996). The efficacy of complaints management training in facilitating benzodiazepine withdrawal. Behaviour Research and Therapy, 34, 149-156.

MEDICATION DISCONTINUATION 173

pressure in the chest by tensing the intercostal and abdominal muscles while pushing air against the closed glottis for 10-15 s. This causes a shortterm reduction in heart rate. These interventions are described in more detail in Chapter 14. Elsesser et al. (1996) compared the efficacy of complaints management training and anxiety management training in helping 44 chronic benzodiazepine users discontinue their drugs. The proportion of patients with panic disorder was not mentioned. Both treatments consisted of nine weekly sessions, and drug withdrawal occurred during the first 4 weeks. Patients in both conditions were given information about benzodiazepine withdrawal symptoms, and were informed that the symptoms were transient. At the end of 9 weeks, Complaints Management Training, compared to Anxiety Management Training, was more effective in terms of the proportion of patients becoming drug free, and in terms of posttreatment measures of anxiety and depression and the number of severe withdrawal symptoms. There was no difference between treatments in terms of benzodiazepine abstinence rate at 6 month followup (overall, 65% were abstinent). Thus, complaints management training was more useful in facilitating withdrawal in the short term but not in the longer term. Both treatments were similar in terms of attrition, with 57% of patients dropping out before the programs ended. The high rate of attrition may reflect the severe nature of the sample; all patients had failed previous attempts to withdraw from benzodiazepines. Additional or alternative interventions may be needed for helping these patients withdraw from their drugs.

Preventing Panic During Benzodiazepine

Withdrawal

An important question is whether complaints management training or anxiety management training are useful in helping panic patients withdraw from benzodiazepines. Clinical experience suggests that methods are beneficial but not sufficient to prevent relapse of panic disorder. Withdrawal symptoms can exacerbate a preexisting panic disorder, particularly when patients catastrophically misinterpret withdrawal symptoms as indications of impending death, insanity, or loss of control. These sensations emerge at a time when patients are more likely to be vigilant to the possible return of their disorder and may be especially prone to respond to sensations of withdrawal with increased arousal and catastrophic interpretations that often culminate in panic attacks. For many patients, benzodiazepine discontinuation removes a conditioned safety

174

COMBINING PSYCHOLOGICALTREATMENTS WITH DRUGTHERAPIES

signal: the expectation that a panic-free period will follow each dose of medication. Hence, these patients face the additional difficulty of increased expectancies of anxiety and panic when a dose is missed. This reaction has been termed pseudowithdrawaland may itself increase anxious apprehension and vigilance to somatic sensations and thereby hasten the return of panic episodes. (Otto et al., 1996a, p. 8, emphasis in original)

Otto et al. (1996a) developed a CBT2 protocol for treating panic disorder during benzodiazepine discontinuation. Treatment consists of eight weekly CBT2 sessions with the option of a further three weekly booster sessions. Drug taper begins during the third week of CBT2. The goal is to gradually discontinue the medication, although patients are permitted an occasional extra dose when needed. Otto et al. begin by rapidly tapering patients from high to low doses of benzodiazepines, and then slowly taper the low doses. The daily dose of alprazolam, for example, is reduced by 0.25mg every two days for doses above 2.0mg/day. When the latter dose is attained, the drug is tapered by 0.125mg every two days. Taper lasts about 5 weeks for patients on 2mg/day, and 7 weeks for those on 4mg/day. A goal of their CBT2 is to use any emerging sensations-including withdrawal symptoms-as opportunities for restructuring catastrophic beliefs. For example, allowing restlessness to build up to test the catastrophic belief that agitation will lead to insanity. Treatment consists of the same procedures used in conventional CBT2; cognitive restructuring, breathing retraining, relaxation training, interoceptive and situational exposure. Treatment also addresses drug withdrawal symptoms. For example, patients are encouraged to label their benzodiazepine withdrawal symptoms as the 'benzodiazepine flu'-uncomfortable but harmless sensations arising when benzodiazepines are slowly withdrawn. Patients are reminded that they have coped with influenza symptoms many times in their lives, even though the sensations are unpleasant. Relaxation training and breathing retraining are used to lessen the arousal associated with drug withdrawal, and these and other sleep hygiene methods are used to help patients cope with withdrawal-related insomnia (Otto et al., 1996a). The results of several recent studies support the use of this and similar CBT2 protocols for helping panic disordered patients discontinue alprazolam or clonazepam (Abelson & Curtis, 1993; Bruce et al., 1999; Hegel et al., 1994; Otto et al., 1993; Spiegel et al., 1994). In these studies patients had responded to drugs and the focus of the research was whether CBT2 helped patients discontinue their medications without relapsing. We will review some of these studies to illustrate the main findings.

c

I I: j

C

MEDICATION DISCONTINUATION 175

Otto et al. (1993) slowly tapered patients off either alprazolam or clonazepam. Patients were more likely to successfully stop regular drug use if they receivecJ. CBT2 during the tapering (76% of patients discontinued drugs) compared to no CBT2 (25% of patients). Type of drug (clonazepam vs. alprazolam) did not affect tapering success. Reasons for failure to complete taper included the emergence of panic attacks, anxiety, or other symptoms. Spiegel et al. (1994) tapered panic patients from alprazolam using a slow, flexible drug taper and supportive medical management, with or without concurrent CBT2. With the help of CBT2, 90% were successfully able to discontinue the drug, compared to 80% of patients who received no CBT2. At 6 month followup, relapse occurred in 50% of patients who received no CBT2, compared to 0% who received CBT2. These results are illustrated in Figure 7.2. The addition of CBT2 did not influence the severity of withdrawal symptoms, possibly because alprazolam was tapered very

10 ---------------------------, 8

-

1· -

Alprazolam with CBT2

- I

Ill

C

Cl)

:.-:;

-

6

IU

~-----7

C. 0

ci

I

L ___________

4

_

Alprazolam without CBT2

z

2

... 8. IU

'-t ~

C.

0

.... Months after Taper

Figure 7.2. Tapering panic patients from alprazolam with or without concomitant CBT2: cumulative number of patients achieving and maintaining drug abstinence From Spiegel, D. A., Bruce, T. J., Gregg, S. F., & Nuzzarello, A. Does cognitive behavior therapy assist slow-taper alprazolam discontinuation in panic disorder? American Journal of Psychiatry, 151, 876-881, Copyright 1994, the American Psychiatric Association. Reprinted by permission

176

COMBINING PSYCHOLOGICAL TREATMENTSWITH DRUG THERAPIES

slowly, thereby limiting withdrawal symptoms. At followup (mean = 3.2 years post-taper), 75% of the CBT2 sample did not seek out further panic treatment, compared to only 30% of patients not receiving CBT2 (Bruce et al., 1999). Bruce and colleagues also conducted a long-term followup (mean = 3.6 years post-taper) of Hegel et al.'s (1994) patients, who were treated with a very similar CBT2 protocol in an open trial. At followup, 76% of patients receiving CBT2 during drug taper did not seek out further panic treatment.

Comment Benzodiazepines have been the focus of drug discontinuation research because these drugs can produce severe withdrawal symptoms. Evidence suggests that complaints management training and anxiety management training can facilitate benzodiazepine withdrawal. Although these protocols are promising, they should be used with caution when treating panic patients. This is because the protocols could protect catastrophic beliefs from disconfirmation (e.g., 'If I hadn't used my distraction exercises when I felt restless, then the agitation could have built up to the point that I'd have gone crazy'). If these methods are used it is important to remind patients that the symptoms are harmless and that the coping methods are simply used to ease discomfort. Otherwise, the coping strategies may become safety behaviours. A further caveat is that the Valsalva manoeuvre (used in complaints management training) can cause fainting and other problems (see Chapter 14). When withdrawing panic disordered patients from their medications, complaints management training and anxiety management training . are unlikely to be sufficient. CBT2 for panic disorder may be required, as described by Otto et al. (1996a) and Spiegel et al. (1994). Detailed tapering schedules are available in Otto et al.'s (1996a) manual. CBT2 could be used in conjunction with complaints management training or anxiety management training, although the combination may confuse patients. CBT2 teaches patients that arousal-related sensations are harmless and need not be controlled. In comparison, the other training programs emphasize the importance of controlling the sensations by various coping strategies. Although several studies have shown that CBT2 can assist patients in tapering off benzodiazepines (alprazolam or clonazepam) without relapsing, it remains to be demonstrated that CBT2 can reduce relapse when patients are tapered off other anti-panic drugs (e.g., imipramine, fluvoxamine, or paroxetine). However, there is no reason to expect that CBT2 would not be helpful in these cases.

SUMMARYAND CONCLUSIONS 177

SUMMARY AND CONCLUSIONS As we saw in Chapter 5, some meta-analyses suggested that combined treatments (e.g., imipramine plus exposure) may be more effective than singular treatments (Mattick et al., 1990; van Balkom et al., 1997). In the present chapter we have taken a closer look at the outcome studies, including many studies appearing since the meta-analyses were conducted. A detailed survey of the literature paints an unimpressive picture of the value of combining drugs with behavioural or cognitivebehavioural therapies. Treatment outcome is not enhanced by combining drugs with either CBT2 or situational exposure. At followup, anti-panic drugs can impair treatment outcome, compared to the outcome for CBT2 alone or situational exposure alone. The studies reviewed in the present chapter consistently found no differences in dropout rates across treatments. This finding also contrasts with the meta-analyses, which found that drug treatments tended to have higher dropout rates. Regardless of whether one places emphasis on the meta-analyses or our narrative review of individual studies, the literature clearly raises concerns about the value of combining drugs and psychological therapies in the treatment of panic disorder. In the long run, drugs may undermine the patient's confidence in overcoming their panic attacks and related symptoms, particularly if patients attribute their gains to medications rather than to their own efforts at using the skills learned in behavioural or cognitive-behavioural therapy. Little is known about the effects of taking anti-panic drugs over very long periods. Short-term behavioural and cognitive-behavioural therapies have long-term beneficial effects, and represent benign alternatives to long-term pharmacotherapy. For patients wishing to discontinue their anti-panic medications, relapse can be prevented by initiating CBT2 during drug tapering.

Chapter 8

''-':'.".'•;;:J_,f{·': ~',_;_{·._:>.,,,--:;.

PREDICTING TREATMENT OUTCOME Although behavioural and cognitive-behavioural therapies are among the most powerful interventions for panic disorder, some patients are unable or unwilling to complete these treatments. Other patients are able to complete therapy but fail to respond or respond but then relapse. Accordingly, it is useful for the clinician to have an early indication of these treatment difficulties-either in the form of pretreatment predictors of poor outcome, or variables that can be assessed early in treatment to predict outcome. Robust predictors can shed light on reasons for treatment failure, and can identify patients at risk for poor response. Treatment protocols may need to be modified to compensate for any anticipated difficulties. This chapter will review the literature on predictors of treatment refusal, treatment dropout, and, for patients completing therapy, predictors of treatment response. The term 'poor outcome' will be used as shorthand for 'relatively poorer outcome.' The latter is a more cumbersome but technically more accurate term. Studies of predictors tend to identify patients who have a poorer outcome relative to other patients. Thus, we will use the term 'poor outcome' in a relative rather than absolute sense. As in previous chapters, 'CBT2' will refer to the second generation cognitive behaviour therapy (e.g., Beck, 1988; Clark, 1989; Clark & Salkovskis, 1987; Craske et al., 1994), and CBTl will refer to earlier therapies (e.g., Beck, 1976; Ellis, 1962; Meichenbaum, 1975). Also, the term 'pa,1ic disorder' will be used as a summary term for 'panic disorder with or without agoraphobia.' Some early predictor studies described their patients as 'agoraphobics'. Most, if not all, of these patients would have met DSM-IV criteria for panic disorder with agoraphobia. Predictors of psychological and pharmacological therapies will be reviewed. It may be that some variables predict poor response for one treatment but not another. This has important implications for treatment selection. For example, if comorbid major depression predicted poor outcome for CBT2 but not for antidepressant medications, then it could be argued that the latter should be used when panic patients are depressed.

I

~

TREATMENT REFUSAL 179

Research on outcome predictors is still in its infancy. Although a great many predictor studies have been conducted, most appear to be post hoc re-analyses of outcome trials, using whatever variables happened to be available. The literature is littered with failures to identify robust (replicable) predictors of treatment outcome. In fact, the majority of candidate variables have failed to emerge as robust predictors. This chapter will not review every predictor that has ever been investigated. Instead, we will review the findings for the main predictors that have been examined (e.g., agoraphobia, depression, marital satisfaction), along with those of theoretical relevance (e.g., anxiety sensitivity), and those that appear most promising for estimating outcome.

TREATMENT REFUSAL Reasons for Refusing Treatment The purpose of this section is to identify why otherwise suitable patients decline effective treatments for panic disorder when such treatments are offered to them. In studies of psychological and pharmacological treatments, patients who refuse treatment have not differed from those accepting treatment in terms of demographics, duration or severity of panic disorder (de Beurs et al., 1995b; Marks et al., 1993a; McNamee et al., 1989). In a detailed analysis of treatment refusal, Emmelkamp and van der Hout (1983) asked 25 panic disordered patients to complete a questionnaire on why they declined to participate in a program of situational exposure. Unfortunately, 9 of the 25 treatment refusers (36%) refused to complete the questionnaire, and it is not known whether results from the 16 respondents were representative of the entire sample. Accordingly, the results must be interpreted with caution. Most respondents attributed treatment refusal to external circumstances (e.g., inconvenient time, no baby-sitter available). This suggests that it is important for the therapist to explore the patient's motivation for seeking treatment, and to identify the patient's beliefs about circumstances that could interfere with therapy attendance. Sometimes external circumstances pose important obstacles to treatment. At other times, however, the obstacle may have more to do with the patient than the situation. For example, 'inability to arrange for a baby sitter' may reflect an assertiveness problem (e.g., difficulty asking one's spouse or relatives to baby-sit). Sometimes it can be useful to offer the patient one or two preliminary treatment sessions to explore the logistics of attending panic therapy, and to find ways of overcoming any obstacles.

180 PREDICTING TREATMENT OUTCOME

Emmelkamp and van der Hout (1983) found that characteristics of the therapist who first interviewed the patient appeared unrelated to treatment refusal. The role of therapist characteristics may become more important later in therapy, as patients become more familiar with their therapists. Preference for one form of treatment over another appeared to be an important factor in treatment refusal. Many treatment refusers were frightened of treatment, presumably because they would be required to confront their fears. Preferences for a particular treatment can arise for a number of other reasons, including the patient's beliefs about the causes of panic disorder and beliefs about the consequences of treatment. Refusal of pharmacotherapy may arise from fear of becoming addicted to drugs, or fear of having one's mind 'controlled' by drugs (Chapter 7). Some patients decline CBT2 because they strongly believe that their panic disorder is due to a 'biochemical imbalance' that can only be rectified by drugs. These and other misconceptions can be sometimes corrected. For patients believing their panic is 'biological,' the clinician can educate the patient about how thoughts and behaviours influence panic attacks (see Chapter 12). In couples therapy for panic disorder, the attitudes of the spouse are important in influencing treatment refusal. The patient is unable to participate in couples treatment if their partner is unwilling to become involved, or if the partner attempts to dissuade the patient from entering treatment.

Comment In summary, treatment refusal arises from a number of factors including practical constraints that make it difficult to attend therapy, and misconceptions about the nature and treatment of panic disorder. Cognitive interventions-directed at either the patient or, in the case of couples therapy, the patient's spouse-can be used to correct misconceptions.

ATTRITION The average attrition (% dropouts) over the typical course of treatment (8-16 weeks) is approximately 16% for situational exposure, 12% for relaxation training, and 8% for CBT2 (Chambless & Gillis, 1994). Although low, these rates are not trivial. Attrition is also a problem for drug treatments (Chapter 5). Sometimes patients drop out because they experience a rapid elimination of symptoms. More often, however, dropouts continue

ATTRITION 181

to experience symptoms of panic disorder (e.g., Aronson & Logue, 1987). Accordingly, there are good reasons for identifying predictors of premature termination and for implementing strategies for circumventing this problem. Several potential predictors have been investigated, as reviewed in the following sections.

Demographics Seven of eight studies show that gender does not predict dropout from psychological treatments or drug therapies for panic disorder (Beck et al., 1994; Carter et al., 1995; Chambless & Mason, 1986; Craske et al., 1997; Klosko et al., 1990; Mavissakalian & Perel, 1989; Michelson et al., 1988; cf. Hafner, 1983, for the sole exception). Similarly, most studies show that age at the time of treatment does not predict attrition (Barlow et al., 1989; Beck et al., 1994; Carter et al., 1995; Craske et al., 1997; Klosko et al., 1990; Michelson et al., 1988, 1996; cf. Mavissakalian & Perel, 1989, for an exception). Ethnicity, marital status, and education level also have failed to predict dropout from psychological therapies (Carter et al., 1995; Michelson et al., 1988, 1996). In a study of antidepressant medication combined with situational exposure, Aronson and Logue (1987) found that dropouts were more likely than completers to be unemployed. The prognostic significance of employment status has not been reported in other studies, although clinical experience suggests that it is of little value in predicting attrition-providing, of course, that treatment is covered through a health care plan. Unemployed patients may have difficulty paying for private treatment. In summary, demographic variables are generally of little value in predicting attrition from either psychological treatments or pharmacotherapies.

Clinical Features of Panic Disorder Age of Onset and Duration of Disorder In a study comparing paradoxical intention, situational exposure, and relaxation training, Michelson et al. (1988) found that dropouts, compared to treatment completers, tended to have an earlier onset of panic disorder. Three studies of panic disorder found that duration of disorder failed to predict premature termination for CBT2 (Barlow et al., 1989; McLean et al., 1998; Michelson et al., 1996). For imipramine treatment, Mavissakalian and Perel (1989) found that dropouts, compared to completers, tended to have longer disorder duration. In summary, one study suggests that age

l 82 PREDICTING TREATMENT OUTCOME of onset may be a useful predictor of outcome. This result needs to be replicated before we can have any confidence in the generalizability of the findings. Duration of disorder appears to be of limited predictive value, particularly for CBT2.

Global Severity Several studies of pharmacological and psychological treatmentsincluding studies of CBT2-have found that pretreatment global severity of panic disorder does not predict who will drop out of treatment (Beck et al., 1994; Carter et al., 1995; Marks et al., 1993a; McLean et al., 1998; Michelson et al., 1988, 1996). Barlow et al. (1989), in a study comparing various forms of CBT2 to applied relaxation, found that dropouts had lower pretreatment global severity than completers. In comparison, in a study of situational exposure combined with either imipramine or placebo, Zitrin et al. (1980) reported that dropouts, compared to completers, had greater global severity at the start of treatment. Thus, despite some inconsistencies, most (5 of 7) studies suggest that pretreatment global severity is not a useful predictor of who is at risk for dropping out of treatment.

Frequency and Severity of Panic Attacks Pharmacotherapy trials of imipramine, alprazolam, or propranololalone or in combination with exposure-have found that dropouts, compared to completers, tend to have greater frequency and severity of panic attacks (Marks et al., 1983; Munjack et al., 1989; Zitrin et al., 1980). In comparison, five of six studies of behavioural or cognitive-behavioural therapies have found that dropouts and completers do not differ in pretreatment severity or frequency of panic attacks (Barlow et al., 1989; Beck et al., 1994; Craske et al., 1997; Kleiner et al., 1987; McLean et al., 1998; Michelson et al., 1996; but cf. Michelson et al., 1988). If patients enrolled in pharmacotherapy trials tend to be more severe than those in behavioural and cognitive-behavioural trials, then the results may simply indicate that patients with more frequent and severe panic attacks are most likely to drop out of treatment. However, there is no evidence that patients in drug trials are more severe. The results therefore suggest that for CBT2 and behavioural therapies, panic frequency and severity have little value in predicting attrition. The findings also suggest that when panic attacks are frequent and severe, dropouts may be reduced by treating patients with CBT2 or with behavioural approaches rather than with pharmacotherapy.

ATTRITION 183

Agoraphobia Several studies have examined whether dropouts differ from completers in terms of the pretreatment severity of agoraphobia, or in terms of components of agoraphobia, such as the severity of the patients' most feared agoraphobic situations ('target phobias'). Most studies of behavioural and cognitive-behavioural treatments have found that measures of agoraphobia fail to distinguish dropouts from completers (Barlow et al., 1989; Carter et al., 1995; Craske et al., 1997; de Beurs et al., 1995a, b; Kleiner et al., 1987; McLean et al., 1998). Michelson et al. (1988, 1996) found that dropouts were more avoidant on some but not all measures of avoidance. Two of three studies of situational exposure combined with drugs (i.e., Marks et al., 1983; Zitrin et al., 1980; but not Wardle et al., 1994) found that measures of agoraphobia failed to distinguish dropouts from completers. Thus, most studies indicate that intensity of agoraphobia is not a useful predictor of treatment dropout. Even severe agoraphobia is. not a reliable predictor of attrition. Once severely agoraphobic patients are able to summon the courage to travel to the first treatment session, they often come to regard the clinic as a source of safety, which may make it easier for them to attend further sessions.

Anxiety Six of eight studies have found that pre-treatment anxiety (assessed as trait anxiety or average severity over the past week) fails to discriminate dropouts from completers, regardless of whether patients receive psychological or drug treatments (Barlow et al., 1989; Carter et al., 1995; de Beurs et al., 1995a; Klosko et al., 1990; McLean et al., 1998; Michelson et al., 1988; but cf. Michelson et al., 1996; Shear et al., 1994). Thus, pre-treatment anxiety is not a useful predictor of attrition.

Axis I Comorbidity Depression The prognostic significance of pretreatment depression has been examined by assessing depression as a dimensional variable (e.g., scores on the Beck Depression Inventory or on the Hamilton Rating Scale for Depression) and by defining depression as a categorical variable (e.g., presence vs. absence of major depressive disorder). Most studies of CBT2 and related therapies have found that depression does not significantly predict attrition, regardless of whether depression is measured dimensionally or categorically (Barlow et al., 1989; Brown et al., 1995; Carter et al., 1995; de

184

PREDICTINGTREATMENTOUTCOME

Beurs et al., 1995a; Kleiner et al., 1987; Klosko et al., 1990; McLean et al., 1998; Shear et al., 1994; but cf. Michelson et al., 1996). Similarly, depression (assessed dimensionally) did not predict dropout from alprazolam treatment (Klosko et al., 1990). One study of combined antidepressant medication and situational exposure found that attrition was predicted by the presence of major depression (Aronson & Logue, 1987). In another study of imipramine combined with exposure, dropouts scored higher than completers on the Hamilton Rating Scale but not on the Wakefield Depression Inventory (Marks et al., 1983). To summarize, pretreatment depression typically fails to predict attrition from psychological and drug treatments. This conclusion is tempered by the fact that severely depressed patients were probably not included in these studies. But on the other hand, severely depressed patients are unlikely to be seeking treatment for panic in the first place; they are more likely to be wanting treatment for their mood disorder.

Worsening of Depression during Treatment Although pretreatment depression is an unreliable predictor of who will drop out from panic treatment, studies of psychological therapies and drug treatments indicate that some panic patients become increasingly depressed during treatment, and as a result they either drop out or are withdrawn from panic treatment by their therapist (de Beurs et al., 1995b; Fux et al., 1993; Lydiard et al., 1987; Sartory et al., 1989; Wade et al., 1998; Welkowitz et al., 1991). These patients are a minority; depression typically improves during panic treatments (Chapter 5). How can the clinician predict which patient is at risk for becoming depressed during treatment and thereby dropping out? Unfortunately, there are no known predictors. A history of mood disorder is not predictive (Fux et al., 1993), nor is the severity of depression at the beginning of panic treatment (McLean et al., 1998). This highlights the importance of ongoing assessment during panic treatment, including the assessment of depression. In some cases the emergence of depression is linked to significant stressors or losses that happen to occur during panic treatment. There is no evidence that panic therapies cause depression. When depression does develop or worsen during panic treatment, the patient may need to receive a different form of treatment. Worsening of depression during treatment has been observed even in trials of 'antidepressant' medications such as fluvoxamine and fluoxetine (de Beurs et al., 1995b; Fux et al., 1993). Thus, the worsening of depression during panic treatment does not necessarily indicate that medication is needed. Erner-

ATTRITION 185

gent depression may be effectively managed cognitive-behavioural therapy.

with Beck et al.'s (1979)

Other Comorbid Axis I Disorders Surprisingly little is known about the effects of other comorbid disorders on attrition from panic treatment. Brown et al. (1995) found that neither concurrent generalized anxiety disorder nor concurrent social phobia predicted dropout. It is unknown whether other Axis I disorders predict dropout.

Personality Disorders Personality disorders are associated with d_ifficulties in establishing and maintaining a sound therapeutic relationship (Beck et al., 1990), and therefore may be predictors of treatment dropout. Consistent with this, Chambless et al. (1992) found that the presence (vs. absence) of paranoid personality disorder predicted premature termination from a treatment program consisting of situational exposure and psychotherapy. Reich (1988) found that the presence of any personality disorder predicted dropout from treatment with alprazolam or diazepam. Further research is needed to establish which personality disorders are the best predictors of attrition.

Cognitive Factors Anxiety sensitivity and worry about panic have both been found to be nonsignificant predictors of who will drop out of CBT2 (Craske et al., 1997; McLean et al., 1998). Further work is needed to examine the predictive significance of other cognitive variables, such as interoceptive acuity.

Therapy Variables Attitudes about Treatment It comes as no surprise that many studies have found that patients are likely to drop out of drug or psychological treatment if they believe their treatment or therapist is ineffective or lacking in credibility (Black et al., 1993; Cross National Collaborative Panic Study, 1992; de Beurs et al., 1995b; Emmelkamp & van der Hout, 1983; Keijsers et al., 1994, 1995; Noyes

186

PREDICTING TREATMENT OUTCOME

et al., 1988; Welkowitz et al., 1991). Preference for a particular treatment can also lead to attrition; e.g., preference for psychological treatments instead of pharmacotherapy, or vice versa (Emmelkamp & van der Hout, 1983; Keijsers et al., 1994; Klosko et al., 1990). Therapist credibility is most likely to be a problem with novitiate therapists. In a study of CBT2 delivered by psychopharmacologically oriented clinicians, for example, Welkowitz et al. (1991) found that 5 of 24 patients dropped out of the 12-session treatment. The authors remarked that 'early dropouts (at sessions 2 and 3) were probably due to a perceived low credibility of the most inexperienced clinician' (p. 474). Attrition may be highest for therapists perceived as 'cold' and inflexible in the way they follow the treatment protocol (Ernmelkamp & van der Hout, 1983).

Side Effects Unpleasant side effects are by far the most commonly cited reason for dropping out of drug treatments. This has been cited as a reason for dropping out of a variety of treatments, including alprazolam, diazepam, phenelzine, propranolol, imipramine, fluvoxamine, and even pill placebo (Black et al., 1993; Cox et al., 1988; Cross National Collaborative Panic Study, 1992; Klosko et al., 1990; Liebowitz et al., 1986; Maddock et al., 1993; Marks et al., 1983; Mavissakalian & Michelson, 1986a; Mavissakalian & Perel, 1989; Mavissakalian et al., 1983; Munjack et al., 1989; Noyes et al., 1984; Pecknold et al., 1994; Sheehan et al., 1980; Telch et al., 1985; Tobefi.a et al., 1990; Woodman et al., 1994). Patients sometimes drop out of CBT2 and exposure therapies for similar reasons. Here, the 'side effects' typically consist of transient increases in anxiety, irritability, or panic frequency. Patients who catastrophically misinterpret the side effects may be at particular risk for dropping out (e.g., those who misinterpret the side effects as an indication that the treatment is making them worse or will cause them to go crazy). Strategies for dealing with this problem are discussed in subsequent chapters. Patients may drop out if they are find situational or interoceptive exposure exercises too frightening (de Beurs et al., 1995a; Ernrnelkamp & van der Hout, 1983; Himadi et al., 1986; Keijsers et al., 1994; Laberge et al., 1993). To illustrate, de Beurs et al. (1995a) reported that 7 of 39 patients dropped out of their cognitive-behavioural treatment. According to the authors, 'it appeared that six out of the seven therapy-related dropouts did so in a very early stage of treatment, in most cases owing to fear of the hyperventilation provocations' (p. 111). de Beurs et al. suggested that when patients are extremely fearful of the sensations provoked by hyperventilation, a different anxiety management technique (pharmacological

ATTRITION 187

or psychological) may be a better option for treatment. Alternatively, one could proceed more slowly with the hyperventilation exercises. For example, beginning by asking the patient to take a single deep breath. Once this is accomplished the patient could be encouraged to attempt two or three deep breaths, and so on.

Other Reasons for Dropping Out The following are other commonly cited reasons for prematurely treatment: •

•

•

•

ending

Practical constraints. These include inability to complete treatment due to financial difficulties (e.g., costs of travelling to treatment, costs of taking time off work), time commitments (e.g., increased job hours or responsibilities), scheduling difficulties (e.g., change in work schedule), or moving out of town (Barlow et al., 1989; Beck et al., 1994; Craske et al., 1997; Hoffart, 1997; Kleiner et al., 1987; Klosko et al., 1990; McLean et al., 1998; Pollack et al., 1994a; Shear et al., 1991a; Wade et al., 1998; Zitrin et al., 1980). Rapid reduction in symptoms. Rapid improvement in panic disorder can lead the patient to conclude that no further treatment is necessary. This has been noted to occur for behavioural and cognitive-behavioural therapies (Barlow et al., 1989; Black et al., 1993) and imipramine treatment (Marks et al., 1983). Dropping out for this reason is a concern if patients are still symptomatic or when they discontinue treatment before the therapist has had an opportunity to implement relapseprevention procedures. Stressful life events and nonpsychiatric medical comorbidity. Patients may drop out of panic treatment because medical problems arise that require immediate attention (e.g., development of cancer), or because of stressful events, such as family crises (Arrindell et al., 1986; Barlow et al., 1984; Beck et al., 1994; Emmelkamp, 1980; Hafner & Marks, 1976; Hoffart, 1997; Jannoun et al., 1980; Keijsers et al., 1994, 1995; Noyes et al., 1984; Pollack et al., 1994a; Shear et al., 1991a, 1994; Tobefi.a et al., 1990; Wade et al., 1998). Stressful life events can add to the anxiety experienced during treatment, thereby making the latter difficult to tolerate. Spouse nonadherence. Premature termination from spouse-assisted behavioural and cognitive-behavioural therapies can occur when the partner refuses to adhere to the treatment protocol (Carter et al., 1995; Himadi et al., 1986; Oatley & Hodgson, 1987). Carter et al. found that although dropouts and completers from spouse-assisted CBTl did not differ in marital adjustment (as rated by the patient and spouse), the

r

188

PREDICTING TREATMENT OUTCOME

spouses of noncompleters rated themselves as less communicative about panic-related issues. Interestingly, the patients did not rate themselves as less communicative. Dropout may have occurred because the partner was reluctant to discuss treatment-related exercises (e.g., reluctant to help plan exposure assignments). Communication difficulties also may add to the stressfulness of therapy, thereby leading the patient to drop out. Sometimes the therapist can identify or anticipate factors leading to dropout, and so can take remedial action. For example, anticipating life events (e.g., the impending death of a terminally ill spouse) and implementing stress management or other strategies to help patients deal with the events. In the case of upcoming medical procedures (e.g., the need for orthopedic surgery) the therapist might opt to delay panic treatment until the other problems have been dealt with.

Comment The most promising predictors of attrition are the occurrence of stressful life events during treatment, comorbid personality disorder, comorbid nonpsychiatric medical disorders, and a range of treatment-related factors such as side effects and practical constraints making it difficult to attend treatment. Predictors of attrition from drug treatments and psychological treatments are generally similar to one another, although there are some exceptions. Pretreatment panic frequency and severity predict attrition from drug treatments but not from behavioural or cognitive-behavioural therapies. Many potentially fruitful predictors remain to be investigated. Much remains to be learned about the relationship between treatment dropout and the patient's motivation for undergoing treatment. Motivation can change over time and circumstance. To illustrate, a patient might be motivated to reduce their panic disorder in order to start dating. Once the patient succeeds at entering a relationship, motivation to work at therapy may be lessened sometimes to the point of prematurely ending treatment.

PREDICTING OUTCOME FOR PATIENTS COMPLETING TREATMENT Predictor studies typically use some measure of panic attacks, agoraphobia, or global severity as their primary outcome measures. Predictorsassessed either before treatment or early in therapy-have been used to

PREDICTING OUTCOMEFOR PATIENTS COMPLETING TREATMENT189

predict a variety of indices of treatment outcome, including: (1) scores on outcome measure(s) at posttreatment and followup, (2) magnitude of change in symptoms from pre- to posttreatment, (3) proportion of treatment 'responders' at the end of treatment, and (4) proportion of patients relapsing during the followup interval. With few exceptions similar results are obtained regardless of the method for assessing outcome. In the reviews in the following sections, we have omitted the many duplicate publications and studies using overlapping samples. If one was to tally up the number of studies supporting a given predictor, then duplicate or overlapping publications could give a misleading impression of the support for a given predictor.

Demographic Variables In 23 out of 25 studies, age at the time of treatment was a nonsignificant predictor of outcome of psychological therapies and drug treatments of panic disorder (Basoglu et al., 1994b; Black et al., 1994; Bowen et al., 1994; Chambless & Gracely, 1988, studies 1 & 2; Craske et al., 1991; Emmelkamp & van der Hout, 1983; Evans et al., 1991; Faravelli & Albanesi, 1987; Fava et al., 1988a; Fischer et al., 1988; Green & Curtis, 1988; Lelliott et al., 1987; Liebowitz et al., 1986; Mavissakalian & Michelson, 1986a; Michelson et al., 1988, 1996; Noyes et al., 1984, 1990b; Otto et al., 1996b; Rijken et al., 1992; Rosenberg et al., 1991b; Wade et al., 1998; but cf. de Beurs et al., 1995a; Woodman et al., 1994). Thus, the weight of evidence indicates that regardless of type of treatment and method of assessing outcome, age is not a useful predictor of treatment response. Although the age ranges in these studies were restricted to exclude children, young adolescents, and the elderly, other .studies have shown that CBT2 is effective for treatment of panic disorder in these populations (see Chapter 16). With regard to gender, two data sets described in three studies (Chambless & Mason, 1986; Hafner, 1981, 1983) indicated that women and men treated with situational exposure tend to be similar in terms of symptom severity at the end of treatment. In both cases, however, women had more frequent panic attacks than men at the end of treatment. Women had more frequent panic attacks at the beginning of treatment, and it appears that both genders had approximately equal reductions in panic over the course of therapy. Using a sample of patients treated with various pharma- . cotherapies, Maier and Buller (1988) similarly found that female gender predicted poorer outcome. In contrast, for two studies it was found that female gender predicted good outcome for situational exposure and imipramine (Mavissakalian, 1985; Mavissakalian & Michelson, 1986a).

l 90

PREDICTING TREATMENT OUTCOME

Another 23 studies found that gender did not predict outcome, regardless of type of treatment and method for assessing outcome (Basoglu et al., 1994b; Black et al., 1994; Bowen et al., 1994; Chambless & Gracely, 1988, studies 1 & 2; Craske et al., 1991; Ehlers, 1995; Evans et al., 1991; Fava et al., 1988a; Fischer et al., 1988; Green & Curtis, 1988; Lelliott et al., 1987; Liebowitz et al., 1986; Michelson et al., 1988, 1996; Noyes et al., 1984, 1989a, 199Gb; O'Rourke et al., 1996; Rosenberg et al., 1991b; Rijken et al., 1992; Wade et al., 1988; Woodman et al., 1994). Thus, most studies indicate that gender does not predict treatment response. Research also generally indicates that other demographic variables do not predict treatment response, regardless of the type of treatment and method of assessing outcome. Variables that appear to have little or no predictive significance include marital status, employment status, education level, social class, and ethnicity (Bowen et al., 1994; Clair et al., 1992; Faravelli & Albanesi, 1987; Fava et al., 1995; Fischer et al., 1988; Mavissakalian & Michelson, 1986a; Michelson et al., 1988; Noyes et al., 199Gb, 1993; O'Rourke et al., 1996; Woodman et al., 1994). It should be cautioned, however, that these studies were restricted in the range of education levels, social classes, and ethnic groups that were sampled. For example, it is not known whether extremely low levels of education or social class predict poor response to CBT2 or other therapies.

Clinical Features of Panic Disorder Age of Onset and Duration of Disorder Chronic, severe panic disorder with agoraphobia can lead the sufferer to become socially isolated and demoralized. Accordingly, the question arises as to whether chronic, early onset panic disorder predicts poor treatment response. The available evidence suggests that the answer is generally no. Six of eight studies of psychological therapies and drug treatments found that age of onset failed to predict outcome (Buller et al., 1991; Faravelli & Albanesi, 1987; Green & Curtis, 1988; Noyes et al., 1984; O'Rourke et al., 1996; Woodman et al., 1994; but cf. Noyes et al., 199Gb; Michelson et al., 1988). Most (24 of 31) studies found duration of panic disorder was also unrelated to treatment outcome (Black et al., 1994; Bowen et al., 1994; Brown & Barlow, 1995; Clark et al., 1997b; Craske et al., 1991; Ehlers, 1995; Emmelkamp & Kuipers, 1979; Emmelkamp & van der Hout, 1983; Evans et al., 1991; Faravelli & Albanesi, 1987; Fava et al., 1988a; Fischer et al., 1988;

PREDICTING OUTCOMEFOR PATIENTS COMPLETING TREATMENT191

Lelliott et al., 1987; Michelson et al., 1996; Munjack et al., 1989; Noyes et al., 1989a; O'Rourke et al., 1996; Otto et al., 1996b; Pollack et al., 1993, 1994b; Rosenberg et al., 1991b; Rijken et al., 1992; Sheehan et al., 1980; Woodman et al., 1994; but cf. Basoglu et al., 1994b; de Beurs et al., 1995a; Mathews et al., 1976; Mavissakalian & Michelson, 1986a; Noyes et al., 1990b, 1993; Scheibe & Albus, 1996). The two groups of studies-the 7 finding duration to be a significant predictor and the 24 finding it to be a nonsignificant predictor-do not reflect differences in treatments; psychological and pharmacological treatments are represented in both groups of studies. Similarly, there is no evidence of systematic differences between the groups of studies in terms of method of assessing outcome, or the mean number of patients included in each group of studies. In summary, duration of panic disorder is not a reliable predictor of treatment outcome at either posttreatment or followup. Similarly, age of onset of panic disorder fails to predict treatment outcome. Global Severity of Panic Disorder Eight studies-examining a variety of psychological and pharmacological treatments-have found that greater global severity at pretreatment predicted greater global severity at posttreatment and at followups of up to two years (Brown & Barlow, 1995; Faravelli & Albanesi, 1987; Hoffart & Hedley, 1997; Mavissakalian & Michelson, 1986a; Michelson et al., 1988, 1996; Pollack et al., 1994b; Rosenberg et al., 1991b). Two other studies found that pretreatment global severity did not predict whether patients will be classified as treatment responders or nonresponders (Black et al., 1994; Emmelkamp & van der Hout, 1983). The two sets of findings can be reconciled if we assume that the efficacy of most treatments is unrelated to the initial severity. For a given treatment, patients with high pretreatment scores on a measure of global severity will, on average, improve by some amount x. Similarly, patients with moderate scores on this measure also will improve by an average of x points (people with low levels of global severity are often excluded from outcome studies or are less likely seek treatment, so we won't consider floor effects here). Thus, pretreatment severity predicts posttreatment severity. If treatment 'success' or 'responder' status is defined by substantial decline in severity scores (x + ox), then the results of Black et al. (1994) and Emmelkamp and van der Hout (1983) suggest that pretreatment severity does not predict the magnitude of treatment response. This suggests the following (tentative) predictive rule: for CBT2 and other treatments, patients who start off treatment with very severe panic disor-

192

PREDICTING TREATMENT OUTCOME

der, compared to those with a milder disorder, are likely to be more symptomatic at the end of a given period of treatment (e.g., 12 weeks), although both groups of patients are likely to improve by much the same amount. Global severity also may predict the length of treatment required in order for the patient to be classified as a treatment success or responder.

Panic Frequency and Severity Out of 19 studies, 9 indicated that pretreatment frequency or severity of panic attacks predicted poor outcome (Black et al., 1994; Ehlers, 1995; Keijsers et al., 1994; Marchand et al., 1998; Mavissakalian & Michelson, 1986a; Michelson et al., 1996; Nagy et al., 1989; Noyes et al., 1990b, 1993; Woodman et al., 1994). Ten studies found that pretreatment panic frequency or severity was unrelated to outcome (Bowen et al., 1994; Chambless & Gracely, 1988, studies 1 & 2; Craske et al., 1995; Faravelli & Albanesi, 1987; Lelliott et al., 1987; Liebowitz et al., 1986; Otto et al., 1996b; Pollack et al., 1994b; Wade et al., 1998). The inconsistent findings do not reflect differences in treatments; psychological and pharmacological treatments are represented in both groups of studies. Similarly, there was no evidence of systematic differences between the groups of studies in terms of (1) method of assessing outcome, (2) method of measuring the predictor variable, or (3) the mean number of patients included in each group of studies. The results suggest that pretreatment measures of panic frequency and severity are unreliable predictors of response to treatment.

Agoraphobia Predictor studies have examined the prognostic significance of measures of the overall severity of agoraphobia and measures of components of this syndrome, such as the intensity of the patient's most severe agoraphobic fear (e.g., intensity of fear of walking a given distance from home). Twenty-one studies found that pretreatment severity of agoraphobia predicted poor response to treatment (Basoglu et al., 1994b; Craske et al., 1995; de Beurs et al., 1995a; Ehlers, 1995; Emmelkamp et al., 1992 [2 samples]; Faravelli & Albanesi, 1987; Fava et al., 1991; Fischer et al., 1988; Keijsers et al., 1994; Maier & Buller, 1988; Mavissakalian & Michelson, 1986a; Michelson et al., 1988, 1996; Noyes et al., 1989a, 1990b; Otto et al., 1996b; Pollack et al., 1990; Scheibe & Albus, 1996; Wade et al., 1998; Williams & Falbo, 1996). Four studies found that pretreatment severity of agoraphobia predicted good outcome (Hafner & Ross, 1983; Jansson et al., 1987; Maier et al., 1991; Stern & Marks, 1973), and 14 studies found that pretreatment agoraphobia was unrelated to outcome (Black et al., 1994; Bowen et al., 1994; Clark et al., 1997b; Emmelkamp & van der Hout, 1983;

PREDICTING OUTCOMEFOR PATIENTS COMPLETING TREATMENT193

Evans et al., 1991; Liebowitz et al., 1986; Mathews et al., 1974; Munby & Johnston, 1980; Pollack et al., 1993, 1994b; Rijken et al., 1992; Thomas-Peter et al., 1983; Tobefla et al., 1990; Woodman et al., 1994). Thus, about half of the studies (21 of 39) suggest that pretreatment agoraphobia predicts poor outcome. The three different categories of results (pretreatment agoraphobia as a predictor of good outcome, poor outcome, or a nonsignificant predictor) do not appear to be due to systematic differences in the type of treatments. Psychological treatments, for example, are proportionally represented in each outcome category. Nor were there any systematic differences between categories in terms of method of assessing outcome (severity after treatment vs. improvement during treatment), method of assessing agoraphobia, or when outcome was assessed (posttreatment vs. followup ). Some but not all of the nonsignificant findings may be due to low statistical power. However, there was no systematic difference between the sample sizes of studies reporting that agoraphobia was a significant versus nonsignificant predictor of outcome. It is possible that the inconsistent results are due to pretreatment differences in the range of severity of agoraphobia. Studies admitting a narrow range of severity (e.g., including only patients with severe agoraphobia) are range-restricted and therefore limited in their predictive power. Unfortunately, it is difficult to examine this possibility from the available research because of the multitude of methods used for assessing and classifying agoraphobia. To summarize, the available evidence suggests that measures of pretreatment severity of agoraphobia are of limited value in predicting treatment outcome. Most studies assessed agoraphobia by sampling part but not all of this syndrome-e.g., composite measures of severity of three agoraphobic fears. These predictors were typically used to predict some global measure of treatment response. It may be that other measures of agoraphobia will turn out to be more useful predictors of outcome. Anxiety With a few exceptions (Faravelli & Albanesi, 1987; Liebowitz et al., 1986; Noyes et al., 1990b), most studies indicate that pre-treatment measures of anxiety fail to predict treatment outcome, regardless of anxiety measure, treatment, and method of assessing outcome (Chambless & Gracely, 1988, studies 1 & 2; Clark et al., 1997b; Craske et al., 1995; Ehlers, 1995; Evans et al., 1991; Mathews et al., 1977; Michelson et al., 1996; Noyes et al., 1984; O'Rourke et al., 1996; Pollack et al., 1994b; Rijken et al., 1992; Scheibe & Albus, 1996; Woodman et al., 1994). Thus, pretreatment anxiety does not appear to be clinically useful for predicting treatment response.

194

PREDICTING TREATMENT OUTCOME

Axis I Comorbidity Presence Versus Absence of Comorbidity Here, comorbidity refers to a disorder (apart from agoraphobia) diagnosed along with panic disorder at the time of the pretreatment assessment. In a study of 126 patients treated with CBT2, Brown et al. (1995) found that pretreatment Axis I comorbidity did not predict treatment outcome. The comorbid disorders were most commonly generalized anxiety disorder, social phobia and mood disorders (major depression or dysthymia). Outcome was assessed at posttreatment and 3 month followup in terms of two measures: the proportion of patients who were panic-free over the past month, and the proportion who had achieved high end-state functioning (defined here as a low rating of panic disorder severity and panic-free over the past month). Brown et al. also assessed whether comorbidity at 3 month followup predicted outcome at 2 year followup. They found no effect in terms of their outcome variables, but did find that comorbid patients were more likely to have sought treatment for panic disorder or other disorders during the interval between 3 and 24 month followup. Other studies have reported generally similar findings (e.g., Tsao et al., 1998). Studies of situational exposure and pharmacotherapies have found that comorbidity-assessed as the presence versus absence of any comorbid disorder-was a nonsignificant predictor of outcome at posttreatment (Pollack et al., 1994b; Wade et al., 1998) and failed to predict relapse of panic disorder over followup periods of two years or more (Ehlers, 1995; Fava et al., 1995; Pollack et al., 1990). In sum, the mere presence of a comorbid disorder has little prognostic significance for panic treatments. Particular comorbid disorders may be of greater prognostic importance. The effects of such disorders may be obscured in studies that simply assess the presence versus absence of any comorbidity. Accordingly, we now will turn to consider the predictive importance of specific types of comorbidity.

Comorbidity of Specific Anxiety Symptoms or Disorders The outcome of either psychological or drug treatments for panic disorder are generally unrelated to the pretreatment presence (vs. absence) of generalized anxiety disorder, social phobia, or social phobia symptoms (Brown et al., 1995; Chambless & Gracely, 1988, studies 1 & 2; Clair et al., 1992; Emmelkamp, 1980; Emmelkamp & Kuipers, 1979; Emmelkamp &

PREDICTING OUTCOMEFOR PATIENTS COMPLETINGTREATMENT195

van der Hout, 1983; Emmelkamp et al., 1992; Fischer et al., 1988; Hoffart, 1997; Noyes et al., 1993; but cf. Pollack et al., 1993). Little is known about the prognostic significance of other anxiety symptoms or disorders. Comorbid Depression Pretreatment depression-assessed either dimensionally or categorically-has been one of the most intensely studied predictors of outcome of panic treatments. Pretreatment depression was a statistically significant predictor of poor outcome in 15 studies (Bowen et al., 1994; Ehlers, 1995; Keijsers et al., 1994; Lelliott et al., 1987; Maier & Buller, 1988; Mathews et al., 1974; Nutzinger & Zapotoczky, 1985; Pollack et al., 1993; Pyke & Kraus, 1988; Rosenberg et al., 1991a; Scheibe & Albus, 1996; van Valkenburg et al., 1984; Watson et al., 1973; Woodman et al., 1994; Zitrin et al., 1980). Pretreatment depression was an inconsistent predictor (i.e., in a given study, significant for some depression measures but not for others) or a uniformly nonsignificant predictor in another 28 studies (Basoglu et al., 1994b; Black et al., 1994; Brown et al., 1995; Chambless & Gracely, 1988, studies 1 & 2; Clark et al., 1997b; Craske et al., 1991; Emmelkamp & Kuipers, 1979; Emmelkamp & van der Hout, 1983; Evans et al., 1991; Faravelli & Albanesi, 1987; Fava et al., 1988a; Fischer et al., 1988; Jansson et al., 1987; Laberge et al., 1993; Liebowitz et al., 1986; Mathews et al., 1976, 1977; Mavissakalian & Michelson, 1986a; McLean et al., 1998; Michelson et al., 1988, 1996; Munjack et al., 1989; O'Rourke et al., 1996; Pollack et al., 1990, 1994b; Rijken et al., 1992; Tobefi.a et al., 1990). The two groups of studies-15 in which depression was a significant predictor and 28 in which it was an unreliable predictor-did not systematically differ in terms of method of assessing outcome, sample size, or means of assessing depression (dimensionally vs. categorically). However, the predictive value of pretreatment depression appears to vary with the type of treatment. Depression was a significant predictor of poor outcome in approximately half of the drug studies and half of the studies using combined treatment (drugs plus psychological treatment). In comparison, depression was a weak or unreliable predictor for psychological therapies. There is no reason to expect that depression in patients treated with psychological panic treatments was milder than depression in patients treated with other panic treatments. However, it is difficult to determine whether depression systematically varied across these studies because various methods were used to measure depression. One inde~ of severity of depression is a clinical diagnosis of major depression. A number of CBT2 studies show that the presence of this disorder

196

PREDICTING TREATMENT OUTCOME

is not a useful predictor of response to panic treatment (Brown et al., 1995; Black et al., 1994; Clark et al., 19976; Craske et al., 1991; Evans et al., 1991; Laberge et al., 1993; McLean et al., 1998). The relationship between pretreatment depression and outcome of panic treatment is illustrated by one of our recent studies (McLean et al., 1998). We treated panic disordered patients with 10 sessions of CBT2. A total of 37 patients had panic disorder with major depression, and 53 had panic disorder without major depression (most patients in both groups had some degree of agoraphobia). Over the course of treatment both groups of patients had equivalent reductions in panic frequency, agoraphobic avoidance, fear of panic, and anxiety. For patients with major depression, depressive symptoms also tended to abate over the course of treatment, although most comorbid patients remained clinically depressed at the end of treatment. We found that patients with comorbid major depression, compared to those without this disorder, tended to have more severe panic disorder at pretreatment, although the groups did not significantly differ at the end of treatment. Both groups had approximately the same proportion of patients achieving high end-state functioning (comorbid: 52%, noncomorbid: 61 %), with no significant difference between the two. We found that the results did not change when we defined major depression in terms of whether it was 'primary' (developing before panic disorder) or 'secondary' (developing after panic disorder). The pattern of results changed slightly when we assessed depression dimensionally. Pretreatment scores on the Beck Depression Inventory (BDI) significantly predicted reductions (assessed by residual gain scores) in global severity of panic disorder and anxiety, but failed to predict reductions in panic frequency, agoraphobic avoidance, or fear of panic. In those cases where the BDI was a statistically significant predictor, the effect sizes tended to be small (TJ2 ranged from 0.07 to 0.12), indicating that pretreatment BDI scores would be of limited clinical value in predicting outcome. Thus, consistent with most other studies, we found that pretreatment major depression failed to predict treatment response and that pretreatment BDI scores were also of limited predictive value.

In summary, for patients presenting for treatment of panic disorder, pretreatment depression is not a reliable predictor of treatment outcome. It may be that very severe major depression significantly hampers all panic treatments, including CBT2. However, patients with severe depression are unlikely to be receiving treatment for panic disorder; they are more likely to be in treatment for depression. Although depressive symptoms may abate over the course of panic treatment (McLean et al., 1998; Sokol et al., 1989), patients with comorbid major depression are likely to have clini-

PREDICTINGOUTCOME FOR PATIENTS COMPLETINGTREATMENT 197

cally significant depressive symptoms at the end of panic treatment (McLean et al., 1998). These symptoms can be effectively treated with Beck et al.'s (1979) cognitive therapy for depression (Woody et al., 1999).

Nonpsychiatric Medical Comorbidity Nonpsychiatric medical comorbidity may work in various ways to impair the efficacy of CBT2 and other treatments for panic disorder. Medical conditions (e.g., chronic obstructive lung disease) and their treatments (e.g., bronchodilators) can exacerbate panic disorder (Rosenbaum & Pollack, 1991). Panic may be exacerbated by the fact that serious medical conditions cause body sensations (e.g., dyspnea), which may make it difficult to convince patients that panic attack symptoms are harmless ('If some of my body sensations are due to serious disease, then maybe the same is true of my panic symptoms'). Nonpsychiatric medical comorbidity also has special relevance for the implementation of CBT2 because some interoceptive exposure exercises may be medically contraindicated for these patients (Chapter 13). Aerobic exercise, for example, is not used with patients with severe heart disease or other diseases that limit physical exertion (e.g., particular spinal injuries). Thus, nonpsychiatric medical comorbidity can limit the number of interventions available to the CBT2 therapist. Findings reported by Schmidt and Telch (1997) support the view that medical comorbidity impairs CBT2. The authors assessed the presence versus absence of nonpsychiatric medical morbidity in a sample of panic patients treated with CBT2. A variety of maladies were classified as medical comorbidity, including hypertension, asthma, arthritis, irritable bowel syndrome, back problems, and heart disease. Not all of these would preclude the use of interoceptive exposure exercises. Nevertheless, these interventions were probably less often used in the comorbid group than in the non-comorbid group. Nonpsychiatric medical comorbidity was unrelated to pretreatment severity of panic disorder, but was negatively correlated with response to panic treatment. Patients with medical problems were less likely to be classified as recovered from panic disorder. These findings suggest that nonpsychiatric medical comorbidity may be an important predictor of response to CBT2. The results of Schmidt and Telch remain to be replicated. Even so, the findings are consistent with clinical experience. It remains to be seen whether medical comorbidity predicts poorer response to other treatments such as pharmacotherapies.

198

PREDICTING TREATMENT OUTCOME

Personality Variables Dimensions of Personality Several personality dimensions have been examined as potential predictors of outcome of psychological therapies and drug treatments of panic disorder. No useful predictors have been identified. Most studies have found that neuroticism does not predict outcome (Clair et al., 1992; Evans et al., 1991; Faravelli & Albanesi, 1987; Mathews et al., 1974, 1976; but cf. Noyes et al., 1993). Hostility proneness, a facet of neuroticism, also failed to predict outcome (Evans et al., 1991; Oatley & Hodgson, 1987). The tendency to direct hostility outward (vs. inward) predicted poor outcome in some studies (Hafner & Ross, 1983; Thomas-Peter et al., 1983) but not in others (Emmelkamp et al., 1992; Evans et al., 1991). Extraversion similarly emerged as an inconsistent predictor of response to behavioural and cognitive-behavioural treatments. Extraversion predicted good outcome on one study (Mathews et al., 1974), poor outcome in another study (Faravelli & Albanesi, 1987), and was nonsignificant in other studies (Clair et al., 1992; Evans et al., 1991). With regard to locus of control (LOC), one study found that poor outcome was predicted by internal LOC (Michelson et al., 1983), another study found the opposite (Bowen et al., 1994), and other studies have found LOC to be a nonsignificant predictor (Emmelkamp & Kuipers, 1979; Michelson et al., 1988).

Personality Disorders Personality disorders-which can be regarded as extremes of normal personality dimensions-are associated with interpersonal problems, such as interpersonal conflicts or social-evaluative anxiety. These problems create arousal-related sensations, which may be catastrophically misinterpreted, thereby leading to panic (Chapter 3). Thus, the presence of personality pathology may increase the risk for panic attacks, and thereby exacerbate panic disorder. This suggests that the presence (vs. absence) of personality disorders may aggravate panic disorder throughout the course of panic treatment, resulting in smaller treatment gains. Personality disorders also may make it difficult for the therapist to establish a sound, collaborative relationship with the patient, which may further undermine the progress of panic treatment. Accordingly, there are a number of reasons why personality pathology might predict poor outcome for panic treatments. Consistent with this, pretreatment global severity of personality disturbance-as measured by the number of personality disorders a person has, or by the total scores on personality disorder inventories-predicted significantly poorer response to treatment in 15 of 20 studies (Black et al.,

PREDICTING OUTCOMEFOR PATIENTS COMPLETINGTREATMENT199

1994; Chambless et al., 1992; Faravelli & Albanesi, 1987; Fava et al., 1995; Green & Curtis, 1988; Hoffart, 1994; Hoffart & Martinsen, 1993; Keijsers et al., 1994; Marchand et al., 1998; Mavissakalian & Hamann, 1987; Noyes et al., 1990b; O'Rourke et al., 1996; Pollack et al., 1992; Reich, 1988; Sokol et al., 1989; but cf. Clair et al., 1992; Dreessen et al., 1994; Hofmann et al., 1998b; Mellman et al., 1992; Ra thus et al., 1995). Thus, regardless of method for assessing outcome, panic disordered people with personality disorders, compared to those without personality pathology, tend to respond less well to a range of panic treatments, including CBT2 and pharmacotherapies. How are specific personality disorders or traits related to treatment outcome? The following is a summary of the main findings. •

•

•

Histrionic personality disorder did not predict outcome in several studies of drug and psychological treatments (Chambless et al., 1992; Green & Curtis, 1988; Hoffart & Martinsen, 1993; Hofmann et al., 1998b; Mavissakalian & Hamann, 1987). Histrionic traits (as assessed by the MMPI Hy scale) predicted poor outcome in one study (Faravelli & Albanesi, 1987), and histrionic personality disorder predicted poor outcome in another study (Reich, 1988). In contrast, Chambless et al. found that the presence (vs. absence) of histrionic traits predicted greater improvement in panic frequency but was a nonsignificant predictor of improvement in agoraphobic avoidance. To summarize, these studies indicate that histrionic personality disorder is an unreliable predictor of treatment outcome. 'Dramatic' traits-characteristic of narcissistic, histrionic, and borderline personality disorders-predicted poor outcome in one study (Hoffart, 1994) but not in another (Hoffart & Hedley, 1997). Mavissakalian and Hamann (1987) found that impulsivity-a trait of these personality disorders-was associated with poor outcome to behavioural and drug treatments. Borderline traits were unrelated to outcome in five studies of psychological and drug therapies (Chambless et al., 1992; Green & Curtis, 1988; Hoffart & Martinsen, 1993; Hofmann et al., 1998b; Reich, 1988). In other words, most (6 of 8) studies suggest that dramatic traits fail to predict the outcome of panic treatment, at least for patients completing treatment. Suspiciousness (as assessed by the MMPI Pa scale) predicted poor outcome in one study of situational exposure (Faravelli & Albanesi, 1987). However, paranoid personality disorder was unrelated to outcome in several other studies (Chambless et al., 1992; Green & Curtis, 1988; Hoffart & Martinsen, 1993; Hofmann et al., 1998b; Mavissakalian & Hamann, 1987; Reich, 1988). These findings suggest that paranoid personality disorder is not a useful predictor of outcome for panic treatments.

200

•

•

PREDICTING TREATMENT OUTCOME

The tendency to subordinate one's needs-a dependent personality trait-predicted poor outcome for situational exposure (guided mastery), CBT2, and combined drug and exposure treatments (Hoffart & Hedley, 1997; Mavissakalian & Hamann, 1987). Dependent personality disorder predicted good outcome for group psychotherapy but not for individual psychotherapy (Chambless et al., 1992). The latter authors suggested that their finding may have been a chance finding (Type I error) due to the large number of statistical tests they conducted. Dependent personality disorder was unrelated to outcome in five other studies (Green & Curtis, 1988; Hoffart, 1994; Hoffart & Martinsen, 1993; Hofmann et al., 1998b; Reich, 1988). Thus, findings are mixed. Most studies suggest that dependent personality disorder fails to predict the outcome of panic treatment. Two studies suggest that outcome may be better predicted by a specific dependent personality trait. In 4 of 5 studies, the presence of avoidant personality disorder predicted poor outcome of psychological and drug treatments of panic disorder (Chambless et al., 1992; Green & Curtis, 1988; Hoffart & Martinsen, 1993; Reich, 1988; but cf. Hofmann et al., 1998b). Elevated sensitivity to criticism, an avoidant personality trait, also predicted poor outcome (Mavissakalian & Hamann, 1987). Hoffart (1994) found that pretreatment avoidant personality disorder traits predicted poorer outcome during the year following treatment, but not 1-2 years after treatment. Chambless et al. (1992) found that avoidant personality disorder predicted poor outcome for panic disorder even after controlling for pretreatment social phobia and depression.

In summary, most personality disorders and traits studied so far have emerged as inconsistent predictors of outcome of treatments for panic disorder. The strongest predictor of poor outcome is avoidant personality disorder and its traits. Several writers have attempted to explain why avoidant personality predicts poor outcome. These accounts are based on the fact that many 'agoraphobic' situations are, in fact, social situations (e.g., shopping malls, line-ups, traveling in buses, walking down public streets). Chambless et al. (1992) observed that 'since overcoming agoraphobia involves going out more into a world populated by others, pervasive anxiety about interactions with other people may make abandoning avoidance behavior seem less rewarding' (p. 208). Hoffart (1994) similarly suggested that 'patients with avoidant traits experience more fear of embarrassment as a consequence of showing anxiety symptoms to persons in their natural environment, and therefore have a more negative course' (p. 340). Thus, although severity of agoraphobia and the presence of comorbid social phobia are weak predictors of treatment response (as discussed earlier), longstanding, pervasive and intense sen-

PREDICTING OUTCOMEFOR PATIENTS COMPLETING TREATMENT201

sitivity to criticism (as characteristic of avoidant personality predicts poorer response to treatment of panic disorder.

disorder)

Avoidant traits also may interfere with interoceptive exposure exercises in CBT2, particularly if the patient feels too embarrassed to perform these exercises in front of the therapist or other group members (in the case of group treatment for panic disorder). In these cases it may be fruitful to combine panic treatment with cognitive-behavioural methods for treating avoidant personality disorder (e.g., Beck et al., 1990). It may come as a surprise that borderline

and paranoid traits and disorders generally failed to predict the outcome of panic treatment. As we saw earlier in this chapter, paranoid traits predict treatment dropout. Borderline traits and disorder also may predict dropout. This was not supported by previous studies (e.g., Chambless et al., 1992), possibly because of the low base-rate of borderline traits and disorder in those studies. Paranoid and borderline traits may fail to predict treatment response (for treatment completers) because few completers have these traits. In other words, paranoid and borderline traits may be more important as predictors of premature termination than as predictors of the extent to which a patient will benefit from a course of treatment for panic disorder.

Cognitive Factors Anxiety Sensitivity Cognitive models of panic suggest that treatment may be less successful for patients with extremely strong beliefs in the dangerousness of arousalrelated sensations (as indicated by extreme anxiety sensitivity). Compared to patients with less strongly held beliefs, it may be more difficult to bring about the necessary cognitive change with either exposure exercises (used as behavioural experiments) or cognitive restructuring. Similarly, elevated anxiety sensitivity at the end of treatment should predict vulnerability to relapse after panic treatment, as well as difficulty in discontinuing antipanic medication. Several studies had examined the relationship between pretreatment anxiety sensitivity (or related constructs) and outcome of behavioural and cognitive-behavioural therapies. Three measures have been used: the Agoraphobic Cognitions Questionnaire (ACQ), the Body Sensations Interpretations Questionnaire (BSIQ), and the Anxiety Sensitivity Index (ASI). These measures are reviewed in Chapter 9. For the present discussion note that the BSIQ and ASI are reliable and valid indices of anxiety sensitivity, with the BSIQ specifically assessing very high anxiety sensitivity. The

T

202

PREDICTING TREATMENT OUTCOME

ACQ has been used to assess anxiety sensitivity, but there are concerns regarding its validity for this purpose (Taylor, 1995). Keijsers et al. (1994) found that higher pretreatment scores on the ACQ predicted poorer response to CBT2 at posttreatment. Michelson et al. (1996) found that high ACQ total scores at pretreatment predicted low end-state functioning at posttreatment but not 3, 6, or 12 month followup. Treatments consisted of either CBTl, situational exposure plus progressive muscle relaxation, or situational exposure alone. Chambless and Gracely (1988; n = 134) found that poorer outcome for CBTl was predicted by higher posttreatment scores on the ACQ physical concerns subscale (i.e., fear of somatic sensations) but not on the ACQ social/behavioural concerns subscale (fear of publicly observable anxiety reactions). Both measures were nonsignificant predictors at 1 year followup. Many patients received further treatment during the followup interval, and so the deleterious effects of elevated anxiety sensitivity may have been overcome with additional treatment. In another study using a smaller sample (n = 30), Chambless and Gracely (1988) treated patients with situational exposure and breathing retraining. The authors reported a nonsignificant trend for higher scores on the ACQ physical concerns subscale to predict greater severity of posttreatment agoraphobia. The ACQ social/behavioural concerns subscale was a nonsignificant predictor. The inconsistencies across studies may be due to differences in sample size. Ost and Westling (1995) examined the predictors of outcome for two treatments; CBT2 and applied relaxation. Data were pooled across treatments to examine predictors of outcome. The authors found that posttreatment scores on the ACQ and BSIQ were significantly correlated with global severity at 1 year followup (r = 0.39-0.58). The correlations became nonsignificant when posttreatrnent global severity was controlled for (partial r = 0.20-0.29). The latter analysis is difficult to interpret because the cognitive models of panic (Chapter 3) suggest that posttreatment global severity is influenced by posttreatment anxiety sensitivity. Thus, partialling out posttreatment global severity amounts to partialling out covariance due to anxiety sensitivity, thereby underestimating the strength of association between posttreatment anxiety sensitivity and followup global severity. In a study comparing CBT2, applied relaxation, and irnipramine, Clark et al. (1994) found that high posttreatment scores on the BSIQ predicted greater severity of panic attacks and anxiety, and higher probability of relapse at 9 month followup. The results held up even after controlling for posttreatment panic and anxiety. Again, this may be an overly stringent test because posttreatment panic and anxiety are presumably influ-

PREDICTING OUTCOMEFOR PATIENTS COMPLETINGTREATMENT203

enced by posttreatment anxiety sensitivity. Even so, it is notable that anxiety sensitivity was significant in these analyses. A number of longitudinal studies have examined maintenance and relapse of panic disorder over periods of 1-2 years after various treatments including drugs and CBT2. Two studies found that ASI scores (measured at the time the patient entered the followup study) predicted greater risk of relapse (Otto et al., 1996b; Pollack et al., 1990). In comparison, Rijken et al. (1992) found that pre- and posttreatment ASI scores were nonsignificant predictors of relapse 18 months after a course of various behavioural or cognitive-behavioural therapies. In a 1 year prospective study of 'treatment as usual' in the community, Ehlers (1995) found that maintenance of panic disorder over the 1 year followup (for those patients who had the disorder on entering the study) was predicted by high ASI scores, as assessed at the beginning of the study. Patients in remission who later relapsed, compared to remitted patients who didn't relapse, did not differ in anxiety sensitivity. Finally, in a sample of panic patients receiving alprazolam with or without CBT2, Bruce et al. (1995, 1999) found that baseline-to-posttaper change in ASI scores was a significant predictor of alprazolam discontinuation; patients who had smaller reductions in anxiety sensitivity during treatment had greater difficulty in discontinuing alprazolam. All patients who reinitiated alprazolam did so after experiencing limited symptom panic attacks. Patients who discontinued the drug either were panic-free or coped with their panics without the need for alprazolam. It is unclear why the magnitude of treatment-related reductions in anxiety sensitivity was used to predict alprazolam discontinuation. Cognitive theories of panic predict that the posttreatment level of anxiety sensitivity should predict the likelihood that the patient would seek out anti·panic medications. The theories make no prediction about baseline-to-posttaper changes in anxiety sensitivity. Unfortunately, Bruce et al. did not report whether posttreatment anxiety sensitivity predicted alprazolam discontinuation. It may be that baseline-to-posttaper change was a marker of posttreatment ASI scores. In other words, if most patients started treatment with high scores (as is characteristic of panic disorder), then the magnitude of baseline-to-taper changes in scores is proportional to (i.e., negatively correlated with) post-taper ASI scores. The larger the changes, the lower the post-taper score. If this is the case then the results are consistent with cognitive theories of panic; high ASI scores at the end of drug taper predict who is most likely to relapse and thereby request further drug treatment. To summarize, there is some evidence to suggest that pretreatment anxiety sensitivity predicts treatment outcome immediately after treat-

204

PREDICTING TREATMENT OUTCOME

ment. The failures to replicate this finding are difficult to interpret because the results were based on the ACQ, which may not be the best tool for assessing anxiety sensitivity. Evidence suggests that anxiety sensitivity predicts risk for relapse. There have been some failures to replicate this finding, which also are difficult to interpret. It may be that people with elevated anxiety sensitivity, compared to those with lower anxiety sensitivity, are more likely to seek extra treatment during the followup period, which allows these people to achieve a clinical outcome similar to those with lower levels of anxiety sensitivity. This would confound any attempt to predict relapse. An important question for further investigation is what aspects of anxiety sensitivity best predict treatment outcome. Beliefs about the harmfulness of somatic sensations (e.g., palpitations) may be a better predictor than beliefs about publicly observable sensations (e.g., trembling) (Chambless & Gracely, 1988). It also may be that extremely strong beliefs about arousal sensations are the best cognitive predictors of poor outcome. That is, there may be a nonlinear relationship between strength of belief and treatment outcome. The following section will review the research on this form of overvalued ideation. Overvalued Ideation

An overvalued idea is 'an unreasonable and sustained belief that is maintained with less than delusional intensity (i.e., the person is able to acknowledge the possibility that the belief may not be true). The belief is not one that is ordinarily accepted by other members of the person's culture or subculture' (American Psychiatric Association, 1994, p. 769). In panic disorder, such ideas may refer to beliefs about body sensations (e.g., 'Dizziness is a sign of a brain tumor'), or they may refer to other sorts of beliefs (e.g., 'People will think I'm a basket case if they see me panic'). Cognitive models suggest that overvalued ideas about arousal-related sensations predict poor response to behavioural and cognitive-behavioural therapies. We will refer such ideation as overvalued arousal beliefs. In comparison to the literature on overvalued ideation in other disorders (e.g., obsessive-compulsive disorder), the literature on panic disorder is exceedingly small. Emrnelkamp and van der Hout (1983) compared 11 treatment 'successes' with 5 'failures.' Treatment consisted of situational exposure. Ratings of overvalued ideation about having a serious disease failed to distinguish the groups. The criterion for defining overvalued ideation was not specified, and details of the types of beliefs were not described. Therefore it is not clear whether the treatment successes and failures differed in terms of overvalued arousal beliefs.

'

1 l t

1

I 1

I

(

I (

E

t (

PREDICTINGOUTCOME FOR PATIENTSCOMPLETINGTREATMENT 205

Two other studies identified overvalued ideation as a source of failure to respond to situational exposure and CBTl. Himadi et al. (1986) examined treatment failures in 14 patients treated with spouse-assisted exposure. One of the two failures was said to have overvalued ideation:

'Despite reassurances from therapists and physicians, this client tenaciously held onto the belief that high anxiety may lead to the inability to breathe and then death. This client also was taking rather high doses of antidepressants and anxiolytics. In combination with the above factors, she would discredit any progress made in therapy by attributing effective coping with anxiety as a result of medication or some random factor. Finally, this client's husband continued to differentially reinforce her by being attentive only when she was panicking' (p. 113).

This description suggests that treatment failure was as least partly due to an overvalued arousal belief. The case also illustrates how attributions for improvement and interpersonal factors can contribute to poor outcome. We will consider these factors later in the chapter. Barlow et al. (1984) observed that 4 of their 28 panic patients seemed to have overvalued ideation because they had great difficulty entertaining the notion that their fears were unrealistic. Three patients had overvalued ideas about the dangerousness of somatic sensations. The fourth believed that anxiety would lead to psychosis. The latter patient was a participant in spouse-assisted group CBTl:

'The one client who deteriorated in the spouse group presented with the common agoraphobic complaint that excessive anxiety or panic would cause her to lose control and "go insane." Nevertheless, unlike most clients who can at least verbalize the irrationality of this fear, this client continued to think that it was a perfectly reasonable possibility. In fact, she did not speak during the entire 12 sessions of group therapy for fear that she would say something that the therapist would diagnose as psychotic. Her husband was forced to do all the interacting in the group. Repeated individual examinations did not reveal any psychotic thinking. However, this client did report growing up with a brother who was diagnosed schizophrenic at an early age and was eventually hospitqlized. This brother had frequent emotional outbursts that frightened her. Losing control emotionally was very clearly associated with the prospect of becoming schizophrenic.' (p. 52)

206

PREDICTING TREATMENTOUTCOME

Overvalued ideation appears to have arisen from a history of learning that intense emotions such as anxiety could lead to psychosis. The example also suggests that her beliefs may have been protected from disconfirmation by well-intended but misguided efforts by her husband. In summary, preliminary evidence suggests that overvalued arousal beliefs predict poor treatment outcome. There are no data on whether this ideation predicts treatment refusal or premature termination from therapy, although these possibilities seem likely. Methods of treating this form of ideation are described in Chapter 13.

Interoceptive Acuity As we saw in Chapter 3, if someone has elevated anxiety sensitivity, then they are likely to be hypervigilant for body sensations. This serves as an early warning mechanism that allows the person to rapidly attempt to avoid or escape the feared consequences of these sensations. Just as people with extreme anxiety sensitivity may respond less well to treatment, it may be that heightened interoceptive acuity also predicts poor outcome and greater likelihood of relapse. To test this possibility, Ehlers (1995) conducted a 1 year prospective study of 39 people who currently had panic disorder, and another 17 former panickers whose panic disorder had remitted for at least six months. Maintenance of panic disorder (i.e., nonremission despite ongoing treatment) was predicted by good heartbeat perception, which is an index of interoceptive acuity. Patients in remission who later relapsed, compared to those who didn't relapse, also had better heartbeat perception. Interoceptive acuity was a significant predictor even after controlling for anxiety sensitivity and panic frequency. These findings were replicated when patients were re-assessed two years later (Anke Ehlers, personal communication, 3 September 1998). In sum, good interoceptive acuity may be a useful predictor of treatment outcome, although more research is needed, including replication of the findings in fresh samples of panic patients.

Stressful Life Events The onset of panic disorder has been linked to the occurrence of stressful life events (Chapter 1). Clinically, stressors occurring before or during treatment often seem to exacerbate panic disorder, possibly because they produce a plethora of arousal-related sensations that can be catastrophically misinterpreted. Stressful life events also may interfere with CBT2 by

PREDICTING OUTCOMEFOR PATIENTS COMPLETINGTREATMENT207

making it difficult to focus on panic treatment. If a panic patient comes to a treatment session with a major crisis, then the latter usually demands the attention of the therapeutic hour. Wade et al. (1993) reanalyzed data from Michelson et al. (1988) to examine the extent to which stressors predicted outcome for three panic treatments: paradoxical intention, situational exposure, and progressive muscle relaxation. The three were found to be equally effective (Michelson et al., 1988) and so Wade et al. pooled the results across treatments. The presence or absence of chronic (>1 month) life stressors occurring before or during treatment was found to significantly predict less global improvement, and a lesser likelihood of patients attaining 'recovered' status. Poorer outcome was associated with each of the following chronic stressors: interpersonal conflict, life-threatening events, and interpersonal losses. In a related study, Michelson et al. (1998) investigated the effects of traumatic stressors on the outcome of treatments for panic disorder. The sample consisted of patients who had participated in the investigators' previous studies (Michelson et al., 1988, 1996). In addition to the treatments mentioned earlier, some patients also were treated with relaxation training combined with either situational exposure or CBTl. Some of the stressors appear similar to those assessed in the Wade et al. study (i.e., lifethreatening events and interpersonal losses), although others were different (e.g., sexual abuse prior to age 14). Michelson et al. (1998) found the occurrence of traumatic stressors was generally unrelated to posttreatment outcome. However, at 1 year followup these stressors more strongly predicted poor treatment outcome. A problem in interpreting these findings is that the assessment of traumatic stressors was conducted 3-5 years posttreatment, and so it is not clear whether some of the traumatic stressors occurred after panic treatment had ended. If they did, then, strictly speaking, these stressors cannot be regarded as predictors of treatment outcome; they are better seen as predictors of the reemergence or exacerbation of a previously treated disorder. In a naturalistic 1 year followup of patients treated with various drugs and/ or psychological treatment, Faravelli and Albanesi (1987) found that a pretreatment measure of recent stressful life events predicted greater global severity of panic disorder at followup. It remains to be seen whether measures of chronic stress predict poor outcome for CBT2. Clinical experience suggests that this is likely to be the case. In these cases it may be useful to incorporate distress management strategies into treatment, such as those described by Linehan (1993).

208

'

PREDICTING TREATMENT OUTCOME

Marital Dissatisfaction Several studies have examined the extent to which marital dissatisfaction (assessed in terms of the patient's dissatisfaction or the dissatisfaction of both partners) predicts response to treatment of panic disorder. In all these studies treatment involved behavioural or cognitive-behavioural therapy, sometimes in combination with pharmacotherapy. Marital dissatisfaction predicted poor outcome in 6 studies (Barlow et al., 1983; Bland & Hallam, 1981; Emmelkamp & van der Hout, 1983; Lelliott et al., 1987; Mathews et al., 1977; Milton & Hafner, 1979) and predicted good outcome in 1 study (Chambless & Gracely, 1988, study 1). Marital dissatisfaction was a nonsignificant predictor for many or all marital satisfaction measures in another 12 studies (Arrindell et al., 1986; Chambless & Gracely, 1988, study 2; Cobb et al., 1984; Craske et al., 1989; de Beurs et al., 1995b; Emmelkamp, 1980; Emmelkamp et al., 1992; Evans et al., 1991; Hafner & Ross, 1983; Himadi et al., 1986; Jansson et al., 1987; Keijsers et al., 1994). Barlow et al. (1983) found that marital dissatisfaction predicted poor outcome in CBTl without spouse assistance, but was a nonsignificant predictor for spouseassisted CBTl. This finding was not replicated by Emmelkamp et al. (1992), in which marital dissatisfaction failed to predict outcome, regardless of whether the spouse was involved in treatment. To summarize, most studies (12 of 19) indicate that marital dissatisfaction is a nonsignificant predictor of treatment outcome-at posttreatment and at followup. Similarly, a recent meta-analysis found that marital dissatisfaction did not significantly predict outcome of behavioural treatments for panic disorder (Daiuto et al., 1998). Although the results do not support the predictive utility of pretreatment marital dissatisfaction, the research is limited in a number of important ways. Most studies included only female patients, and some studies used crude or unvalidated methods for measuring marital dissatisfaction. Other measures of marital interaction seem more promising as outcome predictors. Craske et al. (1989) found that although marital dissatisfaction did not predict response to CBTl, treatment responders, compared to nonresponders, rated themselves and their partners at pretreatment as more communicative about the patient's fears. Responders saw themselves as receiving more spousal encouragement and support for their attempts at overcoming their symptoms. This suggests that measures of spousal communication may be a useful predictor of treatment outcome. Craske et al. concluded that specific communication training may be beneficial for couples who, at pretreatment, report infrequent communication about the patient's fears. This may improve responsiveness to the couple's treatment of panic disorder.

PREDICTINGOUTCOME FOR PATIENTSCOMPLETINGTREATMENT 209

Therapy Variables Attitudes Toward Treatment Six of eight studies of behavioural and cognitive-behavioural therapies found that patient ratings of treatment credibility and expectations of improvement did not predict treatment outcome (Chambless & Gracely, 1988, study 2; Clark et al., 1997b; Edelman & Chambless, 1993; Emmelkamp & Wessels, 1975; Mathews et al., 1974; Stern & Marks, 1973; but cf. Emmelkamp & Emmelkamp-Benner, 1975; Mathews et al., 1976). Treatment credibility may be more important in determining treatment refusal or attrition rather than treatment outcome. Motivation for Treatment Patient motivation, as rated by therapists at the beginning of treatment, was unrelated to treatment outcome in studies of exposure therapies (Mathews et al., 1974) and in a study of CBT2 (de Beurs, 1993). Keijsers et al. (1994) found that patient-rated treatment motivation (assessed before treatment) predicted good outcome for CBT2, as assessed by residualized posttreatment scores on measures of panic attacks and agoraphobia. The motivation measure consisted of a six-item scale measuring 'willingness to participate' in treatment. The items were: willingness to cooperate with treatment, willingness to make sacrifices, willingness to keep appointments, viewing complaints as somatic in nature, patient's perseverance, and their faith in treatment. The content of these items suggests that the scale measured a combination of motivation for treatment, perceived credibility of treatment, and the patient's beliefs about the nature of the disorder. Thus, there is some ambiguity as to what the predictor scale actually measured. In summary, only one out of three studies found that treatment motivation (assessed at the beginning of treatment) predicted treatment outcome, and that study used a dubious measure of patient motivation. At pretreatment, low levels of motivation (as rated by the patient or therapist) may be better predictors of who will refuse or drop out of treatment. For patients who remain in therapy, motivation for treatment may change over time. Patients who initially have low motivation may become increasingly motivated to work at therapy once they begin to experience treatment-related gains. Thus, level of motivation at the beginning of treatment m;3.ybe unrepresentative of the levels of motivation throughout the course of treatment. If that is the case, however, it still means that measures of motivation may be of little help in deciding (at the outset) who is likely to benefit from a course of treatment.

210

PREDICTING TREATMENT OUTCOME

Treatment Adherence It seems intuitively obvious that the degree of treatment adherence should predict outcome. Yet, the literature provides mixed support for this supposition. Bowen et al. (1994) found that the number of CBT2 sessions attended predicted good outcome at 6 month followup. Most other studies have examined the relationship between the degree of patient adherence to situational exposure exercises and treatment outcome, with the latter typically assessed in terms of agoraphobic fear and avoidance. Michelson et al. (1986) found that adherence predicted good outcome. Edelman and Chambless (1993) found adherence to be an inconsistent predictor. Other studies found adherence was unrelated to outcome (Barlow et al., 1984; Lelliott et al., 1987). Schmidt and Woolaway-Bickel (2000) found that patient-reported homework adherence did not predict CBT2 outcome. However, therapist ratings of the quantity and quality of homework was predictive, with quality being the stronger predictor of good outcome. The inconsistencies among studies are likely due to a variety of methodological factors, including the fact that most measures of adherence are unvalidated. Further research is required before firm conclusions can be drawn about the prognostic importance of homework adherence.

Therapist Characteristics and the Therapeutic Relationship Positive therapist behaviours are difficult to disentangle from measures of the quality of the therapeutic relationship. Therapist characteristics include the extent to which the therapist is perceived by the patient as being confident, empathic, caring, understanding, and involved. Such factors contribute to a sound therapeutic relationship. A number of studies have found that combinations of these variables predicted good treatment outcome for behavioural and cognitive-behavioural therapies (Bennun & Schindler, 1988; Emmelkamp & van der Hout, 1983; Keijsers et al., 1991, 1995; Margraf & Schneider, 1991; Williams & Chambless, 1990). However, the studies differ to some extent about which variables were most important in predicting outcome. Of all the variables examined, typically only a subset were significant predictors. Other studies have failed to find that outcome is predicted by positive therapist characteristics or by the quality of the therapeutic relationship (de Beurs et al., 1995a; Gustavson et al., 1985; Keijsers et al., 1994). Some studies used small samples thereby lacking statistical power (e.g., Gustavson et al., 1985). The null results also may have been due to range restriction; most patients describe the therapist and the therapeutic relationship in glowing terms (Keijsers et al., 1994; Williams & Chambless, 1990). It is difficult to determine whether the good ratings reflect the patient's true beliefs about the therapist. Given these difficulties, it appears that patient ratings of the char-

SUMMARYAND CONCLUSIONS 211

acteristics of the therapist or the quality of the therapeutic relationship are likely to be of limited value in predicting treatment outcome.

Rapid Reduction in Symptoms Studies of behavioural and pharmacological therapies have shown that early response to treatment (within the first 1-4 weeks) predicts outcome at the end of 8-12 weeks of treatment (Albus et al., 1990; Basoglu et al., 1994b; Mathews et al., 1974). It seems likely that the same would apply to CBT2 for panic disorder. Interestingly, however, we also saw earlier in this chapter that rapid response to treatment is also a reason that patients sometimes cite for dropping out of therapy. Premature dropout for this reason is a concern if patients have not been trained in methods of relapseprevention.

SUMMARY AND CONCLUSIONS Many treatment predictor studies suffer from methodological weaknesses, such as the use of predictor measures that have weak or unknown reliability and validity. As a result of these limitations, only the strongest prognostic variables are likely to consistently emerge as significant predictors of treatment outcome. Our review of the predictor research, as summarized in Table 8.1, shows that many variables are not clinically useful in predicting outcome. Range restriction may be a contributing factor; panic patients tend to have elevated pretreatment scores on many different indices of psychopathology (Chambless, 1985), which may result in there being an insufficient range of scores for these measures to predict treatment outcome. Most of the useful predictors identified to date are not treatment specific. An exception is that greater frequency and severity of panic attacks predicted attrition from drug treatments but not from psychological treatments. Many other candidate predictors remain to be investigated, such as commonly comorbid disorders (e.g., hypochondriasis, substance use disorders). Research is also needed to examine the optimal combination of predictors for estimating treatment outcome. Keijsers et al. (1994) proposed that 'the chances of becoming a treatment failure increase dramatically when more than one disadvantageous prognostic variable is present' (p. 704). That is, the more poor prognostic factors that are present, the lower the chances of a successful treatment outcome. Research is needed to fully test this hypothesis. Another possibility is that some variables may predict outcome for some patients but not others. In other words, there may be idiographic predictors of treatment outcome. For example, the occurrence of nonpsychiatric medical comorbidity (e.g., heart disease) may be a stronger predictor

212

PREDICTING TREATMENT OUTCOME

Table 8.1

Predictors of outcome for treatments of panic disorder Outcome variables

Candidate predictors

Background variables Demographic variables Stressful life events Marital dissatisfaction Personality variables Personality dimensions (neuroticism, extraversion, locus of control) Hostility proneness and direction of hostility (inward vs. outward) Comorbid personality disorder Pretreatment clinical variables Anxiety Unassertiveness Age of onset of panic disorder Duration of panic disorder Global severity of panic disorder Frequency or severity of panic attacks Severity of agoraphobia Comorbid anxiety disorder (other than agoraphobia) Depression (dimensional severity or presence of mood disorder) Comorbid nonpsychiatric medical disorders Pretreatment cognitive variables Anxiety sensitivity Interoceptive acuity Therapy variables Practical constraints making it difficult to attend treatment Positive therapist characteristics and quality of therapeutic relationship Patient's.negative attitudes toward the treatment offered

Treatment refusal

Attrition

Treatment response (for patients completing treatment)

u

u

? ?

p ?

u p u

?

?

?

?

u u

?

p

p

? ? ?

u

u u

u u

?

*

? ?

u

?

u

u ? u u ** u u u u

?

p

p

? ?

u ?

p p

p

p

?

u

B

p

p

u u

? ?

?

SUMMARY AND CONCLUSIONS 213

Table 8.1

(continued)

Outcome variables

Candidate predictors

Treatment refusal

Low motivation for treatment Treatment side effects Reduction of symptoms early in treatment Nonadherence with homework assignments Spouse nonadherence (for couples' treatment) ?

u p

B

•

••

? ?

p

Attrition

Treatment response (for patients completing treatment)

?

u

p p

? B

?

?

p

?

Not applicable. Insufficient information. Unreliable predictor (nonsignificant or inconsistently significant across studies). Presence (vs. absence) or greater severity of variable tends to predict relatively poorer outcome (less improvement, more attrition). Presence or severity of variable predicts relatively better outcome. Predicts poorer outcome for pharmacotherapies but was nonsignificant for psychological therapies. Predicts severity of symptoms at the end of treatment, but does not predict magnitude of treatment-related changes in symptoms.

of relapse for some patients (e.g., those with strong fears of cardiac sensations) than for other patients (e.g., those with strong fears of other sensations). Predictors of treatment outcome are important for alerting the clinician to potential problems, so that treatment can be modified to deal with these difficulties. A variable that significantly predicts poor outcome may not necessarily cause the outcome, it may simply be a correlate or coeffect. Nevertheless, significant outcome predictors provide a source of hypotheses about how treatment of panic disorder can be improved in particular cases. Our review suggests that in some cases it may be necessary to modify panic treatment in order to help the patient deal with problems other than panic disorder. In particular, helping patients cope with chronic stressors and serious medical conditions, and helping them modify maladaptive personality styles. Therapy-related variables also need to be attended to, such as the patient's attitudes toward treatment. Strategies for dealing with these problems are addressed in later chapters.

Part II

CLINICAL PROTOCOLS AND PROCEDURES

Chapter 9

ASSESSMENT

The purpose of this chapter is to review the targets of assessment relevant to second generation cognitive-behaviour therapy (CBT2), and to review the major instruments and promising new methods for assessing these targets. This chapter is based on the assumption that the reader has the skills needed to conduct a psychiatric evaluation, along with a basic understanding of the central issues in administering and interpreting psychological assessment instruments. Good introductions to these important topics are found elsewhere (e.g., Anastasi, 1988; Groth-Marnat, 1990; Shea, 1988). There are three main goals of assessment: (1) to establish an accurate diagnosis; (2) to obtain information needed for developing a case formulation, including an assessment of variables predicting treatment outcome; and (3) to evaluate treatment progress and outcome. The targets of assessment include the major features of panic disorder and the commonly associated symptoms and disorders. Table 9.1 presents a comprehensive list of the treatment targets. The list is based on the cognitive models of panic (Chapter 3), clinical experience, and the recommendations outlined by Shear and Maser (1994). The latter were the result of a consensus development conference on the assessment and treatment of panic disorder, sponsored by the US National Institute of Mental Health. There is no single best sequence for collecting the information in Table 9.1, and the clinician need not assess every area for every patient. However, covering all the areas will permit a better understanding of the presenting disorder(s), and thereby enable the clinician to develop a treatment package most appropriate for a given patient. A comprehensive assessment is particularly useful for complex (e.g., comorbid) cases, and for atypical or treatment-refractory cases. Space limitations preclude a review of all assessment targets and measures. The chapter will focus on methods for assessing the features of panic disorder that are most important to examine in routine clinical prac-

21 8 ASSESSMENT Table 9.1 Areas to assess in the comprehensive ing for treatment of panic disorder •

• •

•

• •

•

•

•

evaluation of patients present-

Medical evaluation Rule out medical conditions mimicking panic disorder Identify medical conditions that may contribute to, or exacerbate panic disorder Establish current Axes I and II diagnoses and assess degree of impairment in functioning Assess features of panic disorder symptoms (e.g., frequency, severity, pattern of occurrence): Panic attacks Worry about panic Fear and avoidance of specific situations, activities, or substances Assess living circumstances and lifestyle: Social network (including social support) Family functioning (e.g., fulfillment of parental responsibilities) Occupational or school functioning Life stressors (chronic and episodic) Consumption of caffeine, alcohol, nicotine, and street drugs (assess regardless of whether patient has a substance-use disorder) Assess patient's strategies for coping with panic disorder Obtain history: Circumstances in which the disorder(s) arose Previous psychopathology Previous treatment Family history of psychopathology Cognitive assessment of panic disorder and other disorders, e.g., Dysfunctional beliefs (including beliefs about arousal sensations) Catastrophic imagery Body vigilance Identify treatments for panic disorder (or other disorders) the patient is currently receiving: Concomitant psychological therapies Prescription medications Over-the-counter preparations Herbal remedies Other forms of health care utilization (e.g., emergency room visits) Assess motivation for psychological treatment

tice. We will cover some but not all the scales used in treatment outcome studies. Measures of global functioning and disability are commonly used in these studies (e.g., Guy, 1976; Sheehan, 1983), but are too broad to be of much value for the clinician. More specific measures are needed. Rather than obtaining a global rating of, say, disability in occupational functioning, it is more useful for the clinician to assess the precise nature of any impairments. A patient might have a great deal of difficulty traveling to work, but have little difficulty performing the job once in the 'safety' of

DIAGNOSTIC EVALUATIONS 219

the office. This sort of detailed information is more useful for treatment planning than a global measure of occupational impairment. The psychometric properties of the measures will not be discussed in detail. Technical details on reliability and validity are available elsewhere (e.g., see Bouchard et al., 1997, and the source articles for each measure). Instead, we will focus on the clinical utility and comparative strengths and weaknesses of the various measures, and whether they are worth including in a comprehensive assessment package.

DIAGNOSTIC EVALUATIONS Are Diagnostic Evaluations Useful? Diagnostic systems have been criticized for conveying insufficient detail to be useful in understanding and treating psychiatric disorders (Persons, 1991; Schulte, 1996). Diagnostic assessments are insufficient on their own. However, they can provide valuable treatment-related information. The task of assessing DSM-IV Axes I and II encourages the clinician to look beyond the patient's most salient symptoms. This helps the clinician identify psychiatric problems that might be otherwise missed (Wittchen, 1996). A structured diagnostic evaluation might reveal that the patient's panic disorder arises in the context of, say, a long-standing paranoid personality disorder. Armed with this information, the clinician can be alert for potential problems in treating panic disorder (e.g., premature termination from therapy: see Chapter 8), and can prepare to modify treatment accordingly (e.g., proceed more slowly with interoceptive exposure and use interventions for managing paranoid personality disorder, such as those described in Beck et al., 1990).

Screening for Panic Disorder Clinicians, particularly those working in general medical settings, sometimes fail to detect panic disorder (Katon, 1994). A brief screening method may be useful for increasing the chances of detection. Patients screening positive for panic disorder would then receive a more detailed assessment to confirm or refute the diagnosis. Panic disorder can be detected by simply asking patients whether or not they suffer from recurrent, unexpected panic attacks (e.g., 'Have you ever suddenly felt extremely frightened or anxious for no apparent reason?'). Computerized screening methods are available for administering these

220

ASSESSMENT

questions (Kobak et al., 1997; Weissman et al., 1998). Unfortunately, direct questioning about panic attacks can fail to identify cases of panic disorder, particularly in general medical settings. This is because some panic patients believe their symptoms (e.g., palpitations) are due to a serious organic condition (e.g., heart disease), and that their fears of impending doom are natural reactions to a life-threatening disease (Katon, 1988). Katon et al. (1987) found that such patients were accurately detected as having panic disorder when several somatic questions were included in the assessment. The most sensitive question was 'Do you ever have sudden episodes of rapid heartbeat or palpitations?' A promising new instrument was developed by Apfeldorf et al. (1994), who used four items from the Anxiety Sensitivity Index (Peterson & Reiss, 1992) to screen for panic disorder: • • • •

'It 'It 'It 'It

scares scares scares scares

me me me me

when when when when

I feel "shaky" (trembling)' I feel faint' my heart beats rapidly' I become short of breath'

Each item is rated on a five-point scale assessing strength of agreement, ranging from O (very little) to 4 (very much). Apfeldorf et al. found the four-item scale was reasonably effective in identifying panic disorder-as diagnosed by a structured clinical interview-in patients presenting to an outpatient anxiety disorders clinic. On their brief screen, a cut-off score of 11 had a 78% sensitivity, 73% specificity, and 75% overall hit rate. A cutoff score of 8 identified almost all cases of panic disorder (sensitivity = 93%), although it yielded a high proportion of false positives (specificity = 50%), which is what one might expect from a screening instrument. It remains to be seen how well this brief screen performs in other settings, such as community mental health centers and general medical clinics.

Medical Evaluation Medical Conditions that Mimic or Exacerbate Panic Attacks A complete assessment includes a general medical evaluation, as described in the Practice Guidelines for Psychiatric Evaluation in Adults (American Psychiatric Association, 1995). A medical evaluation-including a medical history, review of organ systems, physical examination, and blood tests-is important for identifying general medical conditions that mimic or exacerbate panic attacks or panic-like symptoms. These include

DIAGNOSTIC EVALUATIONS 221

hyperthyroidism, hyperparathyroidism, seizure disorders, cardiac conditions, vestibular disorders, substance intoxication or withdrawal, and (in rare cases) pheochromocytoma (Goldberg, 1988; Raj & Sheehan, 1987). These disorders should be investigated and treated before contemplating a course of CBT2. Suspicion of a medical mimic is raised (1) if the panic disorder begins after age 40, (2) if the attacks are associated with symptoms that are unusual for panic, or (3) if the attacks are very brief or stop abruptly Gacob & Lilienfeld, 1991). Panic-like episodes associated with the following symptoms suggest the possibility that a general medical condition may be responsible: hunger, headaches, loss of consciousness, loss of bladder or bowel control, slurred speech, or amnesia (American Psychiatric Association, 1994; Jacob & Lilienfeld, 1991). Panic attacks accompanied by hunger may be indicative of hypoglycemia; attacks accompanied by pounding headache may be indicative of pheochromocytoma (an adrenaline secreting tumour). Some 'attacks' actually may be hypertensive reactions, possibly tyramine-induced. This can occur if patients taking MAOI medications consume tyraminecontaining goods such as aged cheese or particular wines (McCabe & Tsuang, 1982). Hypertensive reactions are characterized by severe, crushing, throbbing headache, along with other symptoms such as sweating, flushing, nausea and vomiting, stiff neck, and light hypersensitivity (Katon, 1994). As further noted by Jacob and Lilienfeld (1991) in their survey of medical mimics, Panic attacks that are very brief or that have a sudden offset suggest seizure disorder or a cardiac arrhythmia such as paroxysmal atrial tachycardia .... Organic disorders are of greater concern if the patient displays impulsive, aggressive, and interictal behavior (seizure disorder); if the patient displays persistent tachycardia, warm, sweaty hands, and intolerance of warm environments (hyperthyroidism); if the patient has crushing chest pain (angina, myocardial infarction); or if there is a relative absence of avoidance or anticipatory anxiety. (p. 57)

These medical conditions do not invariably cause panic-like episodes; the conditions are sometimes exacerbators rather than sufficient causes of panic. To illustrate, Mr C.'s panic disorder arose shortly after he developed paroxysmal atrial tachycardia (PAT). Mr C.'s cardiologist regarded the PAT as benign. It was considered an exacerbator of Mr C.'s panic attacks because not all patients with PAT develop panic attacks or panic disorder. PAT is most likely to lead to panic disorder when the person catastrophically misinterprets cardiac sensations. Mr C.'s panic attacks

222 ASSESSMENT continued even when his PAT was managed pharmacologically, and so he received a course of CBT2. This case shows how a medical evaluation can dictate the optimal form of treatment, which in this case was a combination of CBT2 and medical management. Medical mimics and exacerbators can be easily overlooked, especially during a cursory medical evaluation. The CBT2 therapist should be alert for this possibility. Ms M. was referred to the author by her primary care physician for treatment of panic disorder. The patient presented with seemingly classic panic attacks. Classic, that is, except for the fact that her panics invariably began with auditory hallucinations. The patient's 'panic attacks' were, in fact, complex partial seizures, occurring without incontinence or the loss of consciousness. Although she had described symptoms of panic to her primary care physician, she failed to mention the hallucinations. When asked why she had not told her doctor about the hallucinations, she replied 'he never asked me'. The patient was referred to a neurologist for further assessment and treatment. Had the patient been treated with CBT2 it is possible that some of the interoceptive exposure exercises (e.g., hyperventilation exercises) would have triggered further seizures. Thus, CBT2 would have exacerbated her seizure disorder, and probably increased her fear of having further seizures. Medical Contraindications A medical evaluation is also important to determine whether there are any contraindications for using particular CBT2 methods. Medical contraindications for exposure exercises are reviewed in Chapter 13.

Diagnostic Evaluations for DSM-IV Axis I Disorders Research suggests that the most comprehensive and accurate diagnostic information emerges when the clinician uses open ended questions and empathic listening, combined with structured inquiry about specific events and symptoms (American Psychiatric Association, 1995). The interview is 'structured' in that it contains a series of questions that systematically cover diagnostic criteria for each disorder. It is up to the interviewer to formulate additional questions to clarify any ambiguities. The interviewer can 'tailor an interview by phrasing questions to fit the subject's understanding, asking additional questions that clarify differential diagnoses, challenging inconsistencies in the subject's account, and

DIAGNOSTIC EVALUATIONS 223

judging whether the subject's description of an experience conforms to the intent of the diagnostic criterion' (Spitzer et al., 1992, p. 624). Structured interviews typically start by asking the patient to briefly describe their main problems. For each problem, the clinician should obtain a detailed description of a recent episode (vvhen, where, how long, how did it end, cognitions, body sensations, emotions, behaviours, reactions of other people), and then check whether the episode was a typical example of the problem (Clark, 1989). Several different structured interviews can be used to diagnose panic disorder, including the Anxiety Disorders Interview Schedule for DSM-IV (ADIS-IV: DiNardo et al., 1994), the Composite International Diagnostic Interview (Robins et al., 1989), the Schedule for Affective Disorders-Lifetime, Anxiety (Mannuzza et al., 1986), and the Structured Clinical Interview for DSM-IV (SCID-IV: First et al., 1996). Two of the most popular structured interviews for DSM-IV are the SCID-IV and ADIS-IV. Each is a revision of its DSM-III-R counterpart and each is available in formats assessing either current or lifetime disorders. There have been few studies of the psychometric properties of the current versions of these interviews. Previous versions have adequate reliability for diagnosing most of the disorders they assess (e.g., DiNardo et al., 1993; Williams et al., 1992). The diagnostic criteria for most disorders changed slightly from DSM-III-R to DSM-IV, and so the psychometric findings from the earlier versions of the interviews probably can be generalized to the current versions. The SCID-IV and ADIS-IV each requires about 1.5-2.0h to complete. The duration of the interview depends on a number of things, including the complexity and number of presenting problems, and the patient's ability to describe their symptoms. The SCID-IV and ADIS-IV should be administered by a suitably trained clinician, typically a psychologist, psychiatrist, or a supervised student of these disciplines. The interviewer should be someone who has enough clinical experience and knowledge of psychopathology and psychiatric diagnosis to conduct a diagnostic interview without an interview guide (Spitzer et al., 1992). The SCID-IV and ADIS-IV each has its pros and cons. Both prompt the interviewer to collect information pertinent to differential diagnosis. The ADIS-IV elicits more information than necessary to establish a diagnosis. This interview contains several dimensional measures of the severity of various symptoms and syndromes. The reliability and validity of most of these measures remain to be established. Compared to the ADIS-IV, the SCID-IV assesses more disorders and assesses them more economically, eliciting just enough information to establish or rule out a diagnosis.

224 ASSESSMENT Personality Disorders Several questionnaires have been developed to assess personality disorders. Among the most widely used contemporary measures are the Millon Clinical Multiaxial Inventory-III (Millon, 1994), the Personality Assessment Inventory (Morey, 1991), the Personality Diagnostic Questionnaire-4 (Hyler, 1994), and the Wisconsin Personality Disorders Inventory (Klein et al., 1993). Among the more widely used contemporary structured clinical interviews are the International Personality Disorders Examination (Loranger et al., 1994), Structured Clinical Interview for DSM-IV, Axis II (also known as the SCID-II, version 2.0: First et al., 1994), and the Structured Interview for DSM-III-R Personality Disorders (Pfohl et al., 1989). Questionnaires have the advantage of being less time consuming (for the clinician) than structured interviews. However, compared to structured interviews, questionnaires tend to over-diagnose personality disorders (Zimmerman, 1994). Thus, although it might appear that the busy clinician can save interview time by assessing personality disorder via a questionnaire, the results may be misleading and therefore result in misdirected treatment plans. Compared to questionnaires, structured interviews have an advantage in that the clinician can elicit additional information to determine whether a given trait is present. The clinician can ask clarifying questions and ask the patient to provide examples of the trait in question. Structured interviews tend to have moderately good inter-rater reliabilities, and are more reliable than unstructured clinical interviews (First et al., 1995; Maffei et al., 1997; Zimmerman, 1994). One of the most efficient structured interviews for DSM-IV personality disorders is the SCID-II, version 2.0. Just like its predecessor (SCID-II, version 1.0: Spitzer et al., 1990), the current version comes with a personality questionnaire in which the patient endorses the presence or absence of personality disorder traits. Although version 2.0 has yet to be extensively evaluated, it is similar to version 1.0, and so its psychometric properties are likely to be much the same. The questionnaire for version 1.0 yields a high rate of false positives (Carey, 1994; Zimmerman, 1994) and a low rate of false negatives (Jacobsberg et al., 1995), and is therefore suitable as a screening instrument. The clinician reviews the completed questionnaire for responses indicative of personality pathology. These responses are then probed with the SCID-II interview. A recent multi-site study of psychiatric patients suggested that the interview (version 1.0) usually requires less than an hour of interview time (mean = 36min, SD= 19min: First et al., 1995). The interview has moderately good reliability, similar to that of most other structured interviews for personality

ASSESSMENTBATTERIES: INTERVIEWSAND SELF-REPORT INVENTORIES 225

disorders (First et al., 1995). The duration and psychometric properties of version 2.0 are probably similar. Because both versions use a screening questionnaire, they are quicker to administer than most other interviews.

Comment To summarize, accurate diagnostic information is important for treatment planning. Preliminary research suggests that a brief screening instrument adapted from the Anxiety Sensitivity Index may be useful in identifying cases of panic disorder. Such cases then may be subjected to more intensive diagnostic evaluations to determine whether a course of panic treatment is warranted. Among the most useful methods for diagnosing DSM-IV Axis I and II disorders are the SCID-IV and the SCID-II, version 2.0.

ASSESSMENT BATTERIES: INTERVIEWS AND SELF-REPORT INVENTORIES The following sections will review other contemporary interviews and self-report inventories to supplement the diagnostic interviews. Because panic disorder is multifaceted, our review will deal mainly with assessment batteries. That is, multi-scale measures of the severity of multiple aspects of panic disorder.

Fear Questionnaire The Fear Questionnaire (Marks & Mathews, 1979) is a small battery containing three five-item subscales, assessing agoraphobic avoidance, avoidance of social situations, and avoidance of blood-injury stimuli. For the present purposes we will focus on the agoraphobia subscale. It requires only a few minutes to complete and is widely used in clinical research (Ogles et al., 1990). It has good psychometric properties (reliability and validity) and is sensitive to treatment-related change (Cox et al., 1991; Mavissakalian, 1986; Oei et al., 1991; van Zuuren, 1988). Detailed norms are also available (Gillis et al., 1995; Mizes & Crawford, 1986; Trull & Hillerbrand, 1990). Despite its popularity, this subscale is of limited value in routine clinical practice. It assesses avoidance of only five agoraphobic situations, and fails to assess variables influencing avoidance; e.g., whether the presence of a trusted companion influences avoidance. Patients can be highly

226

ASSESSMENT

mobile when accompanied by a trusted companion, but severely limited in mobility when they are alone.

Mobility Inventory The Mobility Inventory (Chambless et al., 1985) is another small battery, taking about 5min to complete. It consists of three scales, assessing (1) agoraphobic avoidance when alone, (2) agoraphobic avoidance when accompanied by a trusted companion, and (3) frequency of panic attacks. Chambless and colleagues also constructed scales assessing the degree of anxiety or discomfort experienced when the person enters agoraphobic situations, but found that these scales were highly correlated with the agoraphobic avoidance scales. Accordingly, the anxiety/ discomfort scales were dropped from the battery. Regarding the Mobility Inventory's measure of panic, the patient is provided with a definition of panic, and then is asked to indicate how many attacks were experienced in the past week. Although panic frequency is an important variable, there are other aspects of panic that are important to assess, such as panic intensity, and the types of symptoms and cognitions associated with attacks. None of these are assessed by the Mobility Inventory. Another limitation is that the Mobility Inventory's definition of panic is very brief and may lead patients to confuse panic attacks with episodes of anxiety or startle. The Mobility Inventory's avoidance scales provide a broad assessment of avoided situations. Its avoidance-alone scale assesses the degree of avoidance of 26 situations (e.g., theatres, shopping malls, heights, being at home alone), and the avoidance-accompanied scale assesses the degree of avoidance of the same situations except, of course, being at home alone. Respondents are asked to rate how often they avoid these situations because of anxiety or discomfort. The Mobility Inventory's avoidance scales have performed well on tests of reliability and validity, and are sensitive to treatment effects (Chambless et al., 1985; Craske et al., 1986; Kwon et al., 1990). Norms are also available (e.g., Chambless et al., 1985). Despite their strengths, the avoidance scales have a number of limitations. Although they provide information on whether the presence of a trusted companion influences avoidance behaviour, they fail to assess the multitude of other factors that influence avoidance. Swinson et al. (1992a) modified the Mobility Inventory to better assess the variables influencing avoidance. In addition to the avoidance-alone and avoidanceaccompanied scales, Swinson et al. added a scale assessing the patient's degree of avoidance when they are not taking or carrying any medication

ASSESSMENT BATTERIES: INTERVIEWS AND SELF-REPORT INVENTORIES 227

with them. Although this modification is a step in the right direction, it suffers from a number of important limitations. The 'avoidance without medication' scale does not assess whether or not the patient is accompanied or alone. Therefore, ratings on this scale are ambiguous. It is unclear whether the ratings refer to 'avoidance without medication and without a trusted companion' or 'avoidance without medication but with a trusted companion.' A further limitation is that the revised scale assesses the effects of only two safety signals-medication and a trusted companion. There are many other kinds of safety signals and safety behaviours that influence avoidance (see Chapter 3). Unfortunately, none of the available scales assess the effects of these safety variables. Exposure probes and careful questioning by the clinician are needed to supplement the data obtained from either the Mobility Inventory or from other agoraphobia scales. In summary, the Mobility Inventory provides a useful, albeit partial assessment of agoraphobia. The inventory can be used to assess avoidance before and after treatment. It is not known whether the Mobility Inventory is sufficiently sensitive to detect week-by-week changes during therapy.

Panic Attack Questionnaire The Panic Attack Questionnaire (Cox et al., 1992b; Norton, 1988) is a selfreport battery that takes 10-15 min to complete. It provides a definition of panic attacks and then asks respondents to rate the frequency of their expected and unexpected attacks, and also to rate the duration and intensity of attacks. The questionnaire also contains items assessing medication use, agoraphobic avoidance, and situations in which panics occur. Preliminary reliability and validity data are encouraging (Norton, 1988). However, the questionnaire tends to over-diagnose the frequency of panic, compared to structured interviews (e.g., Norton et al., 1992). This may be because patients confuse panic attacks with episodes of anxiety or startle. A further limitation is that the Panic Attack Questionnaire was not designed to assess weekly changes in symptoms, and so it is not suited for monitoring the patient's progress during treatment.

NIMH Panic Questionnaire One of the most comprehensive self-report inventories is the NIMH Panic Questionnaire (Scupi et al., 1992). This 230 item inventory asks patients to provide information about a variety of areas of functioning relevant to

228 ASSESSMENT panic disorder. The questionnaire has 12 subscales assessing demographic information, past and current medical history, family history of psychiatric disturbance, generalized anxiety, various features of panic attacks (e.g., frequency, severity, and panic triggers), agoraphobic avoidance, worry, depression, and factors that may influence the course of panic disorder (e.g., major life events). A strength of this questionnaire is that it offers a clear definition of panic, including diagrams showing the distinction between panic and anxiety. Unfortunately, the questionnaire also has four important limitations. The first is its length. It requires about SOmin to complete. The second is that much of the information collected in this questionnaire is redundant with information gleaned from structured interviews such as the SCID-IV. The NIMH Panic Questionnaire should not be used in place of these interviews because it was not designed as a diagnostic instrument. The third limitation is that it tends to over-diagnose the frequency of panic attacks, when compared to estimates obtained by trained clinicians administering a structured interview (Scupi et al., 1992). Finally, there is only preliminary data on the questionnaire's reliability and validity, and no norms have been published for any of its clinical scales.

Panic Disorder Severity Scale The Panic Disorder Severity Scale (Shear et al., 1997) is an intervieweradministered measure intended for use with patients already diagnosed as having panic disorder. The scale was designed as a short, comprehensive assessment of the severity of symptoms over the past month. It contains seven items assessing the following domains: • • • • •

• •

Panic attack frequency. A global measure of the frequency of full and limited symptom attacks. Distress during panic attacks. Including the severity of distress during full and limited symptom attacks. Severity of anticipatory anxiety. Severity of panic-related fear, apprehension, or worry. Agoraphobicfear and avoidance. Fear and avoidance of panic-related sensations. Including fear and avoidance of substances, activities, or situations that evoke arousal-related body sensations. Impairment or interference in work functioning due to panic disorder. Impairment or interference in socialfunctioning due to panic disorder.

Each of these items is rated on a five-point severity scale. The entire scale takes 10-15min to complete. Shear et al. (1997) found that the scale

ASSESSMENTBATTERIES: INTERVIEWSAND SELF-REPORT INVENTORIES 229

performs well on indices of reliability and validity, and is sensitive to treatment-related change. Some clinical norms are available, consisting of pretreatment means and standard deviations for each item, obtained from panic patients with little or no agoraphobia (Shear et al., 1997). A self-report version has been developed (Penava et al., 1998), although its reliability and validity is unknown. The Panic Disorder Severity Scale is comprehensive, informative, and concise. Its measure of fear and avoidance of panic-related sensations is a valuable addition to panic assessment, which is not included in other interviews or batteries. Given that the scale was developed to assess symptoms over the past month, it appears best suited as a pre- and posttreatment measure, rather than a measure of how symptoms change from one therapy session to the next. The scale could be adapted to use a shorter time-frame, such as symptom severity over the past week. However, one would need to establish the reliability and validity of such a modification. As it stands, the Panic Disorder Severity Scale might be best incorporated into existing structured interviews such as the SCID-IV.

Panic and Agoraphobia Scale One of the most promising new batteries is the Panic and Agoraphobia Scale (Bandelow, 1995), which is a 13 item scale designed as a short, sensitive measure for treatment outcome studies. It is intended for patients already diagnosed with panic disorder. It comes in self-report and intervieweradministered versions, each of which takes about l0min to complete. The two versions are highly correlated (r = 0.78: Bandelow, 1995). It has five subscales, with 2-3 items per subscale. The patient is asked to rate how severe the following have been over the past week: • •

• • •

Panic attacks. Frequency, severity, and duration of attacks. Agoraphobia. Frequency of avoidance, number of avoided situations, and relevance of these situations (i.e., whether the person is avoiding personally important situations). Anticipatory anxiety. Frequency and severity of worry about having a panic attack. Disability. Impairment in each of three domains of functioning: Family, social relationships, and employment. Worry about the health-related implications of panic. This scale contains two items. One measures the frequency that the patient worries about the consequences of panic (e.g., having a heart attack or collapsing and being injured). The other item assesses the frequency of worry that the panic attacks are caused by an undiagnosed medical condition.

230

ASSESSMENT

The subscales of the Panic and Agoraphobia Scale show good performance on various indices of reliability and validity (Bandelow et al., 1995). They also are sensitive in detecting treatment-related effects (Bandelow et al., 1998). Preliminary norms for panic patients are available, and the scale is available in several different languages (Bandelow, 1995). The Panic and Agoraphobia Scale does not distinguish between full and limited symptom panic attacks, or among the types of panics (unexpected, situational, and situationally bound attacks). When asked to recall their attacks, patients may have difficulty making these distinctions. Prospective (ongoing) monitoring is needed to provide this information (Rapee et al., 1990). The agoraphobia subscale includes a list of situations adapted from the Mobility Inventory. Patients are asked to rate how frequently they avoid these situations when alone. Avoidance when alone tends to be greater than avoidance while accompanied by a trusted companion (Chambless et al., 1985), and so avoidance-alone provides an index of the severity of the patient's avoidance. Unfortunately, however, the agoraphobia subscale does not assess the frequency of avoidance when the patient refrains from other safety-seeking behaviours. In summary, the Panic and Agoraphobia Scale provides a brief survey of a number of important facets of panic disorder. The scale does not assess these areas in depth, but nevertheless is useful for monitoring progress throughout the course of treatment. A copy of the scale appears in the appendix of Bandelow (1995).

Panic-associated

Symptom Scale

The Panic-associated Symptom Scale (PASS: Argyle et al., 1991) combines prospective monitoring of symptoms with an interview assessment. The PASS contains a preliminary interview in which patients are given a definition of panic attacks and anticipatory anxiety. Patients are then asked to rate various panic-related symptoms over a period of one week. This is done by asking them to record, in a panic attack diary, the details of each panic attack shortly after it occurs. The PASS is completed when the interviewer rates the severity of the symptoms, based on the information provided by the diary. Scores on the following variables are obtained: • • • •

Number Average Number Average

of situational panics. intensity of situational panics. of unexpected panics. intensity of unexpected panics.

ASSESSMENT BATTERIES: INTERVIEWS AND SELF-REPORT INVENTORIES 231

• • • • •

Number of limited symptom panic attacks. Average intensity of limited symptom panic attacks. Average % of waking hours worried about panic. Average intensity of anticipatory anxiety. Overall phobia score (global severity of fear and avoidance).

Preliminary data suggest adequate performance on indices of reliability and validity, and the scale is sensitive to treatment effects (Argyle et al., 1991). Some normative data is available in the original article, but nothing else has been published. A problem with the PASS is the unusual rating scheme used for scoring the frequency of full and limited symptom panic attacks. The number of attacks in the past week is coded on the following four-point scale: 0 (no attacks), 1 (one attack), 2 (2 to 7 attacks) and 3 (more than 8 attacks). Bandelow et al. (1995) observed that the PASS is likely to be insensitive for detecting change in the range of 2-7 attacks per week. It would be straightforward to correct this problem by simply scoring the items according to the number of attacks per week. A greater problem, however, is that copies of the PASS interview and panic diary are difficult to obtain. The PASS was not published in the original article and is not described in sufficient detail to reconstruct it.

Comment Which of these are most useful for evaluating treatment outcome? Five criteria are important for choosing a battery: (1) breadth of assessment (i.e., the number of dimensions of panic disorder assessed by the instrument), (2) the amount of time required to complete the battery, (3) whether the battery is sensitive to week-by-week changes in symptoms, so that the therapist can closely monitor symptoms over the course of therapy, (4) whether the battery is psychometrically sound, and (5) whether copies of it can be readily obtained. On these criteria the Panic and Agoraphobia Scale (P&A) is one of the most useful batteries; it is as broad or broader than most other batteries, is quick to administer, is sensitive to weekly changes in symptoms, has good reliability and validity, and copies of the battery can be readily obtained. Like all batteries, the P&A is not sufficient. A clinical interview is needed to fully assess the various aspects of panic disorder, such as the safety signals and safety behaviours used by the patient. Careful questioning is also useful for identifying the causes of feared sensations. One can ask the patient to demonstrate how they were doing things just before they panicked (Barlow & Cerny, 1988); e.g., 'Show me how you were breathing when you started to panic.' An interview can also be used to detect covert

232

ASSESSMENT

avoidance, such as the tendency to distract oneself when in feared situations. One patient, for whom air travel was highly fear-evoking, managed to occasionally take flights, but while on the plane she would close her eyes and attempt to imagine herself safely in her living room armchair. A clinical interview can identify the complex motivations that shape avoidance behaviour, particularly in comorbid cases. For patients with agoraphobia and major depression, avoidance may be motivated by a combination of fear of panic, lack of energy, and hopelessness about ever becoming comfortable in feared situations.

ASSESSING GENERAL ANXIETY General anxiety refers to the person's typical level of anxiety over a given period of time (e.g., the past week). Anxiety may arise from panic-related thoughts (e.g., worry about having a panic attack) and from other thoughts of threat (e.g., worries about other issues). Therefore, measures of general anxiety reflect the anxiety associated with panic disorder as well as anxiety associated with other, comorbid conditions. General anxiety is important to assess not only because it is commonly elevated in panic patients, but also because it is composed of body sensations that patients catastrophically misinterpret. All the body sensations that characterize panic attacks can occur in general anxiety. This is apparent when one compares the symptoms of generalized anxiety disorder with the symptoms of panic attacks, as described in DSM-III-Rand DSMIV (American Psychiatric Association, 1987, 1994). The phenomenological difference between general anxiety and panic is largely a difference of form rather than substance; in panic attacks the sensations erupt suddenly with great intensity, whereas in anxiety the sensations are more gradual in their rise and fall, and are usually not as intense. There are a multitude of self-report measures of general anxiety, the most widely used being the State-Trait Anxiety Inventory (Spielberger, 1983) and the Beck Anxiety Inventory (BAI: Beck & Steer, 1993). Both perform well on indices of reliability and validity, and both are sensitive to treatmentrelated change (Beck & Steer, 1993; Jorm, 1989; Spielberger, 1983). Detailed norms are available for both scales (Beck & Steer, 1993; Gillis et al., 1995; Spielberger, 1983). The BAI is a purer measure of anxiety than the StateTrait Anxiety Inventory, because the latter contains items assessing depression or dysphoria. The BAI is also more useful for assessing progress during treatment because it assesses anxiety over the past week. In comparison, the State-Trait Anxiety Inventory assesses anxiety 'right now' (state anxiety scale) and anxiety proneness (trait anxiety scale).

PROSPECTIVEMONITORING

233

The most commonly used interviewer-rated measure of general anxiety is the 14 item Hamilton Anxiety Rating Scale (HARS: Guy, 1976; Hamilton, 1969). The HARS has generally good psychometric properties and is sensitive to treatment-related change (Bruss et al., 1994; Maier et al., 1988; Moras et al., 1992). Recently, a structured interview guide was developed (Bruss et al., 1994), which should further improve the reliability and ease of administration of the HARS. This scale can be used to assess the severity of general anxiety over the past week or past month (Moras et al., 1992; Riskind et al., 1987). Revisions of the HARS have been proposed to further improve its validity (Riskind et al., 1987; Snaith et al., 1982). However, the original 14 item version remains the most widely used (Maier et al., 1988). Both the HARS and BAI can be used to assess the severity of general anxiety before and during treatment. An advantage of the BAI over the HARS is that it is quicker to administer. A disadvantage of both measures is that neither distinguish between anxiety and panic attacks (Bandelow et al., 1995; Cox et al., 1996; Steer & Beck, 1996). Scores on these scales represent the average severity of general anxiety symptoms over the past week, regardless of whether the symptoms occurred during episodes of anxiety or panic. More specific measures could be constructed. For example, two versions of the BAI could be created. In one version the patient would be asked to rate the severity of anxiety symptoms as they occurred during panic attacks. In the other version the severity of symptoms would be rated as they occurred when the patient is not panicking. A problem with this approach is that it is not known whether patients can reliably and validly make this distinction using retrospective measures like the BAI or HARS. Clinically, a global measure is sufficient for assessing the average severity of anxious arousal, regardless of the pattern of symptom onset (gradual vs. paroxysmal). To assess the symptoms occurring during panic attacks, one could use any of a number of measures, such as the Panic and Agoraphobia Scale and prospective monitoring.

PROSPECTIVE MONITORING Panic Attacks Panic Attack Records To gain a more detailed assessment of panic attacks, clinicians and clinical researchers are increasingly including some form of prospective monitoring in their assessment packages (Shear & Maser, 1994). The most widely used are the panic attack records. The patient is provided with a definition of a panic attack and then given a pad of panic attack records

234

ASSESSMENT

that can be readily carried in a purse or pocket. The patient is instructed to carry the records at all times and to complete one sheet for each fullblown or limited symptom attack, soon after the attack occurs. One could ask patients to attempt to complete the record during the attack, although the act of monitoring may disrupt the attack by serving as a distraction. It is better to ask patients to complete a panic attack record as soon as possible after an attack. The records are then reviewed during the next therapy session to learn about the causes and consequences of the panics, and to assess treatment progress. Many different types of panic attack records have been developed (Barlow & Cerny, 1988; Barlow & Craske, 1994; Clark, 1989; Clum, 1990; Franklin, 1996; Hecker & Thorpe, 1992; Rapee et al., 1990; Sheehan, 1983; Taylor & Arnow 1988). These records can be modified as needed for particular patients. Computerized versions are also available, using hand-held 'palm-top' computers (Taylor et al., 1990). These are not as widely used as the paper versions of the diaries, although this may change as palm-top computers become less expensive. Variants on the paper-and-pencil panic diaries developed by Barlow and colleagues (e.g., Barlow & Craske, 1994; Rapee et al., 1990) are among the most informative and easy to use. A version slightly adapted from Rapee et al. (1990) is shown in Figure 9.1. This form has been modified simply by adding a section in which patients can list thoughts and images occurring during the panic. This simple record provides a good deal of clinically useful information about each panic attack and the context in which it occurs. When completing the records, patients often fail to describe their catastrophic thoughts in much detail. Sometimes they simply list thoughts like 'I'm going to suffocate' or 'I have to get out of here.' Sometimes they list body sensations instead of thoughts; e.g., 'My heart is pounding.' Despite these problems, the records are still useful. Even a meagre recording provides a useful reminder or mnemonic cue to help the patient recall, during treatment sessions, what sensations triggered their attacks, and what cognitions occurred before and during the attacks. Mr J. brought in the panic attack record shown in Figure 9.2. Although the record seems to convey little information, it helped him recall the sequence of events leading up to his panic attack. The record also furnished enough information for the therapist to identify links between sensations and catastrophic thoughts, and to identify escape behaviours. Just before the attack, Mr J. was involved in a heated argument with a friend. He recalled feeling angry and was speaking loudly and rapidly in an effort

PROSPECTIVE MONITORING 235

PANIC ATTACK RECORD NAME: _________ DATE: __

_

WITH: SPOUSE __

TIME: __ _

_

FRIEND __

DURATION: ___ _

STRESSFUL SITUATION: YES /NO

STRANGER __

_

(min.) _

ALONE __

_

EXPECTED: YES / NO

MAXIMUM ANXIETY (CIRCLE) 0 -----NONE

I ---

2 -------- 3 ----4 --5 ---MODERATE

6 -----

7 ----

8 EXTREME

SENSATIONS (CHECK): POUNDING HEART

SWEATING

HOT/COLD FLASH

TIGHT/PAINFUL CHEST

CHOKING

FEAR OF DYING

BREATHLESS

NAUSEA

FEAR OF GOING CRAZY

DIZZY

UNREALITY

FEAR OF LOSING CONTROL

TREMBLING

NUMB/TINGLE

THOUGHTS OR MENTAL IMAGES AT THE TIME (DESCRIBE):

Figure 9.1. A form for prospectively assessing panic attacks From Journal of Anxiety Disorders, 4, Rapee, R. M., Craske, M. G., & Barlow, D. H., 171-181, Copyright (1990), with permission from Elsevier Science

to get his· point across. Recall from Chapter 4 that people sometimes hyperventilate while speaking, particularly when discussing distressing topics. This seemed be the case with Mr J. He felt hot, sweaty, and tense during the argument. As the argument grew increasingly intense, he felt himself becoming increasingly dizzy and lightheaded. He interpreted these sensations as an indication that he was about to 'lose control'. Mr J. feared he would collapse and start crying uncontrollably. On further questioning, Mr J. admitted that the worst thing about collapsing would be that his friend would regard him as 'pathetic wimp'. As his panic escalated, his mind was flooded with thoughts and images of physical collapse and its consequences. Desperate for an excuse to escape the situation, he shouted at his friend, 'I don't have to put up with this!' and fled to his car. He drove a few blocks and then pulled over and took a lorazepam tablet. Mr J. was convinced that if he had not fled the situation

236

ASSESSMENT

PANIC ATTACK RECORD NAME:

24_

DATE _fkute

TIME

WfTT-1:SPOUSE ___

_//t,fft,_

FRIEND

§

STRESSFUL SITIJATION:

,l____

NO

_s, 9,____

_

DURATION: _/O_(min.) STRANGER ___ EXPECTED: YES

e

ALONE __

_

MAXIMUM ANXIETY (CIRCLE) NO~--

I -----

2 -----

3 - MOD~~TE

5 -----

6

---0-ix-rkME

SENSATIONS (CHECK): POUNDING HEART TIGHT/PAINFUL CHEST

L L

SWEATING

HOT/COLD FLASH

L

CHOKING

FEAR OF DYING

✓

✓

BREATHLESS

✓

NAUSEA

FEAR OF GOING CRAZY

DIZZY

L

UNREALITY

FEAR OF LOSING CONTROL

TREMBLING

L

NUMBfTINGLE

THOUGHTS OR MENTAL IMAGES AT THE TIME (DESCRIBE):

_____ Figure 9.2.

1 'm ~ta~~------Sample of a completed panic attack record

and taken a tablet, his dizziness would have escalated to the point that he would have 'collapsed in a pitiful heap'. Sampling the Heterogeneity of Panic

It is fruitful to use panic attack records to assess limited symptom attacks. This can provide important clues about how patients attempt to dampen their panic attacks. Assessing these episodes often provides useful information about the patient's adaptive and maladaptive coping strategies, including safety behaviours intended to avert the feared consequences of body sensations. The panic records also provide information on the symptoms and cognitions associated with various types of panic attacks; expected, unexpected, and situationally predisposed. This can help the clinician identify the various causes of these attacks. Panic attack records are a good way to assess within-patient variability in panic symptom profiles (Rapee et al., 1990). One day a person might experience an attack characterized by intense chest pain and palpitations.

PROSPECTIVEMONITORING

237

The next day they might have an attack associated with intense depersonalization and derealization. By sampling panics the clinician can identify the various links between body sensations and catastrophic beliefs. Nonadherence with Panic Monitoring The panic attack record in Figure 9.1 could be modified to assess more information, such as the level of arousal before, during, and after panic attacks, along with information on how the patient coped with the attacks (e.g., whether or not avoidance, coping statements, or breathing exercises were used). But any increase in complexity must be balanced with the increase in time and effort required to complete the record. Adherence to the requirements of prospective monitoring tends to decrease as the recording process becomes more complicated, especially for patients who have many panics each day. The simple form shown in Figure 9.1 provides a good balance between ease of completion versus information yield. Nonadherence in completing the panic attack records can occur when the patient believes that monitoring will increase the likelihood of anxiety or panic. This belief can be tested by asking the patient to monitor their panics on a trial basis. Typically, the belief is disconfirmed. As noted by several writers (e.g., Barlow et al., 1984; Beck & Zebb, 1994), adherence can be improved if the therapist does the following: •

• • • •

Provide the patient with a convincing rationale for using the panic attack records, emphasizing that treatment is most likely to be effective when it is based on accurate information. Elicit an agreement from the patient to complete the records. Identify beforehand any obstacles to completing the records, and develop strategies for circumventing these difficulties. Train the patient in the use of the records. Review the records during each session, and ensure that you acknowledge or praise the patient's efforts at monitoring their panic attacks.

Monitoring Other Symptoms In addition to monitoring panic attacks, one could also ask patients to monitor other clinically important variables, such as their degree of worry about panic. The latter can vary markedly on a day-to-day basis (Shear & Maser, 1994), and so it may be useful to assess it frequently (e.g., % of

238

ASSESSMENT

waking hours spent worrying about panic, recorded at the end of each day). On the other hand, detailed prospective monitoring can be onerous for the patient, and may result in nonadherence. An alternative for many clinical variables is to simply ask patients to make a rating of these variables each week, at the beginning of each therapy session. The Panic and Agoraphobia Scale can be used for this purpose.

Monitoring Homework Assignments Prospective monitoring is often the most useful way of assessing adherence with homework assignments, and for assessing the effects of these assignments on catastrophic beliefs. The information recorded depends on the nature of the assignment. The therapist might ask the patient to record each trip to the shopping mall, noting the day and time the trip was made, along with the duration of the journey. The patient also could be asked to record whether a panic attack occurred, and to list their thoughts before, during, and after the journey. This information provides valuable material for cognitive restructuring exercises.

Comment Prospective and retrospective assessment methods both have their pros and cons. Retrospective assessments, compared to prospective monitoring, tend to result in the overestimation of the frequency and severity of panic attacks (de Beurs et al., 1992; Margraf et al., 1987, 1988; Rapee et al., 1990). This could be due to any of a variety of reasons. It could be that retrospective methods are distorted by memory biases and errors. On the other hand, prospective monitoring methods may be more reactive than retrospective methods (Barlow et al., 1984; Beck & Zebb, 1994; Taylor, 1985). The act of monitoring might alter the frequency and severity of panic attacks. Panic frequency would be reduced by prospective monitoring if the latter interrupted the chain of events leading to panic (Kirk, 1989). The act of monitoring may cause patients to become more aware of situations that evoke panic, thereby leading to avoidance of these situations. Panic severity also may be reduced if patients attempt to complete the panic attack records during their attacks. The act of completing the record may allow patients to distance or distract themselves from their attacks. It is not known which of retrospective and prospective methods provide the most accurate methods. Reactivity, if it occurs, may be reduced by

COGNITIVEASSESSMENT 239

asking patients to complete their panic attack records after, rather than during their attacks. Despite possible problems with reactivity, prospective monitoring provides a wealth of information and is an important component of the assessment package.

COGNITIVE ASSESSMENT Cognitive models of panic (Chapter 3) suggest that several cognitive variables are important to assess in panic disorder. Among the most important variables is the strength of the person's arousal beliefs. That is, beliefs about the dangerousness of arousal-related sensations (as indexed by scores on measures of anxiety sensitivity). It is also important to assess the person's degree of body vigilance, which is the tendency to focus attention on one's body. The greater the person's body vigilance, the higher the probability of detecting body sensations. For people who believe that arousal-related sensations are dangerous, the higher the degree of body vigilance, the greater the chances of making catastrophic misinterpretations and thereby panicking. Other cognitive variables are important in agoraphobia (Chapter 3). These include tendencies to pessimistically predict that aversive events will occur in particular situations; e.g., predicting that the odds of suffocating are great if one enters a confined space such as a crowded elevator. It is up to the therapist to identify these beliefs. This can be done when assessing the patient's reasons for fearing and avoiding particular situations. Measures of expectancies can be constructed as needed. To illustrate, one might ask an agoraphobic patient to rate on a 0-8 scale the likelihood of collapsing in a supermarket (0 = extremely unlikely, 8 = extremely likely). Such ratings can be used to assess the patient's progress in treatment. Unfortunately, there have been few studies of the psychometric properties of scales measuring cognitive aspects of agoraphobia. The following sections will focus on scales most'important for assessing panic-related cognitions. These measures are brief, each taking about 5-l0min to complete.

Assessing Anxiety Sensitivity and Arousal Beliefs Recall that anxiety sensitivity is the fear of arousal-related sensations, arising from arousal beliefs. The latter are beliefs that the sensations have harmful consequences. Measures of anxiety sensitivity are often used

240

ASSESSMENT

as indices of the strength of arousal beliefs. There are several self-report measures of these variables, as described in the following sections.

Agoraphobic Cognitions Questionnaire and Body Sensations Questionnaire This pair of scales was developed by Chambless et al. (1984) as measures of 'fear of fear,' a forerunner to the concept of anxiety sensitivity. The Body Sensations Questionnaire asks respondents to rate their fear of each of 17 arousal-related body sensations (e.g., palpitations). The questionnaire is widely used in research and clinical practice, and has generally performed adequately on various tests of reliability and validity (Arrindell, 1993; Chambless, 1988; Chambless & Gracely, 1989). However, a recent study found it was only modestly correlated with the Anxiety Sensitivity Index (r = 0.38 in each of two samples: Asmundson et al., 1996). A limitation of the Body Sensations Questionnaire is that it provides no information on the reasons why the person is frightened of body sensations. This omission is understandable because the Body Sensations Questionnaire was based on a conditioning theory, which stated that fear of fear arose as a conditioned consequence of panic attacks (Goldstein & Chambless, 1978). In later writings, Chambless and Goldstein (1988) emphasized the role of cognitive factors in fear of fear, to the point that their theory is quite similar to anxiety sensitivity theory. The Agoraphobic Cognitions Questionnaire asks respondents to rate how often each of 14 threat-related thoughts occur when the person is feeling anxious. Examples include thoughts pertaining to physical threat (e.g., 'I must have a brain tumor') and those to do with social threat (e.g., 'I'm going to act foolish'). This questionnaire is commonly used as a measure of arousal beliefs. Some studies suggest the Agoraphobic Cognitions Questionnaire has adequate reliability and validity (e.g., Arrindell, 1993; Bouchard et al., 1997; Chambless et al., 1984; Chambless & Gracely, 1989). However, two studies failed to support its criterion-related validity (Craske et al., 1986; Thorpe et al., 1987). A further limitation of the Agoraphobic Cognitions Questionnaire is that its scores are ambiguous. The scale is based on the assumption that high scores indicate a greater tendency to become frightened by anxiety; thoughts like 'I have a brain tumor' are assumed to be triggered by anxiety. However, it may be that the items assess thoughts that cause anxiety (Taylor, 1995). Furthermore, Costello (1992) observed that the Agoraphobic Cognitions Questionnaire fails to clearly distinguish thoughts from sensations. Consider the item 'I am going to pass out.' By reporting that this 'thought'

COGNITIVEASSESSMENT 241

occurs frequently, respondents might really be saying that they frequently feel dizzy (a sensation). On the other hand, they might be indicating that they expect they will become very dizzy (a thought). Similarly, it is unclear whether the item 'I will faint' measures a thought ('I expect I will faint') or a sensation ('I'm feeling very lightheaded'). These problems do not apply to some of the other indices of arousal beliefs, such as the Anxiety Sensitivity Index and the Body Sensations Interpretations Questionnaire, which both assess the links between sensations and cognitions (e.g., 'When I can't concentrate, I worry that I could be going crazy'). The Anxiety Sensitivity Index, unlike the Agoraphobic Cognitions Questionnaire, specifically assesses anxiety that arises as a consequence of body sensations (e.g., 'It scares me when I feel faint'). To summarize, the Agoraphobic Cognitions Questionnaire and Body Sensations Questionnaire both have important limitations. In the 1980s and early 1990s they played a valuable role in advancing our knowledge about anxiety sensitivity and related constructs. These days, better scales are available.

Anxiety Sensitivity Index The Anxiety Sensitivity Index (Reiss et al., 1986) consists of 16 items. High scores indicate a strong tendency to catastrophically misinterpret arousalrelated body sensations. Moderate scores indicate a tendency to believe these sensations have harmful but not necessarily catastrophic consequences. People with low scores believe that arousal-related sensations are harmless. Norms to aid interpretation are available (Peterson & Reiss, 1992; Taylor, 1999). This brief scale is the most widely used method for assessing anxiety sensitivity. It has performed well on numerous tests of reliability and validity, and is sensitive to treatment-related effects (see Taylor, 1999). Posttreatment scores on the scale also predict who is likely to relapse after panic treatment (Otto & Reilly-Harrington, 1999). Despite its popularity, the scale is not without its limitations. Due to its brevity, it may not adequately sample all the important domains of anxiety sensitivity. For instance, it has no items assessing fear of depersonalization. Expanded versions of the Anxiety Sensitivity Index have been published (Taylor & Cox, 1998a,b), but these are still in the early stages of development.

Body Sensations Interpretations Questionnaire This scale specifically measures the tendency to catastrophically misinterpret body sensations (Clark et al., 1997c). In each item the respondent

242

ASSESSMENT

is asked to respond to a scenario ('You feel short of breath. Why?') by writing down the first interpretation that comes to mind. The respondent is then presented with three explanations for the scenario, one of which refers to a catastrophic outcome; e.g., (1) 'You are developing the flu,' (2) 'You are about to suffocate or stop breathing', and (3) 'You are physically "out of shape'". The respondent is asked to rank-order the explanations in terms of the likelihood that they would come to mind if the scenario actually occurred. Other ratings are also obtained. The Body Sensations Interpretations Questionnaire comes in 14 and 27 item versions. Each version contains items assessing interpretations of arousal-related body sensations as well as items containing other (control) scenarios (e.g., 'You have visitors round for a meal and they leave sooner than expected. Why?'). Both versions have adequate reliability and validity, and preliminary norms are available (Clark et al., 1997c, 1999). Both are sensitive to treatment-related change (Clark et al., 1994, 1997c, 1999). Both versions of this questionnaire yield a good deal of clinically useful information. The short version is limited, however, in that it has only seven items assessing interpretations of arousal-related body sensations. The longer version is more comprehensive, assessing interpretations of a range of sensations, including palpitations, lightheadedness, chest tightness, concentration difficulty, and dyspnea.

Other Questionnaire Measures Several other measures have been developed, although none are as widely used or evaluated as the above-mentioned scales. The Panic Belief Questionnaire (Greenberg, 1989) assesses erroneous beliefs about the consequences of panic attacks, and beliefs about the inability to cope with panic. Little is known about the scale's psychometric properties. One study provided mixed support for its criterion-related validity; Brown et al. (1992, unpublished, cited in Bouchard et al., 1997) found that the scale was able to discriminate between 'good' and 'poor' responses to panic treatment, although scores were not significantly correlated with panic frequency. The Agoraphobic Cognitions Scale (Hoffart et al., 1992) is a 10 item measure of feared consequences of anxiety (e.g., dying, fainting, making a scene). The scale has performed well on tests of reliability and validity (Hoffart et al., 1992) and is sensitive to treatment-related effects (Hoffart & Martinsen, 1990). The main shortcoming is that this scale is too brief to provide a detailed assessment of the patient's arousal beliefs.

COGNITIVE ASSESSMENT 243

The latest version of the Catastrophic Cognitions Questionnaire (Khawaja et al., 1994) contains 21 items assessing catastrophic panic-related beliefs. This scale performed poorly on a measure of criterion-related (knowngroups) validity, in that it was unable to distinguish among patients with different anxiety disorders (Khawaja et al., 1994). Using the Clinical Interview to Elicit Arousal-related Beliefs

The assessment of arousal beliefs is not limited to self-report questionnaires. Valuable information can be obtained by other means. Among the most useful interview methods for eliciting catastrophic beliefs is the downward arrow method (Burns, 1981). This uses a series of questions to identify catastrophic beliefs. To use the downward arrow method, the therapist can ask the patient to describe a distressing event, such as a recent episode of panic. Systematic questioning is then used to identify what the patient regards as the worst part of the event, and why they think that is bad. Questions such as the following are asked: • • •

•

'What was most upsetting about __ ?' 'Supposing_ did happen, why would that be bad?' 'If_ was true, what would that mean to you?' 'What could happen if _ did occur?'

The following transcripts illustrate the downward arrow method. The first shows how this method was used to identify the catastrophic misinterpretations during Mrs A.'s panic attacks. The second uses the downward arrow to identify catastrophic beliefs that appeared to contribute to Mr B.'s agoraphobia. Mrs A's Panic Attacks

Therapist (T): You mentioned earlier that the most frightening sensations you get are tingling in your fingertips and dizziness. Patient (P): That's right. They really scare me. T: What runs through your mind when you notice these sensations? P: Nothing really, I just get really scared. [This is a cue for the therapist to use the downward arrow method to assesswhether the patient catastrophically misinterprets the sensations.] T: Imagine for a moment that you're at home alone and you suddenly notice the dizziness and tingling. Would you worry something bad was going to happen?

244

ASSESSMENT

P: I'd worry I was going to panic. T: And if you did panic, what would be bad about that? P: It would feel awful. T: Would that be the worst of it, or would you worry that something even worse could happen? P: I'd worry that all the anxiety was damaging my nerves. The tingling sensations must mean that my nerves are being damaged. T: Anxiety can't damage your nerves. But it sounds like you believe that while you're panicking. Supposing that anxiety could damage your nerves, what would be bad about that? P: It would mean that I would lose control of my mind and my body1'd wind up as a vegetable in a nursing home. T: And if that did happen, what would be the worst thing about that? P: Well, I'd no longer be able to do anything and I'd be suffering for the rest of my life. T: So, in summary, it sounds like when you feel dizzy and tingling you interpret these sensations as signs that your nerves are being seriously damaged by anxiety, and you'll wind up incapacitated and suffering. P: Yes, that sums it up.

Mr B's Fear and Avoidance

T: You mentioned that driving over bridges is one of your biggest problems. When was the last time you tried to drive across a bridge? P: I tried last week, but it was too difficult. T: How far did you get? P: I didn't even get my car out of the driveway. I was too anxious. T: When you were sitting in your car, what was running through your mind? P: I was scared I might panic while driving on the bridge. T: And if that happened, what would be the worst thing about that? P: I might have to stop the car and leave it on the bridge. T: Supposing that did happen, what would be bad about that? P: Other drivers would be angry at me. T: And supposing that happened, what would be so bad about that? P: Well, when I think about it, I guess it wouldn't be so bad really. I'd live to see another day. T: So to summarize what we've been talking about here, it sounds like there are two beliefs. One is that if you panicked you'd have to stop

COGNITIVE ASSESSMENT 245

the car. The other is that if you stopped the car you'd make other drivers angry. It also seems that when you think about it, you can see that it wouldn't be a catastrophe to make other drivers angry. P: Yes, that's right. [The discussion continued to review the evidence for and against these beliefs.]

Comment

During the treatment sessions, the downward arrow method is a useful strategy for eliciting detailed information about catastrophic misinterpretations. The information obtained from the interview can be supplemented by information obtained from standardized self-report measures, for which norms are available. The most useful self-report measures are the Anxiety Sensitivity Index and the Body Sensations Interpretations Questionnaire. Other scales are either too brief or have weak validity. It remains to be seen whether the versions of the Body Sensations Interpretations Questionnaire are more useful than the Anxiety Sensitivity Index for treatment planning and evaluation. Each would make a valuable addition to an assessment battery, although it would be redundant to use both. The Anxiety Sensitivity Index has an advantage of being more widely used and therefore having more information on norms. It is also quicker to administer and score. On the other hand, the Body Sensations Interpretations Questionnaire has the advantage of providing a direct measure of arousal beliefs; the Anxiety Sensitivity Index assesses fear of arousal sensations, and is therefore an indirect measure of arousal beliefs.

Assessing Other Panic-related Cognitive Constructs Panic Appraisal Inventory

This questionnaire contains three 15 item scales (Telch, 1987): •

•

The Anticipated Panic Scale asks the respondent to estimate the likelihood of panicking in various situations (e.g., while riding on a bus or train). The Panic Consequences Scale lists 15 thoughts that may occur during panic (e.g., 'I may do something uncontrolled like jump out a window'). The respondent is asked to rate 'how troubled or concerned' they would be about each thought during a panic attack.

246

•

ASSESSMENT

The Panic Coping Scale assesses the person's perceived efficacy in controlling or enduring panic attacks under various circumstances. The scale does not assess the efficacy of particular coping strategies. Rather, it assesses the person's general belief about their ability to cope with panic.

The three scales have performed adequately on tests of reliability and validity, and are sensitive to treatment effects (Feske & de Beurs, 1997; Telch et al., 1989). Unfortunately, however, little is available in terms of norms. The scales have several other shortcomings. A limitation of the Anticipated Panic Scale is that it assesses the probability of panic rather than the probability of catastrophic outcomes that would lead the person to panic. For the purposes of CBT2, it is more important to assess expectations about catastrophes than to assess expectations of panic. A patient may be accurate in predicting the probability of panicking in a given situation. It is more important to know why the patient expects to panic; what catastrophic event does she or he expect to happen that would give rise to panic? A limitation of the Panic Coping Scale is that it fails to distinguish between adaptive and maladaptive coping, and fails to tell us why the patient engages in particular coping strategies. Some patients may actively engage in coping strategies because they believe arousal-related sensations are dangerous. This is seen, for example, in patients who use breathing exercises to escape dreaded feelings of dyspnea or dizziness. Maladaptive strategies (e.g., escape, avoidance, or other safety behaviours) provide short-term relief from panic and anxiety, but prevent the person from learning that arousal-related sensations are harmless.

Anxiety Control Questionnaire Panic patients commonly report that they are troubled by their lack of control over their anxiety and panic symptoms. Accordingly, Rapee et al. (1996) developed the Anxiety Control Questionnaire to assess the respondent's perceived control over anxiety-related events. The scale has performed well on various tests of reliability and validity (Rapee et al., 1996). However, it remains to be determined whether it makes a useful addition to the clinician's assessment battery. Although it can be useful to assess the patient's perceived control, it is more important to know why the person strives for control in the first place. Panic patients attempt to control (reduce) anxiety-related events because they believe anxiety has harmful consequences. Unfortunately, the Anxiety Control Questionnaire only tells us about the patient's perceived efficacy in controlling anxiety-

EXPOSUREPROBES 247

related events. It does not tell us why the person seeks control or whether the control strategies are adaptive or maladaptive. Coping Strategies Questionnaire

To assess panic-related coping strategies, the patient's reasons for using these strategies, and the effects of the strategies on arousal beliefs, one . could use Clum's (1990) Coping Strategies Questionnaire supplemented by a clinical interview. The questionnaire lists 32 ways that people cope with panic. The respondent rates the frequency of use and efficacy of each strategy. The questionnaire assesses passive distraction strategies (e.g., 'I sit or lie down'), cognitive strategies (e.g., 'Tell myself it will pass or it is nothing to worry about'), physical distraction strategies (e.g., 'I splash water on my face'), and other strategies. Although there appear to be no published psychometric data on this scale, it is clinically useful for discovering how the patient copes with panic. The clinician can follow up responses to this scale to assess whether the strategies are being used as safety behaviours that prevent arousal beliefs from being disconfirmed. Body Vigilance Scale

Until recently, little was available for assessing a person's tendency to be hypervigilant for arousal-related sensations. A promising development is the Body Vigilance Scale, which is a short measure with encouraging data on its reliability and validity (Schmidt et al., 1997b; Schmidt & Trakowski, in press). Schmidt and et al. (1997b) have reported preliminary norms for panic patients and normal controls. Comment

To summarize, the Body Vigilance Scale and the Coping Strategies Questionnaire make unique and important contributions to the panic battery. The Panic Appraisal Inventory and the Anxiety Control Questionnaire are less useful, for the reasons described above.

EXPOSURE PROBES The assessment methods reviewed so far can be usefully supplemented by exposure probes. The latter are methods enabling the therapist to collect data on beliefs, behaviours, and emotions by exposing the patient to feared stimuli. Such stimuli include environmental events or situations

248

ASSESSMENT

(e.g., supermarket queues) and internal stimuli (e.g., arousal-related body sensations). These exposures can be used for therapeutic purposes or as assessment probes. When these exercises are used for assessment purposes, the patient is provided with little information about what to expect. The patient is simply told that the tasks may elicit anxiety and body sensations, and can help the therapist better understand the nature of the patient's problems. When the exercises are used for therapy they incorporate cognitive restructuring methods; e.g., the patient is asked to make catastrophic and noncatastrophic predictions about what will happen. Assessment probes are often integrated with therapeutic exposure exercises, which are described in Chapter 13. The following sections will review two types of exposure probes: behavioural avoidance tests (BATs) and interoceptive exposure tests.

Behavioural Avoidance Tests BATs were originally developed to assess circumscribed fears and phobias (Lang & Lazovik, 1963). In these tasks the person is asked to approach a feared stimulus until he or she feels too frightened to advance any closer. The distance from the stimulus is used as a measure of avoidance, and self-reported levels of distress at particular distances are used to measure fear. A subjective units of distress (SUD) scale is commonly used to measure fear. Here, the person provides a rating of fear or distress on a 0-100 scale, ranging from 0 = no fear/anxiety, to 100 = extreme fear/anxiety. Three types of BAT have been developed to assess agoraphobia, differing in terms of the number and type of stimuli to which the patient is exposed (Agras et al., 1968; de Beurs et al., 1991; Mathews et al., 1981). BATs can be readily incorporated into therapeutic exposure exercises, thereby cutting down on assessment time. The main types of BATs are summarized as follows. •

•

Standardized single-task BAT. Here, each patient is asked to complete the same task, such as walking from the clinic toward a busy city center until they feel too anxious to go on. The advantage of this taskat least for research purposes-is that performance can be directly compared across patients. The disadvantage is that the standardized task may be fear-evoking for some patients but not for others. If the task is irrelevant to the patient's agoraphobia, then it will be of little value .in treatment planning and evaluation. Idiosyncratic single-task BAT. The exposure task in this BAT is tailored to the patient's fears. One might select a task or activity for which the

EXPOSUREPROBES 249

•

patient reports the greatest fear and avoidance, and then break the task into a number of steps; e.g., fear of heights can be assessed in terms of the number of floors the patient can ascend. Progress in treatment can be assessed in terms of the number of steps completed. A disadvantage of this form of BAT is that only a single task is assessed. Although the patient may make good progress on this task during treatment, they may be unable to complete other tasks. Multi-task BAT. Given the multifaceted nature of agoraphobia, a measure of multiple tasks would provide a better indication of the patient's fear and avoidance. A multi-task BAT typically consists of 3-5 tasks. These can be either standardized or tailored to the specific fears of the patient. For clinical purposes the latter are more useful.

Studies of fears, phobias, and agoraphobia indicate that BAT measures of fear and avoidance have acceptable test-retest reliability and good convergent validity with other measures of fear and avoidance (Nietzel et al., 1988). However, their validity may be reduced when performance demands are high. If the clinician strongly encourages the patient to approach the feared stimulus, then this may not provide an accurate measure of naturally occurring avoidance. Low demand BATs are likely to be better measures of avoidance (Kern, 1983). BATs are useful for sampling the patient's cognitions. This can be done in a variety of ways. Before attempting the BAT, patients could be asked to write down their expectations about what they think will happen during the task. Afterwards they can be asked whether the predictions came true. This can be used to assess and challenge any catastrophic predictions they might hold. The time required to complete the BAT is also useful information. Some patients attempt to hurry through the task in an attempt to limit their exposure to 'danger' (Salkovskis et al., 1997). Mr E. feared that stressinduced nausea would lead him to vomit and thereby humiliate himself. When asked to walk through a shopping mall, Mr E. hurried as quickly as he could, believing that the more he lingered the greater the chance of vomiting. Therapist-accompanied BATs can be used to identify subtle safety behaviours, which the patient may not be aware of doing. During a walk to a busy city center, Ms J. stayed very close to the buildings rather than walking in the middle of the pavement. Later discussion revealed that she did this in order to have the 'safety' of a wall to lean on, in case she felt

250

ASSESSMENT

overwhelmingly dizzy. Ms J. believed that 'dizziness can lead to physical collapse' and that 'leaning against a wall can help me get my bearings and stop the dizziness from becoming too intense'. Identifying these safety behaviours proved important, because in order to challenge her beliefs, it was necessary for her to drop these behaviours. In doing so she learned that when she did feel dizzy, this sensation was transient, with no ill effects. When conducting therapist-accompanied BATs it is important to bear in mind that the therapist may become a source of safety to the patient, thereby making the BAT easier to complete than if the patient were performing the task alone. If this happens, then the patient could attempt the BAT alone (as a homework assignment), or the therapist or other trained observer could follow at a distance behind the patient.

Interoceptive Exposure Tests Interoceptive exposure tests are a series of exercises that induce arousalrelated sensations. These can be performed in the therapist's office for either assessment purposes or to therapeutically challenge catastrophic beliefs. When used for assessment, they serve as exposure probes. If the patient catastrophically misinterprets the sensations, then a full or limited symptom panic attack would ensue, or the patient would prematurely terminate the exercise. Thus the probes can be used to assess the patient's interpretations of body sensations. There are many different interoceptive exposure tests. Those included in a battery assembled by Antony et al. (1998) are as follows. • • • • • • • • • • • • •

Shake head rapidly from side to side (30s). Place head between knees for 30s and then lift head quickly up to a normal (upright) position. Spin around with arms out while standing up (lmin). Hold breath (30s). Hyperventilate (1 min). Breathe through a narrow straw without breathing through nose (2min). Stare continuously at a ceiling fluorescent light (1 min). Stare continuously at reflection in mirror (2min). Stare continuously at spot on wall (3min). Tense all muscles in body while sitting in a chair (1 min). Jog ori. the spot (1 min). Face a heater so that hot air blows on one's face (5min). Place a tongue depressor at the back of one's throat (30s).

EXPOSUREPROBES 251 Table 9.2

Interoceptive exposure exercises for eliciting arousal-related sensations

Sensations

Exercises

Numbness or tingling in face or extremities

•

Trembling or shaking Dizziness, lightheadedness, unsteadiness, or faintness

Pounding or racing heart Breathlessness or smothering sensations Chest pain or tightness Choking Sweating Hot flushes or chills Depersonalization

or derealization

Nausea or abdominal distress

• • • • • • • • • • • • • • • • • • • • • • • • • • • • • • •

Place head between knees, then move head to upnght position Hold breath Tense muscles Tense muscles Breathe through straw Hyperventilate Spin Place head between knees, then move head to upright position Hold breath Breathe through straw Jogging on the spot Breathe through straw Hold breath Breathe through straw Hold breath Jogging on the spot Breathe through straw Jogging on the spot Hold breath Tongue depressor Breathe through straw Hold breath Jogging on the spot Face heater Jogging on the spot Face heater Spin Spin Stare at reflection in mirror Stare at spot on wall Tongue depressor Spin

Reproduced by permission of the authors from Antony, M. M., Roth, D. A., & Swinson, R. P. (1998). Symptom induction exercises in panic disorder with agoraphobia and non-anxious individuals. Unpublished data, Department of Psychiatry, McMaster University.

Antony et al. (1998) collected pilot data from 10 panic patients on the sensations typically evoked by these exercises. The exercises best suited for inducing arousal-related sensations are shown in Table 9.2. The table shows that there are exercises for inducing all kinds of arousal-related sensations. The following exercises most often induced panic attacks: breath holding (70% of patients panicked), breathing through a straw (60%), and

252

ASSESSMENT

hyperventilation (50%). The other exercises induced panics in 20-30% of patients (Antony et al., 1998). It is important to instruct patients to focus on the sensations produced by the exercise and not to distract themselves. The patient's reactions are influenced by the context in which the exercise is conducted. For instance, a period of voluntary hyperventilation tends to be more fear evoking when the patient is standing up (vs. sitting down), and when the task is performed alone (vs. performing the task while the therapist is present). Understandably, patients are often reluctant to perform these exercises in an 'unsafe' context. This need not be a problem because even moderately fear-evoking exercises can provide useful information on the patient's beliefs about body sensations. After each exercise is completed, the therapist can ask the patient to rate the intensity of the sensations experienced, to list thoughts or images arising during the exercise, and to indicate whether or not a panic attack occurred. The patient also can be asked how similar the sensations were to naturally occurring panic symptoms. The therapist also can observe whether the patient is attempting to avoid completing the exercise. That is, terminating the task before the allotted time, or trying to avoid the sensations evoked by the task (e.g., by taking shallow breaths during the hyperventilation exercise). The small amount of available information indicates that interoceptive exposure tests show generally good convergent validity; anxiety or panic evoked by these exercises is significantly (but by no means perfectly) correlated with the paper-and-pencil indices of the strength of arousal beliefs (Stein & Rapee, 1999). Responses to the tasks also change as a result of treatment, with panic attacks and catastrophic thoughts becoming less likely after a full or partial course of CBT2 (Clark, 1993; Schmidt et al., 1997c). The usefulness of the probes is limited by the lack of norms. Ongoing work, such as that by Antony et al. (1998), should soon rectify this problem. Another limitation is that the probes may not be safe to use on patients with serious medical problems, such as seizure disorders, cardiac conditions, and respiratory diseases (see Chapter 13). For these people, other methods of assessment must be used. Interoceptive exposure tests are useful for following up on unusually low scores on paper-and-pencil indices of arousal beliefs. Mr R. obtained a score of 9 on the Anxiety Sensitivity Index, which is a score lying in the lower end of the norm range. Yet his description of panic suggested that his attacks typically occurred because he misinterpreted dizziness as a

PSYCHOPHYSIOLOGICALMONITORING

253

sign he was about to go crazy. When I raised this possibility with him, he seemed convinced that thoughts had nothing to do with his attacks. To gain further information, I asked him to hyperventilate for 1 min. He began the exercise but stopped after 30s because he was starting to panic. Mr R. observed that as he hyperventilated he had difficulty thinking clearly and became increasingly frightened he was losing control of his mind. This case shows that paper-and-pencil measures are not always accurate. Fortunately, this is not often the case. Nevertheless, interocep• tive probes provide important additional information. Interoceptive exposure probes not only serve as assessment tools, but also help educate patients about the panic attacks. Mr R.'s experience with the hyperventilation probe led him to conclude that 'Maybe my thoughts do have an effect on my panic attacks.'

Comment Exposure probes make a useful contribution to the clinical assessment package. However, compared to other methods, there is less published information on their reliability, validity, norms, and sensitivity to treatment effects. This is partly because exposure probes-particularly the BATs-are often idiosyncratic; adapted to assess the specific fears of specific patients. Sometimes the clinician has to be highly creative to come up with appropriate BATs. This is well worth the effort, because they can be readily developed into therapeutic exercises to challenge catastrophic beliefs (Chapter 13). Interoceptive exposure exercises seem more amenable to standardization, because the list of feared stimuli consists of only about a dozen arousal-related sensations (see Table 9.2). A small set of exercises can be developed to induce these sensations, as the table shows.

PSYCHOPHYSIOLOGICAL MONITORING Psychophysiological recording can take the form of either monitoring in the office (e.g., while the patient completes an interoceptive exposure exercise) or ambulatory monitoring, in which the patient wears a portable physiological recording device while completing an exposure probe. The variables that can be monitored include heart rate, respiratory rate, and skin conductance. These are commonly used as indices of physiological arousal, although they are also influenced by other factors.

254

ASSESSMENT

Psychophysiological monitoring is not commonly used in clinical settings, mainly because it often does not tell us much beyond the fact that the patient is anxious or panicky. Such information can be usually obtained via self-report methods. Psychophysiological measures often fail to distinguish panic patients from controls, and can be insensitive to detecting treatment-related change, even when the patient has improved on a host of other measures (Arena et al., 1983; Holden & Barlow, 1986). Psychophysiological recording methods may be more useful when they are used to provide patients with feedback about their levels of arousal, as part of a behavioural experiment. For patients who believe that increased heart rate is dangerous, the clinician can ask the patient to voluntarily hyperventilate for 1 min while the therapist monitors the patient's heart rate. This can be done by using a portable heart-rate monitor, which clips onto the patient's ear lobe. These are inexpensive devices available at many sporting goods stores. When patients hyperventilate, their hearts' rate typically increase to over 100 b.p.m. Showing patients the reading on the monitor conveys information that heart rate can dramatically increase with no adverse consequences. This can effectively challenge catastrophic beliefs about tachycardia. Hofmann and Barlow (1996) similarly used ambulatory monitoring of respiration rate to prove to a patient that her panic attacks were triggered by hyperventilation. That is, hyperventilation produced feared sensations, which were catastrophically misinterpreted.

OTHER TARGETS OF ASSESSMENT

Comorbid Clinical Problems Given that panic disorder is often comorbid with other disorders, it is important to conduct a broad assessment of the patient's psychopathology. Structured interviews, such as the SCID-rv, are useful tools for this purpose. There are a number of other interviews and self-report scales that provide dimensional measures of comorbid clinical problems. A detailed review of these instruments would take us well beyond the scope of this chapter. Numerous recent reviews are available for measures of depression, pervasive worry, substance abuse, and other clinical problems that commonly co-occur with panic disorder. See Fischer and Corcoran (1994) for reviews and information on how to obtain a wide variety of clinical scales. Also see recent reviews of specific clinical problems (e.g., Sobell et al., 1994; Margolin et al., 1988; Taylor et al., in press).

OTHERTARGETSOF ASSESSMENT 255

Scales assessing cognitive aspects of comorbid problems are also available, and can be useful in planning treatment. To illustrate, the Automatic Thoughts Questionnaire (Hollon & Kendall, 1980) and the Dysfunctional Attitudes Scale (Weissman, 1980) can be used to assess cognitive aspects of depression. These and other cognitive measures are reviewed elsewhere (e.g., Fischer & Corcoran, 1994; Heimberg, 1994; Obsessive Compulsive Cognitions Working Group, 1997; Segal & Swallow, 1994). If panic disorder can develop as part of a general neurotic syndrome (Chapter 2), then it is important to assess whether the patient has a heightened proneness for experiencing aversive emotions in general (i.e., elevated neuroticism). This could be assessed by the revised NEO-PI (Costa & McCrae, 1992) or the Eysenck Personality Questionnaire (Eysenck & Eysenck, 1975). The value of assessing neuroticism has yet to be examined in depth, although as we saw in Chapter 8, pretreatment neuroticism is a weakpredictor of outcome of panic therapies. This may be because most panic patients tend to have elevated neuroticism, compared to general population norms (Chambless, 1985). Range restriction in scores may prevent neuroticism from being useful in predicting treatment outcome. Neuroticism might be most usefully assessed at the end of panic treatment. If the patient's panic disorder has been reduced but neuroticism remains elevated, then the person may be at risk of relapse or development of another neurotic-spectrum disorder.

Personal and Family History Information on the patient's personal and family history of psychopathology is typically obtained by clinical interview. The clinician should assess what was happening in the patient's life at times when there were any notable changes in symptoms or disorders (e.g., at times of onset, exacerbation, remission, or relapse of a given disorder). One should also assess the patient's history of previous treatments. If the patient has received partial or temporary benefit from previous behavioural or cognitive-behavioural therapy, information on why the response was incomplete can guide planning of the current course of treatment (Anderson et al., 1996). The Timeline Follow-Back Procedure (Sobell & Sobell, 1992) can be used to assess the course of the patient's panic disorder and other disorders. This procedure facilitates recollection of historical information by having the patient draw a timeline marking the major milestones in their life (e.g., graduation from high school, marriage, birth of children, start of a new

256 ASSESSMENT job). The patient then marks key psychopathologic events on the timeline (e.g., the first panic attack). The milestones help patients recall the approximate dates of key psychopathologic events (e.g., whether the first panic attack occurred before or after high school graduation). Milestones in the development of arousal beliefs also can also be marked on the timeline (e.g., 'In June 1984, the ER doctor told me my attacks were just anxiety. After that I started to believe the sensations meant I was going crazy').

Living Circumstances An unstructured interview can be used to assess the patient's living circumstances, including social support and patterns of interpersonal relationships. An important question is whether avoidance behaviours are facilitated by a caregiver, such as a parent or spouse. Obtaining collateral information can be useful in this regard. A caregiver might take over all the daily chores and errands, such as shopping. This may decrease the patient's motivation for working on reducing agoraphobic avoidance. Such information is important for treatment planning, such as deciding whether or not to include the patient's significant other(s) in treatment.

Stressful Life Events Stressful life events are important to assess because they can exacerbate panic disorder and interfere with treatment (Chapter 8). Life events can be assessed when the clinician is recording details of the patient's history and living circumstances. Alternatively, a standardized measure can be used. There are many instruments available for measuring stressful life events, including questionnaires (e.g., Holmes & Rahe, 1967; Sarason et al., 1978) and structured interviews (e.g., Dohrenwend et al., 1978). The interviews are thorough but time consuming, often taking 60-90 min to complete (Roy-Byrne et al., 1986). For the CBT2 practitioner, self-report measures are typically sufficient, and can be followed up, if necessary, by further questioning. The Life Experiences Survey (Sarason et al., 1978) is a comprehensive, psychometrically sound self-report measure. Its 76 items cover a range of positive and negative life events (e.g., marriage, death of a spouse, change in residence, legal problems). For each event, the respondent is asked to rate whether it occurred in the past 0--6 months or past 7-12

OTHERTARGETSOF ASSESSMENT 257

months, and to rate the impact of the event on a scale ranging from -3 (extremely negative) to +3 (extremely positive). Measures of life events assess major stressors in the patient's life, but fail to assess common day-to-day stressors. The latter include problems getting along with coworkers, traffic congestion, and misplacing or losing things. These have been termed hassles (Lazarus & Folkman, 1984) and can be measured by the Daily Hassles Scale (Lazarus & Folkman, 1989). Although this scale can be a useful addition to the clinical battery, it is not essential. This is because panicogenic hassles can be identified from the patient's panic attack records. In other words, one can readily identify hassles by having the patient describe the situations in which the attacks occurred, and whether or not the attacks were preceded by or accompanied by minor stressors. If the Daily Hassles Scale is not an essential measure for CBT2, then it may be asked whether it is useful to administer the Life Experiences Survey to assess major life events. The survey is worth including because patients sometimes fail to recall or fail to report clinically relevant life events. A relationship breakup eight months ago may have led to an increase in panic frequency. The patient may forget to report the breakup, focusing instead on immediate life problems. The breakup would probably be detected by the Life Experiences Survey because the latter specifically asks the patient about such events.

To gain a more complete picture of the patient's life stressors, the clinician should inquire about whether any of the stressors are chronic (e.g., protracted legal difficulties). The clinician also can ask the patient to review the events listed in the Life Experiences Questionnaire to see if there are any foreseeable life events. Does the patient, for example, have an ailing loved one who may pa·ss away during the course of the patient's panic treatment? Identifying these events is the first step toward modifying the CBT2 protocol to include strategies for helping the patient deal with upcoming stressors.

Therapy Variables Several therapy variables predict treatment outcome. These include practical constraints making it difficult to attend treatment, negative attitudes held by the patient toward treatment, treatment side effects, rapid reduction in symptoms, and spouse nonadherence (for couples' treatment) (Chapter 8). These can be assessed by clinical interview either

258

ASSESSMENT

before or during treatment. Patients with extremely negative attitudes toward CBT2 may not be suitable for this form of treatment, unless they can be persuaded to try a couple of weeks of therapy on a trial basis. Treatment motivation is influenced by a variety of factors, including characteristics of patients and their life circumstances, and characteristics of the therapy and the therapist. Motivation may increase when the patient begins to experience the benefits of treatment, or may decrease if the patient perceives that progress is slow. Accordingly, it is useful to periodically assess the patient's attitude toward treatment. This can be done by asking the patient for feedback. When patients fail to make expected treatment gains, it is important to assess the factors responsible. These may include negative beliefs about homework assignments ('It won't work,' 'It's too difficult'), hopelessness ('Trying is the first step toward failure'), secondary gains ('If I get over my agoraphobia, I'll have to return to the job I hate'), and misunderstandings about what homework entailed (Kirk, 1989).

SUMMARY AND CONCLUSIONS For patients presenting for treatment of panic disorder, the clinician needs to complete a comprehensive assessment in order to make an accurate diagnosis, develop a case formulation, and to monitor treatment response. Although assessment should be tailored to the specifics of the patient, there are several measures that can be usefully administered to all patients. A comprehensive package is suggested in Table 9.3. The interviews requite 2-3 h in all, which may be spread out over two sessions. The questionnaires can be completed as homework. The therapist can give the patient feedback about the results, as part of presenting the patient with an explanation of the problems and a plan for treatment. Exposure probes can be employed during treatment as part of situational and interoceptive exposure treatments. Comprehensive assessment takes time, but usually pays off in terms of developing a good treatment plan. If the clinician cannot spare the amount of time suggested, then the SCID-II, version 2.0, could be dropped. However, in doing so the clinician runs the risk of failing to detect important personality disorders that may interfere with treatment. For assessing symptoms throughout the course of treatment, it is useful for patients to monitor their panics using the panic attack records (Figure 9.1). Full and limited symptom attacks can be monitored. The patient can

Table 9.3 A comprehensive assessment package for treatment planning and evaluation

Type of measure

Assessment domain

Before treatment

SCID-IV for Axis I

IV

*

SCID-11,version 2.0 Clinical interview

IV IV

Clinical interview Behavioural Avoidance Tests Interoceptive exposure Tests

IV EXP EXP

Panic and Agoraphobia scale

SR

Panic Attack Record Anxiety Sensitivity Index

SR SR

Body Vigilance Scale Coping Strategies Questionnaire Beck Anxiety Inventory Beck Depression Inventory-II Life Experiences Survey Other self-report measures

SR SR

Axis I disorders, current level of functioning, living circumstances, and treatment history Personality disorders Therapy variables (e.g., practical constraints making it difficult to attend treatment) Treatment side effects Agoraphobic fear and avoidance Anxiety sensitivity and underlying arousal beliefs Panic attacks, worry about panic, and agoraphobic fear and avoidance Panic symptoms and cognitions Anxiety sensitivity and underlying arousal beliefs Body vigilance Strategies for coping with panic (including safety behaviours) General anxiety Depression Stressful life events Other symptoms, disorders, or cognitive variables (as indicated)

Measure

SR SR SR SR

During treatment

After treatment *

*

* * * *

* *

* *

* *

* *

* * * *

* *

IC s (/l

It~ )>

*

z0 n 0

zn r-

C (/l

EXP= exposure probes (administered throughout treatment as part of therapeutic exposure exercises). IV =interview. SR= self-report questionnaire. Assessment of each domain is supplemented, as needed, by information collected by clinical interview.

0

z (/l

N

u, --0

260

ASSESSMENT

complete the Panic and Agoraphobia Scale just before each treatment session. These measures also can be completed during the week or two before treatment in order to obtain a baseline measure of severity. Treatment-related factors (e.g., treatment motivation and homework adherence) should be periodically monitored throughout treatment. At posttreatment, the patient can complete some of the self-report measures in Table 9.3 and, if there is time, the SCID-IV This will provide the clinician with a reasonably good picture of the patient's response to treatment, as well as an indication of any problems requiring further attention. Periodic followup several months after treatment also is useful for assessing long-term outcome (Shear & Maser, 1994). Alternatively, one could ask the patient to return to the clinic should a relapse occur.

DEVELOPING A CASE FORMULATION A case formulation consists of a model of the causes of the patient's problems, and a plan for overcoming them. The formulation attempts to explain the causes of long-standing problems (e.g., chronic panic disorder) as well as the causes of acute problems (e.g., a given panic attack). Formulations seek to explain the four Ps of clinical causation: the predisposing, precipitating, perpetuating, and protective factors in clinical problems. Predisposing factors are diatheses or vulnerability factors, such as maladaptive beliefs laid down early in life. Precipitating factors are those stimuli or circumstances that trigger the problems. For example, stressful life events may trigger panic attacks in a person who believes arousalrelated sensations are dangerous. Perpetuating factors are those that maintain the problems; e.g., agoraphobic avoidance may perpetuate panic attacks because avoidance prevents the person from learning that arousalrelated sensations are harmless. Protective factors either prevent problems from developing, or prevent them from getting worse; e.g., occupational or other pressures may compel a panic disordered person to be repeatedly exposed to fear-evoking situations such as driving to work, thereby preventing agoraphobia from becoming severe. Protective factors may not be present in every case. This chapter has two aims. First, to consider the arguments for and against the use of case formulations to guide treatment. When treating patients with panic disorder, the therapist could simply follow the steps laid out in a treatment manual. Alternatively, the therapist could try to develop an understanding of the causes of the patient's problems, and to use this understanding to guide treatment. As we will see, treatment manuals for second generation cognitive-behaviour therapy (CBT2) (e.g., Clark & Salkovskis, 1987; Craske et al., 1994; Otto et al., 1996a) are important, but insufficient. Case formulations are needed to adapt these protocols to the characteristics of the patients and their problems. The second aim is to describe a method for developing cognitivebehavioural case formulations. Many authors have written on this topic

262

DEVELOPINGA CASE FORMULATION

(e.g., Bruch, 1998; Persons, 1989; Persons & Tompkins, 1997). The approach described in this chapter was adapted from Persons (1989; Persons & Tompkins, 1997), who developed one of the most systematic, practical methods for developing a formulation. In this chapter we adapt Person's method specifically to patients presenting with panic disorder, with or without comorbid problems.

THE UTILITY OF CASE FORMULATIONS Formulations are an Integral Part of Treatment In describing their CBT2 protocol, Craske and Barlow (1993) observed that It ... may seem that this structured, protocol-driven treatment is applied in a very standard fashion across individuals. Nothing could be further from the truth. The clinical art involved in this ... required a careful adaptation of these treatment strategies to the individual case. (p. 42)

Yet Craske et al.'s (1994) CBT2 manual, and those of others, offers few guidelines about how and when to select interventions. They also offer little guidance about how to treat comorbid or other challenging cases of panic disorder. There are only a handful of manuals for treating comorbid disorders (see Chapter 16). These manuals deal with only a fraction of the comorbidities seen in clinical practice. There are no manuals for treating many of the most common comorbid problems, such as panic disorder combined with alcohol-use disorders. Should the therapist start by implementing an alcohol treatment protocol or start with panic protocol? Should both problems be treated at the same time? Sometimes it is necessary to treat the alcohol problem before one can treat the panic disorder. However, as we saw in Chapter 6, alcohol abuse sometimes remits when the therapist begins by treating the panic disorder. The case formulation approach provides a method for developing treatment plans for such cases. Even in non-comorbid cases of panic disorder, formulations are needed to determine the pacing, difficulty level, and timing of interventions. Formulations are also useful for selecting the choice of treatment targets, and for predicting and preparing for obstacles to successful therapy (Bruch & Bond, 1998; Persons, 1989; Turkat, 1985).

Pacing With regard to treatment pacing, consider Craske et al.'s (1994) CBT2 protocol. Here, treatment is divided into 15 therapist-administered 'lessons'

THEUTILITYOF CASE FORMULATIONS 263

or modules, organized in a particular sequence. The patient completes one lesson and then moves on to the next. The lessons are quickly covered by some patients, while progress is slower for others. Several sessions are sometimes required for the patient to benefit from a lesson. Some patients may refuse some interventions (e.g., exposure exercises) or even drop out of treatment if the therapist attempts to move too quickly through the protocol. .A case formulation can help the therapist select the most appropriate pacing. To illustrate, a patient might hold the beliefs that 'Exposure exercises are dangerous' and 'Therapists can't always be trusted.' These beliefs would be taken into consideration when planning formulation-driven treatment. The therapist might devote more time to develop a trusting therapeutic relationship, and proceed more slowly with exposure assignments. The therapist might also spend more time modeling the experiments. For example, the therapist might hyperventilate for 2min to show how the exercise is done. This would convey two important messages to the patient: (1) 'I wouldn't ask you to do something that I wouldn't do myself,' and (2) 'the exercises are harmless.' Difficulty Level

A formulation can help the therapist identify the optimal difficulty level of interventions. Difficulty can be defined in various ways, such as by the amount of fear evoked by a given task. Would an easy or difficult behavioural experiment be more useful in challenging a person's dysfunctional beliefs? One patient might dismiss an easily completed behavioural experiment as 'irrelevant'; a more challenging experiment might have a greater effect on challenging beliefs. For another patient the opposite might be true; an easier experiment might be the best place to start because the patient might prematurely abort a more difficult task. Timing and Targets of Treatment

A formulation provides guidance on which problems should be tackled at which particular time. This can help the therapist determine, for example, which maladaptive beliefs are most important to challenge, and when should they be challenged. Consider a patient with multiple, interrelated catastrophic beliefs. The patient might have a very strong belief that 'tachycardia leads to cardiac arrest,' and a moderately strong belief that 'nausea leads to loss of control.' Should the therapist first challenge the most strongly held, and presumably most debilitating belief? Or would it be better to begin by challenging the weaker and presumably

264

DEVELOPING A CASE FORMULATION

more malleable belief? Does the therapist need to challenge both beliefs or would challenging one lead to a reduction in the strength of the other? A formulation could be developed to address this issue. To construct such a formulation the therapist might review the patient's history of strongly held beliefs. Has the patient previously held a strong, maladaptive belief that they no longer hold now? Mr Z. recalled that some years earlier he had an episode of severe depression. During that time he was unemployed and 'absolutely convinced' he was unhireable. At the time, Mr Z.'s therapist urged him to apply for several jobs. The patient regarded this as a futile exercise, but followed through with it anyway. To Mr Z.'s surprise, he was offered not one but two jobs. He recalled that this experience profoundly changed his belief about his unhireability. This information proved useful in selecting cognitive interventions some years later, when Mr Z. was in treatment for panic disorder. His panic therapist hypothesized that Mr Z.'s strongly held, maladaptive beliefs tended to be 'brittle,' and that they could be readily shattered by disconfirmatory experiences. This led the therapist to treat Mr Z.'s panic disorder by starting with the most strongly held belief ('tachycardia leads to cardiac arrest'). To test the belief, the patient completed several aerobic exercise classes. Tachycardia was induced on several occasions with no ill effects, and the 'cardiac arrest' belief was weakened considerably. The belief that 'nausea leads to loss of control' was also weakened. The patient reasoned that 'if tachycardia is harmless, then nausea is probably harmless too.' Dealing with Treatment Obstacles

CBT2 manuals provide little guidance about what to do when treatment obstacles are encountered. The case formulation can be used to anticipate, understand, and help overcome these obstacles. Sometimes difficulties can be predicted, whereas others emerge unexpectedly, such as a relapse after an initially favourable response. The case formulation approach encourages the therapist to develop hypotheses about the causes of the difficulties, and to revise the treatment plan accordingly.

Criticisms of Formulation-Based Approaches to Treatment G. Terence Wilson (1996b, 1997) raised several concerns about the utility of case formulations and about the treatments derived from them. The following are Wilson's four main arguments against the routine use of formulations, along with my responses.

THE UTILITYOF CASE FORMULATIONS 265

1. Caseformulations can be difficult to devise, even with appropriate training. This is true for some cases. However, for the reasons stated above, there is no escaping the need for case formulations. Work is underway to find ways of making it easier to develop formulations. Persons (1989; Persons & Tompkins, 1997) devised a structured step-by-step approach, which therapists can learn in a few hours' training. Unfortunately, formulations derived from this approach may not be reliable when constructed by individual clinicians. Inter-rater reliability can be improved to acceptable levels by using consensus ratings from multiple (e.g., five) clinicians (Persons & Bertagnolli, 1999; Persons et al., 1995). Thus, formulations may be most reliable when a case-conference format is used, where clinicians are using the same method to develop consensus formulations. With further refinement of protocols for teaching formulations, reliability may be increased even further. 2. Clinical reasoning is flawed; therefore case formulations and formulationbased treatment plans are flawed. Wilson's (1996b) second criticism arose

from findings that the clinician's judgment can be biased or otherwise flawed, particularly under conditions of uncertainty or incomplete information (Kahneman et al., 1982; Meehl, 1960; Nisbett & Ross, 1980). Clinicians sometimes find what they expect to find. This may be a result of attitudes they hold (e.g., stereotypes) or reasoning strategies they use (e.g., selectively looking for evidence that confirms an expectation). As a result, clinicians may perceive illusory correlations between variables, and fail to detect unexpected relationships. Wilson (1996b, 1997) argued that if judgment is biased, then clinicians should simply follow a treatment manual rather than develop formulation-based treatment plans. There are several problems with Wilson's argument. First, as we have seen, there is no escaping the need for clinical judgment, even when the therapist follows a treatment manual. Second, treatment manuals are the product of clinical judgment; clinicians developed manuals as a result of their judgments about what interventions should be used and in what sequence. Although manuals are refined on the basis of empirical research, clinical judgment plays a prominent role in their development. CBT2 manuals are the product of clinical reasoning. Therefore, the clinician's judgment may not be as bad as Wilson fears. Third, if the case formulation is properly used it should be self-correcting; ineffective treatments should prompt the clinician to consider reasons for the lack of treatment response and thereby develop a revised formulation and treatment plan (Persons, 1989). Finally, there are several methods for avoiding or correcting biases in clinical judgment. Turk and Salovey (1986) suggested several guiding prompts or reminders that clinicians can use to 'debias' their reasoning. A sampling of these is shown in Table 10.1.

266

DEVELOPINGA CASE FORMULATION

Table 10.1 reasoning

• • • • • • • •

Reminders that therapists can use to reduce biases in their clinical

My a priori theories and stereotypes might influence what questions I ask, and thereby influence the information I obtain. To overcome this bias I need to look for evidence that might challenge my theories and stereotypes. I need to be aware of the possibility of selectively reinforcing the patient for providing some types of information and withholding others. My behaviour and interpersonal style could influence the patient's responses. My mood may influence what questions I ask and what I remember about the patient. My needs or conflicts may influence how I perceive the patient. I need to be aware that it is easier to recall vivid or otherwise salient information about my patients. Over-reliance on this sort of information could bias my judgment. My questions, interpretations, and predictions could be influenced by the patient's characteristics, such as gender, age, religion, socioeconomic status, and physical appearance. I need to be alert to these biases. I need to consider the base-rate (frequency) of relevant variables.

Adapted from: Turk, D. C., & Salovey, P. (1986). Clinical information processing: Bias inoculation. In R. Ingram (Ed.), Information processingapproachesto clinicalpsychology(pp. 305-323). San Diego, CA: Academic.

3. The empirical literature fails to support the superiority of formulation-based treatment over 'standardized' (manual-driven) treatment. This third argument of Wilson (19966) is based on only a handful of studies examining the potential benefits of treatment tailoring. Fewer still have investigated the benefits of tailoring panic treatments. The panic studies are summarized as follows. Emotional reactions can be decomposed into three 'systems': emotional/physiological, verbal/ cognitive, and behavioural/motoric systems (Lang, 1968). A crude form of formulation-driven treatment is to match interventions to the most active system. Cognitive restructuring would be used for patients with numerous or frequently occurring maladaptive cognitions as their 'most salient' problem. Situational exposure would be used for patients who have excessive avoidance as their main problem, and relaxation training for patients with hyperarousal as their most severe problem. Three studies of panic patients found that treatment outcome tended to be improved by this kind of matching (Mackay & Liddell, 1986; Michelson, 1986; Ost et al., 1984). Some of the results were nonsignificant trends, which may have reached conventional levels of significance if larger samples had been used. Treatmentmatching may have been even more effective if contemporary CBT2

THE UTILITYOF CASE FORMULATIONS 267

methods had been used, with interventions tailored to the patient's specific maladaptive beliefs. For example, interoceptive exposure tasks could be selected that induced sensations that the patient feared the most. In the only other major study of treatment tailoring in panic disorder, Schulte et al. (1992) found that 'standardized' behaviour therapy was more effective than formulation-driven treatment. The sample consisted of patients with various phobias, mostly agoraphobia with panic disorder. Greater treatment efficacy was associated with greater use of situational exposure. Formulation-driven and standardized treatments did not differ in efficacy when matched in terms of the amount of exposure that was used. Several methodological problems seriously weaken the conclusions that can be drawn from this study. Schulte and colleagues did not describe how their therapists devised their formulations. Nor is it clear whether therapists were using the same methods to formulate their cases. It is also unclear whether panic attacks were considered in the formulation, or whether the full range of CBT2 methods were considered for use. Finally, there were problems with treatment fidelity, suggesting that all treatments were formulation-driven to some extent: Standardization on the level of therapeutic methods was only moderately successful. Even in the manual-guided therapies the therapists made some adaptations to the individual case-on the pretext that they were necessary. (Schulte et al., 1992,p. 85) In summary, Schulte et al.'s study tells us little about the merits of formulation-driven versus manual-driven treatment. The treatmentmatching studies reviewed earlier suggest, contrary to Wilson's (1996b) claim, that formulation-based treatments tend to be most effective. It remains to be seen whether formulation-driven CBT2 for panic disorder, using Persons' method (or some variation thereof), is more effective than 'manual-driven' CBT2. A difficulty in testing such a comparison is that all treatments are formulation-driven to some extent.

4. Formulation-based treatment is better suited for treatment-resistant cases, complex disorders, and for clinical problems for which there are no empirically validated manuals. Wilson (1996b, 1997) acknowledged that some degree of tailoring is required even to implement manual-driven treatment. However, he cautioned that clinicians might be too quick to deviate from an empirically supported manual. Wilson argued that manual-driven treatment should be the mainstay of therapy, and that formulation-based treatment should be reserved for treatment-resistant, complex disorders,

268

DEVELOPINGA CASE FORMULATION

and those for which there are no empirically a given problem.

supported

protocols for

I agree that case formulations are especially important in devising treatments for complex (e.g., polycomorbid) and other challenging panic disorders. For 'simple' or 'uncomplicated' cases, formulations may play a lesser role in shaping treatment. Observe, however, that a formulation is still needed to determine whether a case is truly 'uncomplicated,' and whether the therapist should simply follow the treatment manual. Even cases that seem simple on the surface can turn out to be complex (e.g., see Anderson et al., 1996). And even if the formulation suggests that one should follow a treatment manual, the formulation is still relevant for helping the clinician determine how to implement the manual (e.g., timing and pacing of interventions).

Comment To summarize, treatment manuals for panic disorder provide useful templates for developing treatment plans. The protocols outlined in treatment manuals typically don't prohibit the use of case formulations, but nor do they encourage their use. Importantly, CBT2 manuals provide no guidelines for developing a formulation or for integrating the formulation with the protocol. Even if therapists choose to follow a treatment manual, they will need to make decisions about how to conduct treatment. Some sort of formulation is needed; 'the omission of a case formulation can leave the clinician, and the patient, with an amorphous blur that has no direction and can have no clean conclusion' (Rachman, 1985, p. ix). Formulationbased treatment may tend to be most effective when the clinician selects from a set of empirically validated interventions (Wilson, 1996b). Accordingly, when treating panic disorder the therapist should use, whenever possible, the procedures set out in empirically validated CBT2 manuals (e.g., Clark & Salkovskis, 1987; Craske et al., 1994). Formulations can be used to adapt these manuals to individual cases.

DEVELOPING A CASE FORMULATION Persons' (1989; Persons & Tompkins, 1997) method for developing cognitive-behavioural case formulations can be applied to many kinds of clinical problems. The following sections will describe an adaptation of this approach intended specifically for cases in which panic disorder is one of the main reasons for treatment. The components of the formulation are summarized in Table 10.2. Each component is described and illustrated in the following sections.

DEVELOPINGA CASE FORMULATION 269

Table 10.2

Components of the case formulation

Component

Description

1. Problem list

A list of the patient's difficulties, beginning with the chief problem. Maladaptive beliefs about self, world, or future. Some of these beliefs may be causing the patient's current problems. Objects, people, events, or situations associated with the patient's problems. Some of these may contribute to (e.g., trigger) the patient's problems. A model specifying links between components 1-3 that describes the predisposing, precipitating, and perpetuating factors for all the problems on the problem list. Protective factors (if any) are also described. The working hypothesis emphasizes cognitive and behavioural mechanisms, although it also considers other factors. Derived from the working hypothesis, the treatment plan has two components: (1) a specification of treatment goals, and (2) a description of the methods for attaining these goals. A list of predicted or actual obstacles to successful treatment, and strategies for overcoming them. Strategies for overcoming the obstacles are based on either the working hypothesis or, if the obstacles arise unexpectedly, a specific formulation of these new difficulties.

2.

Dysfunctional beliefs

3.

Problem context

4.

Working hypothesis

5.

Treatment plan

6. Treatment obstacles

Adapted from: Persons, J. B., & Tompkins, M. A (1997). Cognitive-behavioral case formulation. In T. D. Eells (Ed.), Handbookof psychotherapycaseformulation (pp. 314-339). New York: Guilford.

Problem List The task of constructing a case formulation begins by assembling a list of the patient's major problems. It is useful to initially develop an overinclusive list of problems because this will provide information for developing a good case formulation (Persons, 1989). In order to gain a good understanding of the breadth of the patient's problems, it is important to examine each of the following domains: psychological/psychiatric, interpersonal, occupational, medical, financial, housing, legal, and leisure (Persons & Tompkins, 1997). Sampling from these domains can yield important information about the type and causes of problems that are amenable to CBT2. The problems should be described as specifically as possible in terms of their cognitions, behaviours, and emotions. It is useful to break problems

270

DEVELOPING A CASE FORMULATION

down into their components or sub-problems. This is done in order (1) to identify relationships among problems and thereby shed light on underlying mechanisms, (2) to plan how to treat the sub-problems, and (3) to mobilize the patient's hope by showing how overwhelming problems can be broken down into manageable units (Persons, 1989). To keep the list within manageable limits, only the 10 most serious problems are retained on the list, beginning with the chief complaint (Persons & Tompkins, 1997). The following is the problem list of Ms W., a 27 yearold bank clerk presenting for treatment of panic disorder: 1. 2. 3.

4. 5. 6. 7. 8.

Recurrent panic attacks. Frequent worry about having panic attacks. Agoraphobic fear and avoidance. In particular, fear of traveling to and from work and fear and avoidance doing the weekly grocery shopping. Depression. Obesity (14kg (301b) overweight). Binge eating and purging. Irritable bowel syndrome. Social isolation; few close friends or confidants.

Patients are not always aware of the full extent of their problems. Some regard their agoraphobic avoidance as a partial solution to their panic problems, rather than a problem in itself (Persons & Tompkins, 1997). Some patients are reluctant to acknowledge their problems, such as patients who abuse alcohol to dampen their panic symptoms. Patients are sometimes afraid to discuss their problems, such as patients who believe that their panics will get worse if they think about them. Thus, the therapist faces several challenges in assembling the problem list. Sometimes it takes several sessions before all the problems have been identified. Development of an accurate problem list is facilitated by a detailed assessment (Chapter 9) conducted in the context of a sound therapeutic relationship. The task of simply listing the problems can be therapeutic, particularly if the patient can be led to understand that the problem list is an important step toward overcoming the problems (Persons & Tompkins, 1997). Sometimes, however, patients become anxious or demoralized when they see their list of problems, particularly when they discover they have more problems _than they previously imagined. When this occurs the therapist should attempt to identify and challenge any associated dysfunctional beliefs, such as 'I'll never get over all these problems,' or 'I must be really screwed up to have so many problems.' Cognitive restructuring methods

DEVELOPINGA CASE FORMULATION 271

such as those described by Beck and colleagues (Beck & Emery, 1985; Beck et al., 1979) can be used to overcome this obstacle. The therapist need not attempt to remedy all the problems on the problem list. The therapist, in consultation with the patient, might decide to address only a few of the major problems (Persons, 1989). The decision depends on a number of factors, including the patient's goals, the number of treatment sessions available, and whether the problems are amenable . to CBT2. Interrelationships among problems, as specified in the case formulation, also determine which problems are most important to treat. Sometimes the successful treatment of one problem (e.g., panic attacks) can lead to reductions in other problems (e.g., depression). For patients who catastrophize about having 'so many' problems, it can be therapeutic for them to see that it may not be necessary for them to tackle all the problems on the list.

Dysfunctional

Beliefs

A further step in developing a case formulation is to identify the patient's dysfunctional or maladaptive beliefs, some of which may be causing the patient's current problems. Cognitive models of psychopathology (Chapters 2 and 3) provide useful guidelines as to which beliefs are likely to be associated with which particular problems. Dysfunctional beliefs leading to depression, for example, are typically pessimistic attitudes about oneself, one's future, or one's environment (Beck et al., 1979). Important beliefs in panic disorder typically have to do with the dangerousness of arousal-related sensations. Panickers who believe that one arousal-related sensation is dangerous also tend to believe that others are dangerous (Taylor, 1999). However, there are often exceptions to this rule. Sometimes, panickers believe that only some arousal-related sensations are dangerous, and that others are aversive but harmless. In order to assess the relevant dysfunctional beliefs in panic disorder it is therefore necessary to survey the patient's beliefs about a variety of arousal-related sensations. The therapist should be alert to the possibility that the patient may have important dysfunctional beliefs that are not predicted by the cognitive models outlined in Chapters 2 and 3. The assessment methods described in Chapter 9 should provide sufficient information for identifying the dysfunctional beliefs likely to be associated with the patient's problems. For example, one could ask the patient to record the thoughts that occur during emotional episodes (e.g., during panic attacks). The therapist and patient then can review the list of thoughts to identify themes suggestive of underlying beliefs.

272

DEVELOPING A CASE FORMULATION

Ms W., whose problem list we reviewed earlier, held the following maladaptive beliefs, which seemed relevant to many of her problems: Beliefs about arousal-related sensations: 'Dizziness leads to insanity,' 'Jumbled thinking is a sign of madness,' 'Palpitations indicate that I'm falling apart-physically and mentally,' 'I can only endure so much anxiety; too much will drive me crazy,' 'Avoiding stressful situations is essential for my mental health.' 2. Beliefs about self-worth being contingent on successful, independent functioning: 'In order to be acceptable I must be self-reliant,' 'I shouldn't rely on others for support,' 'I'm a failure if I ask others for help.' 3. Beliefs about the inability to overcome emotional problems: 'When a person has emotional problems, there isn't much they can do to help themselves,' 'I will never get over my problems,' 'Eating and avoiding stress are the only ways I can feel better, at least for a short while.' 1.

Problem Context The problem context includes stimuli associated with the patient's problems. Elements of the context may play a role in precipitating or perpetuating the problems. For example, stressful life events may form a context in which panic disorder develops. Stressors induce body sensations, which may be catastrophically misinterpreted. The context in which each panic attack occurs provides important information for understanding the underlying causes. One patient reported that some of her panics occurred in the washroom at her workplace. On further inquiry it became apparent that the attacks occurred while she was looking into the washroom mirror, which was illuminated by fluorescent lights. Staring at her reflection induced depersonalization (see Chapter 4). One of her main dysfunctional beliefs was that she was at risk for going crazy. Accordingly, the trigger stimulus (her fluorescent-illuminated reflection) led to catastrophic thoughts of going crazy, which in turn led to panic. Given the number of problems on Ms W.'s list, it comes as no surprise that the problem contexts were many and varied. The following are some examples. Ms W.'s attacks began at age 18, about six months after she had moved out of her parent's home to attend university. Ms W. recalled that her first year at university was lonely and stressful; she felt overwhelmed with her studies, had ongoing financial problems, and no one with whom to share her concerns. Specific stressors included course exams, difficult class assignments, and reminders of her poverty (e.g., overdue notices for unpaid bills). Panic attacks often occurred during these stressful experi-

DEVELOPING A CASE FORMULATION 273

ences. Ms W. was also likely to panic if she went several hours without eating, which she did in an attempt to lose weight. In summary, there were two main contexts in which Ms W.'s panics occurred: (1) during or shortly after stressful events, arising in the absence of social support, and (2) during periods of food deprivation.

Working Hypothesis The working hypothesis is the heart of the formulation, where the therapist brings all the strands of information together to create a model of the predisposing, precipitating, and perpetuating factors responsible for the patient's problems. Protective factors, if present, are also described. The working hypothesis is an approximate account, which may change over time as information accrues. The hypothesis should attempt to account for all the problems, while being as parsimonious as possible. The hypothesis is guided by the cognitive models of panic and other disorders (Chapters 2 and 3). Mechanisms that might account for the problems include dysfunctional beliefs, misinformation, Clark's (1986) vicious cycle of panic, belief-maintaining behaviours (e.g., safety behaviours), operant reinforcement contingencies, and skills deficits (Persons, 1989; Wells, 1997). Cognitive and behavioural mechanisms might not account for all the problems. Biological factors also may contribute, for example, to the frequency and intensity of sensations that the patient may catastrophically misinterpret. The process of establishing DSM-IV diagnoses enables the clinician to begin to organize or group together the patient's problems, and to consider differential diagnoses. As we saw in Chapter 9, this can help determine whether the patient's 'attacks' are episodes of panic, or whether they are symptoms of a general medical condition. The problem list should be examined to identify themes among problems. This can help identify common underlying factors (Persons, 1989). One should also look for correlations among the problem contexts, cognitions, body sensations, emotions, and behaviours (Persons & Tompkins, 1997). These links can provide clues about the factors that precipitate and perpetuate the problems. Ms W.'s efforts at dieting led to recurrent hypoglycemic episodes, characterized by dizziness, palpitations, and concentration difficulties. She catastrophically misinterpreted these sensations as signs of impending insanity, and thereby panicked. Thus, her efforts to control one problem (her weight) exacerbated another problem (panic).

27 4

DEVELOPING A CASE FORMULATION

PredisposingFactors The working hypothesis should describe the predisposing factors in the patient's problems. Predisposing factors can be identified by reviewing the patient's developmental history, including important learning experiences. Maladaptive beliefs may be acquired by verbal instruction, observational learning, or traumatic experiences. For example, a history of unpleasant medical illnesses may lead the person to believe that body sensations indicate that one's physical or mental health is in jeopardy. Failures to acquire important skills (e.g., social skills) also can be predisposing factors. Ms W.'s learning history gave rise to three sets of beliefs that appeared to predispose her to many of her current problems. Recall that these cognitions were (1) beliefs about the dangerousness of dizziness, palpitations, and concentration difficulties, (2) beliefs that her self-worth was contingent on being autonomous, and (3) beliefs about the inability to overcome emotional problems. All of these beliefs were acquired early in life, and appeared to predispose her to many of her current problems. These beliefs appear to have arisen from several learning experiences. Ms W. recalled her mother had often complained of palpitations, dizziness, 'feeling unreal,' and difficulties controlling her thoughts. These complaints became increasingly more frequent as she progressed into what was later recognized as an agitated depression with psychotic features. Ms W. was often warned by her father not to do anything that would upset her mother. This reinforced the patient's belief that arousal-related sensations were harmful. Ms W. also recalled that her father was proud of the fact that he was a 'self-made man', and often lectured Ms W. about the importance of 'making it on your own'. Initially, he chided Ms W.'s mother for having emotional problems, but then conceded that 'there was nothing anyone could do' to help her. The mother later committed suicide when Ms W. was a teenager. This further strengthened Ms W.'s belief that there is little one can do to overcome emotional problems.

Precipitating Factors The formulation should also specify the precipitants of the patient's problems, including the precipitants of panic attacks. Or, to be more accurate, the factors that precipitate arousal-related sensations that in tum lead to panic. Ms W.'s problems were precipitated by a cascade of events. With the increasing stressors during her first year at university, Ms W. began to more frequently experience arousal-related sensations, including dizzi-

DEVELOPING A CASE FORMULATION 275

ness, palpitations, and 'jumbled thoughts.' She misinterpreted these sensations as signs she was going crazy, which therefore led her to panic. Her beliefs about these sensations, along with her belief that there was nothing that she could do to overcome emotional problems, also precipitated agoraphobic avoidance and depression. She increasingly turned to binge eating as a way of temporarily improving her mood, which in turn led to the development of weight problems. Binge eating also precipitated bouts of diarrhea (related to her irritable bowel syndrome), which triggered acute exacerbations in agoraphobia. Ms W. dieted in an effort to lose weight, but this led to hypoglycemic symptoms, which were catastrophically misinterpreted and thereby led to further panic attacks.

Perpetuating Factors Panic disorder can be perpetuated by factors that prevent the patient from learning that arousal-related sensations are harmless. These factors include avoidance and escape behaviours, as well as the more subtle safety behaviours discussed in Chapter 3. Interpersonal factors also may be relevant to the maintenance of panic disorder. In some cases, the patient's avoidance behaviour is positively reinforced by significant others. For example, spousal attention or nurturance may be obtained only when the patient complains of feeling unable to cope. Ms W.'s maladaptive beliefs appeared to play roles as predisposing and perpetuating factors. Her belief about not relying on others prevented her from seeking social support from her few friends, and delayed her from seeking professional help. In fact, it was not until several years later, when her panic disorder had become quite severe and debilitating, that she felt compelled fo seek veatment. By that time she had dropped out of university and returned home to live with her father. Ms W.'s panic disorder was also perpetuated by enduring beliefs about the dangerousness of arousal-related sensations. In turn, these beliefs were maintained by (1) reluctance to discuss her problems with others, which prevented her from learning that the feared sensations were harmless, (2) avoidance of arousal-inducing stimuli and situations, and (3) recurrent dizziness, palpitations, and thinking difficulties, which 'confirmed' her view that she was on the verge of insanity. She was unaware that these symptoms were the results of hyperarousal and dieting. Her agoraphobia was perpetuated by her belief that avoidance of arousing stimuli and situations prevented her from going crazy. (Note that avoidance not only was caused by this belief, but also maintained it, because avoidance prevented the belief from being disconfirmed.)

27 6

DEVELOPING A CASE FORMULATION

Depression was perpetuated (and precipitated) by her ongoing social isolation in conjunction with her belief that she was a 'failure' because she had dropped out of university and because she had failed to control her weight. In turn, her obesity was maintained by a combination of overeating and avoidance of physical exercise. The latter was avoided because it produced feared sensations (dizziness, palpitations). Ms W.'s binge eating was maintained by maladaptive attempts to improve her mood by overeating. Purging was maintained by the belief that this would prevent her from gaining weight. To summarize, there appeared to be three interlocking factors perpetuating Ms W.'s problems: dysfunctional beliefs, maladaptive coping (e.g., avoidance, overeating), and social isolation. Her beliefs played roles as predisposing and perpetuating factors. This underscored the importance of addressing these cognitions in therapy.

Protective Factors Protective factors either prevent the occurrence of particular symptoms or prevent them from become severe. For many psychiatric disorders, social support is a protective or buffering factor; the greater the support, the lesser the severity of the problems. Social support exerts its effects in many ways. For example, it lessens isolation and feelings of stigmatization that may arise from having a psychiatric disorder. Social support also may be protective because it enables the person to be exposed to corrective information (e.g., information that people don't die or go crazy from panic attacks). Ms W. had low social support, and therefore did not have the benefit of this protective factor. Clinically, protective factors are important to assess. They help the therapist understand what keeps the patient's problems from getting worse. In the treatment of panic disorder, the therapist can help the patient maintain, if not strengthen, existing protective factors. Ms W. was counselled to widen her social circle, and thereby increase her social support. Apart from the supportive atmosphere of group therapy (see below), she was encouraged to become active at a local community centre. To this end, she eventually enrolled in a fitness program as well as a drawing class. Although Ms W. had little in the way of protective factors for most of her problems, there was a notable protective factor for her mood disorder. Although Ms W. felt depressed and had suicidal ideation, she denied any urge, plan, or intent to harm herself. Ms W. said she could never kill herself because that would devastate her father, who had already suffered the tragic loss of his wife. Thus, Ms W.'s bond with her father was a pro-

DEVELOPINGA CASE FORMULATION 277

tective factor against suicide attempts. This is important to know because should the situation change, then the protective factor might be removed. If her father passed away then Ms W.'s reasons for living would be undermined, and she would therefore be at greater risk of suicide. A goal of treatment was to increase her reasons for living. Treatment of her panic disorder and other problems was one step in this direction.

Treatment Plan Goals and Interventions The treatment plan has two components; a statement of goals and a description of how to achieve these goals (Persons & Tompkins, 1997). Both are derived from the working hypothesis. Goals should be specific and clearly defined in terms of cognitions, behaviours, and emotions. Clearly stated goals better tell the therapist how to intervene and make it easier to monitor treatment progress. As mentioned earlier, treatment goals might include only a few of the problems on the problem list. The relative severity of problems and the patient's purpose for seeking treatment are important considerations. If the purpose of treatment is to correct the causes of the problems then the working hypothesis provides important information about which goals to pursue (Persons, 1989). If several problems are due to a single belief, then correcting the latter would be an important goal. CBT2 for panic disorder tends to focus mainly on reducing the strength of beliefs in the dangerousness of arousal-related sensations. However, other beliefs and behaviours can be important targets. For example, suicide risk would be the first priority for a patient who has the means and intent for inflicting serious self-harm. To help patients make better informed decisions about which goals to pursue, it is useful to share the formulation with him or her. This can strengthen the collaborative relationship between patient and therapist. Feedback from the patient is also useful in identifying flaws in the formulation. The working hypothesis provides indications about which treatment format is likely to be most useful; individual, group, couples, or other formats. Recall that Ms W. believed that there is little that people (including herself) could do to overcome emotional problems. She also believed she shouldn't rely on others for support. These beliefs could potentially interfere with treatment adherence, and so they were important to consider in the treatment plan. Group treatment-focused primarily on panic disorder-seemed appropriate because it would provide Ms W. with the opportunity to observe that other patients could accept help for their

278

DEVELOPINGA CASE FORMULATION

problems. This form of modeling was expected to make it easier for Ms W. to accept help herself. She also would have the opportunity to observe other group members gradually gaining mastery of their emotional problems, thereby making her more hopeful of overcoming her own difficulties. Given Ms W.'s numerous problems, including problems not shared by most other patients in her panic treatment group, concurrent individual treatment was also considered appropriate. The individual and group therapies were conducted by the same therapist. Individual treatment focused mainly on problems not addressed in the group. The working hypothesis guides the selection of interventions. The following are some examples from the treatment plan for Ms W. Hyperventilation exercises (Chapter 13) were considered useful for teaching her that dizziness does not lead to insanity. Monitoring stressful events, food intake, feared symptoms, and binges also were considered useful in helping her understand the relations among these variables. Monitoring enabled her to see that the feared sensations-dizziness, palpitations, and concentration difficulties-were not signs of impending insanity, but were the harmless result of dieting and stress. Ms W. also came to realize that binges were also more likely to occur when she went for long periods without eating, because the longer she went without food, the hungrier and more dysphoric she became. The working hypothesis predicted that hypoglycemic symptoms and binges would be less likely to occur if she ate more frequent, smaller meals throughout the day. Although Ms W. was skeptical that her problems could be reduced by merely changing her eating habits, she agreed to test this prediction. The prediction was supported, which further increased her confidence in mastering her problems.

Managing the Therapeutic Relationship The working hypothesis can lead to predictions about how the therapist's behaviours are likely to influence treatment outcome. Thus, the hypothesis can provide guidance on how to manage the therapeutic relationship. Some therapist behaviours are likely to be therapeutic for most patients. These are listed in the first column of Table 10.3. Other behaviours are likely to be helpful for some patients but not for others. AuBuchon and Malatesta (1998) suggested a list of these behaviours, a sample of which appear in the second and third columns of Table 10.3. The distinctions between therapist behaviours in Table 10.3 are based largely on clinical experience. Therapist behaviours have failed to reliably

DEVELOPING A CASE FORMULATION 279

Table 10.3 Formulation-based sets of therapist behaviours

management

Therapist behaviours used for maintaining the therapeutic relationship with all patients

Therapist behaviours that are systematically varied, on a case-by-case basis, to maintain the therapeutic relationship

• • • • • • • • • •

Shows respect Exhibits trustworthiness Displays interest Shows caring Conveys understanding Acceptance Accurate empathy Appears competent Instills expectation of change Genuineness

• • • • • • •

of the therapeutic relationship: Two

Therapist availability Length of sessions Frequency of sessions Degree of structure in session Directiveness Provision of information Limit setting

• • • • • • •

Praise/ social reinforcement Nurturance Encouragement Humour Models coping behaviours Confronts maladaptive behaviours Self-disclosure

Adapted from: AuBuchon, P. G., & Malatesta, V. J. (1998). Managing the therapeutic relationship in behavior therapy: The need for a case formulation. In M. Bruch & F. W. Bond (Eds.) (1998), Bl?!Jond diagnosis:Caseformulation approachesin CBT (pp. 141-165). Chichester: Wiley.

predict treatment outcome (Chapter 8). This may be because a specific therapist behaviour can either facilitate, impair, or have no effect on treatment, depending on the nature of the patient and the patient's problems. Thus, the research to date on the effects of therapist behaviours may not have identified strong predictors of treatment outcome because the studies have failed to examine interactions between patient and therapist factors. Further research is needed to examine the effects of using case formulations to manage the therapeutic relationship. Clinical impressions, despite their limitations, suggest it is worthwhile to use case formulations for selecting therapist behaviours (AuBuchon & Malatesta, 1998; Persons, 1989). To illustrate the role of the working hypothesis in managing the therapeutic relationship, consider the amount of structure provided by the therapist. Highly structured sessions might begin by setting an agenda, followed by a systematic review of homework assignments and problems of the previous week, and then followed by exercises completed during the session (e.g., interoceptive exposure exercises). Finally, homework assignments are set. Typically, the early sessions are highly structured. As the patient learns more about the treatment procedures, the need for structure lessens and the patient takes an increasingly active role in organizing how therapy time is spent.

280

DEVELOPING A CASE FORMULATION

The case formulation can guide the therapist in determining how to best depart from this approach. A high degree of structure throughout treatment may be useful for patients who are unable to approach their problems in a structured way. Such a patient might jump from problem to problem, raising one issue and then moving onto the next. As a result, the patient may feel overwhelmed by their problems, and fail to accomplish anything. A high degree of structure can help such a patient tackle the problems in systematic manner. Similarly, for a patient with comorbid panic disorder and borderline personality disorder, the therapist might begin therapy by implementing a course of CBT2, combined with suicide risk management strategies. This may be followed by a protocol for treating the patient's personality problems in greater depth (e.g., Linehan, 1993). The amount of structure during CBT2 might remain high throughout panic treatment, as part of the commonly used principles for managing severe personality disorders (Kernberg et al., 1989; Linehan, 1993). For other patients, the working hypothesis might suggest that less structure is needed for CBT2. A patient with comorbid panic disorder and dependent personality disorder, for example, might benefit by being advised at the outset to be active in treatment, assuming that the formulation indicated that such an approach would help the patient become less diffident (Turkat, 1990). Further examples are provided by Persons (1989): Patients frightened of loss of autonomy need more interpersonal distance than those feeling unable to cope without support. Those who fear they will be rejected if they say the wrong thing need, at least at first, more structure and less open-ended time in the session than patients who feel the need to 'just say what's on my mind.' Those frightened of humiliation may be unable to tolerate any teasing or joking by the therapist, whereas others may find it a welcome relief. Warmth is important for most patients, but ... severely depressed patients, who feel worthless and unacceptable, may feel they are duping someone who seems to like and respect them. (p. 165)

For Ms W., the therapist encouraged her to play an active role in developing her homework assignments. This was done in order to implicitly challenge her belief that there was little she could do to overcome her problems. She completed a hierarchy of exposure assignments, progressing from easy to difficult. This helped Ms W. realize that she could overcome many of her difficulties. One of her other important dysfunctional beliefs-that she shouldn't rely on others-was not addressed until later on. In the early stages of therapy,

DEVELOPING A CASE FORMULATION 281

the therapist used Ms W.'s desire to be independent as a motivator to help her work on overcoming her panic disorder and other problems. As treatment progressed, attention was turned to challenging Ms W.'s belief about reliance on others. By that time the belief was easier to challenge because she had observed from her panic group that asking for help was, in fact, a way of becoming less reliant on others.

Treatment Obstacles Sometimes obstacles can be predicted from the working hypothesis and treatment plan. If obstacles arise unexpectedly, then the therapist attempts to develop a hypothesis specifically to explain the difficulties. Expectations the patient holds about therapy can be a source of treatment obstacles. When expectations are unfulfilled, the patient may become demoralized and drop out of treatment. These include expectations about one's performance (e.g., 'I must be completely successful in all my homework assignments') and expectations about the effects of therapy (e.g., 'It is possible to be completely free of anxiety') (Persons, 1989). As we saw in Chapter 8, practical difficulties in attending treatment sessions are also important because they predict treatment refusal and dropout. Accordingly, these need to be identified and overcome early in treatment. This can be done by problem-solving with the patient, as illustrated by the following case. Ms R.-a depressed, panic disordered single mother-was able to travel to the clinic for an intake evaluation. But when treatment was described to her, Ms R. thought she wouldn't be able to attend the sessions because she couldn't arrange for child care. The working hypothesis suggested that her depressogenic beliefs probably played a role in perpetuating her agoraphobic avoidance (e.g., 'What's the point in going out; it will never get easier to travel alone'), and may have played a role in her other problems. The therapist hypothesized that similar pessimistic beliefs may play a role in her difficulty arranging a baby-sitter. A discussion of this issue revealed that Ms R. believed that 'nobody cared' about her, and so she expected that her neighbours would refuse to help with child care. She was persuaded by her therapist to put these expectations to the test. Fortunately, the expectation proved wrong and child care was arranged. The case formulation approach can be used to identify environmental contingencies-also known as 'secondary gains' or 'reinforcement traps'that maintain problems. and discourage patients from successfully

282

DEVELOPINGA CASE FORMULATION

completing therapy (Persons, 1989). One patient, for example, was frequently ridiculed by her husband whenever she tried something new. This led the therapist to predict that the husband would ridicule the patient's efforts to do homework assignments (e.g., interoceptive exposure). A treatment plan was devised to deal with this potential problem. This involved a session with her husband and a number of conjoint sessions with the patient and husband. Cognitive models of panic disorder (Chapter 3) suggest further reasons why therapy may not be having its intended impact. Exposure exercises, for example, may fail if the patient continues to use safety or avoidance behaviours. A patient might make repeated trips to the supermarket, yet continue to believe he could 'lose control' if he became anxious while standing in the checkout queue. Several factors could prevent this belief from being disconfirmed, such as (1) distraction while in the queue (e.g., imagining himself somewhere safe), (2) escape from the queue whenever his anxiety started to rise, and (3) selecting fast-moving queues (e.g., an 'express checkout'), which would provide insufficient time to test his catastrophic belief. A review of the evidence for and against these possibilities can help the therapist develop a working hypothesis about why behavioural experiments are not having their intended effect. Finally, the working hypothesis can be used to predict the conditions under which the patient is likely to relapse. The therapist can implement relapse prevention strategies to address this problem (Chapter 15).

Testing the Formulation Testing the accuracy of the formulation is an ongoing process, in which the therapist looks for evidence for and against the working hypothesis (Bruch, 1998; Persons, 1989). One of the first steps is to share the formulation with the patient. To limit the chances that the patient will simply acquiesce with the therapist's hypotheses, the patient should be asked to think of specific examples that either support or challenge the formulation. The therapist also can test the formulation by reviewing naturally occurring changes in the patient's problems as well as the patient's responses to interventions, to see if these are consistent with the formulation. If Ms W.'s binge eating exacerbated her irritable bowel syndrome, then treatment of the former should improve the latter. If this did not occur, then the working hypothesis and treatment plan would need to be revised.

DEVELOPINGA CASE FORMULATION 283

The therapist should be alert for phenomena that cannot be explained by the formulation. If Ms W. one day reported that her dizziness worsened to the point that she collapsed and lost continence, then that would be clearly inconsistent with the hypothesis that she catastrophically misinterprets the symptoms of mild hypoglycemia. In this example the working hypothesis would be revised in light of a neurological investigation. The value of the case formulation depends on whether or not it leads to successful treatment; if the case formulation does not improve treatment outcome, then it is a poor formulation, regardless of how much information it provides (Persons & Tompkins, 1997). If the treatment proves unsuccessful, then it is important for the therapist to revise the working hypothesis and the treatment plan.

Comment There is no single 'correct' method for developing a cognitive-behavioural case formulation. A useful method is one that provides a systematic way of developing a model of the causes and cures of the patient's problems. The method outlined in this chapter is one approach for understanding and treating panic disorder. Readers are encouraged to experiment with this and other methods for formulating their cases. The process of developing the formulation is a collaborative venture, in which the therapist uses cognitive-behavioural principles (e.g., guided discovery, collaborative empiricism) to help the therapist and patient arrive at a shared understanding of the causes of the problems. The formulation can be presented to the patient verbally and in writing. The latter is more likely to be remembered by the patient. The formulation can be prepared after a couple of assessment sessions, although it is not uncommon to revise formulations during treatment, as data accumulates. If the therapist is unable to develop a formulation-or if the patient is in great distress and in need of immediate assistance before the therapist has time to complete a formulation-then one could initiate symptomfocused treatment (Persons & Tompkins, 1997). For example, the therapist could start by simply following Craske et al.'s (1994) manual, or by implementing procedures that produce rapid relief, such as relaxation training. As more information is collected, a formulation can be developed. Then in later sessions the therapist could begin to implement formulation-based treatment.

284

DEVELOPINGA CASE FORMULATION

SUMMARY AND CONCLUSIONS Case formulations are an important part of treatment, even when treatment is based largely on the steps laid out in a treatment manual. The case formulation approach requires more than technical competency in CBT2; it requires the clinician to formulate and test hypotheses about the predisposing, precipitating, perpetuating, and protective factors in the patient's problems, and to select interventions based on these hypotheses. This approach goes beyond matching symptoms to treatments. The case-formulation approach described in this chapter emphasizes the role of dysfunctional beliefs, although other factors are also considered. The formulation should have a sound rationale, drawing on the empirical literature and on the cognitive models of psychopathology (Chapters 2 and 3). The formulation not only helps guide treatment, but also provides a method for understanding and overcoming obstacles. When a case formulation is used, each difficulty in treatment becomes a problem to be analyzed, understood, and resolved.

i

,tr

I

COGNITIVE-BEHAVIOUR THERAPY: AN OVERVIEW This and the following chapters will review the second generation cognitive-behavioural (CBT2) protocols and techniques for treating panic disorder. The discussion will draw on a number of published and unpublished sources to present some of the most useful methods. Sources include the work of David M. Clark's Oxford group (e.g., Clark, 1989; Clark & Salkovskis, 1987; Salkovskis & Clark, 1991; Wells, 1997) and David H. Barlow's Albany group (e.g., Barlow & Craske, 1994; Craske et al., 1994).1 The chapters will consider the strengths and limitations of the various interventions to help the reader make informed choices about which approaches to use and when to use them. The current chapter will present an overview of the major treatment protocols, and describe their main components. Patient suitability also will be considered. Details of specific interventions-derived from the Oxford and Albany protocols, and from other sources-are provided in Chapters 12-14. The treatment strategies described in these chapters were developed for treating panic disordered adults in psychiatric outpatient settings. As we will see later in this chapter, the treatments can be delivered by other means, such as by telephone, book, or computer. The application of these treatments to special populations such as children or the elderly will be discussed in Chapter 16.

WHO IS SUITABLE FOR COGNITIVE-BEHAVIOURAL TREATMENT OF PANIC DISORDER? In Chapter 8 we saw that there are few reliable predictors of treatment outcome, and so far we have been largely unable to identify variables that predict who is most likely to benefit from one treatment (e.g., CBT2) 1

The terms 'Oxford group' and 'Albany group' refer to where the treatment protocols were developed and tested.

286

COGNITIVE-BEHAVIOUR THERAPY:AN OVERVIEW

versus another (e.g., drugs). CBT2 can effectively treat patients from a wide range of ethnic, educational, and socioeconomic backgrounds. A case formulation and patient preference are used to decide who is suitable for CBT2. If the patient has panic disorder as a major problem, and is sufficiently motivated to complete a course of CBT2, then this therapy may be indicated. If the patient also has other major problems, such as severe depression or substance dependence, then panic treatment may be deferred until the more pressing problems have been addressed. Before commencing a course of CBT2, the patient should have a physical examination in order to rule out medical mimics of panic disorder. Many patients are on psychotropic medication when they present for CBT2. For patients on fixed (scheduled) doses of medication, it is important that they do not increase the dose of medication during the course of CBT2. This is so that the therapist can assess the effects of CBT2 separately from the effects of medication. Ideally, the patient would either keep the dose of medication constant throughout CBT2, or taper off the medication prior to starting CBT2. Patients often taper themselves off their medication during CBT2, suggesting that CBT2 is exerting its beneficial effects even when the panic-suppressing effects of drugs have been removed. Drug tapering during the course of CBT2 is unlikely to be a problem so long as the patient tapers the medications slowly rather than abruptly, in order to limit the likelihood of rebound panics and other withdrawal effects. The CBT2 therapist and prescribing physician should be informed about any attempts at drug tapering, so withdrawal effects can be monitored and adjunctive interventions can be implemented as needed. Patients taking PRN (as needed) medications should be asked to either refrain from using the drugs unless absolutely necessary, or to take their medications on a fixed schedule. This is so that the use of PRN medication does not interfere with interoceptive or situational exposure exercises. Patients who take a dose of, say, lorazepam (Ativan) prior to an interoceptive exposure exercise may either not experience the feared sensations (pharmacologic suppression of arousal) or may believe that the drug prevented some feared outcome from occurring. Either way, the efficacy of exposure may be compromised.

l

f f

TREATMENT COMPONENTS

5

The purpose of CBT2 is not to simply teach patients to relabel their arousal sensations as harmless; the goal is to reduce the frequency and severity of these sensations and related symptoms. This is accomplished by a variety of strategies, delivered in a coordinated, integrated fashion. Inter-

1 \

j

C

l

TREATMENTPACKAGES 287

ventions are administered in the context of ongoing assessment of sensations, behaviours, and cognitions. The main components of CBT2 are as follows. • • •

•

Psychoeducation. Information about panic disorder (including agoraphobia), cognitive models of this disorder, and treatment methods. Cognitive structuring. Evaluating and replacing catastrophic beliefs with more realistic, noncatastrophic ones. Exposure strategies. Interoceptive and situational exposure exercises, performed during treatment sessions and as homework assignments. The exercises can be performed as 'behavioural experiments' to test catastrophic and noncatastrophic beliefs about body sensations and other events. Other interventions. Breathing retraining, relaxation training, relapseprevention strategies, and other interventions used as needed (e.g., couples therapy, assertiveness training).

TREATMENT PACKAGES Several empirically supported CBT2 packages have been developed. The protocols typically range from 8 to 16 weekly sessions, conducted either individually or in groups of 6 to 8 patients. Patients typically receive at least 12 sessions of treatment. Individual sessions are usually 45-60 min, whereas group sessions tend to be 90-120 min. For the average patient, group and individual formats appear to be equally effective (Chapter 5). There are several considerations in deciding which format to use. Health maintenance organizations prefer group treatments because they are more cost-effective, whereas clinicians in private practice may prefer individual treatment, because it may take too long to obtain a sufficient number of panic patients to form a treatment group (Craske et al., 1994). Patient preference also is an important consideration; some prefer not to discuss their problems in front of a group, whereas other patients prefer group treatment because they are reassured by seeing that they are not alone in their struggle with panic disorder. The two most widely used protocols are those developed by the Oxford group (e.g., Clark, 1989; Clark & Salkovskis, 1987) and the Albany group (e.g., Craske et al., 1994). Treatment is often effective when the therapist simply follows the steps laid out in the manuals. However, a case formulation is useful in determining how to implement the various interventions and when to depart from the protocol (see Chapter 10). The Albany and Oxford protocols share many similarities in terms of the style of therapy and the structure of sessions. These are described in the next

288

COGNITIVE-BEHAVIOUR THERAPY:AN OVERVIEW

section. This is followed by a summary of the Albany and Oxford protocols, and a discussion of their strengths and weaknesses.

Style of Therapy and Structure of Sessions The style of the Albany and Oxford protocols is collaborative and interactive. Beliefs are framed as hypotheses, and the therapist and patient work together to identify evidence for and against the beliefs. Panic attacks are framed as opportunities to learn about the cause and control of these attacks (Craske et al., 1994). The therapist encourages the patient to ask questions and express any doubts about what is discussed during treatment. For example, the therapist might ask 'Does the cognitive model of panic fit with your experience?' This helps the therapist identify idiosyncratic, catastrophic beliefs, and to devise convincing disconfirmations of these beliefs (Salkovskis & Clark, 1991). Sessions tend to be highly structured, particularly at the beginning of therapy. Sessions typically start by a review of the patient's symptoms over the past week and a review of any assigned homework. This gives the therapist an idea of what is important to cover during the session. Then an agenda is set for the remainder of the session. The agenda typically includes a discussion of specific problems along with planning or actual implementation of treatment strategies. Some interventions are implemented during the session, whereas others are assigned as homework. Sessions typically end with homework assignments. The importance of continued practice of the various exercises is emphasized (Craske et al., 1994). Throughout each session feedback is elicited to check the patient's understanding of the material; e.g., 'Can I ask you to summarize what we've just discussed so I can check that we're both on the same track?' (Clark & Salkovskis, 1987). During each session the therapist provides periodic capsule summaries of important topics that have been discussed, and asks whether the patient agrees with the summaries. Capsule summaries help educate the patient, help the therapist check the accuracy of his or her understanding of the patient's problems, and help the patient and therapist focus on the most important issues (Clark & Salkovskis, 1987). Sessions are often audiotaped and patients are asked to listen to the tapes in between sessions. This is to consolidate learning. Listening to the audiotapes is also a distancing strategy, in which the patient can listen to the catastrophic cognitions discussed during the session, and reflect on their accuracy (Clark, 1989). Learning is further facilitated if the patient writes out a summary of the most important things learned from the tape. As an

l

TREATMENTPACKAGES 289

alternative to audiotaping the session, a patient workbook can be used, where the key points covered in the sessions are reiterated in the workbook.

Albany Protocol Description

Treatment materials for this protocol include a therapist manual for panic disorder with or without agoraphobia (Craske et al., 1994), a patient workbook for panic disorder and mild agoraphobia (Barlow & Craske, 1994), and a patient workbook for moderate or severe agoraphobia (Craske & Barlow, 1994). The workbooks were not designed as stand-alone bibliotherapies but rather to be used in conjunction with face-to-face therapy. Thus, learning during therapy is facilitated by reading the workbooks in between sessions. Each section of the workbooks contains a selfassessment quiz, which tests the patient's understanding of key concepts, and thereby helps correct any misconceptions about panic attacks, agoraphobia, and their treatment. In order to apply the Albany protocol, the therapist requires the therapist manual and the patient workbooks. The manual provides an outline of the sessions, sample vignettes, and guidelines for dealing with some of the problems that might be encountered. The manual, however, provides very little information on the actual interventions; these are described in the patient workbooks. The protocol for treating panic disorder and mild agoraphobia consists of 15 'lessons', which provide patients with exercises to practice each week. Typically, one lesson is covered in each weekly therapy session. The content of this protocol is summarized in Table 11.1. Here, lesson 1 corresponds to the first treatment session, which is preceded by a pre-treatment assessment. The core program consists of lessons 1-12 and 15. Lesson 13 (mild agoraphobia) and lesson 14 (medication discontinuation) are optional, depending on the needs of the patient. Craske et al. (1994) recommend that the core program be completed in its entirety even if the patient improves after a couple of weeks of treatment. Patients who discontinue before the end of the program may be at increased risk for relapse. Each patient's partner (or significant other) is invited to read the workbook(s), and patients are encouraged to involve their partners in particular treatment exercises such as interoceptive exposure. The partner can be trained to serve as a 'coach,' thereby supporting the patient's efforts at completing the various exercises.

Table 11.1 Outline of the Albany protocol Lesson

Content

1

•

2

• •

3

• • • • • •

4

• • • •

5

• •

•

6 7, 8

• • •

9

• •

10, 11

12 13 14

15

• • • • • • . •

Psychoeducation about panic disorder and maladaptive coping (e.g., avoidance). Description of treatment program. Instruction in observing symptoms and introduction to prospective monitoring. Discussion of the distinction between panic and anxiety. Discussion of stimuli that trigger panic attacks. Further information about the nature of anxiety and panic; e.g., the components of anxiety and panic (cognitive, physiological, and behavioural). Discussion of some of the etiologic factors (e.g., role of stressful life events). Discussion of how the treatment methods address each component of anxiety and panic. Prospective monitoring to identify the sequence of events during panic attacks, including panic triggers. Discussion of the physiological basis of anxiety and panic. Discussion of the role of cognitive factors in panic. An account similar to Clark's (1986) model is presented. Review of common catastrophic misinterpretations. Continued prospective monitoring, with instructions to pay particular attention to thoughts occurring during panic attacks. Discussion of the role of hyperventilation in panic. Assessment of whether hyperventilation plays a role in the patient's panics. This includes a short hyperventilation exercise to assess whether the patient's hyperventilation symptoms are similar to symptoms of their naturally occurring panics. Instruction in breathing retraining or progressive muscle relaxation. Patients typically receive only one of these interventions. Patients who appear to hyperventilate during panic attacks are treated with breathing retraining. Otherwise, patients receive relaxation training. Breathing retraining or relaxation. Breathing retraining or relaxation. Cognitive restructuring, with an emphasis on two cognitive errors: (1) overestimating the probability of aversive events, and (2) overestimating the cost (badness) of events. Breathing retraining or relaxation. Cognitive restructuring; patients are taught to consider their beliefs as predictions to be tested. Instruction in identifying panic triggers. Interoceptive exposure. Interoceptive exposure. Discussion of the causes of the patient's worst panics. Planning situational exposure to reduce agoraphobia (exposure is set as homework). Discussion of medication discontinuation. Review of progress made so far, planning for continued progress, and discussion of relapse preverition.

From Craske, M. G., Meadows, E. A., & Barlow, D. H. (1994). Masten; of your anxiety and panic II and agoraphobia supplement: Therapist guide. San Antonio, TX: Psychological Corporation.

l

TREATMENT PACKAGES 291

For patients with moderate or severe agoraphobia, an 8 lesson agoraphobia supplement can be used. Situational exposure is the mainstay of this treatment. Exposure can be either massed (i.e., flooding) or involve working up a hierarchy of increasingly fear-evoking situations. Imaginal rehearsal and cognitive restructuring are also used. Imaginal rehearsal is used to prepare the patient for entering feared situations. Cognitive restructuring is used to address maladaptive cognitions occurring before, during, and after each exposure exercise. For example, cognitive restructuring can be used to challenge catastrophic expectations about what could happen during an exposure exercise. Versions of the panic manual and workbook are also available for panic disordered patients who have difficulty discontinuing benzodiazepines (Otto et al., 1995, 1996a). This protocol aims to reduce panic disorder while also reducing the distress associated with benzodiazepine discontinuation. The protocol differs from the standard Albany protocol (Barlow & Craske, 1994; Craske et al., 1994) in the following ways: (1) a drug tapering schedule is used, commencing after the third therapy session, (2) benzodiazepine withdrawal symptoms are monitored, and (3) these symptoms are described to patients as being a benign 'benzodiazepine flu.' The protocol can be readily adapted for tapering patients off other anti-panic medications.

Strengths and Limitations The Albany protocol in its various forms offers a comprehensive program for treating panic disorder. The manuals and workbooks are detailed and provide rating forms necessary for treatment. The materials are commercially available and therefore readily obtained by patients and therapists. Barlow, Craske, and colleagues have shown the Albany protocol to be efficacious in several studies (see Chapter 6). It could be objected that the protocol places too much emphasis on breathing retraining, the use of which is based on the assumption that hyperventilation contributes to panic attacks. Recent studies suggest that hyperventilation is not as important as previously thought, and that for many patients breathing retraining may be, at best, a placebo and, at worst, a safety behaviour that maintains their catastrophic beliefs (see Chapter 6). To limit the unnecessary use of breathing retraining, the Albany protocol includes procedures to assess the importance of hyperventilation, such as an assessment of naturally occurring hyperventilatory symptoms, and a hyperventilation exercise to investigate whether hyperventilatory symptoms are similar to the patient's naturally occurring panics. These assessments may ensure that breathing retraining is only used when needed.

292

COGNITIVE-BEHAVIOUR THERAPY: AN OVERVIEW

A strength of the Albany protocol is that it includes strategies for eliminating safety signals. Recall from Chapter 3 that these are stimuli associated with a sense of safety. These include medication containers and the presence of 'safe' others. The Albany protocol systematically prompts patients to not rely on their safety signals in order to better increase their autonomy and better convince themselves of the harmlessness of body sensations. The protocol also discourages patients from using coping statements, breathing retraining, and relaxation exercises as safety behaviours. Patients are instructed not to use breathing ret;aining and relaxation techniques to avoid catastrophic outcomes, but instead to simply use them as ways of reducing unpleasant, harmless arousal. The Albany protocol could be improved by addressing safety behaviours in more detail. Safety behaviours can be difficult for the therapist to identify, and so it would be useful to contain a lesson devoted to them (see Chapter 13 for details on assessing and eliminating these behaviours). For example, some patients may engage in a variety of strategies to avoid depersonalization, such as avoiding fatigue and avoiding buildings illuminated by fluorescent lighting. Patients sometimes fail to report that they engage in these forms of avoidance. Identification and elimination of these behaviours is important for eliminating catastrophic beliefs. A further limitation is that the protocol places a heavy emphasis on didactic presentations. The therapist's manual (Craske et al., 1994) suggests that a good deal of the material is covered in mini-lectures. In fact, 6 of 15 lessons are primarily didactic. Some of this material may not be needed, and some of it is presented in a way that is too technical for some patients to understand. The latter is illustrated by the following transcript, which appears on p. 60 of the manual.

Patient (P): Once I have a panic attack, the feeling stays around for weeks afterward. How can that happen? Therapist (T): The actual panic attack arousal is generally very brief, as sympathetic nervous system exacerbation is counterbalanced by parasympathetic nervous system activation. However, after a panic attack, you likely experience a lot of generalized arousal and vigilance for signals of another panic attack. That state of generalized anxiety can be accompanied by various symptoms that are similar to some of the symptoms experienced during panic, except that they are present more chronically as opposed to acutely.

TREATMENT PACKAGES 293

Here's another way of answering the same question, using guided questioning (Socratic dialogue) and less technical language:

P: Once I have a panic attack, the feeling stays around for weeks afterward. How can that happen? T: Why do you think it happens? P: I guess it could be the after-effects of panic. T: In the days after a bad panic, do you worry about anything in particular? P: Yes, I worry that my panics might be getting worse. T: Do you think your worrisome thoughts might cause the anxious feelings to stay around for weeks? P: Yes, I guess they could.

Socratic dialogue is a good way of having patients come up with the answers to their problem. This method is described in detail in Chapter 12.

Oxford Protocol Description The Oxford protocol is set out in an unpublished therapist's manual (Clark & Salkovskis, 1987). and described in several articles and book chapters by Clark and colleagues (e.g., Clark, 1989, 1996; Salkovskis & Clark, 1991; Wells, 1997). A series of self-study booklets have been developed for patient use (Clark et al., 1999), and the therapist's manual contains patient handouts and rating forms (Clark & Salkovskis, 1987). Handouts and forms are also available in Clark (1989) and Wells (1997). Treatment consists of 10-12 weekly 60min individual sessions, followed by up to three booster sessions over three months (Clark et al., 1994, 1999). Although the Oxford protocol primarily focuses on panic attacks, it also addresses agoraphobia. Like the Albany protocol, the Oxford protocol makes extensive use of interoceptive and situational exposure exercises. These are conducted as behavioural experiments to test beliefs and to test the effects of safety behaviours. Exposure exercises are performed during the session and as homework assignments. One of the key differences between Oxford

294

COGNITIVE-BEHAVIOUR THERAPY:AN OVERVIEW

and Albany protocols is that over the years Clark and colleagues have decreased the amount of time devoted to breathing retraining, to the point that they now see it as a very minor or unnecessary component of treatment (David M. Clark, personal communication, 4 December 1998). Relaxation training is also infrequently used. The focus of treatment is on helping patients make noncatastrophic interpretations of feared sensations and situations, with the aim of replacing maladaptive, catastrophic beliefs with noncatastrophic ones. Compared to the Albany protocol, the Oxford approach contains less didactic material. The first few sessions of the Oxford treatment are devoted to (1) educating the patient about Clark's (1986) cognitive model of panic, (2) identifying the relationship between the patient's body sensations and catastrophic cognitions, and (3) devising and carrying out interoceptive exposure exercises to help the patient learn that the feared sensations are innocuous. Later sessions implement interoceptive and situational exposure exercises (behavioural experiments) to further test catastrophic and noncatastrophic cognitions. Exercises are refined and repeated as necessary. Patients are asked not to enter feared situations until their fifth treatment session. This is so that they have time to develop confidence that treatment is helping them reduce their panics. Socratic dialogue and cognitive restructuring exercises are also used throughout, including strategies for modifying catastrophic imagery. Treatment sessions are gradually faded out, with the patient having the option of attending booster sessions. Toward the end of the program, the therapist and patient review the progress made so far, and develop plans for continuing progress after treatment ends. Strategies for relapse prevention are also discussed.

Strengths and Limitations Several studies by Clark and colleagues have shown the Oxford protocol to be effective in reducing panic disorder (Chapter 6). A strength of this protocol is its emphasis on using exposure exercises as behavioural experiments, including experiments that teach patients that they don't need to rely on safety signals and safety behaviours. A further strength is that the protocol makes extensive use of Socratic dialogue. An unfortunate disadvantage of the Oxford protocol is that the therapist manual and patient self-study booklets are currently unpublished. The treatment approach is described in various book chapters and journal articles (e.g., Clark, 1989; Wells, 1997) but do not provide the degree of detail and structure provided by the Albany therapist manuals and patient workbooks.

THEMEAND VARIATION:ALTERNATIVE WAYSOF DELIVERING TREATMENT295

Comment The Oxford and Albany protocols are the leading CBT2 protocols, representing state-of-the-art treatments for panic disorder. Although the two protocols have not been directly compared in treatment outcome trials, they probably produce similar results in reducing panic disorder. Neither protocol is effective for all patients, and no doubt both could be improved. Which of these protocols should the therapist use? The writings of the Oxford group (e.g., Clark, 1989; Clark & Salkovskis, 1987; Wells, 1997) provide more informative guidelines and examples of how to do cognitive restructuring and conduct behavioural experiments. However, an Oxford manual has yet to be published. The treatment manuals and patient workbooks for the Albany protocol have been published, and so are more readily available to patients and trainee therapists. There is also a commercially available set of training videotapes for the Albany protocol-available from the Psychological Corporation-which further helps therapists learn the protocol. One approach to learning CBT2 is to begin with the Albany protocol. Once this has been mastered the therapist can incorporate aspects of the Oxford protocol into the Albany protocol, or even replace some components of the Albany protocol with those used in the Oxford protocol. For example, depending on the case formulation, breathing retraining and relaxation training could be dropped from the Albany protocol, and greater use could be made of Socratic dialogue and other methods of cognitive restructuring such as imagery modification. We have successfully used this approach in training our therapists (e.g., Taylor et al., 1996), although it remains to be empirically established whether this is any better than simply following either the Albany or Oxford protocol.

THEME AND VARIATION: ALTERNATIVE WAYS OF DELIVERING TREATMENT The previous sections examined protocols for delivering the 'full' CBT2 packages, typically consisting of 12 or more face-to-face sessions conducted either individually or in groups. Such protocols are among the most commonly used methods for delivering CBT2. However, these protocols are •not always feasible or appropriate. The following sections review the ways in which these treatments can successfully be delivered in alternative formats.

296

COGNITIVE-BEHAVIOUR THERAPY:AN OVERVIEW

Abbreviated Protocols CBT2 is labor-intensive, requiring months of therapy delivered by suitably trained clinicians. In an attempt to develop more economical treatments, several investigators have evaluated short forms of standard protocols. Some short protocols have proved disappointing in that they appear to be far less effective than the standard Oxford or Albany protocols (e.g., Black et al., 1993). One of the promising new developments is a five-session Oxford protocol, which appears to be as effective as the full protocol (Clark et al., 1999). The brief treatment requires Sh of therapist time (five lh sessions) compared to 12h for the full protocol. Patients receiving either full or brief treatment also receive two booster sessions (I.Sh in total). Brief treatment is implemented in conjunction with self-study modules, which patients read between sessions. The modules contain written exercises and homework assignments, which are reviewed in the therapy sessions. A different module is studied before each of the first four sessions. The modules cover the following material:

•

•

•

•

Module 1. This module illustrates Clark's (1986) vicious cycle of panic and provides patients with a series of self-assessment questions about the thoughts and feelings arising during their panics. This is done to help patients understand panic attacks in terms of the vicious cycle. The module also contains strategies for challenging catastrophic beliefs about body sensations, and discusses ways that avoidance, catastrophic images, and attention to sensations can maintain catastrophic beliefs. Module 2. This focuses on the patient's worst fears about the sensations experienced during panic attacks. The module also helps the patient come up with alternative, noncatastrophic explanations for the sensations. Module 3. This module introduces the concept of safety behaviours, and discusses ways that these behaviours prevent the patient from learning that the feared sensations are harmless. The module also helps patients identify their own safety behaviours. Module 4. Here, behavioural experiments are introduced in which patients are advised to drop their safety behaviours in feared situations and during panic attacks. The module also helps patients identify triggers for their panics, and strengthens noncatastrophic explan·ations of arousal-related sensations. Strategies for relapse prevention are also outlined.

THEMEAND VARIATION:ALTERNATIVEWAYS OF DELIVERINGTREATMENT 297

Handouts describing common catastrophic cognitions are also provided at the therapist's discretion. Sessions are increasingly spaced apart. Sessions 1-3 are separated by 2 weeks. Three weeks later the patient receives session 4, and session 5 is completed three weeks after that. Situational exposure is implemented in session 5. An even shorter treatment was developed by Clark (1996), who reported a single-case study showing that panic disorder could be successfully treated in a two session version of the Oxford protocol, supplemented by self-study modules like those mentioned above. The first session was 4h. Three weeks later the patient completed a 1 h session. Other brief protocols are also effective. Based on the Oxford and Albany protocols, Cote et al. (1994a,b) developed and evaluated a CBT2 protocol consisting of seven treatment sessions (mean session duration= 76min) plus eight telephone contacts with the therapist (mean duration of each call = 11min). At posttreatment and 3 year followup the brief treatment was just as effective as a longer (17 session) protocol. In summary, effective brief protocols have been developed, consisting of two-seven sessions, supplemented by brief telephone consultations or self-study modules. The outcome data for these protocols are generally impressive, but further research is needed. Although the brief treatments are effective for many patients, they may not be sufficient for severe panic disorder (characterized by very frequent or intense panic attacks, and/ or severe agoraphobia). In these cases a full protocol is probably needed.

Treatment in Emergency Room Settings Most panic patients are treated in psychiatric outpatient clinics or in the private offices of CBT2 therapists. Yet, many panickers first seek help at hospital emergency departments, thinking that they may be having a heart attack or suffering some other somatic or psychological calamity. Early intervention, conducted in the emergency room, can be helpful. Unfortunately, the interventions patients often receive in these settings are not helpful. Pollard and Lewis (1989) described the problem as follows: A significant portion of panic disorder patients leave the ED [emergency department] more discouraged, confused, or apprehensive than before they arrived. Our patients' reports suggest a somewhat common scenario. A series of medical tests are conducted. Patients are informed that results

298

COGNITIVE-BEHAVIOUR THERAPY:AN OVERVIEW

are negative or that they may have a relatively benign condition such as mitral valve prolapse. Though momentarily relieved, most patients remain anxious to discover the source of their symptoms. Unfortunately, explanations provided are often incomplete, with vague references to a 'nerve condition' or 'stress'. Too often, they recall being given no explanation at all. Instead of being referred to a mental health professional, many patients are treated with short-term drug therapy, which is frequently ineffective ... or are referred back to a primary care physician. (pp. 547-548).

Panic patients are often alarmed by negative test results, concluding that they must be going crazy because the doctor could find nothing physically wrong with them. In place of the unhelpful advice patients often receive, Pollard and Lewis (1989) suggested three simple interventions that could be usefully implemented by emergency room staff: •

•

•

Staff could instruct panickers in breathing retraining as a way of controlling their attacks. (This is most useful in controlling panic attacks associated with hyperventilation.) The staff member should redirect the patient's attention to what is actually occurring, rather than what the patient fears will happen. For example, rather than focus on fears of going crazy, the patient would be asked to observe and objectively describe the feared sensations. 'This redirection of attention from the future to the present, from the feared to the actual, helps interrupt the preoccupation with impending doom that exacerbates panic symptoms ... The clinician can also be helpful by clarifying distinctions between sensations (the actual) and anticipations (the feared)' (Pollard & Lewis, 1989, p. 550). Once the patient has sufficiently calmed down, then they can be provided with corrective information, presented verbally and in written form. This could include information that the body sensations are harmless, and education about the vicious cycle of panic. A followup appointment can then be scheduled a week or two later with the patient's primary care physician to check whether the panics have abated.

Swinson et al. (1992b) demonstrated that another simple protocol is effective for treating panickers presenting to the emergency room. These authors assessed 33 patients attending emergency rooms because of panic attacks. Patients were first seen by an emergency room physician who ruled out general medical conditions. Patients were then seen in a single 1 h session by a psychiatrist. The session took place either before the patient left the emergency room or within the next two days. Patients

l

l

THEMEAND VARIATION:ALTERNATIVE WAYSOF DELIVERING TREATMENT299

received either (1) instructions for self-directed exposure, or (2) reassurance that their panics were harmless anxiety reactions and that no serious physical disorder was present. When assessed 6 months later, patients in the exposure condition were improved on measures of panic frequency and agoraphobia. Patients receiving reassurance were not improved on any measure, and in fact had become more avoidant. In summary, brief protocols implementing one or two CBT2 techniques can be useful in treating panic patients presenting to emergency rooms. These protocols also could be used by primary care physicians when panic patients present directly to them. The protocols seem best suited for mild, uncomplicated cases of panic disorder. If panics persist, then a referral to a CBT2 therapist could be arranged.

Telephone-administered

Treatment

Some patients are unable to attend a clinic to obtain treatment. Some are unable to travel to the clinic because of severe agoraphobia, while others live too far to realistically travel for appropriate help. Telephoneadministered treatment is an option for these patients, assuming the patient has already been assessed by their primary care physician to rule out medical mimics of panic disorder. The CBT2 therapist can conduct a structured diagnostic assessment over the phone to establish a diagnosis of panic disorder, followed (if indicated) by treatment over the phone. Telephone-administered treatment has been evaluated in several studies. Taylor (1984) described a case study of a patient with panic disorder (with severe agoraphobia) who was treated with 11 weekly lOmin phone calls. During the calls the patient was instructed in how to go about completing situational exposure exercises. Treatment focused on bus travel, the patient's major problem, but also addressed other problems such as shopping in crowded stores and walking increasingly further from home. During each phone call, homework from the previous week was reviewed and the patient received appropriate feedback and encouragement from the therapist. The patient was reassured that many of his physical symptoms were manifestations of anxiety, which he should ignore as much as possible. Treatment led to reductions in agoraphobia and anxiety, with gains maintained at 2 month followup (data on panic frequency data were not reported). Although this treatment was helpful in reducing some of the patient's problems, Taylor (1984) observed that the patient's motivation to complete the exposure exercises steadily waned over the days between phone

300

COGNITIVE-BEHAVIOUR THERAPY:AN OVERVIEW

calls. More frequent calls may have reduced this problem. The author concluded that telephone therapy is useful for well-motivated patients, although many patients will require a good deal of prompting and encouragement. Later studies using larger samples found that situational exposure administered over the telephone is effective at posttreatment and at 6 month followup, and is more effective than telephone-administered relaxation training (McNamee et al., 1989; Swinson et al., 1995). In these studies telephone exposure therapy was supplemented by bibliotherapy. Treatment consisted of 8-10 calls over 10-12 weeks. The duration of calls ranged from 12 to 60 min. Studies so far have delivered via telephone only one CBT2 component; viz., situational exposure. Treatment efficacy might be improved by using other components of CBT2, such as cognitive restructuring and interoceptive exposure. These could be delivered by telephone or by Internetbased videoconferencing. It remains to be seen whether these treatments are useful for complex conditions, such as panic disorder with comorbid disorders. If a panic patient suffers from depression with serious suicidal ideation, then telephone therapy is not sufficient. Face-to-face contact with a clinician is essential.

Bibliotherapy Bibliotherapy refers to the use of written instructional materials, often in the form of a self-help book, to guide the patient through a course of treatment. Such materials are commonly used as adjuncts to behavioural or cognitive-behavioural therapies (e.g., Barlow & Craske, 1994). However, bibliotherapy also has been used as the primary intervention, with therapist contact limited to a clinical assessment and just enough help to get the patient started in the program. The latter is important for encouraging patients to attempt the program (Ghosh et al., 1988). Thus, bibliotherapy involves more than simply giving the patient a book to read. It requires some therapist contact, although not as much contact as in the abbreviated protocols described earlier. Situational Exposure Protocols A number of bibliotherapies have been developed for helping patients implement situational exposure exercises (e.g., Marks, 1978; Mathews et al., 1981). Some self-help guides advise the patient to enlist the sup-

THEMEAND VARIATION:ALTERNATIVE WAYS OF DELIVERINGTREATMENT 301

port of a significant other in treatment, with the latter reading the materials and working as a coach to help the patient to implement the interventions. Bibliotherapy-administered situational exposure should provide patients with detailed and concrete guidelines for implementing the exercises: Self-exposure does not consist of merely telling patients to expose themselves to the situations they fear. Rather, they are told how to do it within an ordered framework. ... The patient requires careful instruction in how to (i) identify agoraphobic target problems that have to be systematically dealt with one by one, (ii) practice self-exposure to each target situation regularly and for hours at a time whenever necessary, (iii) record tasks performed and monitor reduction in panic level in specially designed diaries, (iv) anticipate and deal with setbacks, and (v) recruit significant others as cotherapists if feasible. (Ghosh & Marks, 1987, p. 12, emphasis in original)

Several studies have tested the efficacy of bibliotherapy-directed situational exposure, when used as the primary intervention. Such bibliotherapy has been shown to be effective for many patients with panic disorder, and more effective than waiting-list controls (Ghosh & Marks, 1987; Jannoun et al., 1980; Mathews et al., 1977, 1981; McNamee et al., 1989). However, it may be ineffective for patients with severe agoraphobia (Holden et al., 1983). Such patients may require more extensive contact with a clinician, such as telephone-administered therapy or face-to-face treatment.

CBT2 Protocols A number of studies have examined CBT2 bibliotherapy as the primary intervention. None of the studies have looked at whether this treatment is effective for severe panic disorder. Nonetheless, the available results are suggestive. Hecker et al. (1996) compared self-directed and therapistdirected versions of the Albany protocol. In the self-directed condition, therapists met with patients three times over 12 weeks to simply assign readings and answer questions. Treatments were equally effective at the end of the 12 weeks, and at 6 month followup. Several studies have compared Clum's (1990) CBT2 bibliotherapy to other interventions, such as an 8 session clinician-administered version of the protocol described in Clum's self-help book. The clinician-administered version is similar to the Albany and Oxford protocols, although it places greater emphasis on the use of coping strategies (e.g., distraction) and less

302

COGNITIVE-BEHAVIOUR THERAPY:AN OVERVIEW

on the disconfirmation of catastrophic beliefs. Studies by Gould et al. (1993) and Lidren et al. (1994) suggest that the clinician-administered and bibliotherapy versions of Clum's treatment are equally effective, and both more effective than a waiting-list control. In contrast, Febbrano et al. (1999) recently found that Clum's bibliotherapy was no more effective than a waiting list. It is possible that this bibliotherapy is not as powerful as others. Clum's bibliotherapy was recently expanded to include (1) a short videotape containing a description of the etiology of panic disorder (including the role of catastrophic cognitions) and instruction in diaphragmatic breathing, and (2) an audiotape guiding the patient through a course of progressive muscle relaxation. The expanded version was shown to be as effective as therapist-led treatment, and more effective than a waiting-list control (Gould & Clum, 1995). Several other CBT2 bibliotherapy programs have been developed (e.g., Beckfield, 1998; Franklin, 1996; R.R. Wilson, 1996), although their efficacy remains to be demonstrated. Self-help materials are also available over the Internet, although it is not known whether they are helpful.

Comment Currently, the most promising bibliotherapies are the situational exposure guides developed by Marks (1978) and Mathews et al. (1981), Clum's (1990) CBT2 self-help book, and the patient workbook from the Albany protocol (Barlow & Craske, 1994). These are effective in treating many patients with panic disorder. Having a therapist available for periodic advice and encouragement appears to improve treatment adherence and outcome. Bibliotherapy may not be sufficient for patients with severe panic disorder, and may not be sufficient with complex cases, such as panic disorder comorbid with other serious psychiatric conditions. Bibliotherapy is also unsuited for patients with limited reading abilities or insufficient motivation to follow a largely self-directed program. If the patient undertakes a program in a half-hearted manner, then treatment is likely to fail. Bibliotherapy may be best suited for mild, uncomplicated cases of panic disorder in which the patient is well-motivated and can readily comprehend the written materials. For such patients, panic disorder could be treated with a stepped-care approach. Patients would initially receive the self-help materials, with periodic contact from the therapist (e.g., via telephone) to assess progress, to solve any difficulties, and to reinforce continued adherence to the program. If this fails to reduce panic disorder, or

J 1

l (

t

I

THEMEAND VARIATION:ALTERNATIVE WAYSOF DELIVERING TREATMENT303

if progress is too slow, then the next step is to implement a series of faceto-face sessions with the therapist in order to more directly help the patient implement the treatment procedures (Ghosh et al., 1988). Although the stepped-care approach seems economical, a problem with this approach is that if bibliotherapy fails, then the patient may become disheartened with treatment and assume that therapist-directed exposure or CBT2 also will fail (McMullin, 1986; Salkovskis et al., 1997). This problem may be averted by ensuring that the patient understands, at the outset, that if bibliotherapy fails then therapist-directed treatment is an important and effective backup.

Computerized

Treatment

With the increasing availability of cheap, powerful, portable computers and sophisticated software, clinical investigators have begun to explore the benefits of delivering treatment via computer. Effective computerized treatment holds the promise of being more economical and more readily available than treatment administered in face-to-face sessions with a clinician. Virtual Reality (VR) Therapy

VR exposure therapy has been used for treating some phobias, although work is still in the embryonic stages (Marks et al., 1998). It may be some time before VR protocols are available for treating panic disorder. VR driving simulators could be readily adapted as one component of this treatment. The simulators not only approximate situational exposure, but also induce arousal-related body sensations in many people, such as dizziness and nausea. (These sensations typically habituate with repeated VR exposure.) Thus, VR could be used to create a mild form of combined situational and interoceptive exposure. Driving simulators could be a useful first step for panickers who wish to overcome their driving fears but are too frightened to get behind the wheel of a real car. Palmtop-assisted Treatment

Another development in computerized treatment is to implement CBT2 by means of 'palmtop' computers. These small computers are easily carried by patients and used whenever they feel anxious or want to practice CBT2 techniques. Newman et al. (1997) developed a treatment package in which palmtop-assisted treatment was used in conjunction

304

COGNITIVE-BEHAVIOUR THERAPY:AN OVERVIEW

with brief CBT2. The latter consisted of 4 sessions with a live clinician, totaling 6h in all. The computer was carried by the patient at all times, and used for 8 weeks after completing the 4 sessions. The computer was used to prompt patients to practice breathing retraining and cognitive restructuring, and to keep track of evidence for inaccurate predictions. The computer also prompted patients to conduct interoceptive and situation exposure exercises (for details, see Newman et al., 1996, 1997). Compared to 10 sessions of CBT2 with a live clinician, the palmtopassisted brief CBT2 was less effective at posttreatment, but no less effective at 6 month followup. Although this suggests that palmtop computers may play a role in treatment, a limitation of Newman et al.'s (1997) study is that it does not tell us whether the palmtop contributed to treatment efficacy. It may be that their 4 session CBT2 protocol is just as effective as their 10 session protocol. Recall that Clark et al. (1999) found that 5 sessions of an abbreviated Oxford protocol ( administered without a palmtop) was just as effective as 12-15 sessions of the standard Oxford protocol. Palmtop computers may not be necessary. Computerized Bibliotherapy

Computer programs have been developed to deliver interactive bibliotherapy. These teach patients to construct exposure hierarchies and to apply strategies for coping with panic attacks and agoraphobia. Ghosh and Marks (1987), for example, developed a computerized version of the self-help book Living with Fear (Marks, 1978), which instructs patients how to conduct situational exposure. The computer program was found to be as effective as the book, and both were as effective as treatment delivered by a live therapist (Ghosh & Marks, 1987). Note that in both bibliotherapies a therapist was used to instruct patients how to work through the book or computer program. Without this initial direction, patients may not adhere to either intervention (Ghosh et al., 1988; Marks et al., 1998). Extending this work, Shaw et al. (1999) developed a computer program called Fear Fighter, which is a 12 session protocol that instructs patients how to set up and implement, in step-by-step fashion, a program of exposure to feared situations. Fear Fighter uses sound and photographs to accompany text-based screens. The program provides patients with feedback and printouts on their progress. Pilot studies by Shaw and colleagues suggest that the program is useful for treating panic disorder, although these investigators believe that contact with a clinician probably improves outcome.

L

THEMEAND VARIATION:ALTERNATIVE WAYS OF DELIVERINGTREATMENT 305

Comment Currently, the computerized treatments are best regarded as aids rather than replacements for live therapists. The caveats for computerized treatment are likely to be the same as those for bibliotherapy; computerized therapy may be best suited for mild, uncomplicated panic disorder, occurring in patients who are sufficiently educated and motivated to complete the program. As in bibliotherapy, there is the risk that if patients fail at computerized treatment, they may become disillusioned with all behavioural and cognitive-behavioural therapies, including therapistadministered versions.

Inpatient Treatment As we have seen, most innovations in CBT2 involve ways of abbreviating treatment and reducing therapist contact time. While many of these developments are important, different sorts of innovations are required for treating especially severe or complex forms of panic disorder. Pollard et al. (1987) observed that inpatient treatment is sometimes needed for panic patients who have medical, psychological, or situational problems that substantially impede the initiation or implementation of outpatient treatment. Such problems include severe (housebound) agoraphobia, suicidal behaviour, serious comorbid medical conditions, and chaotic home environments that discourage the patient from becoming mobile and autonomous. To overcome these obstacles, Pollard and colleagues developed a multidisciplinary inpatient treatment for panic disorder, integrating pharmacotherapy, CBT2, social skills training, and family therapy. These interventions were administered in a structured and strategically reinforcing environment. In this setting, complicating medical conditions could be safely evaluated and treated, and pharmacotherapies could be closely monitored when there were problems or concerns about medication abuse, nonadherence, or intolerance. Either treatment of the complicating problems was integrated into panic treatment, or problems were treated sequentially, depending on which problems required immediate attention. Nursing staff were trained to administer most of the psychological treatment. Patients received as much as 4-Sh/day of exposure, 7 days/week. This included situational, interoceptive, and imaginal exposure. Situational exposure involved a combination of therapist-assisted and selfdirected exposure. Staff members were trained to positively reinforce patients for completing therapeutic tasks. Reinforcers included verbal

306

COGNITIVE-BEHAVIOUR THERAPY: AN OVERVIEW

praise, increased privileges, day passes, or other things that the patient enjoyed (e.g., trips to the ice cream parlour). Reinforcement was faded out as treatment progressed, and patients were encouraged to take increasing responsibility for practicing on their own the skills they had learned. In an uncontrolled trial, Pollard et al. (1987) reported outcome data for 25 patients who completed this program. The mean length of inpatient admission was 35 days, followed by outpatient aftercare. Although patients had failed to respond to previous treatments, at the end of their inpatient stay, most (76%) were rated as significantly improved on measures of panic attacks and agoraphobia. A 3 year followup of 13 of these patients revealed that many (63%) had either maintained their gains or continued to improve (Pollard & Pollard, 1993). Maintenance of gains was associated with quality of outpatient follow-up and adequacy of family support. These findings suggest that intensive inpatient treatment is worth considering for severe, treatment refractory panic disorder. However, in an era of reduced funding for health care, such programs are unlikely to be implemented unless it can be shown-in controlled outcome studiesthat inpatient treatment is more effective than standard outpatient treatment.

THERAPIST TRAINING AND EVALUATION Although a good deal of work has been done on how to best treat patients with CBT2, comparatively little has been done on the best way to train therapists to deliver this treatment. Training typically consists of an apprenticeship, where novitiate therapists receive coursework and workshops on cognitive-behavioural therapies in general, and CBT2 for panic disorder in particular. Treatment manuals and other treatment-related writings contribute to this training. Training videotapes further help the trainee learn how to conduct treatment. Competency in CBT2 is further developed by applying this therapy under the supervision and guidance of an experienced therapist. Training can be facilitated by providing the trainee with feedback on their therapy sessions, obtained by directly observing the sessions (e.g., via a one-way mirror) or by audio- or videotapes of the sessions. The clinical supervisor can provide systematic, structured feedback by rating the trainees' therapy sessions on a competency checklist. Clark and Salkovskis (1987) devised such a measure, consisting of 17 items. Their checklist assesses general therapeutic skills (e.g., agenda setting, dealing with questions, clarity of communication, pacing and efficient use of

l

SUMMARYAND CONCLUSIONS 307

time), and the use of specific, appropriate techniques (e.g., review of panic diary, review of homework, use of feedback and summaries, use of Socratic dialogue, and selection of appropriate strategies for cognitive change and control of body sensations). Each item is rated on a 7-point scale, ranging from O (poor) to 6 (excellent).

SUMMARY AND CONCLUSIONS This chapter outlined the steps used in the Albany and Oxford protocols, which are state-of-the-art CBT2 packages. Although effective, these protocols can be improved by adapting them to the specific needs of the patient and the treatment setting. Treatment can be administered by telephone for patients who are unable to travel to the therapist's office. Treatment also can be delivered in the settings where panic patients first come for help, such as the hospital emergency ro.om. Recent findings suggest that abbreviated protocols are useful, and that some patients can be effectively treated by instructing them to read a self-help book or follow a computer program. However, for severe or complicated cases of panic disorder, it seems likely that a full protocol will be required, delivered in face-to-face sessions. Treatment manuals and other instructional aids are available for training therapists in these procedures, although direct supervision from an experienced CBT2 therapist is needed to ensure adequate training.

Chapter 12 . .

COGNITIVE INTERVENTIONS Cognitive restructuring of maladaptive beliefs 1 and images entails several methods applied in the context of a collaborative relationship between patient and therapist. Psychoeducation plays an important role, and patients are encouraged to view their beliefs as hypotheses rather than statements of fact. Many strategies have been developed for restructuring maladaptive cognitions in the treatment of panic disorder (e.g., Beck & Emery, 1985; Clark, 1989; Clark & Salkovskis, 1987; Craske et al., 1994; Salkovskis & Clark, 1991). Some of the most useful interventions are described in this chapter. Space limitations preclude a discussion of cognitive interventions for other disorders that may be comorbid with panic disorder, such as major depression. Several texts provide excellent descriptions of cognitive interventions for those disorders (e.g., Beck & Emery, 1985; Beck et al., 1979, 1990; Hawton et al., 1989; Wells, 1997). Instead, this chapter will focus on strategies for restructuring catastrophic cognitions in panic disorder. These include beliefs or images about the dangerousness of arousal sensations (e.g., 'Palpitations lead to heart attacks') and about the danger associated with agoraphobic situations (e.g., 'If I panic while driving, I'll crash the car'). The chapter begins with a discussion of Socratic dialogue because this method can be used for administering most cognitive interventions, including psychoeducation.

SOCRATIC DIALOGUE What is Socratic Dialogue and Why is it Important? Although many writers have summarized the Socratic method, few have described it in a way that trainee therapists can readily learn. This is why

1

For ease of exposition, the term 'beliefs' will be used as a shorthand term encompassing beliefs, assumptions, attitudes, and interpretations.

SOCRATICDIALOGUE 309

we will devote some space to describing the aims, components, and processes of this important approach. The following discussion draws on the work of Overholser (1993a,b, 1994, 1995, 1996) to show how Socratic dialogue can be used in the treatment of panic disorder. Socratic dialogue is a form of guided questioning that helps patients examine the validity and usefulness of their beliefs, and helps them arrive at more adaptive beliefs. Socratic dialogue can be used in all stages of treatment, from the initial session onwards. In contrast to the lecture approach, in which the patient is the passive recipient of information provided by the therapist, the Socratic approach induces patients to do most of the work in questioning their beliefs and coming up with alternatives. The goal is not to provide patients with all the answers, but instead to help them think for themselves. This approach improves retention of material discussed during therapy (cf. Anderson, 1990). Thus the Socratic method can be more effective than simply lecturing the patient. Like all powerful interventions, however, Socratic dialogue can be misused. Excessive questioning can make the patient feel interrogated. Socratic questioning is best used in short sequences, interspersed with other kinds of dialogue. The latter include (1) capsule summaries of the material discussed so far, provided by either the patient or therapist, (2) short sequences in which the therapist provides information, and (3) simply letting the patient recount some incident (e.g., a panic attack) while the therapist engages in reflective listening. The following sections will describe and illustrate the three elements of Socratic dialogue: inductive reasoning, use of definitions, and systematic questioning. Systematic questioning will receive particular attention because it is the vehicle by which the patient's reasoning and definitions are examined and restructured.

Inductive Reasoning Inductive reasoning involves drawing general inferences from specific experiences. Reasoning errors can lead to maladaptive beliefs. Accordingly, Socratic dialogue is used to examine the reasoning behind the patient's beliefs. The following questions illustrate how patients are prompted to examine the logic underlying their catastrophic beliefs. • •

How does trembling lead to insanity? If your symptoms are due to a brain tumor, then why would they come and go?

310

•

COGNITIVE INTERVENTIONS

If your chest pain is caused by heart disease, then why would distraction make the pain go away?

There are numerous errors of inductive reasoning, such as overgeneralization and selective abstraction. These errors have been described in detail by Beck and colleagues (e.g., Beck & Emery, 1985; Beck et al., 1979). To identify and correct reasoning errors, the therapist can provide the patient with a list of common errors (e.g., Burns, 1981, pp. 40--41). The therapist also can ask patients to look for evidence for and against their beliefs, and to evaluate the quality of the evidence. Specific strategies are illustrated throughout this chapter.

Use of Definitions Arriving at definitions is a good way of identifying reasoning errors such as all-or-nothing thinking and overgeneralization. Patients are asked to define personally important concepts. For people with panic disorder these typically include concepts such as 'losing control', 'going crazy', and 'unable to cope'. To arrive at definitions, patients are asked to view their problems from a broad perspective that goes beyond the specifics of their life circumstances. For a patient who viewed herself as helpless, for example, the process of defining 'helpless' led her to realize that she has absolutely no control over some things (e.g., the weather), but a great deal of control over other things (e.g., the situations she chooses to avoid), and partial control of yet other things (e.g., her panic attacks). The process of defining 'helpless' revealed to the patient that she had been mislabeling herself as 'completely helpless', which in the past had led her to give up all attempts to overcome her panic attacks and agoraphobia. With the realization that she had partial control over her problems, she became more motivated to work at overcoming them.

Systematic Questioning Systematic questioning is used to examine and correct the patient's reasoning, and thereby replace maladaptive beliefs with more adaptive ones. The therapist can use many different kinds of questions to lead the patient to engage in different types of thinking. The following are some examples. Interpretation questions help the patient discover relationships among two or more things, such as relationships among stimuli and themes shared by ideas. To illustrate:

L

SOCRATICDIALOGUE 311

Therapist (T): What would you think if you suddenly felt dizzy? Patient (P): I'd worry I could pass out. T: And what would you think if, instead, your heart suddenly started pounding? P: That would really scare me. I'd worry I was having a heart attack. T: OK, so we have two interpretations; an interpretation of dizziness and an interpretation of a pounding heart. How are these similar to one another?

As the dialogue continued the patient realized he had a tendency to catastrophize about many kinds of body sensations. This helped him realize that his problem wasn't the occurrence of body sensations, rather it was the meaning he attached to them. Application questions ask patients to apply information or skills to specific tasks or problems; e.g., 'How could you test that belief?' Application questions include just enough direction to ensure that the patient can identify the steps in solving a particular problem. Analysis questions ask the patient to solve a problem by breaking it into parts; e.g., 'What kind of thinking might have led you to panic while you were waiting for the bus?' Synthesis questions encourage problem solving through the use of creative or divergent thinking. The questions should suggest many possible solutions; 'What else might have caused you to feel "spacey" [depersonalized]?' Questioning is unlikely to be fruitful if the discussion wanders aimlessly from topic to topic. A more productive approach is to identify a maladaptive belief that seems to play an important role in the patient's problems, and then to systematically challenge the belief and replace it with a more adaptive belief. To achieve this goal it is useful to follow the following steps in the questioning process (Overholser, 1993a): 1. 2. 3. 4.

guiding questions; the explication; the defense; sequential progression.

Guiding Questions These contain an implied assumption (i.e., a 'correct' answer), such as the assumption that arousal-related sensations are not dangerous. However, guiding questions often offer two alternatives; e.g., 'Do you think your

312

COGNITIVE INTERVENTIONS

panic attacks are dangerous, or are they simply unpleasant?' Alternatives are presented so that (1) the patient can think about which alternative is likely to be most accurate, instead of simply providing a 'yes' or 'no' response, and (2) the patient does not feel unduly compelled to agree with what the therapist says. The various sorts of questions described earlier in this chapter can be used as guiding questions (interpretation questions, application questions, etc.). The Explication

f

This occurs when the patient does not understand the guiding question or gives a 'don't know' response. The therapist provides an explication, which can involve rephrasing the question to make clearer the implied assumption; e.g., 'If panics were dangerous, what do you think would happen after an attack? ... And what actually does happen after you panic?' If the therapist is asking questions at the right level of difficulty, then explications should occur only occasionally.

• • •

The Defense

•

Here patients reflect on the accuracy of their beliefs. 'Why' questions can be used to help the patient reason through this process. The therapist should conduct the questioning so that the patient does not feel interrogated, distrusted, or demeaned. Questions should be asked in the spirit of mutual discovery; e.g., 'To help me understand things, tell me why you believe that panic attacks could cause your heart to stop.' It can be unpleasant for patients to recognize errors in their thinking, and so the questioning should be laced with compassion (Overholser, 1995).

• •

Sequential Progression

This occurs when additional logue closer to the intended your heart will stop? ... Do mentioned, Socratic dialogue with non-Socratic dialogue.

guiding questions are used to move the diagoal; e.g., 'What evidence do you have that you have any evidence to the contrary?' As is best used in short sequences, interspersed

•

•

Comment

•

Socratic dialogue is a style of guiding the patient's thinking. Many of the cognitive interventions discussed later in this chapter are based on this approach. Although Socratic dialogue should not be overused, it is usefully included in most aspects of treatment, including psychoeducation.

•

I

l

PSYCHOEDUCATION 31 3

PSYCHO EDUCATION Patient education, also known as psychoeducation, is the first step toward correcting misconceptions and filling information gaps that the patient has about the nature and treatment of panic disorder. Treatment protocols differ in the amount of information they present to patients. The Albany protocol (Barlow & Craske, 1994; Craske et al., 1994) provides patients with a great deal of information, probably more than necessary. Patients may feel overwhelmed if they are given too much information. It is useful for the patient to receive the information summarized in Table 12.1, which Table 12.1

• • • • •

•

• • • • •

•

Psychoeducation:

useful information to convey to patients

Panic attacks and panic disorder are not fatal maladies, nor do they lead to psychosis. Panic attacks are very common in the general population, and panic disorder is one of the most common anxiety disorders. This information can help the patient feel less isolated and stigmatized. Distinction between adaptive and maladaptive emotions: panic and anxiety are often helpful in alerting the person to danger. However, recurrent unexpected panic attacks are maladaptive 'false alarms'. People have 'noisy bodies'-we all commonly experience arousal-related sensations. (The therapist should describe some of the causes of these sensations.) Panic attacks aren't an inevitable consequence of these sensations; everyone experiences arousal-related sensations, but only some people develop panic disorder. The tendency to make catastrophic misinterpretations appears to contribute to panic attacks. Provide patient with a working hypothesis of his or her problems. This will typically include an explanation of panic attacks as arising from catastrophic misinterpretations of benign, arousal-related sensations. (Illustrate how this can account for the patient's panics.) Agoraphobic avoidance, other safety behaviours, and abuse of anxiolytic substances (e.g., alcohol) are common ways of avoiding feared sensations. These avoidance strategies reduce arousal in the short term, but create problems in the longer term. (If applicable, review how avoidance has created problems in the patient's life.) CBT2 interventions can effectively reduce panic disorder. (Briefly describe some of the interventions and explain how they work. Emphasize the collaborative nature of treatment.) Provide information about the requirements of treatment; e.g., use of reading materials, self-monitoring of panic attacks, importance of regularly practicing exposure exercises. Offer information about treatment side-effects, such as transient increases in panic or anxiety when severely agoraphobic patients begin situational exposure assignments. (This information is presented in order to limit the likelihood that patients will catastrophize about side effects.) Address any other questions raised by the patient, such as the role of biological factors.

314

COGNITIVE INTERVENTIONS

includes the information conv,:yed in the Oxford protocol (Clark, 1989; Clark & Salkovskis, 1987). This corrective information can be presented during the first few sessions, along with patient's handouts and other materials for the patient to read between therapy sessions. The information can be revisited as needed throughout treatment.

Socialization

Strategies

Corrective information can be introduced by means of socialization strategies, which are also used to engage patients in therapy. The strategies are intended to demonstrate the role of cognitive factors in panic disorder, and how second generation cognitive-behaviour therapy (CBT2) can help them. The following are the commonly used socialization strategies. Using Clark's Cognitive Model to Explain Recent Panic Attacks To illustrate Clark's (1986) cognitive model of panic, the therapist can sketch out the sequence of events that occurred during one of the patient's recent attacks. As a start, the therapist can introduce the vicious cycle, showing the links between arousal-related sensations, catastrophic beliefs, and feelings (anxiety, panic attacks).

T: What was the first sensation you noticed when you began to panic? P: I felt really dizzy. T: OK, and what did you think about that? P: I worried I could pass out. T: How did that make you feel? P: It made me really anxious. T: When you felt anxious, what was happening in your body? P: My heart was pounding and I began to shake. T: And what did you think about that? P: I was sure that something really bad was happening. I thought I was going to die. T: How did that make you feel? P: By that time I was panicking really badly. T: It sounds like a very intense experience. Notice how your body sensations led to frightening thoughts, which led to anxiety, which led to more sensations, and so on in a vicious cycle.

l

PSYCHOEDUCATION 315

Sensations (dizziness, palpitations)

Anxiety or Panicky Feelings

Thoughts ('I'm going to pass out,' 'Something really bad is happening; I could die')

Figure 12.1. Patient education: using a recent panic attack to illustrate the relationships among arousal-related sensations, catastrophic misinterpretations, and emotions •• •

The therapist then sketched on a chalkboard the vicious cycle for this panic attack, as shown in Figure 12.1. Once the patient understands the concept of the vicious cycle, the next step is to illustrate the role of triggering stimuli, thereby illustrating how Clark's full model can account for panic. Identifying the triggers can make the attacks more understandable and less threatening (Clark & Salkovskis, 1987). The full model is illustrated in Figure 12.2, using labels that the patient could readily understand. Patients should sketch out other examples of their panic attacks to show how the attacks can be explained in terms of the model. A variation on this approach is for the therapist to show how the model accounts for one of the patient's cued panics, and then ask the patient to explain how the model could account for an unexpected panic. With a little bit of prompting, patients are often able to come up with the idea that a body sensation, catastrophic image, or other event may have been the triggering stimulus. This can then be taken a step further. The patient can be asked to explain how the model can account for nocturnal panics (if the patient has experienced such attacks). Patients can often successfully apply the model to such attacks, particularly if they are helped to understand that people monitor their external and internal environments while they sleep. A parent might sleep through the noise of outside traffic, but selectively awaken to a personally important stimulus, such as the cry of one's child. In the same way panickers may detect body sensations while they sleep, which may lead them to awaken and catastrophically misinterpret the sensations.

31 6

COGNITIVE INTERVENTIONS

Trigger Stimulus: Internal or External (Dizziness caused by standing up quickly)

Perceived Threat ('I could pass out')

Catastrophic Misinterpretations

Anxiety or Panicky Feelings

('Something really bad is happening; I could die")

Sensations (Palpitations, stronger dizziness)

Figure 12.2. Applying Clark's (1986) full cognitive model to further illustrate the mechanisms of panic

For patients who express doubts about their ability to monitor events while they sleep, the following strategy can be useful (Paul M. Salkovskis, personal communication, 1997). Patients can be asked whether they always wake up in the same position in which they fell asleep. Typically, the answer is 'no'. The therapist can ask how this could be. Patients typically report that they move during sleep as a result of a number of factors, including the desire to become comfortable. The therapist then can ask the patient how they know they're uncomfortable while they're asleep. The ensuing discussion should lead the patient to understand that they monitor their bodies even while they sleep. Sleeping on one's side for several hours, for example, can lead to pain or discomfort in the arm, which the brain detects and corrects by shifting sleeping position (e.g., rolling over to the other side). Other examples of detecting sensations during sleep include waking up in the middle of the night to urinate (detection of bladder sensations while asleep) and waking up to get a glass of water (detection of thirst sensations). Experiential Methods

Another socialization strategy is to ask patients to complete tasks that illustrate how cognitions lead to anxiety or panic attacks (Clark, 1989; Wells, 1997). These can be completed during the first session or two. The tasks tend to elicit anxiety rather than panic, partly because the therapist and therapy setting are powerful safety signals. Nevertheless, the tasks

PSYCHOEDUCATION 317

are usually sufficient to illustrate the role of cognitive factors. Examples are as follows: •

• •

•

•

Examine changes in emotion occurring in the therapy session. When the patient appears to become increasingly anxious while discussing feared sensations, the therapist can explore what thoughts or images might have occurred ('I notice that you seem upset. What went through your mind just then?'). Patients can be asked to read a list of word pairs (e.g., 'breathless-suffocate,' 'palpitations-dying') (Clark, 1989). Many panic patients find this anxiety evoking. This task shows how anxiety can be induced by purely cognitive methods, and also can be used to educate the patient about the role of catastrophic beliefs in producing panic. A similar approach is to ask patients to vividly imagine anxietyevoking scenarios. They might be asked sit back with their eyes closed and imagine their worst panic attack. After imagining the scenario, patients are asked to scan their bodies and report any emotions and sensations induced by the exercise. Typically patients feel anxious and experience arousal sensations (e.g., palpitations, chest tightness). This exercise is useful for three purposes. First, it shows that anxiety or panicky feelings can be elicited by a purely cognitive means. Second, it helps patients understand that arousal sensations are a normal part of the anxiety response. Third, the exercise can help patients identify their catastrophic beliefs. A focus of attention exercise can be used to demonstrate how the direction of one's attention can influence the likelihood of experiencing sensations (Clark, 1989). This exercise-which is actually a form of 'behavioural experiment' (Chapter 13)-is particularly useful for patients who think they experience more sensations than other people, and believe this indicates that something is seriously wrong with them ('I'm more aware of my heart than most people. This means I have an abnormal heart'). In this exercise the patient is asked to close their eyes and focus on the body for a few minutes, and then describe any sensations arising. Then the patient is asked to look around the therapist's office and describe, out loud, the contents of the room. Then the patient describes whatever sensations were experienced. Patients typically report that they notice more body sensations when they focus on their bodies than when their attention is directed outwards. This exercise can help patients understand that the more frequently they focus on their bodies, the more often they will notice body sensations. In turn, this can help them understand why they seem to experience more body sensations compared to other people.

318

•

COGNITIVE INTERVENTIONS

Brief interoceptive exposure exercises, such as a 2min hyperventilation exercise, also can be used to assess the relationships between sensations and thoughts (see Chapter 9). Patients are presented with little information before attempting the exercise. They are simply told that the task is harmless, and that they may experience some body sensations. After completing the exercise, the patient and therapist review the patient's thoughts (interpretations) about the sensations, along with a review of how the interpretations may have influenced any resulting emotions, such as anxiety or panic.

Not all socialization strategies work for all patients, and so it can be useful to try a number of approaches. These can be introduced as tests to help the therapist better understand the patient's panics. Socialization strategies are among the first steps in the process of cognitive restructuring, and are usually sufficient to engage the patient in treatment. As such, they set the stage for interventions used later in therapy.

Troubleshooting Contradictory Advice Sometimes patients are given advice by their primary care physician or other health care professionals that contradict what is taught in CBT2. Contradictory advice is not only confusing to the patient, but also can interfere with CBT2 and perpetuate panic disorder. For instance, Alexander (1991) offered the following advice, which could prevent patients from learning that arousal-related sensations are harmless. Patients should be informed that caffeine and other stimulants can aggravate anxiety and cause panic attacks. Therefore, panic patients should be told to avoid the consumption of caffeine, in drinks such as coffee, tea and soda, and in large amounts of chocolate. (p. 330)

Contrary to this admonition, the CBT2 therapist would recommend that patient consume caffeinated foodstuffs as a way of learning that the resulting sensations are harmless. If patients are receiving CBT2 from one therapist and another treatment (e.g., pharmacotherapy) from another therapist, then it is important that the clinicians do not contradict one another with the advice they give. Periodic communication among the treating clinicians is important for achieving this goal. If this does not succeed, then another option is to suspend one treatment (e.g., CBT2) until the patient has completed a course of the other treatment.

%

f

PSYCHOEDUCATION 31 9

Strong Beliefs in Biological Causation

The therapist should openly discuss any doubts the patient has about the cognitive model, including any perceived discrepancies between the model and the patient's panic attacks (Clark & Salkovskis, 1987). Most patients are willing to explore the possibility that their panics are a product of their thought processes. However, some patients strongly believe that their panics are entirely due to a 'biochemical imbalance', over which they have no personal control. These patients typically believe that panic disorder can only be treated with drugs. Such beliefs can lead to CBT2 nonadherence and treatment dropout, and are best addressed early in therapy. To this end, it is useful to inquire, early in treatment, about the patient's beliefs about the causes of panic, and to correct any misconceptions. Craske et al. (1994) recommend dealing with this problem by informing patients about the research on the role of cognitive factors in panic disorder, and by reviewing the effects of safety signals and distraction on their panic attacks. If panic were simply the misfiring of some 'purely biological' system, then the frequency of panic would not be reduced by, say, the presence of safety signals such as a cellular phone or pill container. A discussion of these factors can be very helpful in engaging the patient in treatment. The following is another strategy, which uses Socratic dialogue to show patients that cognitive factors are important, even if biological factors contribute to their panics.

T: What do you think causes your panic attacks? P: My doctor tells me I have a biochemical imbalance that requires medication. T: Has medication cured your problem? P: Well no, it's helped a little but not much. T: So why do you think your doctor suggested you come see me? I use psychological therapy not drugs to help people overcome their panic attacks. P: I never thought about it. Maybe there's a psychological component to my problems. But I don't see how it could help. My panic attacks are physical-they happen in my body. T: Do you think that psychological processes, such as your thinking, could influence your body? P: I don't see how. T: Let's explore this possibility. Can you think of a time recently when you craved something really bad, like an intense craving for a particular sort of food?

320

COGNITIVE INTERVENTIONS

P: Well, yeah. I'm trying to lose weight, so I get cravings all the time. I walked past a pizza shop the other day. The sight and smell of pizza nearly drove me mad. T: What happens when you vividly imagine eating a pi=a? I'd like you to spend a few moments imaging yourself eating your favourite kind of pizza ... [The therapist then asked the patient to provide a vivid description of eating, including information on visual, olfactory, tactile, and taste sensations.] P: Hey, I'm starting to salivate ... and my stomach is churning! T: So what's happening here? What's your thinking doing to your body? P: I guess my thoughts can influence my body.

Other sorts of evocative imagery-such anger, or other arousing emotions-also effects of cognitions on body sensations.

as imagery evoking anxiety, can be used to illustrate the

Catastrophizing About the Cognitive Explanation

While educating patients about the role of cognitive factors, it is important to assess what patients make of this information. Some patients are relieved to learn of what might be causing their attacks (i.e., catastrophic cognitions). Other patients catastrophize about this information, taking it as evidence that they're going crazy. If the latter occurs, it can be addressed by an intervention such as the following:

P: So from what you're telling me, I have something wrong with my thinking. [Patient starts to cry.] T: It looks like it's really upsetting for you to hear that ... Tell me, what are you thinking? [Therapist begins searching for catastrophic beliefs.] P: So many doctors have told me there's nothing physically wrong with me. It makes me wonder whether I'm losing my mind. And now you tell me it's true. T: You're not losing your mind-even if you tried to make your panics as bad as they could be, you wouldn't lose your mind. [Therapist provides corrective information. Patient starts to calm down.] T: Think of it this way. Have you ever believed something that later turned out to be untrue? For example, did you ever come out of an exam convinced you'd failed, only to discover later on that you'd actually passed? [This scenario was selected because the therapist

GOAL SETTING 321

P:

T: P:

T: P:

T: P:

T: P:

learned during the pretreatment assessment that the patient was anxious about her performance at university, but had always obtained good grades.] Well, yes, it happened last semester. I was sure I'd failed my statistics course. And did you fail? Um, no, I got an A minus. [Smiles sheepishly.] So there you had a false belief; the belief that you'd failed. Were you losing your mind because you had that belief? Well, no, I was just thinking the worst. That's right. Now let's compare 'exam thinking' to 'panic thinking'. Do you tend to think the worst while you're having a panic attack? I can see what you're getting at. During my really bad panics I always think I'm having a heart attack. I guess I'm probably wrong in thinking my heart is going to stop. [Patient begins to see that the problem is one of mal-learning, not madness.] So panic thinking arises from mistaken beliefs; from having learned to be frightened of sensations that are really harmless. Does that mean you're losing your mind? No, I guess not.

GOAL SETTING Goal setting is a cognitive intervention in which patients are encouraged to make realistic, attainable choices about what they would like to achieve from therapy. Goal setting best occurs after the patient has been socialized into treatment, because patients are likely to make better-informed choices about their goals once they have been educated about the probable causes and cognitive-behavioural treatment of panic disorder. Patients will have learned, for example, that anxiety and panic attacks are normal, adaptive reactions that sometimes become excessive. This will help patients understand that permanent freedom from panic attacks is an inappropriate goal. Patients choosing this goal may do so because they believe panics are dangerous. Psychoeducation about the cognitive factors in panic can help patients understand how their catastrophic thinking can lead them to select unrealistic goals. Appropriate goals are those that are specific, unambiguous, and measurable. That way both the therapist and patient know when the goals have been attained. Treatment goals include those pertaining to outcome (e.g.,

322

COGNITIVE INTERVENTIONS

reducing the frequency of panic attacks) and those to do with performance (e.g., appropriate practice of homework assignments). Specific examples of appropriate goals are as follows: • • • • •

to practice my interoceptive exposure exercises each day; to reduce the frequency of my panic attacks from 4 per week to no more than 1 a month; to reduce the number of hours I spend worrying about panic from Sh/day to less than lh/day; to be able to drive over the Lion's Gate Bridge during rush hour; to discontinue my anxiety medications.

Multiple goals should be ranked in order of importance. Goal setting is facilitated by teaching the patient to take each long-term goal (e.g., 'Taking the bus all the way to work') and to identify all the steps required to attain that goal. These can be medium-term goals (e.g., 'Taking the bus half the distance to work') and short-term goals (e.g., 'Taking the bus for 1 block'). Reducing goals to a series of subgoals teaches patients how to break problems down into manageable steps.

VERBAL DISPUTATIONS FOR CORRECTING CATASTROPHIC BELIEFS Verbal interventions help patients identify and objectively evaluate their catastrophic beliefs, and to generate plausible, noncatastrophic alternatives. These interventions provide patients with tools for use during therapy assignments and in their daily lives. Tools consist of questions they can ask themselves, statements of noncatastrophic beliefs, and short lists of evidence for and against particular beliefs. Each of these can be written on flashcards that patients carry with them.

Distancing

Strategies

Panic patients tend to strongly believe their catastrophic beliefs, particularly when they are feeling anxious or panicky. Distancing strategies (Beck & Emery, 1985; McMullin, 1986) are methods for helping patients view their beliefs objectively, as hypotheses rather than facts. There are several kinds of distancing strategies. •

Patients can be asked to record the sequence of events during their panic attacks; objectively observing, as if they were scientists,

VERBALDISPUTATIONSFOR CORRECTINGCATASTROPHIC BELIEFS 323

•

•

•

the sequence of symptoms, thoughts, feelings, and behaviours. The task is simply to observe and describe these events, without evaluating them as good or bad, and without trying to influence them. Perspective-taking exercises: Patients are asked to examine their catastrophic beliefs as if they belonged to someone else. Example 1: 'Imagine a friend told you that every time she felt dizzy she was worried about going crazy. What would you advise her?' Example 2: 'Imagine that I (the therapist) suddenly noticed my heart pounding. What would I probably think about that?' In vivo labelling of cognitions: Here, patients are taught to label their beliefs. For example, if a patient believed 'I'm going to lose control if I stay here any longer,' then the following self-statement could be used: 'I'm simply having another catastrophic thought. I don't need to take it seriously.' This intervention is particularly useful for patients who have too many different catastrophic cognitions to individually challenge. The patient also can be asked to write out all the catastrophic beliefs that occurred during their panic attacks in the past few weeks. This information could be collected from the panic attack records (Chapter 9). The patient and therapist then review the list to look for themes representing underlying or core beliefs. The evidence for and against the beliefs is examined, along with evidence for alternative, noncatastrophic beliefs. The following is a list of beliefs reported by one patient over a period of several weeks, along with the core belief identified by the patient and therapist:

Beliefs: • • • • • •

My nasal infections could spread to my brain, resulting in a brain tumor. My heart will stop if my pulse becomes rapid. Dizziness means I could be going crazy. Nausea means that something very bad is about to happen to me. If I get too anxious this feeling might never stop. Chest tightness means that I could suffocate.

Core belief: •

My health is extremely fragile.

324

COGNITIVE INTERVENTIONS

Empirical Disputes: Reviewing the Evidence Empirical disputes are exercises in which the patient and therapist review the available evidence for and against the patient's catastrophic beliefs. Evidence is collected from the patient's own past experiences and from other sources. For each catastrophic belief, the therapist and patient develop an innocuous, plausible, and accurate alternative belief statement (Salkovskis & Clark, 1991). The evidence for and against each belief is then listed. There are several questions that patients can ask themselves to facilitate this process: • • •

What evidence do I have for this belief? Are my beliefs based on facts? Is there another explanation or alternative way of looking at things?

The example in Table 12.2 illustrates the evidence for and against a catastrophic belief and its noncatastrophic alternative. Once a list like this is generated, the patient and therapist review the evidence to decide which belief is best supported. To determine whether this exercise is persuasive, the patient can be asked to rate the strength of the catastrophic and noncatastrophic beliefs before and after reviewing the evidence. A 0-100 scale can be used, ranging from 0 = do not believe that_ is true, to 100 = completely believe it is true. If the exercise is effective, then it should produce a substantial drop in the strength of the catastrophic belief, and a corresponding increase in the strength of the noncatastrophic alternative. The key points from the exercises can be distilled into a pithy coping statement written on a card, which the patient carries and reviews as needed. For the example in Table 12.2, the statement might be 'My chest pain is due to harmless muscle tension.'

Challenging Beliefs by Reviewing the Sensations To develop realistic interpretations of feared sensations, the patient and therapist can review the differences between the symptoms of potentially hazardous events versus the symptoms of harmless events. The patient's personal experience can be used to make this distinction. Ms D. had a history of recurrent panic attacks in which dizziness was the sensation she feared the most. She catastrophically misinterpreted dizziness as a sign that she was about to faint, which (she believed) would cause her to fall, hit her head, and die. A noncatastrophic alternative belief was that the dizziness was due to stress-induced acute hyperventilation, which would not lead to fainting. To discriminate between the symptoms of

VERBALDISPUTATIONSFOR CORRECTINGCATASTROPHIC BELIEFS 325

Table 12.2 Review of evidence for and against a catastrophic belief and its noncatastrophic alternative Evidence for the belief

Catastrophicbelief:My chest pain indicates that there is something seriously wrong with my heart

•

•

Heart disease runs in families; my father died of a heart attack and therefore I could have something wrong with my heart. The pain is in my chest and arms, which is what happens when people have heart attacks.

Evidence against the belief •

•

•

•

Noncatastrophicalternative belief:My chest pain is due to harmless, stressrelated tension in my chest muscles

•

The pain tends to get worse after I've been feeling stressed out.

•

The pain eases off when I'm focusing on things I enjoy, such as when I'm watching a good movie. The pain gets worse when I worry about my heart. Worry causes muscle tension, not heart attacks.

•

•

I've had extensive medical evaluations. All the tests indicate that my heart is healthy. Biology is not destiny. A family history of heart disease does not mean that I will develop heart problems. If the pain is due to heart disease, then it should get worse when I exert myself. Yet, my pain tends to go away when I exercise. Pain location is not a reliable indicator of its cause. All sorts of things can cause pain in the chest and arms, including muscle spasms. Sometimes the pain occurs even when I'm not feeling stressed. (But the pain could be a carryover effect from a prior stressful episode.)

hyperventilation versus fainting, Ms D. was asked to recall an episode in which she had fainted, and then to compare this experience to the way she felt when she was dizzy and panicking. The two differed primarily in that a feeling of sleepiness occurred just before she fainted, but never occurred when she was dizzy and panicking. A further distinction was that palpitations occurred when she was dizzy and panicking, but not on the occasion when she actually fainted. This information helped Ms D.

326

COGNITIVE INTERVENTIONS

develop more accurate beliefs about her dizziness (e.g., 'If my heart is pounding and I'm not feeling sleepy, then I'm not going to faint'). This intervention was supplemented by corrective information. The therapist informed her that fainting is associated with a drop in blood pressure, whereas panic is associated with an increase in blood pressure. Thus, if her dizziness occurred while she was anxious or panicky, her blood pressure would be increased, not decreased (Clark, 1989).

Challenging

the Components

of Catastrophic Beliefs

Catastrophic beliefs can be challenged by breaking them down into components, such as overestimations of the probability and overestimations of the cost (badness) of unpleasant events. The components are then challenged separately. Whenever Mr C. felt dizzy in a department store, he overestimated the probability of collapsing. He also overestimated how bad it would be if he did collapse. Mr C. believed that if he did collapse, then (1) he would definitely injure himself, and (2) everyone would think he collapsed because he had some terrible disease (e.g., HIV), and would avoid him thereafter. There are several strategies for challenging these overestimations (Craske et al., 1994). The therapist can guide the patient to examine questions such as these: • • •

•

Arn I overestimating how likely_ would be? Has it happened in the past? Realistically, how bad would it be if_ occurred? Arn I overestimating how bad things would be? Arn I overestimating how long_ would last? Can I cope with a few minutes of _ in my day? (These questions are particularly helpful for patients who fail to realize that intense arousal sensations or panic attacks are short-lived events.) Arn I exaggerating how uncomfortable_ would be? Have I been able to tolerate discomfort in the past? (These questions are useful when patients catastrophize about the unpleasantness of body sensations or panic attacks.)

Another useful strategy for challenging overestimations is to evaluate whether the patient's belief in the probability of a feared event is consistent with his or her past experience (Craske et al., 1994). This is illustrated in the following transcript.

VERBALDISPUTATIONS FOR CORRECTINGCATASTROPHIC BELIEFS 327

T: Based on all the evidence, what are the odds that the next time you P:

T: P:

T: P:

T: P: T: P: T: P: T: T:

P:

felt 'spaced out' [the patient's term for depersonalization] you'd wind up going crazy? The odds aren't high, but I'm still worried about it happening. I guess the chances are 100 to 1 . Would you buy a lottery ticket if the chances of winning were 100 to 1? Yes, I would. So 100 to 1 are pretty good odds. Yes, especially in the long run. It sounds like you believe that the odds of going crazy are also fairly high. That's why the fear of going crazy is always at the back of my mind. Tell me, how many times in your entire life have you felt 'spaced out'? What's your best guess? It's happened almost every day for the past 20 years .... I guess it's happened about 6000 times. If the odds were 100 to 1, that would mean you've gone crazy 60 times. Is that how many times it's happened? Well no, I've never gone crazy. So does that mean the odds of going crazy are much lower than you thought, like less than 1 in 6000? [Patient nods.] I wonder if it would be worth writing this down on a card, so that you could remind yourself of the odds the next time you felt spacey. Yes, I think that would be helpful.

Certain beliefs are better challenged by some strategies than others. For example, the belief 'I will have a heart attack and die if my heart beats too quickly' is better challenged by questioning the probability than the cost of the event. Conversely, the belief 'I could collapse if I feel dizzy' may be better challenged by questioning the cost rather than the probability (Otto et al., 1996a). For patients expressing fears of dying, the therapist can explore their beliefs about what will happen after death. This can sometimes reveal what are probably overestimations of the badness of death. Some patients believe they will remain aware of their surroundings and body after death (Wells, 1997); e.g., 'I'll be all alone in the cold earth, as my body decom-

328

COGNITIVE INTERVENTIONS

poses'. To challenge such beliefs, patients can be asked about the evidence they have that this will occur, and whether there are less aversive possibilities of what will happen when they die (e.g., eternal sleep, Heaven, reincarnation as a puppy).

]:;

t I

Challenging the Demand for Certainty Even when the odds of an aversive event are very low, patients may still worry about the possibility of it happening. For example, Mr J. worried that one day his chest pain might be due to a real heart attack instead of harmless spasms in his chest muscles. Of course, his doctors could not guarantee that this would never happen. In these cases it can be useful to assess whether the patient has an unrealistic demand for certainty (Craske et al., 1994). The latter is reflected in beliefs such as 'I can never relax so long as I know that_ could happen'. To challenge demands forcertainty, the patient can consider questions such as these: • • •

Is it useful for me to worry about _ or are my worries spoiling my life? What is so bad about the uncertainty? Am I making a mountain out of a molehill? Have I learned to tolerate other uncertainties? If so, then I can learn to tolerate this one.

The patient and therapist can review the daily low-probability 'risks' that the patient already takes; e.g., using electric appliances (a potential source of shocks), using scissors (a potential source of wounds), and walking about their apartment (where one could possibly be hit by meteors falling through the roof). These examples can help patients learn that they already tolerate all kinds of uncertainties, and therefore can learn to accept other low-probability uncertainties.

Devil's Advocate Method Once the patient has become familiar with the process of questioning beliefs, the therapist can consolidate belief change by using the Devil's Advocate method. This is a form of role play in which the therapist voices one of the patient's catastrophic beliefs and the patient is asked to come up with evidence and arguments to refute the belief. The therapist then comes up with counter-arguments, which the patient tries to refute. Then the patient voices a noncatastrophic alternative to the catastrophic belief, which the therapist attempts to challenge. The patient's task is to defend this

a

1 b d a

p tl s

4·

s

V

!

Vi b£ UJ

d1

U]

p,

bE bE ar di

VE

ra Tl ti,

P, di

Ir

ar

VERBALDISPUTATIONSFOR CORRECTINGCATASTROPHIC BELIEFS 329

belief. Through this process the patient is induced to think deeply about the evidence for and against the catastrophic and noncatastrophic beliefs.

Identifying Causes of 'Bad Days' Panic patients commonly report that they have 'good days' and 'bad days,' and sometimes claim to know as soon as they awaken whether the day will be good or bad. Bad days are characterized by chronic anxiety and worry that something bad will happen. Panic attacks may be particularly frequent and severe on bad days, and agoraphobic avoidance may be greater. Patients often have difficulty identifying the causes of bad days. In order to make panic attacks and anxiety more understandable and therefore less threatening, the patient should generate a list of benign potential causes of bad days, and evaluate the evidence for and against these. Possibilities include the many different causes of arousal-related sensations (e.g., caffeine, lack of sleep, a mild physical illness: see Chapter 4), as well as cognitive factors such as pessimistic expectations (Clark & Salkovskis, 1987).

Troubleshooting Vague Beliefs

Verbal interventions are most effective when they focus on clearly defined beliefs. This permits the therapist and patient to identify evidence that unambiguously supports or refutes the beliefs. Sometimes, however, despite the therapist's best efforts, patients may have difficulty coming up with clearly defined beliefs. This can be for a number of reasons. Some patients have clearly defined beliefs that they describe vaguely, possibly because they are embarrassed about the beliefs, or because discussing the beliefs makes them feel anxious or panicky. Deliberate evasiveness (avoidance) may be suspected when the patient is visibly distressed about discussing the beliefs and therefore attempts to change the topic of conversation. The therapist can address this problem by directly but tactfully raising it with the patient and then collaboratively looking for solutions. The therapist should be mindful of the quality of the therapeutic relationship and the way in which the verbal interventions are implemented. Patients will be reluctant to disclose if they feel attacked or criticized for disclosing catastrophic beliefs. In other cases the beliefs appear to be genuinely vague. Here too, avoidance may play a role; upsetting beliefs may be vague because the patient

330

COGNITIVE INTERVENTIONS

tries not to think about them. Mrs E. believed that 'something bad' would happen if she became very jittery. She was not sure what would happen, and tried not to think about it because that brought on the very experience she feared. A review of the evidence for and against this vague belief would probably be ineffective because 'something bad' is so ambiguous that it would be difficult to disprove. Instead, Mrs E. was asked to think of examples of 'something bad', and if necessary to guess what this might be. This yielded two specific beliefs ('I could go crazy' and 'I could have a stroke'), which became the targets of cognitive restructuring. The exercise had a further beneficial effect in that it made her very jittery, thereby allowing her to see that this had no harmful consequences.

Patients Who 'Know' Their Sensations are Harmless Appropriate targets for verbal interventions also can be difficult to identify in patients who claim they 'know' that arousal-related sensations are harmless, and yet continue to experience panic attacks. These patients may hold catastrophic beliefs without recognizing them as such. The beliefs may not be about dying, going crazy, or losing control; rather they may be about the intolerability of anxiety or panic (e.g., 'I can't stand feeling anxious') or beliefs that anxiety or panic will never end (e.g., 'Next time I panic I could wind up suffering in agony for the rest of my life'). Once identified, these beliefs can be addressed by using the verbal interventions described earlier in this chapter. Another approach is to assess whether the patient avoids any situations or activities (Craske et al., 1994). A discussion of the reasons for avoidance can elicit catastrophic beliefs, and thereby show patients that they hold such beliefs. Exposure probes can be used for the same purpose (see Chapter 9).

Misuse of Coping Statements Coping statements (e.g., 'Racing thoughts are harmless; they don't mean I'm going crazy'), often provided on written cards, can serve as useful aids for challenging catastrophic beliefs during the patient's daily life. However, sometimes the rehearsal of these statements becomes a safety behaviour (Wells, 1997). The misuse of coping statements may be suspected when the patient still believes in the catastrophic beliefs despite frequent use of coping statements. Ms A. used a number of coping statements that she gleaned from a self-help book. She never left home without her list of statements, and would select one to rehearse, mantra-like, whenever she felt nervous. This was calming, at least in the short term. In the longer term this activity prevented her from testing her catastrophic belief that 'nervous feelings lead to nervous breakdowns'.

c

1 t

t I

t

a

1

I Ii

a ti

S'

IMAGERY METHODS 331

Coping statements also may be misused when they are used for wishful thinking (Andrews et al., 1994); e.g., 'I'm not going to panic.' Wishful thinking sets patients up for unexpected aversive experiences when they enter feared situations. To identify potentially problematic wishful thinking, patients should be asked to describe what their coping statements mean to them. This can reveal catastrophic beliefs. A coping statement like 'I'm not going to panic' may be shorthand for 'I'm not going to let myself panic, because that would be dangerous'. When coping statements are misused in these ways, the patient should be advised to refrain altogether from using them. Behavioural experiments (Chapter 13) can then be used to disconfirm the catastrophic beliefs. Overvalued Beliefs Some patients have very strongly held, sustained beliefs in the dangerousness of arousal-related sensations. These overvalued beliefs can be difficult to challenge with verbal disputations alone. Behavioural experiments used in conjunction with verbal disputations are the most effective way of challenging these beliefs (see Chapter 13). Irrational Beliefs About Setbacks Patients may catastrophize about setbacks in treatment. After being panicfree for several weeks, Mr F. had a full-blown panic attack during a particularly stressful day at work. At his next therapy session he felt despondent and expressed beliefs such as 'I can't stand these panic attacks' and 'I should be able to control my feelings by now'. Two strategies were used to help Mr F. deal with the setback. First, he was guided to objectively examine and challenge his beliefs about the setback. Second, the therapist re-framed the occurrence of panic as an opportunity for Mr F. to work further at challenging his catastrophic beliefs. Mr F. came to see that he still believed to some extent that 'nausea leads to loss of control,' and that this belief contributed to his panic attacks. Ways of further challenging this belief were explored in the session.

IMAGERY METHODS Imagery methods such as imaginal flooding, graded imaginal exposure, and systematic desensitization, have long been used in behavioural treatments of panic disorder. Patients are asked to imagine fear-evoking situations, panic attacks, or a combination of the two, and to rehearse the

332

COGNITIVE INTERVENTIONS

image until fear habituates. Such methods are not very powerful by themselves (Chambless et al., 1979; Emmelkamp & Wessels, 1975; Gelder et al., 1973; Mathews et al., 1976; Stern & Marks, 1973; Watson et al., 1973). However, when used with other CBT2 interventions, imagery methods can be useful adjuncts for (1) eliciting and restructuring catastrophic thoughts and images, and (2) facilitating interoceptive and situational exposure.

Imagery-assisted

Cognitive Restructuring

Episodic Memories One way to elicit catastrophic beliefs is to ask the patient to vividly imagine a recent panic attack, including the sensations occurring before and during the attack, and the interpretations of these sensations. A variation on this approach is to focus on memories of the patient's worst panics. Catastrophic beliefs about the panic sensations are then elicited, and then compared to pleasant memories in which the patient experienced the same sensations; e.g., palpitations during a panic attack versus palpitations during a rollercoaster ride (Otto et al., 1996a). The two memories are then contrasted to highlight that the sensations in both cases were harmless; the two differed in the interpretations of the sensations, and in the emotions resulting from the interpretations. Eliciting Catastrophic Images Panic patients often describe frightening cognitions in the form of fleeting images. These are sometimes evoked in response to body sensations, thereby leading to panic. The images are likely to trigger panic attacks if the patient believes the images portray realistic catastrophic outcomes. Thus, frightening images can be forms of catastrophic misinterpretation. Palpitations, for example, may trigger the image of the patient collapsing and being rushed away by ambulance. Derealization may trigger an image of burly hospital orderlies hauling the patient off to a psychiatric hospital. Trembling or blushing may trigger an image of the patient cowering within a circle of jeering faces. Simply helping the patient to identify images can make the panics more understandable and less frightening. This can be done by asking questions such as the following:

I l:

•

~ a

•

When you started to panic did any images or pictures run through your mind? When you started to feel_ [feared body sensation], what did you think would happen? ... Was this a thought or an image?

1

f f

t,

IMAGERYMETHODS 333

•

Let's spend a few moments imagining that you feel·_ [feared body sensation] right now. Do you have any thoughts or images of what could lead to?

Image Suppression

Patients sometimes try to suppress or distract themselves from frightening images. The problem with this approach is that it can paradoxically increase image frequency (Wegner, 1994). An important step in modifying these images is to counsel the patient to refrain from attempting to avoid these cognitions. If the patient is reluctant to do so, then a behavioural experiment can be used, where the patient suppresses images on some days and does not suppress them on other days. The consequences of suppression versus non-suppression can then be contrasted in terms of their effects on image frequency and panic attacks. This strategy has proved useful in helping patients with obsessive-compulsive disorder to learn that suppression of intrusive thoughts is counterproductive (Salkovskis, 1999). The same can be done for panic disorder. Note that this strategy can be confusing for patients if the therapist trains them in distraction or thought-stopping techniques. Further problems with distraction and thought stopping are discussed later in this chapter. Image Modification Strategies

One way to reduce the emotional impact of catastrophic images is to use imaginal exposure. This can help patients learn that the feared images have no harmful consequences. This experience can alter patients' beliefs about the meaning of the images. Images that were formerly seen as harbingers of disaster are reinterpreted as 'mental garbage'. • Cognitive restructuring can be used to challenge the likelihood that the event depicted in the image would come true, and to challenge the cost (badness) of the imagined event if it did occur. This can help the patient reappraise the meaning of the image. Guided imagery exercises also can be used to restructure catastrophic images. One method involves imaginally altering the image into something innocuous or amusing (Clark, 1989). In response to feeling derealized, Mr L. had the image of being taken away by 'men in white coats'. This image frightened him so much that it triggered panic attacks. During therapy Mr L. was asked to call up the image and then deliberately modify it. The men in white coats were transformed from hospital attendants to bakery assistants, carrying trays of pastries for Mr L. to enjoy. As Mr L. practiced this exercise, he gradually learned to take the feared image less seriously.

334

COGNITIVE INTERVENTIONS

Another method is to '_finishout' the image, in which the patient is asked to provide a realistic, positive conclusion to a frightening image (Clark, 1999; Wells, 1997). This strategy is based on the observation that catastrophic images often stop at the worst point in the scenario (Beck & Emery, 1985): for example, imagining oneself feeling dizzy while driving, but failing to imagine that the dizziness will shortly pass so that driving can be resumed. The 'finishing out' strategy is implemented by asking the patient to recall the feared image, then hold it in mind until anxiety is experienced, and then continue the image until there is a positive resolution. The patient may spontaneously come up with a realistic, positive outcome, or the therapist can guide the patient to arrive at such an outcome.

Sl

Imaginal Adjuncts to Interoceptive and Situational Exposure Using Imaginal Exposure to Induce Feared Sensations

Imaginal exposure also can be used as a form of interoceptive exposure. The patient can be asked to imagine being in a feared situation (e.g., standing in an elevator), performing a feared activity (e.g., running up a flight of stairs), or experiencing some other feared event (e.g., their worst panic attack). When these scenes are vividly imagined, arousal-related body sensations are induced. As the patient continues to imagine the scene, they begin to learn that the sensations have no harmful consequences. Body sensations evoked by imaginal exposure tend to be milder than those evoked by other interoceptive exposure exercises, such as aerobic exercise or voluntary hyperventilation. But imaginal exposure can be a good place to start for patients who are too frightened to actually perform more vigorous interoceptive exercises. For extremely fearful patients, a hierarchy of anxiety evoking situations can be established, from the least distressing to the most distressing, and patients can gradually work up the hierarchy. To conduct imaginal exposure of a given scene, the patient can be asked to close their eyes and attempt to recall a distressing situation. Alternatively, the patient could write out and read over the feared scene, or make an audiotape description that they repeatedly listen to. As in all imagery exercises, it is important that the patient try to imagine all the components of the feared situation as vividly as possible, including visual, auditory, olfactory, and somatic stimuli, and the thoughts, behaviours, and feelings imagined to occur in the situation.

im fe ce ag in m COi

fea

In)-'

hai

Tn S01 SID

as1 rec car ing thE str;

IMAGERY METHODS 335

Jmaginal Rehearsal

Imaginal exposure can be used as a form of mental rehearsal to prepare for situational exposure assignments. Imagining a trip through a supermarket can help the patient plan exactly how she or he would complete this assignment (e.g., planning when to go, what to buy, which queue to stand in, and so forth). This can help identify points at which the patient is most likely to avoid or escape actual situations, along with associated catastrophic thoughts. The patient and therapist can then plan how to overcome these difficulties. For example, when actually entering a feared situation, patients could carry a cue card to prompt them to use cognitive restructuring in the situation; e.g., 'Am I expecting a catastrophe? Realistically, what is most likely to happen?' Coping Imagery

To facilitate situational exposure, some therapists ask their patients to imagine feeling calm and functioning competently in the feared situations (Andrews et al., 1994; McMullin, 1986). Although the use of coping imagery can be useful, the therapist needs to ensure that this intervention is compatible with other strategies used to treat the patient. Coping imagery involves the patient imagining having little or no distress in the feared situation. This contrasts with other interventions, such as interoceptive and situational exposure exercises, in which the patient is encouraged to fully experience arousal-related sensations. Combining these interventions with coping imagery can confuse patients. It sends a mixed message about what to do when the feared sensations occur. A further concern is that coping imagery may not change catastrophic beliefs about feared sensations. Patients may use imaginal coping as a means of avoiding the feared sensations, and thereby fail to learn that the sensations are harmless.

Troubleshooting Some patients have difficulty generating sufficiently vivid images by simply imagining a feared scenario. In these cases the patient could be asked to write or draw a description of the image, or produce a tape recorded description. Patients should try to include all the elements that contribute to image vividness-all the sensory elements, thoughts, feelings, and behaviours that make up a given scenario. If these methods fail then some other intervention may be used instead, such as the verbal strategies for cognitive restructuring or the exposure methods.

336

COGNITIVE INTERVENTIONS

OTHER COGNITIVE STRATEGIES

Distraction Diverting tasks such as playing loud music, watching television, or playing computer games can effectively distract patients from catastrophic cognitions and feared sensations. Distraction can be a useful socialization strategy to show patients how their attention and thinking influences their feelings. Thus, distraction can be useful in challenging a patient's beliefs that they have no control over their anxiety or panic. Distraction also can be useful in situations where it is difficult to challenge catastrophic cognitions (Clark, 1989). For example, when a patient's mind is filled with catastrophic images while driving over a bridge, it may be better to distract him- or herself from the images by focusing on driving. Although useful on occasion, there are many kinds of ways that distraction can be misused. Distraction can become a safety behaviour; e.g., a patient in a supermarket queue might attempt to read a magazine rather than focus on the fact that she is in a crowded store. Given these concerns, the therapist should be cautious about recommending distraction.

Thought Stopping A procedure similar to distraction is thought stopping, which has been used to increase the patient's control over upsetting thoughts and images. The procedure was used in an early version of the Albany protocol (Barlow & Cerny, 1988) but not in the latest version (Craske et al., 1994). It is, however, used in Franklin's (1996) CBT2 protocol. The procedure is as follows. As soon as a patient notices an upsetting thought or image-such as the thought 'I am about to go crazy'-the patient imagines loudly yelling 'STOP!-STOP!-STOP!' and also imagines a flashing stop sign. Then he or she immediately shifts attention onto some other diverting activity until the upsetting thought or image has passed. Presumably this procedure blocks panic attacks by removing the catastrophic cognitions that cause panic. The procedure might also indirectly teach the patient not to take the thoughts or images seriously. However, the risk is that patients may use thought stopping as a safety behaviour. Indeed, patients sometimes discover thought stopping on their own, and use it in an attempt to ward off catastrophic consequences. Mrs K.

SUMMARY AND CONCLUSIONS 337

panicked whenever she found herself dwelling on morbid thoughts, such as thoughts of her son's suicide several years earlier. She believed the thoughts would drive her insane if they became more frequent. Mrs K. compulsively used thought stopping in an effort to banish the unwanted thoughts. This prevented her from learning that the thoughts were not dangerous. In fact, during therapy she discovered that thinking about her son's death could serve a useful purpose; it could help her make sense of the tragic loss. A further problem with thought stopping is that it can lead to the paradoxical rebound of unwanted intrusive thoughts and images (Wegner, 1994). Rebound appears to be triggered by distracters used during thought stopping. The distracters act as retrieval cues, causing the unwanted thoughts or images to be accessed from memory. To deal with this problem, Franklin (1996) recommended that patients avoid returning to the situation that originally provoked the negative thought. A simpler strategy would be to refrain from using thought stopping altogether. Thought stopping does not directly change catastrophic beliefs or images, and can create more problems than it solves.

SUMMARY AND CONCLUSIONS There are many cognitive interventions for correcting catastrophic cognitions in panic disorder. To assess their efficacy the therapist can periodically obtain ratings of the strength of catastrophic and noncatastrophic beliefs, or ratings of the extent to which the patient believes that catastrophic images portend danger. Socratic dialogue plays an important role in most cognitive interventions. Among the first cognitive interventions that patients receive is psychoeducation about the nature and treatment of panic disorder. Socialization strategies are used to increase the impact of corrective information and to engage patients in treatment. Verbal disputations and imagery methods are other useful strategies for producing cognitive change. Coping statements, distraction training, and thought stopping should be used with caution because they can create more problems than they solve. To consolidate the learning that goes on during therapy sessions, cognitive interventions typically entail some form of homework. Patients might complete a workbook summarizing the session's material, or make notes from an audiotape of the session. Other written assignments are also commonly completed as homework, where, for example, the patient might write out a disputation of a catastrophic belief and consider the evidence for a noncatastrophic alternative.

338

COGNITIVE INTERVENTIONS

Cognitive interventions are most commonly used in conjunction with behavioural experiments. The combination of the two increases the odds that treatment will reduce catastrophic cognitions. In the next chapter we will consider how behavioural experiments are constructed, implemented, and integrated with cognitive interventions.

fl

f ' ,

l

t

t

Chapter 13

INTEROCEPTIVE AND SITUATIONAL EXPOSURE Exposure exercises are those in which the patient is presented with feared stimuli in a systematic, controlled manner. Although behavioural therapies have long used exposure for purposes of 'desensitizing' or 'deconditioning' emotional reactions, exposure also can be seen as a way of providing the patient with corrective information (Foa & Kozak, 1986). In other words, exposure exercises are important vehicles of cognitive change. In second generation cognitive-behavioural therapy (CBT2), exposure exercises are carried out as 'behavioural experiments,' which involve putting to the test predictions derived from cognitive restructuring (Salkovskis & Clark, 1991). The goal of exposure in CBT2 is not to simply reduce anxiety or panic reactions, but rather to: • • •

produce potent tests of catastrophic beliefs; help patients discover the likely noncatastrophic causes of their feared sensations; test whether patients' coping strategies (e.g., safety behaviours) ameliorate or exacerbate their problems.

The patient plays an active role in choosing which exposure exercises to attempt and when to attempt them. These exercises are set up as 'no lose' assignments; something important is learned regardless of the outcome (Clark & Salkovskis, 1987). The exercises can involve interoceptive exposure, situational exposure, or a combination of the two. This chapter will review these exposure methods, describe how they can be integrated with the cognitive interventions described in Chapter 12, and discuss solutions to common problems in implementing exposure.

• 340

INTEROCEPTIVEAND SITUATIONAL EXPOSURE

INTEROCEPTIVE EXPOSURE Exercises Used in the Clinic Interoceptive exposure exercises are designed to induce arousal-related body sensations. The exercises are performed in the therapist's office and then as homework assignments. Some of the commonly used exercises were introduced in Chapter 9, where we reviewed how they could be used as exposure probes for the purposes of assessment. The exercises also can be used as behavioural experiments, in which they are performed over several trials until corrective learning takes place. When used for treatment purposes, the choice of exercise depends on which sensations they evoke (see Chapters 4 and 9). Exercises that induce a given sensation (e.g., dizziness) are used to test beliefs about that sensation (e.g., 'Dizziness leads to physical collapse'). Table 13.1 provides a summary of interoceptive exposure exercises commonly used in treatment, and examples of beliefs they can test. These and the other exposure exercises described in this chapter were gathered from several sources (e.g., Andrews et al., 1994; Antony et al., 1998; Clark, 1989; Clark & Salkovskis, 1987; Craske & Barlow, 1993; Craske et al., 1994; Franklin, 1996; Otto et al., 1996a; Wells, 1997), in addition to my own clinical experience. The list in Table 13.1 is intentionally long; the more exercises a therapist has at her or his disposal, the greater the chances of finding an exercise that will effectively challenge the patient's catastrophic beliefs. The table shows that interoceptive exposure exercises tend to be quite short, requiring at most a few minutes each. The exercises are usually practiced repeatedly in the therapy session and as homework. A patient might perform three trials of two of these exercises in a 60 min treatment session (six trials in all), and then practice the six trials four times in the following week. Most of the exercises in Table 13.1 are self-explanatory, although the procedure for conducting the hyperventilation exercise requires some comment. It is among the most commonly used exercises because it induces a host of arousal-related sensations. Patients are asked to pant deeply and vigorously through the mouth and nose at a rate of 1-2 breaths/ s. The therapist can mark time to ensure that the patient is breathing at the desired rate. Predominantly upper-chest movement is necessary to maintain a high rate of respiration (Bonn et al., 1984). This can be demonstrated by the therapist. Patients sometimes try to avoid hyperventilatory sensations by taking very shallow breaths. This should be dis-

Table 13.1 Interoceptive exposure exercises

Examples of exercises • • • • • •

Shake head rapidly from side to side, or roll head in circles (30s) Place head between knees for 30 s and then lift head quickly up to a normal (upright) position Spin around while standing up with arms stretched out (1min) Hold breath (30s) Hyperventilate (1min)

•

Breathe through a narrow straw without breathing through nose (2min) Stare continuously at a ceiling fluorescent light (1min)

•

Stare continuously at reflection in mirror (2min)

• •

Stare continuously at spot on wall or at one's hand (3min) Stare at Wilkin's visual grid (Figure 4.1) (3min)

• • •

Tense all muscles in body while sitting in a chair (1min) Jog on the spot or run up stairs (1min) Face a heater: heater blowing hot air at the face (Smin)

• • • •

Tongue depressor: place tongue depressor at back of throat (30s) Drink hot coffee (2-3 cups) To induce throat tightness, one can ask the patient to start to swallow and then hold the throat in the 'mid-swallow' position for 5-l0s. To induce chest pain, ask the patient to interlock their fingers and place hands behind the head while stretching the elbows backwards. The patient then takes a deep breath and tries to chest breathe at a rate of 1 breath/s for lmin.

Examples of catastrophic beliefs that can be tested* 'Dizziness leads to insanity' 'When I feel lightheaded it means I could be having a stroke' 'I will throw up if I let myself feel nauseous' 'Chest tightness means I'm having a heart attack' 'If I start to feel unsteady it means I will physically collapse' 'Choking sensations are dangerous' 'If my surroundings start to look weird it means I'm going psychotic' 'If I let myself feel spacey I could permanently lose touch with reality' 'Feeling unreal is a sign that I'm having a stroke' 'Staring at visual illusions or other unsettling images could "tip me over the edge" into permanent insanity' 'People will laugh at me if I start to tremble' 'I will have a heart attack if my heart starts pounding' 'People will ridicule me if they see I'm having hot flushes' 'If my stomach gets upset I'll vomit uncontrollably' 'I could go crazy if I get too jittery' 'If my throat feels tight it means I'm about to choke to death' 'Chest pain means I'm having a heart attack'

* These exercises are also used to test noncatastrophic alternative explanations of the sensations; e.g., 'Palpitations are simply due to my lack of physical fitness.' See Chapters 4 and 9 for descriptions of the sensations evoked by these exercises.

z-l

m

;;o

0()

m

2'l < m m X

"TI

0(/l

C ;;o

m

w .i,.

342

INTEROCEPTIVEAND SITUATIONAL EXPOSURE

couraged. In later trials the effects of the hyperventilation can be enhanced by having the patient overbreathe while standing up with their eyes closed. The standing position introduces new sensations (e.g., body sway), which can be useful in testing catastrophic beliefs (e.g., 'Unsteadiness leads to physical collapse').

• •

Homework exposure exercises are presented to patients as important elements of treatment. Homework should be reviewed in each session-in terms of what happened and what was or wasn't learned-and any difficulties should be addressed. Homework assignments should be clearly explained beforehand so that the patient and therapist can determine whether or not the tasks were properly completed. Patients should be reinforced (and encouraged to self-reinforce) for their efforts. Patients are advised to practice the exercises regularly, regardless of how they feel. They should attempt the exercises even on 'bad days,' when they might feel particularly anxious or panicky. Indeed, exercises can be most fruitful when conducted on those days (e.g., 'Even when I was feeling really anxious, I still didn't go crazy when I hyperventilated').

•

Adherence can be improved by asking patients to record each exercise on a homework rating form. On the form the patient notes the day the exercise was attempted, its duration, peak anxiety (0-100), and the predictions derived from the catastrophic belief and its noncatastrophic alternative. The patient can also note on the form what they learned from each exercise.

•

•

• •

•

Naturalistic Exercises Naturalistic interoceptive exposure refers to exposure to daily tasks or activities that have been avoided or endured with dread because of the associated sensations (Craske & Barlow, 1993). The following is a list of examples that can be set as homework assignments. The list is drawn from the same sources used to generate the list of office-based exposure exercises. As before, the present list is lengthy; better to have too many than too few suggestions for naturalistic interoceptive exposure.

•

• • • • •

Wear a lot of clothes on a warm day (to induce hot flushes, sweating, and palpitations). Wear a tight collar or necktie (to induce choking sensations). Watch arousing videos at home (e.g., thrillers or horror films). Engage in sexual intercourse (this produces a range of arousal-related sensations).

A p, p,

INTEROCEPTIVE EXPOSURE 343

• •

• •

• • •

•

•

•

Sit in a sauna (hot flushes, sweating, palpitations). Take a hot shower with the windows and doors closed and the ventilation fan turned off (dyspnea, choking and smothering sensations, chest tightness). Drink two or three cups of strong, hot coffee at home (trembling, palpitations, hot flushes, sweating). Household chores: cleaning, vacuuming and picking up clothes or other items from the floor. Vigorous cleaning and vacuuming can induce palpitations. Frequent bending over to pick up objects can induce dizziness. Go on carnival rides such as a rollercoaster or Ferris wheel (dizziness, nausea, faintness). Go to nightclubs with strobe lights (derealization, depersonalization, dizziness). Perform activities that intensify sensations during naturally occurring panic attacks (e.g., deliberately hyperventilate, focus on one's body, run up a flight of stairs). By increasing the intensity of panic sensations, the activities provide strong tests of beliefs about the consequences of arousal-related sensations. Participate in an exercise program; e.g., cycling, jogging, exercise classes (aerobics or weight training), dancing classes, swimming classes, or water aerobics. These activities induce a variety of arousalrelated sensations, such as palpitations and dyspnea. Water aerobics is useful for patients whose physical condition prevents them from participating in other exercise programs (i.e., conditions such as joint diseases, lower back pain, or severe obesity). Exercise programs not only disconfirm catastrophic beliefs (e.g., 'My heart is weak') but also are very useful for testing the noncatastrophic alternative that the feared sensations are due to lack of physical fitness (e.g., 'I'm short of breath because I'm out of shape'). If the noncatastrophic alternative is accurate, then the sensations should abate as fitness improves (Clark & Salkovskis, 1987). Lie on one's back in a hot bath, head submerged to the point that both ears are under water. This provides an environment in which the patient becomes aware of their breathing and heartbeat pulsing in the ears. The heat induces palpitations. Video arcade games, such as driving simulators, provide a combination of visual, vestibular, and auditory stimulation, which can induce dizziness, derealization, palpitations, and nausea.

As the patient's fear of arousal-related sensations abates, their preoccupation with these sensations should also diminish. In other words, the patient is less likely to notice the sensations. This is a beneficial side-effect

344

INTEROCEPTIVE AND SITUATIONALEXPOSURE

for patients who regard their heightened awareness of sensations as evidence that there is something seriously wrong with them (e.g., 'I often notice my heart skipping a beat; other people don't notice their hearts skipping a beat. That means there's something wrong with my heart').

Eliminating Safety Signals and Safety Behaviours Recall from Chapter 3 that safety signals are stimuli that the person associates with the absence of feared outcomes. Safety signals include medicine containers, cellular phones, and good luck charms. Safety behaviours are things the person does to ward off feared events. Such behaviours include escape, avoidance, and other 'coping' behaviours. Safety signals and safety behaviours can prevent catastrophic beliefs from being disconfirmed; e.g., 'If I hadn't distracted myself, my anxiety would have escalated to the point that I'd have gone crazy'. Safety signals and behaviours are often related to one another; a cellular phone can be a safety signal, and making a call to a significant other can be a safety behaviour. By definition, interoceptive and situational exposure lead the patient to eliminate the more obvious safety behaviours (i.e., frank escape or avoidance) and to eliminate the more prominent safety signals (e.g., presence of a trusted companion). However, this may not eliminate the more subtle safety signals and behaviours. There are a great many of these that could be used in any given situation. A person frightened of showering with the doors and windows closed, for example, might take a cool drink, anxiolytic medication, or cellular phone into the bathroom because these things confer (for him or her) a sense of safety. Safety signals and behaviours are important to eliminate from all kinds of exposure-interoceptive, situational, and their combination. There are three steps in achieving this goal: •

•

Identify obvious and subtle safety signals and behaviours. Observe patients during exposure exercises and ask them whether they did anything to cope with the situation or to prevent something bad from happening. Different safety signals or behaviours may be used in different situations, or on different occasions in the same situation (cf. Williams, 1990), and so the therapist needs to assess in detail what patients did during each of their exposure exercises. Show patients the effects of safety signals and behaviours by comparing what happens when patients use them and when they don't.

(

r'

I l

't

C

r

l F l F r

e

s f

I:

e

s

INTEROCEPTIVE EXPOSURE 345

•

Mrs W. habitually tensed the muscles in her arms, hands, and torso in an effort to prevent herself from 'shivering' (trembling) when she felt nervous. She believed this was partially effective in reducing her tremulousness. To test the effects of this safety behaviour, she was asked to compare the effects of increasing the behaviour (i.e., tensing even more) versus dropping the behaviour (not tensing). She learned that the safety behaviour made her trembling worse, not better. Encourage patients to drop safety behaviours altogether, and not to seek out or carry safety signals with them. Safety signals and behaviours can be gradually faded out over successive exposure trials. This can involve fading out the use of relaxation or breathing exercises if these become safety behaviours.

It can be instructive to look at the differences between the sensations and feelings evoked by exposure exercises versus those occurring in naturally occurring panic attacks (Greenberg, 1989). This can help the therapist identify contextual factors (e.g., presence vs. absence of safety signals) that are important to consider in order to develop powerful tests of catastrophic beliefs and their noncatastrophic alternatives. Safety behaviours can take the form of reassuranceseeking or other forms of checking. Ms F. frequently checked her pupils in the mirror to see whether they were equally dilated. Although she was physically healthy, Ms F. feared she was at risk for having a stroke, and believed that unequal pupil dilation was the first sign of a cerebrovascular event. Thus, Ms F. used mirror checking as an 'early warning' indicator to determine whether she should call an ambulance. Ms F.'s checking simply perpetuated her health concerns, and prevented her from learning that the checking was unnecessary. She also read medical texts and consulted medical sites on the Internet to reassure herself that she was unlikely to have a stroke. Ironically, her searches turned up alarming evidence of rare, poorly understood cerebrovascular disorders. This further strengthened her belief that she was at risk for having a stroke. Treatment of these problems involved exposure and response prevention, as used in treatments of obsessive-compulsive disorder (Steketee, 1993). Ms F. was exposed to situations and activities she feared would bring on a stroke (e.g., doing pushups while at home alone) and asked to refrain from checking (e.g., refraining from checking her pupils and from checking the Internet). As a result, she came to learn that checking was unnecessary. She also became less convinced that she was at risk for having a stroke.

346

INTEROCEPTIVEAND SITUATIONAL EXPOSURE

Medical

Contraindications

for Interoceptive

Exposure

Interoceptive exposure exercises are safe to perform for the majority of people with panic disorder. However, there are some medical contraindications for certain procedures. Examples are shown in Table 13.2. When in doubt about using a given exercise, the therapist should either consult with the patient's physician or refrain from using the exercise. When these exercises cannot be used, the therapist can still demonstrate the exercise to the patient. For example, the therapist could hyperventilate for 1-2min, and then describe the sensations experienced. This could lead to a discussion of how naturally occurring hyperventilation might cause the patient's sensations (Clark, 1989).

Procedures for Conducting Interoceptive Exposure Interoceptive exposure exercises are typically first performed in the therapist's office. Choice of exercise depends on which sensations are feared the most, as discussed earlier in this chapter. The following is a protocol for conducting interoceptive exposure exercises. •

•

•

•

•

Identify any medical contraindications for interoceptive exposure. Information obtained from the referral source and from the pretreatment assessment may be sufficient. Identify a catastrophic belief that appears to play an important role in the patient's panic attacks. For example, although medical evaluations had shown that Mr L. was physically healthy, he held the catastrophic belief, 'Palpitations indicate that there is something wrong with my heart; I could die if my heart starts pounding.' In collaboration with the patient, generate a plausible, noncatastrophic alternative belief statement; e.g., 'Palpitations indicate that my heart is working properly; a healthy heart should speed up if I'm physically active or feeling stressed out.' Rate the strength of belief of the patient's catastrophic cognition and its noncatastrophic alternative (using a 0-100 scale, ranging from 0 = do not believe that _ is true, to 100 = completely believe it is true). The therapist and patient then select an interoceptive exercise (e.g., from Table 13.1) that will test predictions arising from the catastrophic belief and its noncatastrophic alternative. Mr L. was asked to vigorously jog on the spot for 1 min. Patients will have completed several of these exercises for the purposes of assessment (Chapter 9), and so they will have an idea of the sensations that are induced. In prepara-

Table 13.2 Contraindications for using interoceptive exposure exercises

Exercise •

Shake or roll head

•

Place head between knees and stand up

• • •

Spin around Hold breath Hyperventilate

• • • • •

Breath through a narrow straw Stare at fluorescent light Stare at reflection in mirror Stare at spot on wall or at one's hand Stare at Wilkin's visual grid

•

Tense all muscles

• Jog on the spot or run up stairs • Heater blowing hot air at face • Tongue depressor at back of throat • Drink hot coffee • Hold throat in 'mid-swallow.' • Hands behind head while stretching elbows backwards ...

Potential contraindication Cervical pain or disease (e.g., whiplash injury), history of falling due to dizziness or balance disorder.* Postural hypotension, lower-back pain, history of falling due to dizziness or balance disorder.* Pregnancy, history of falling due to dizziness or balance disorder.* Chronic obstructive lung disease. Chronic obstructive lung disease, severe asthma, cardiac conditions, epilepsy, renal disease, pregnancy. Chronic obstructive lung disease. History of seizures caused by staring at flickering lights. No apparent contraindications. No apparent contraindications. History of seizures or migraine headaches (these can be triggered by the grid: see Chapter 4). Pain disorders. If pain is localized, patients could tense all but the afflicted region. Cardiac conditions, severe asthma, lower-back pain, pregnancy. No apparent contraindications. Prominent gag reflex (stimulation of which will induce vomiting). History of severe insomnia. Prominent gag reflex. Chronic obstructive lung disease, severe asthma, cardiac conditions, epilepsy, pregnancy, pain disorders. If pain is localized, patients could tense all but the afflicted region.

*Some forms of vertigo habituate to these exercises. Preparation of this table was facilitated by the helpful advice of Jonathan Fleming M.D.

z-I

m

;u

0 ()

m ~

< m ~

0 (/)

C ;u m

w

.Is,

'-I

348

INTEROCEPTIVEAND SITUATIONAL EXPOSURE

tion for the exercise, the patient and therapist can review the sensations likely to be experienced, and discuss the catastrophic thoughts (or images) that might occur. • Two predictions are set up. One prediction, based on one of the patient's catastrophic beliefs, is that there will be a catastrophic outcome (e.g., death, losing touch with reality, loss of control over one's actions). The other prediction, based on the noncatastrophic alternative, is that the outcome will be harmless (e.g., the patient will simply panic or experience some other discomfort, without any enduring harmful effects). Predictions are made as specific as possible. For example, instead of testing the prediction that the patient will 'go crazy,' one should test a more specific prediction based on the patient's conception of exactly how they would go crazy (e.g., 'I will lose the ability to think clearly'). The prediction derived from Mr L.'s catastrophic belief was 'Jogging will give me a heart attack, which will lead me to collapse and lose consciousness'. The prediction derived from the noncatastrophic alternative was 'Jogging will make me anxious and tired, but nothing harmful will occur'. • The chosen exercise is initially performed in the therapist's office. The duration of the exercise is determined, in part, on what the patient thinks they are able to perform. Ask the patient to rate their confidence about performing the task for the required duration (0 = not at all confident, 100 = completely confident). To ensure the task is performed, select a duration that has an 80 or higher confidence rating. • The therapist demonstrates (models) the task and describes what sensations were experienced. • The patient then performs the task, attempting to make the sensations as intense as possible. Mr L. described palpitations and sweating while performing the exercise, and felt dizzy and tremulous afterwards. During the exercise he had the fleeting image of collapsing from a heart attack. • The therapist and patient then review how similar the sensations were to those experienced during naturally occurring panics. This is to check whether the exercise induced the feared sensations. The patient and therapist then review whether the outcome supported the prediction from the catastrophic belief or its noncatastrophic alternative. Ratings of belief strength are also obtained. If the strength of the catastrophic belief remains high, and if the patient believes the exercise didn't represent a strong test, then the patient and therapist review why this was so. The therapist should look for safety signals and safety behaviours that might have undermined the exercise. Mr L. believed the exercise wasn't a strong test because he had taken med-

•

•

INTEROCEPTIVE EXPOSURE 349

•

•

•

ication (Ativan) before attending the session. He believed that this protected his heart from beating too rapidly. Accordingly, he was asked to refrain from taking Ativan before his next session, when the exercise was repeated. Repeat the exposure exercises as often as needed, increasing the 'difficulty' level with each successive trial. Difficulty can be raised by increasing the duration or intensity of the exercise, and by successively eliminating safety signals and safety behaviours. For Mr L., the duration of jogging (without Ativan) was gradually increased during the next session (2min on the first trial, 4min on the second trial, 6min on the third trial). Strength of the catastrophic belief and strength of the noncatastrophic alternative were recorded at the start of the session and after each trial. Over trials the strength of Mr L.'s catastrophic belief gradually fell to 10/100, and the strength of belief in the noncatastrophic alternative rose to 80/100. Assign the exposure exercise or a similar task as homework. Mr L. was asked to cycle (alone) on his wife's stationary exercise bicycle for 10 min three times each day. Based on his catastrophic belief (i.e., 'Palpitations indicate that there is something wrong with my heart; I could die if my heart starts pounding'), Mr L. predicted that bicycling would cause him to experience angina. Based on the noncatastrophic alternative ('Palpitations indicate that my heart is working properly; a healthy heart should speed up if I'm physically active or feeling stressed out'), it was predicted that bicycling would only make him feel anxious and fatigued, and that these feelings would soon pass. Note that the context can be important; Mr L. felt safer bicycling when his wife was present because she could phone an ambulance 'at the first sign of danger'. Accordingly, to strongly test his beliefs, he was asked to bicycle alone, so he would not be able to so easily 'escape' the situation by having his wife call an ambulance. Mr L. was asked to keep a record of his progress, noting (1) the duration of each exercise, (2) sensations experienced, (3) strength of the catastrophic belief and strength of the noncatastrophic alternative (from 0-100) before and after the exercise, and (4) his peak anxiety level during the exercise (0 = not at all anxious, 100 = extremely anxious). Mr L. was also asked to write down any thoughts or images that occurred during the exercise. During the next therapy session, the therapist reviewed the homework, looking for whether safety signals or behaviours prevented the catastrophic belief from being disconfirmed. Advise the patient to continue practicing the exercises so long as the strength of the catastrophic belief continues to decrease, and the strength of the noncatastrophic alternative continues to increase.

350

•

•

•

•

INTEROCEPTIVE AND SITUATIONALEXPOSURE

Look for opportunities for introducing naturalistic interoceptive exposure that tests the same beliefs (i.e., build in generalization). It was suggested to Mr L. that he take the stairs to his fifth floor office instead of taking the elevator. Stair climbing induced palpitations, without ill effects, thereby providing further evidence to challenge the catastrophic belief and strengthen the noncatastrophic alternative. Integrate the interoceptive exercises with situational exposure exercises. Do this on an 'as needed' basis, depending on whether the patient has agoraphobia. Mr L. graduated from the stationary exercise bicycle to jogging around the hospital grounds, and finally to jogging through wooded parkland trails far from the nearest medical facility. These exercises simultaneously address two fears arising from his belief in the weakness of his heart: fear of palpitations and fear of being far from medical aid. Encourage patients to chart their progress in terms of tasks completed and reactions to those tasks. Review these records with the patient during therapy sessions. As treatment progresses, one should lead the patient to take increasingly more responsibility for deciding which interoceptive exercises to undertake. Thus, treatment becomes increasingly 'patient-directed,' and can thereby continue after the end of the formal course of therapy.

Interoceptive exposure exercises can be used to show patients how they can turn their sensations 'on' and 'off'. For example, a 1 min hyperventilation followed by a few minutes of slow diaphragmatic breathing (Chapter 14) can show patients how hyperventilatory sensations can be induced and reduced. By manipulating the sensations in this way, patients can learn that their sensations are easily controlled and therefore harmless. However, this does not convince all patients. Some believe the sensations would be dangerous if uncontrolled. The therapist needs to be alert for this possibility, and for the related problem of patients using some strategies (e.g., breathing exercises) simply as ways of avoiding the feared sensations. There are several strategies for demonstrating to patients that they can 'function' or 'cope' even while experiencing feared sensations. While experiencing these sensations, patients should try to continue with ongoing activities (e.g., gardening, household chores, walking about, interacting with others) (Barlow & Cerny, 1988; Otto et al., 1996a). Interoceptive exposure exercises are particularly useful for patients who believe they 'won't be able to function' if they experience the feared sensations.

i

l

i

J

''

c1

(

1 i ( (

1 (

C

1

I t i

1

f (

I t (

SITUATIONAL EXPOSURE 351

With practice, patients find that they can carry on with daily activities even if they experience intense arousal-related sensations. To promote generalization, patients should perform the interoceptive exercises under a wide range of circumstances and conditions. This helps patients learn that the sensations are safe in general, rather than safe only under specific circumstances (Otto et al., 1996a). It also can be useful to have patients complete exercises that induce 'novel' sensations. That is, sensations that the patient has not experienced before (Otto et al., 1996a). Such sensations may arise in their daily lives at some point in the future. Panic patients are likely to catastrophically misinterpret sensations that they have never experienced before, and so exposure to novel sensations is a useful preventive intervention.

SITUATIONAL EXPOSURE

Indications Situational exposure is used primarily for reducing agoraphobia. In situational exposure the patient repeatedly encounters feared external stimuli (objects or places) in order to challenge maladaptive beliefs about these situations, such as beliefs about catastrophic outcomes (e.g., dying, going crazy, losing control). The situations are typically those in which patients believe they would panic (e.g., crossing bridges or tunnels), or situations in which they believe it would be dangerous or embarrassing to panic (e.g., 'I would make a fool of myself if I panicked in the grocery store,' 'I could crash the car if I panicked on the freeway'). The distinction between 'situational exposure' and 'naturalistic interoceptive exposure' is arbitrary but clinically useful. Both entail some degree of exposure· to arousal-related sensations. They differ in the intent of the exercise and, to some degree, in the nature of the experience. Interoceptive exposure is designed to induce intense arousal-related sensations. Situational exposure emphasizes encounters with situations that are feared and avoided, regardless of whether the situations induce intense sensations. The distinction between the two types of exposure is useful to underscore the fact that it is important to reduce patients' fear and avoidance of situations in which the feared sensations might occur. In CBT2, situational exposure takes the form of behavioural experiments that test beliefs about what could happen in the situation. As in interoceptive exposure, the experiments involve testing predictions from catastrophic beliefs (e.g., beliefs such as 'People will laugh at me if they see

352

INTEROCEPTIVE AND SITUATIONALEXPOSURE

me panic') and their noncatastrophic counterparts (e.g., 'People either won't notice or won't care if I panic'). Situational exposure exercises are also used to test beliefs about the usefulness of safety signals and safety behaviours (e.g., 'I will crash the car if I don't grip the steering wheel tightly'). Exposure exercises are also used to help the patient understand why exposure might not have worked in the past. Safety signals or safety behaviours may have prevented the patient from learning that the situations are safe.

Medical Contraindications Situational exposure exercises are safe for the great majority of patients. The only medical contraindications are serious diseases that prevent the patient from entering the situations. A history of seizures poorly controlled by medication would preclude a patient from operating a motor vehicle. Medical conditions that seriously limit mobility-such as severe obstructive lung disease-can prevent the patient from completing some forms of situational exposure (e.g., walking long distances from home). In practice, however, it is rare to encounter a patient who is medically unsuited for at least some form of situational exposure.

Procedures for Conducting Situational Exposure General Guidelines

There are several protocols for conducting situational exposure. Rather than describe them all, we will review the general guidelines for conducting various forms of situational exposure in CBT2. Here, exposure is based on a cognitive rationale: 'Remain in the situation until you can determine whether the outcome supports your catastrophic belief or the noncatastrophic alternative.' Many of the empirically derived guidelines for conducting exposure were based on a deconditioning rationale ('Remain in the situation until your fear subsides'). Despite the differences in rationales, the guidelines are broadly applicable to exposure exercises used in CBT2. Situational exposure exercises should be specific and well-defined, in terms of duration, situation, and what the patient must do (or not do). The exercises should be written down so that they are not forgotten by either patient or therapist. Predictions based on the catastrophic belief and its noncatastrophic alternative are made beforehand. The patient and therapist might generate an assignment such as the following:

SITUATIONAL EXPOSURE 353

'The purpose of this exercise is to see which prediction is most likely to be true: (1) "If I get anxious or panicky, I will lose control and jump off the bridge" (catastrophic belief) or (2) "If I get anxious or panicky I'll simply feel lousy; my feelings can't force me to jump off the bridge" (noncatastrophic alternative). On three occasions during the coming week-preferably during daylight hours-walk at a casual pace from one end of the Burrard Street bridge to the other, and then back again. Try not to distract yourself. Pay attention to how you feel and also observe what it's like to walk on the bridge. Notice how the passing cars cause the bridge to vibrate. When you reach the midpoint of the bridge pause for l0min and look over the railing. Observe the activities below, such as the yachts gliding under the bridge and the small ferries shuttling passengers from one side of the inlet to the other. When the l0min is up, continue walking to the other end of the bridge. Then repeat the exercise walking back the other way. Remember to complete the homework rating form and write down what you learned from the experience.'

The goal of situational exposure is to test beliefs, not to be free of anxiety or panic. In fact, panic can be a good thing because it provides further information for testing beliefs (e.g., 'During the height of panic I was shaking and crying, but the panic didn't make me jump off the bridge'). Thus, episodes of anxiety and panic are framed as valuable opportunities for learning (Craske et al., 1994). Situational exposure exercises can be practiced during therapy sessions and as homework assignments, using similar procedures and rating forms to those used for interoceptive exposure. Situational exposure exercises should be ones that patients can realistically accomplish, given their current levels of functioning. Patients should not be asked to attempt exercises that they are too frightened to complete. In selecting the appropriate difficulty level, the therapist can describe the exercise and ask the patient to rate their confidence about completing the task (0 = not at all confident, 100 = completely confident). Patients can be asked to complete tasks that they are quite confident of performing (confidence rating of 80 or higher). Exercises should be of sufficient duration and frequency to clearly test the catastrophic belief and its noncatastrophic alternative. The patient should actively participate in the exposure situation in order to be fully exposed to the various stimulus elements (Barlow & Cerny, 1988).

If

354

INTEROCEPTIVEAND SITUATIONAL EXPOSURE

Ms N. was asked to park her car at the back of the shopping mall parking lot so she would have to walk some distance to the mall and would not have rapid access to the 'safety' of her vehicle. Instead of racing through the mall, she was encouraged to stroll through, taking time to look through a number of stores, ask questions of sales clerks, and buy one or two items, and then have a snack at a mall cafe. Contrast this therapeutic exposure to her brief exposures prior to treatment, in which she would park her car as close as possible to the mall entrance, then race in to purchase what she needed, and then race out. Situational exposure should be comprehensive, with different exercises covering different manifestations of the feared situations. A patient who fears and avoids shopping at supermarkets, for example, would be advised to perform this activity in many different shopping conditions, including different stores (e.g., small vs. large stores), at different locations (e.g., near vs. far from home), at different times (e.g., times in which the store was busy vs. quiet), and with different purchases (e.g., few vs. many items). This ensures that learning generalizes across settings and difficulty levels. Frequency and Duration of Situational Exposure

Long, continuous periods of exposure tend to be more effective than short or interrupted periods (Chaplin & Levine, 1981; Marshall, 1985; Stern & Marks, 1973). Although short exposure sessions (e.g., 10-15min) can be effective if they strongly disconfirm catastrophic predictions, longer sessions (20-60min) are often necessary. The longer the exposure, the greater the opportunity of learning that the feared consequences do not occur. If the exposure session is short (e.g., 10-15min), the patient may conclude that the feared consequence could have occurred if exposure had lasted longer. With regard to the frequency of exposure exercises, the research has produced conflicting results. Some studies have found that the higher the frequency of exposure sessions, the greater the rate of dropouts (Emmelkamp & Ultee, 1974; Ernrnelkamp & Wessels, 1975), the greater the short-term response for treatment completers (Foa et al., 1980), but the higher the rate of relapse later on (Hafner, 1976; Jansson & Ost, 1982). However, none of these findings were replicated by Chambless (1990). Some situational exposure is better than none, and so agoraphobic patients should be counselled to practice the exposure exercises as frequently as they reasonably can, depending on the type of exercise and on the logistic constraints. Patients frightened of riding on buses might practice this activity four times per week. Patients frightened of riding as a passenger in a car might be able to practice this activity only one

SITUATIONALEXPOSURE 355

or two times each week, depending on the availability of someone to drive them. Graded versus Intensive Exposure

Situational exposure can be graded such that the patient works up a hierarchy of increasingly more frightening exercises. To facilitate hierarchy construction, the patient is asked to generate a list of fear-evoking exercises and to rate each in terms of the amount of fear or anxiety it evokes (e.g., using a 0-100 scale, 0 = no fear/anxiety, 100 = extreme fear/anxiety). The hierarchy is usually organized in terms of thematically related exercises. To illustrate, a patient might work through a hierarchy of driving tasks, ranging from mildly frightening situations (e.g., sitting in the driver's seat of a stationary car with the engine running) to intensely frightening ones (e.g., driving in evening rush-hour traffic along a particular stretch of freeway). Once an exercise is no longer fear-evoking the patient attempts the next step in the hierarchy. Notice here that the steps in a hierarchy can be subtasks of the exposure exercise at the top of the hierarchy, which in this case consisted of driving to and from work during rush hour. In terms of cognitive models of panic attacks and agoraphobia (Chapter 3), graded exposure enables patients to conduct increasingly stronger tests of their catastrophic beliefs and noncatastrophic alternatives, starting with mildly threatening exercises and gradually moving to more threatening (but objectively harmless) exercises. Table 13.3 gives an example of a hierarchy for graded situational exposure.

Table 13.3

Sample exposure hierarchy: Fear of travelling far from home

Exercise: Walking to ... and then returning 1. End of street (two blocks away from house)-accompanied by spouse 2. End of street-unaccompanied 3. End of street and one block around the corner (so house is not visible)-accompanied 4. End of street and one block around the cornerunaccompanied 5. To the nearby park (six blocks from house)-accompanied 6. To the nearby park-unaccompanied 7. To the grocery store (eight blocks from house)accompanied 8. To the grocery store-unaccompanied * Ranging from 0 = no fear/ anxiety, to 100 = extreme fear/ anxiety.

Distress level* 10 25 35 50 60 75 80 90

356

INTEROCEPTIVEAND SITUATIONAL EXPOSURE

The advantage of using graded situational exposure is that it teaches patients a skill that they can continue to use after finishing a course of CBT2. That is, the use of a hierarchy teaches patients a structured, stepby-step method for overcoming seemingly insurmountable fears. After the formal course of treatment ends, patients can continue to devise hierarchies for overcoming any remaining fears. A disadvantage of graded situational exposure is that it can be time consuming. If time is of the essence, then intensive exposure (flooding) can be attempted. This treatment involves situational exposure to stimuli at the very top of the patient's hierarchy (or hierarchies). Intensive approaches can be highly effective for patients who can tolerate intense anxiety or panic. Although quicker than graded exposure, intensive exposure does not teach patients a skill that they can use on their own. Patients often find it too fear-evoking to put themselves through a self-directed program of flooding. A further concern is that patients may be more likely to refuse or drop out of very intensive programs, compared to less demanding programs. Assisted Exposure There are numerous ways that patients can be assisted in completing situational exposure exercises. Among the more commonly used are (1) therapist-assisted exposure, conducted either from the clinic or from the patient's home, and (2) exposure in which the patient is assisted by a friend, partner, or family member. Assistance consists of initially accompanying the patient, and by providing prompting and encouragement for attempting the exposure exercises. As we saw in Chapters 5 and 6, the available evidence indicates that such assistance does not enhance treatment outcome for the average patient. However, there are cases in which some form of assistance is helpful, particularly for severe (e.g., housebound) agoraphobia in which previous efforts at self-directed exposure have failed. Assistance can be particularly encouraging during the initial attempts at exposure. Therapist-assisted exposure is also a good way for the therapist to learn about the patient's subtle avoidance or safety behaviours. There are several ways that the therapist can directly help the patient complete situational exposure exercises (Williams, 1990): •

The therapist can provide just enough assistance for the patient to perform the exercise, and then the assistance is faded out. If the task is to walk around a 10 block circuit from the patient's house, the therapist might accompany the patient all the way on the first trial. On the second trial the therapist might accompany the patient for 5 blocks, and

INTEGRATING INTERVENTIONS 357

•

•

on the third trial the patient might complete the circuit alone, with the therapist waiting outside the patient's home. On the fourth trial the patient might complete this task without the therapist's presence. The therapist uses the least amount of help needed so that patients attribute gains to their own performance rather than to the therapist's assistance. The therapist can demonstrate (model) how the exercises are performed. Here, the patient observes the therapist perform the exercise in real life or on videotape. For example, the patient and therapist might travel to the foot of a bridge, and then the patient would observe the therapist walk to the midpoint of a bridge without holding onto the railing, pause to lightly touch the railing (instead of gripping it tightly) and look over the edge. The therapist would then return to the foot of the bridge and the patient would perform the task. Alternatively, each step of the task could be modelled. For example, the therapist could walk to the midpoint, and then the patient would do so. The therapist would then lightly touch the railing and look over the edge. The patient then follows suit. The therapist can provide any physical and mechanical support that may be required. The therapist might offer his or her arm for assistance when the patient is first approaching a bridge railing. A therapist in a hospital setting who is helping a patient overcome fear of elevators might reserve a service elevator for some of the exposure session. The therapist and patient would have exclusive control of the elevator, and so it could be used for various exposure assignments (e.g., standing in a stationary elevator with the doors open, then with the doors closed, then traveling to the next floor, etc.).

A friend of the patient or significant other can be recruited as a 'coach' to provide prompting, encouragement, and guidance (Craske et al., 1994). This person also can be used as a stimulus in the exposure hierarchy, as illustrated in Table 13.3. Later in treatment, all forms of assistance should be faded out, along with other safety signals and safety behaviours. Otherwise, patients may attribute their gains to the aid they received, rather than to their own efforts (cf. Basoglu et al., 1994a).

INTEGRATING INTERVENTIONS Combining Interoceptive and Situational Exposure The full CBT2 package integrates interoceptive exposure, situational exposure, and cognitive restructuring. In this and the following sections we will discuss ways in which this is done. Interoceptive and situational

358

INTEROCEPTIVEAND SITUATIONAL EXPOSURE

exposure can be integrated by finding ways of deliberately inducing feared sensations in feared situations. Here are some examples: • • • • •

•

• • •

Wear a lot of clothes (to induce hot flushes, sweating, and palpitations) while out walking. Wear a tight collar or necktie (to induce choking sensations) while standing in an elevator or other enclosed space. Watch arousing movies (e.g., thrillers or horror films) while sitting in the middle of the aisle of a crowded cinema. Drive far from home on a warm day with the heater on and the windows rolled up (to induce hot flushes, sweating, and palpitations). Drink two or three cups of strong hot coffee (to induce trembling, palpitations, hot flushes, sweating) in a shopping mall cafe and then go for a walk through the mall. Go for walks during humid or cold weather. Walking from a warm environment (e.g., one's house) out into cold weather can induce dyspnea. This sensation also is induced by going from a cool environment out into humid weather. Scan labels on the supermarket shelf (to induce dizziness, derealization). Tense all one's muscles in a social situation (e.g., while waiting at a bus stop) to induce trembling that could be observed by others. Restrict nose breathing (Franklin, 1996). To induce dyspnea, occlude one nostril while in a feared situation (e.g., riding on a bus) for 1-2min, and then partially occlude the other nostril. This is done by pressing fingers against each side of the outside of the nose.

Integrated interoceptive and situational exposure exercises are performed according to the same guidelines as for other forms of exposure. Specific predictions are tested, derived from catastrophic beliefs and their noncatastrophic alternatives, and patients are to focus on both the situation and the sensations. As in all forms of exposure, patients should be encouraged to wean themselves off their safety signals and safety behaviours. The following homework assignment illustrates exposure without safety signals or behaviours:

'Walk seven blocks from home to the local running track. Don't take your Ativan, cell phone, or Medi-Alert bracelet. Sprint up and down the track five times. Make sure you exercise vigorously enough to make your heart pound, but please refrain from repeatedly checking your pulse. Then, as you walk home, pay attention to your body sensations and to your surroundings.'

INTEGRATING INTERVENTIONS359

Adding Cognitive Interventions to Exposure Exercises Preparing for Exposure

Cognitive interventions (Chapter 12) can be used before, during, and after all types of exposure exercises, whether they be interoceptive, situational, or combined exposure. When used for preparatory purposes, cognitive interventions are used (1) to help the patient make specific predictions for the forthcoming exposure, and (2) to help manage anticipatory anxiety (e.g., 'This exposure exercise is short-term pain for long-term gain'). Patients also can be asked to recall pleasant experiences with the exposure stimuli (Otto et al., 1996a). This is used to help the patient view the stimuli as harmless. Prior to performing an exercise to induce dizziness, for example, a patient might recall a childhood memory of deliberately spinning around to induce this sensation.

During Exposure

When used during exposure, cognitive interventions can be used in the form of coping statements or self-instructions (e.g., 'Just let the sensations happen, don't try to fight them'). Patients often have difficulty thinking clearly during their panic attacks, and so it is useful to have the statements or instructions written down. If coping statements distract the patient from the task at hand, then she or he should simply observe the fear-evoking stimuli, without resorting to coping statements or selfinstructions. However, an important exception is when the patient is about to escape the fear-evoking situation. The point of escape can be used as a cue for using self-instructions to help the patient remain in the situation (Craske et al., 1994); e.g., 'If I'm about to leave I must be predicting something awful will occur. I'll stay here for another minute to see what happens'. Paradoxical intention is another cognitive intervention used to facilitate exposure (Ascher, 1981; Ascher et al., 1986; Dattilio, 1987; Frankl, 1975; Michelson & Ascher, 1984; Michelson et al., 1988; Neeleman, 1992). It involves asking the patient to deliberately give up attempts to suppress or reduce feared sensations. Instead, patients are asked to make the sensations as strong as possible in an effort to bring on a feared consequence; e.g., 'Go ahead, try and make yourself have a heart attack'. Paradoxical intention can be used whenever sensations occur during the patient's daily life. For example, 'If you notice yourself feeling depersonalized, try to make the unreal feeling even stronger-try to make yourself go crazy.'

360

INTEROCEPTIVE AND SITUATIONALEXPOSURE

The use of humour can facilitate paradoxical intention; e.g., 'For homework, die at least three times a day of a heart attack' (Gerz, 1967, p. 202). Humour is useful because it helps patients distance themselves from their catastrophic beliefs, and thereby view their cognitions more realistically. If the paradoxical instructions are followed, the patient soon learns that feared consequences do not occur even when they try to bring them on. Paradoxical intention is useful in treating anxiety or panic triggered by relaxation sensations (Dattilio, 1994). It is also useful for treating fears of publicly observable arousal reactions, such as fears of blushing, trembling, sweating, or stammering. The patient learns that by bringing on the feared sensations, (1) feared outcomes, such as ridicule or rejection by others, typically do not occur, and (2) even if a feared outcome does occur, the experience is not as bad as the patient feared. After Exposure

When used after an exposure exercise, cognitive interventions can help patients objectively examine their performance during the exercise; e.g., taking the belief 'I failed at the assignment' and framing it more adaptively: 'I didn't stay in the supermarket for as long as we planned, but at least I tried. I can learn from this experience so I can do better next time'. Post-exposure cognitive interventions also can help the patient review the evidence for and against the catastrophic belief and its noncatastrophic alternative, and can be used for planning future exposure assignments; e.g., 'What do you think would have happened if you'd stayed in the supermarket longer?' If the patient attributed the absence of catastrophe to a 'lucky escape' then a more rigorous exposure exercise could be devised.

Treating Overvalued Beliefs Integrated cognitive and exposure interventions are particularly useful when patients have overvalued (i.e., strongly held and sustained) beliefs in the dangerousness of arousal-related sensations. Mr G. had frequent panic attacks that appeared to be triggered by catastrophic misinterpretations of chest pain. He was convinced the pain was due to an undiagnosed heart condition, and that he was 'fated' to die of a heart attack, just like his older brother. Reluctantly, Mr G. permitted his cardiologist to refer him for a course of CBT2. Although he had undergone many medical evaluations that all failed to detect any abnormality, Mr G. continued to believe he had undiagnosed heart disease. Accordingly, he avoided all

INTEGRATING INTERVENTIONS 361

forms of physical exertion so that his heart would not beat too fast. He also frequently checked his pulse for tachycardia. Mr G. surreptitiously obtained a supply of nitroglycerin, which he took whenever he thought his heart was beating too fast. This safety behaviour prevented him from learning that tachycardia was harmless. Mr G. was unwilling to drop this safety behaviour or engage in any form of interoceptive exposure. To treat a case like this it can be fruitful to draw on cognitive-behavioural approaches for treating delusions (Alford & Beck, 1994; Chadwick & Lowe, 1994; Watts et al., 1977) and on those used for treating hypochondriasis (Salkovskis, 1989; Warwick & Salkovskis, 1990). The strategies for treating overvalued beliefs in panic disorder are not greatly different from treating other catastrophic beliefs. The main difference is in the way the strategies are introduced and implemented. The following approaches can be used for treating overvalued beliefs: •

For patients who strongly believe they have an undiagnosed medical condition, CBT2 is presented as a 'no lose' option. Patients are told that if a trial of CBT2 is effective, then they will have overcome their problems. If CBT2 is ineffective, then further medical evaluation can be pursued. • The therapist attempts to identify all the relevant catastrophic beliefs, including ancillary beliefs. As per standard CBT2, the strength of each belief is rated on a 0-100 scale (0 = do not believe that_ is true, to 100 = completely believe it is true). • Unless the case formulation suggests a different approach, begin by applying cognitive restructuring to beliefs that are least-strongly held. These are likely to be most amenable to change. Then proceed up a hierarchy from weak to strongly held beliefs. Mr G. agreed to initially focus on his belief that 'Frequent pulse-checking is vital to my health'. • Do not directly confront the beliefs. Direct confrontation can strengthen overvalued beliefs. This is known as 'psychological reactance' (Brehm, 1962). Instead, ask the patient to simply explore alternative, noncatastrophic possibilities. • Challenge the evidence for the beliefs, not the beliefs themselves. Instead of challenging Mr G.'s belief that he had heart disease, the therapist challenged the notion that chest pain is evidence for this belief; 'What sorts of harmless things could be causing your chest pain?' Socratic dialogue can be used throughout; e.g., 'Did your cardiologist find evidence of heart disease?' Mr G. replied 'no.' The therapist replied 'Why do you think the cardiologist was so sure his tests ruled out heart disease?' • Look for naturally occurring fluctuations in belief strength. Was there a time in the past week when the patient doubted-even for a

362

•

•

INTEROCEPTIVEAND SITUATIONAL EXPOSURE

moment-the veracity of a given catastrophic belief? Explore and build on these doubts during the therapy session. During one session, Mr G. remarked that 'the last time I had chest pain I noticed that my chest muscles were twitching. This made me think that the pain came from these muscles, not from my heart'. This observation led to a review of the evidence for the hypothesis that his pain was stress-related. Once the patient has begun to doubt the veracity of a given belief, the therapist can suggest some simple, easily accomplished behavioural experiments. Mr G. began by walking up a flight of four steps to see what effect this had on his chest pain. He noticed that this produced no increase in pain, which was inconsistent with the belief that the pain was due to a cardiac condition. The following week he was willing to walk up eight steps, which similarly had no effect on this pain. Thus, behavioural experiments were gradually implemented at a pace acceptable to Mr G. Patients may feel embarrassed when they come to realize that their overvalued beliefs are false. Therefore, the therapist 'must be exceptionally sensitive to avoid threats to the patient's self-esteem ... and must apply standard cognitive therapy to restructure the negative self-concept' (Alford & Beck, 1994, p. 375).

IMPLEMENTING EXPOSURE: TROUBLESHOOTING Adherence Problems Adherence problems are the most common difficulties encountered when using interoceptive and/ or situational exposure. Patients may refuse to perform the exercises. Other, more subtle, adherence problems consist of using distraction during the exercise, or limiting the intensity or duration of exposure; e.g., a patient might avoid deep, rapid breathing during a hyperventilation exercise, and may stop the task as soon as any sensations are noticed. The best strategy for dealing with adherence problems is to avoid them before they occur. The following can help avoid these difficulties: • •

Ensure that patients know why the exercises are important. Periodically check their understanding. The interventions should not be too complicated. Understandably, patients might become confused if asked to do many things at once in a given situation. Patients also may get confused if the interventions appear to be at cross-purposes with one another, such as the

IMPLEMENTING EXPOSURE:TROUBLESHOOTING

363

combination of an interoceptive exercise (e.g., 'Take a shower with the doors and windows closed to test whether or not you will suffocate') and a coping exercise (e.g., slow, diaphragmatic breathing). Simpler is often better. During exposure to fear-evoking stimuli, patients could focus on just the sensations and the situation, using cognitive interventions judiciously, such as when they feel compelled to escape (e.g., 'Am I making a catastrophic prediction? What evidence do I have?'). • Ensure that the patient understands that the selection of exercises is a collaborative venture, negotiated between patient and therapist. Tasks can be readily modified in such a way that the patient can perform them. If the patient is too frightened to hyperventilate for 1 min, then a shorter duration can be used. Mrs J. was extremely frightened of performing this exercise. For her, interoceptive exposure was a slow process, starting with exercises that simply consisted of taking two deep breaths and then observing the effects. With time and patience, however, she was eventually able to complete a series of 2min hyperventilation exercises. • Perform the exercises yourself so the patient can see they are safe. As a rationale, the therapist might tell the patient, 'I wouldn't ask you to do something that I wouldn't do myself, so I'm going to show you how this exercise is done'. • Adherence can be maintained at an adequate level if the patient is sufficiently motivated to complete the various treatment exercises. To sustain motivation it can be useful to (1) reinforce (e.g., verbally praise) the patient for their efforts in completing each therapy assignment, and (2) have the patient self-reinforce for these efforts (e.g., praising oneself or using other incentives). Initially, reinforcement . from the therapist should occur frequently. As therapy progresses, reinforcement from the therapist can be faded out as the patient continues to use self-reinforcement.

If adherence continues to be poor then it is important to identify the source of the problem. Is the exercise a credible test of the patient's beliefs? If not, then the patient may regard the task as pointless. Is the timing right for using exposure? More time may need to be spent on developing a trusting therapeutic relationship. Also, more time may be needed for cognitive restructuring (Chapter 12) to address beliefs about the exercises and about the sensations they induce. If the patient is using safety signals or safety beha_viours then a hierarchy can be constructed, where the sources of safety are gradually faded out. For patients on PRN (as needed) medication, it is important to ensure that medication isn't used as a way of dampening the sensations.

364

INTEROCEPTIVE AND SITUATIONALEXPOSURE

When Exposure Induces Panic Attacks If a panic attack occurs during an exposure exercise, the patient should attempt to complete as much of the exercise as they can. Afterwards, the therapist and patient should conduct a 'post-mortem' of the experience, tracing the sequence of events in terms of Clark's cognitive model of panic. Links between body sensations, catastrophic cognitions, and emotions are identified, and cognitive restructuring can be used as needed (e.g., 'What evidence do you have that you were about to pass out?'). It is also important to identify whether the patient did anything to avoid the feared consequences. Patients who have panic attacks during interoceptive exposure often terminate the exercise as soon as they start to panic, believing the sensations would have catastrophic consequences. Aborted exercises should be discussed with the patient, pointing out that by discontinuing the exercise they failed to learn the true consequences of the sensations. The patient should attempt the exposure exercise again, preferably soon after the panic attack. If the patient is reluctant to do this, then alternative strategies can be explored, such as performing a shortened version of the exercise. While repeating the exercise, patients should simply observe any sensations arising, without trying to make them less intense. A related, useful strategy is for patients to relax passively while experiencing the sensations (e.g., 'Try to relax the muscles in your jaw, face, and neck while observing your heart pounding'). Otto et al. (1996a) referred to this as relaxing with the sensations.

When Interventions

Become Safety Behaviours

Patients sometimes misuse interoceptive exposure as a means of avoiding some feared consequence. Mr Y. initiated a program of regular jogging, ostensibly to challenge his catastrophic belief that 'palpitations lead to heart attacks'. However, two weeks of regular running was not successful in reducing panic frequency. A review of this problem revealed that the patient went jogging whenever he noticed his heart pounding. Mr Y. believed that jogging 'burned off' harmful stress hormones that were causing his heart to pound. He believed that this prevented him from having a heart attack. To challenge this belief, Mr Y. was advised to refrain from jogging and instead to simply sit and pay attention to his heart whenever he noticed it was pounding. Although he found this experience unpleasant, he soon learned that (1) his heart slowed down on its own without the need for jogging, and (2) palpitations did not lead to heart attacks.

SUMMARY AND CONCLUSIONS 365

Self-defeating

Efforts

at Managing

Anticipatory

Anxiety

In helping patients deal with anticipatory anxiety, it is important to identify and reduce any self-defeating strategies they might use. One is to attempt to suppress anxiety-evoking thoughts of an upcoming exposure exercise (e.g., via distraction). Suppression can lead to a paradoxical increase in thought frequency (see Chapter 12). Another self-defeating strategy for avoiding anticipatory anxiety is to impulsively launch into an exposure exercise, instead of implementing it in a planful fashion. Ms W. attempted to use the impulsive approach to overcome her fear of enclosed spaces. While shopping in a department store one day she suddenly decided, without forethought, to enter a crowded elevator and ride to the top floor. She merged with a crowd of shoppers pouring into the elevator. As the doors closed and people jostled against her, Ms W. became overwhelmed with fear that she would suffocate. Her fearful thoughts rapidly spiraled into panic. Stricken with terror, she fled the elevator at the next floor, none the wiser for her experience.

SUMMARY AND CONCLUSIONS Exposure exercises are potent ways of producing cognitive change. Their efficacy may be greater still when interoceptive and situational exposure are integrated, and when cognitive interventions (Chapter 12) are judiciously used. There are numerous protocols for implementing exposure exercises. This chapter reviewed and illustrated the widely used procedures, and described solutions to common problems in implementing these methods.

ADJUNCTIVE COGNITIVEBEHAVIOURAL INTERVENTIONS Cognitive interventions and exposure exercises are the main strategies used in second generation cognitive-behaviour therapy (CBT2). However, other cognitive-behavioural interventions also can be useful, depending on the nature of the patient's problems. This chapter begins by describing breathing retraining and relaxation training, which are often usedand misused-in the treatment of panic disorder. Other interventions reviewed in this chapter are vagal innervation methods, interpersonally focused interventions (e.g., behavioural couples therapy), and cognitivebehavioural strategies for improving quality of life.

BREATHING RETRAINING Indications and Caveats Breathing retraining instructs patients in slow diaphragmatic breathing. It was once widely used in panic protocols, because it was assumed that hyperventilation played a major role in producing sensations that are catastrophically misinterpreted in panic disorder. Eliminating the feared sensations was used as a way of reducing panic frequency. As it became apparent that hyperventilation was not as important as previously thought (Chapter 6), training in diaphragmatic breathing became less important. These days, breathing retraining should be reserved for the minority of panic patients who have one or more of the following: (1) chronic hyperventilation, (2) a tendency to episodically hyperventilate (e.g., when under stress), or (3) recurrent chest tightness or chest pain due to a tendency to breathe with the chest rather than the diaphragm. In each of these cases breathing retraining can be used to alleviate unpleasant but harmless sensations, including those due to hyperventilation (e.g., dizziness, dyspnea) and those due to chest breathing (e.g., chest pain). Training in diaphragmatic breathing is used to slow the respiration

1

BREATHINGRETRAINING 367

rate, thereby eliminating hyperventilation, and to replace the habit of chest breathing with the tendency to breathe from the diaphragm. With repeated practice, slow diaphragmatic breathing becomes habitual. Acute hyperventilators can be taught to identify stimuli that trigger hyperventilation (e.g., stressful events) and to implement slow, diaphragmatic breathing at these times. For acute hyperventilators who overbreathe mainly while speaking, they can be taught to talk more slowly and to avoid gasping between sentences (Salkovskis & Clark, 1986). There are several ways of assessing the role of hyperventilation in panic disorder (e.g., Barlow & Craske, 1994; Clark, 1989). One is to ask patients to demonstrate how they breathe when they are panicking. Another is to ask patients to hyperventilate for 1-2min and then assess their emotional reactions and the similarity of the resulting sensations to sensations of naturally occurring panics. A third approach is to ask patients if they often feel short of breath (even when not panicking), and whether they tend to yawn or sigh a lot, or take in gulps of air. An important caveat in using breathing retraining is that breathing exercises can become a safety behaviour. That is, patients may use them to avoid or escape feared sensations, and thereby fail to learn that the sensations are harmless. Related safety behaviours include procedures to abort hyperventilation, such as breathing into a paper bag or through cupped hands. If breathing retraining is used in CBT2, it is important that the patient learn that the feared sensations are harmless. Breathing retraining can be incorporated into CBT2 in a number of ways. A common approach is to start out by teaching breathing retraining as a means of providing quick relief from feared sensations. These rapid benefits can enhance treatment credibility and thereby increase the patient's motivation to work at therapy. Patients are informed that breathing retraining is used only because the sensations might be unpleasant or unwanted; not because they are dangerous. As treatment progresses, cognitive interventions and exposure exercises (Chapters 12 and 13) are implemented to reduce the tendency to catastrophically misinterpret arousal-related sensations. Then the patient can be asked to refrain from using the breathing retraining exercises and to deliberately hyperventilate in order to fully learn that the sensations have no harmful effects. As noted in Chapter 13, the mere ability to turn the feared sensations 'off' (via a breathing exercise) can help persuade some patients that the sensations are harmless ('How could the sensations be dangerous if they are so easily removed?'). If the patient continues to have respiratory problems (recurrent hyperventilation or recurrent pain from chest breathing), and

368

ADJUNCTIVE COGNITIVE-BEHAVIOURAL INTERVENTIONS

if the patient regards the resulting sensations as unwanted but harmless, then a breathing exercise could be reinstituted.

Breathing Retraining Procedures Diaphragmatic breathing consists of a number of simple exercises that can be taught in a couple of sessions. Some of the exercises are introduced in the first session and set as homework. In the second session the homework is reviewed, problems are addressed, and then patients are taught to decrease their respiration rate while diaphragmatically breathing. Details on the procedures are as follows. These have been synthesized from several sources (Andrews et al., 1994; Barlow & Craske, 1994; Bonn et al., 1984; Clark, 1989; Clark & Salkovskis, 1987; Clark et al., 1985; Craske et al., 1994; Otto et al., 1996a; Rapee, 1985; Rapee & Barlow, 1991; Timmons, 1994). •

•

•

•

In the first session of breathing retraining, patients are educated about the distinction between 'chest breathing' and 'diaphragm breathing.' Patients are informed about how chest breathing can produce harmless but uncomfortable sensations such as chest tightness or pain, and is associated with hyperventilation. The therapist should emphasize that hyperventilation is harmless, but can cause unpleasant sensations, such as dizziness or shortness of breath. An overview of breathing retraining is presented, along with a written description. The therapist demonstrates the two types of breathing, and demonstrates how to best observe the difference between the two. Placing a hand to one's chest and the other on one's stomach (in order to notice differences in movement), the therapist demonstrates how the ribcage moves upwards and outwards during chest breathing. Then, the therapist demonstrates how the stomach but not the chest moves in and out during diaphragmatic breathing. Patients then practice the exercise, placing their hands over their own chest and stomach, practicing the two types of breathing and observing the differences. Patients are then asked to practice diaphragmatic breathing for 10min in the therapist's office. Respiration should be at a comfortable rate, breathing through the nose rather than the mouth. The goal is to achieve a slow, smooth, shallow pattern of breathing. Patients can practice diaphragmatic breathing either by sitting in a chair or by lying in a supine position. If the patient has a longstanding habit of chest breathing, then the abdomen may have become deconditioned and they may have difficulty breathing via the diaphragm. In this case the patient can be asked to recline with a tele-

BREATHING RETRAINING 369

•

•

•

•

•

phone book placed on the abdomen and then to attempt abdominal breathing. This weightlifting exercise strengthens the diaphragm and increases the patient's awareness of what it is like to breathe with the diaphragm (Timmons, 1994). The patient should observe the abdomen rising and falling with each breath, while breathing with the diaphragm and keeping the chest still. Patients are instructed to notice the cool air slowly coming in through the nostrils as they inhale, then pause for 1-2 s, and then notice the warm air slowly 'oozing' out as they exhale (Rapee & Barlow, 1991). Initially, the target respiration rate is about 12 breaths/min (i.e., inhaling for 2s, pausing for ls, and exhaling for 2s). A meditational component also can be included, where the patient imagines saying 'one' on inspiration, and 'relax' during the slow expiration (Barlow & Craske, 1994; Rapee, 1985). Then 'two' on inspiration, and 'relax' on expiration. Counting continues up to 'ten' and then recommences at 'one,' and so forth for l0min. Alternatively, the patient can imagine saying 'reeee ... laaax.' The patient says the first syllable on inhalation, then pauses, and says the second syllable on exhalation (Otto et al., 1996a). For homework, the patient is requested to sit or lie down in a quiet place at home, free from distractions, and practice the diaphragmatic breathing two or three times/day for l0min each time. Breathing should be at a comfortable rate. Patients are instructed to use a monitoring form to record the duration of each practice period. Other ratings can also be made, such as the degree of relaxation, ease of breathing, types of sensations experience (Barlow & Craske, 1994; Clark, 1989). The importance of regular practice is emphasized. A Smin pacing tape can be recorded during the session to be used as a homework aid to help the patient breathe at the desired rate (Clark, 1989). The pacing on the tape is at the patient's usual rate of respiration or, if possible, a little slower. The patient follows the tape for 5 min, and then for the next Smin practices without the tape. Such tapes typically consist of the therapist saying 'in' for 2 s, pausing for 1 s, and then 'out' for 2s (12 breaths/min). Pacing can be faster if this is too slow for the patient. By prolonging the articulation of 'in' and 'out,' gentle and extended inspirations and expirations are achieved, in contrast to the sharp gasps that characterize hyperventilation. Application training usually commences in the second treatment session. To train patients to abort episodes of hyperventilation, they are aske.d to hyperventilate for 1-2min and then practice controlling their respiration by implementing slow diaphragmatic breathing. This is practiced several times during the therapy session. Other forms of application training also can be used, such as practicing slow

370

•

•

ADJUNCTIVE COGNITIVE-BEHAVIOURAL INTERVENTIONS

diaphragmatic breathing while standing up, and while walking around (Otto et al., 1996a). Patients should be advised that their breathing rate should always feel comfortable; if they are engaging in some physical activity then their respiratory rate should increase to match the body's oxygen requirements. For homework, the patient continues practicing slow diaphragmatic breathing while sitting at home for l0min two or three times/day. The respiration rate can be slowed down to 8-10 breaths/min. A slower pacing tape can be recorded for use on alternate days for the following week, and then the tape is discontinued. Patients are also asked to complete application training homework. This consists of practicing slow diaphragmatic breathing in a variety of everyday situations; while watching TV, waiting in queues in stores, sitting in the car at traffic lights, sitting in crowded cinemas, and so forth. Patients gradually practice in increasingly challenging situations. Various stimuli can be used as cues or reminders to practice a breathing exercise, including particular situations (e.g., stressful circumstances), and sensations (e.g., dizziness, chest tightness, or other sensations suggesting that respiration may have increased).

Troubleshooting As noted earlier, one of the main problems with breathing exercises is that they may become safety behaviours, thereby preventing the patient from learning that the feared sensations are harmless. If this is the case then breathing retraining should be discontinued. Even when breathing retraining is being used for the right reasons-to avoid unpleasant but harmless sensations-difficulties may still arise. Common problems and suggested solutions are as follows. •

•

•

Problem: Dizziness or breathlessness are induced or worsened by breathing retraining. Solution: Increase the rate or depth of respiration until the patient feels comfortable. Problem: Difficulty finding time to practice. Solution: Help the patient examine their goals and priorities. Is it important to be rid of unpleasant but harmless sensations? If the answer is 'no,' then there is no harm in discontinuing breathing retraining. If the answer is 'yes,' then the patient can be helped to schedule fixed times to practice. Problem: Difficulty concentrating during the exercises. Solution: Continued practice, using a pacing tape for the first 5 min and then practicing without the tape for the remaining Smin. Patients

RELAXATIONTRAINING 371

should be advised that it can take time and practice to become proficient at the breathing exercises, and that they should not give up if they don't immediately succeed. • Problem: Breathing exercises become a source of frustration. Solution: The patient may be striving to do the breathing exercises 'perfectly'; e.g., forcing the abdominal muscles out and pulling them in, rather than letting the abdomen naturally fall outward during inhalations and relax back during exhalations (Otto et al., 1996a). If this occurs then the patient can be instructed to simply observe, without trying to control, the abdomen going in and out during respiration. Patients should be informed that the goal is to be 'good enough' at the exercises, not to do them perfectly. Homework assignments may involve the instruction to perform the exercise imperfectly. 'Counterdemand' instructions can be used; 'You'll need a few weeks of practice before you notice any beneficial effects'. • Problem: Focusing on breathing induces anxiety or panic. Solution: The associated beliefs should be examined. Some patients become frightened or panic because breathing exercises require them to focus on feared sensations, which leads them think about all the catastrophes that could befall them (e.g., choking, suffocating). Other patients fear that slowing their breathing may cause them to underbreathe. Ms W. believed that breathing retraining would cause her to lose automatic control of her respiration, leading her to stop breathing when she fell asleep. Once beliefs such as these are identified, they can be challenged via cognitive interventions and behavioural experiments (Chapters 12 and 13). The patient can be encouraged to continue to practice breathing exercises as a way of testing catastrophic beliefs (e.g., 'I will lose automatic control of my breathing') and noncatastrophic alternatives (e.g., 'My body will continue to breathe regardless of what I do'). Breath-holding is another exercise that patients can use to test the belief that breathing retraining will cause them to stop breathing (Clark & Salkovskis, 1987). Patients will soon learn that the physiological respiratory drive compels them to breathe.

RELAXATION TRAINING Indications and Caveats Several relaxation methods have been used in conjunction with other CBT2 interventions to treat panic disorder. Some relaxation methods, such as progressive muscle relaxation, have been used mainly to reduce tension and arousal, including arousal arising from benzodiazepine withdrawal (Craske et al., 1994; Otto et al., 1996a). Other methods, such as applied

372

ADJUNCTIVECOGNITIVE-BEHAVIOURAL INTERVENTIONS

relaxation, have been used to reduce anxiety and to abort panic attacks (Ost, 1987). As we will see in the following sections, there are pros and cons for using relaxation for each of these purposes.

Using Relaxation to Reduce Tension and Anxiety Progressive muscle relaxation training is effective in reducing muscle tension and anxiety (e.g., Barlow et al., 1989). Thus, relaxation training can be used to treat panic patients who are chronically tense or hyperaroused. This approach is also useful for panic patients who have comorbid psychophysiological conditions such as pain disorder or irritable bowel syndrome, which may be exacerbated by increases in anxiety or muscle tension. Relaxation methods can be used to directly treat these conditions, and to offset the transient increases in anxiety and tension that arise from exposure interventions for panic disorder. Relaxation exercises also can be used to replace maladaptive strategies for coping with anxiety or panic, such as abuse of alcohol or other sedatives. A relaxation exercise can be used to calm the patient down if she or he is too anxious to concentrate on whatever cognitive interventions are being used in the therapy session. Relaxation also can be used if the patient feels too anxious to undertake in-session exposure assignments, such as 1-2 min of hyperventilation in the therapist's office. Similarly, relaxation exercises can be used to facilitate homework exposure assignments, particularly when anticipatory anxiety is so high that it interferes with the patient's willingness to attempt the assignments. In these applications, a short (e.g., lOmin) relaxation exercise can be used, such as release-only relaxation, and then a cognitive restructuring or exposure exercise is attempted. Relaxation training can sometimes lead patients to see that arousal-related sensations are harmless. This may happen because relaxation exercises can convince some patients that the feared sensations are simply the harmless by-products of tension-sensations that can be simply turned 'off' if desired. However, sometimes patients use relaxation methods as safety behaviours, where they strive to reduce arousal because they fear its consequences. This is why relaxation training is often omitted from CBT2 protocols (e.g., Wells, 1997). Instead, cognitive interventions are used to challenge beliefs causing anxiety or tension. A problem with this approach is that cognitive interventions are not as quick as relaxation exercises in reducing arousal. Rapid relief is necessary to prevent some patients from prematurely dropping out of treatment.

If relaxation training is used, it is important to check whether or not the relaxation exercises interfere with the reduction in catastrophic beliefs. If

RELAXATIONTRAINING

373

interference occurs, then the patient can be asked to periodically refrain from using the relaxation exercises in order to prove to themselves that anxiety and tension do not have catastrophic consequences. It is also important to inform patients that some degree of anxiety is normal and adaptive. By allowing themselves to periodically experience anxiety-in full-patients are reminded that this is emotion is harmless. Using Relaxation to Abort Panic Attacks The indications and caveats for using relaxation training to abort panic are similar to those mentioned above. Relaxation methods can be used early in treatment to rapidly block panic attacks. This can be useful in motivating patients to complete treatment, particularly those who are skeptical about psychological treatments. However, even under these circumstances it is useful to eventually fade out or temporarily suspend use of the relaxation exercises, so patients can learn that arousal-related beliefs are harmless.

Protocol for Progressive Muscle Relaxation Progressive muscle relaxation (Bernstein & Borkovec, 1973; Wolpe & Lazarus, 1966) is a component of the Albany CBT2 protocol (Craske et al., 1994) and a component of applied relaxation (Ost, 1987). Progressive muscle relaxation involves training in a series of exercises, administered in the following order. Tense-release Relaxation This involves working through various muscle groups, as described in Table 14.1. The exercise can be practiced in a comfortable chair. Each muscle group is tensed for 5 s and then relaxed for 5-15 s. Muscle tensing and releasing is used to deepen relaxation, and to sharpen the patient's ability to identify the difference between tension and relaxation. The better the ability to detect tension, the greater the chances of implementing relaxation when tension first develops. Otherwise, tension may build up, undetected, until it become so intense that it produces headaches, muscle cramps, or other tension-related problems. The goal of the tense-release relaxation is not to tense the muscles until they hurt. If pain or cramps occur-or if the patient has a history of some sort of recurrent pain (e.g., low-back pain)-then the afflicted muscles should be weakly tensed or not tensed at all. People wearing contact lenses should be instructed not to tightly squint their eyes.

37 4 ADJUNCTIVE COGNITIVE-BEHAVIOURAL INTERVENTIONS Table 14.1

Sample protocol for tense-release relaxation

Muscle group

Activity

Session 1 1. Fingers and hands 2. Wrists and forearms 3. Biceps 4. Shoulders 5. Forehead 6. Eyes 7. Jaw 8. Tongue 9. Throat 10. Neck

Clench each hand into a fist, one hand at a time Bend wrists back toward forearms Tense both biceps ('Strong Man' act) Hunch shoulders Raise eyebrows and then frown Squint eyes Jut lower jaw outward Push tongue against roof of mouth Yawn Gently rotate neck: left, right, back, forward

Session 2 1. Chest cavity 2. Chest and upper back 3. Abdomen 4. Lower back 5. Thighs and legs-I 6. Thighs and legs-II 7. Toes and feet-I 8. Toes and feet-II

Take a deep breath and then slowly exhale Pull shoulders back and push chest outwards Push out stomach and then pull it all the way in Arch lower back Knees locked, feet pointing upwards Knees locked, feet pointing down Toes curled down Toes curled upward

From Wolpe, J., & Lazarus, A. A. (1966). Behaviourtherapytechniques:A guide to the treatment of neuroses.Oxford: Pergamon.

Tense-release relaxation begins by having the patient work through all the muscle groups (Table 14.1) during the therapy session and as homework. Then the patient practices by relaxing more rapidly by using fewer and fewer muscle groups. The assumption is that relaxation will spread to the remaining groups. Homework consists of practicing for 15-30min, twice per day, for at least two weeks. As with other relaxation methods, the exercises can be audiotaped to facilitate homework practice. The audiotape is then faded out as the patient learns to relax without the tape. The importance of regular practice is emphasized, and patients receive homework monitoring forms to record each practice session. The forms record the number of practices per day, and ratings of the degree of relaxation and the degree to which the patient is able to concentrate on the exercise. Release-only Relaxation

The next step is to repeat the protocol for tense-release relaxation, but this time omitting the tensing portion of the exercise. Patients are simply asked

RELAXATIONTRAINING 375

to focus on releasing tension from the various muscle groups, starting at the top of the head and working down to the toes. If patients notice tension in a particular muscle group, then they can briefly increase the tension of those muscles, and then release the tension. As before, the exercise is practiced in the session and as homework. The latter should be done twice per day for at least a week. The protocol can be audiotaped to facilitate practice, and then the tape is faded out. The following is a sample script of release-only relaxation, which the therapist uses to guide the patient through the exercise.

'Breathe with calm, regular breaths and feel how you relax more and more with every breath ... Just let go ... Relax your forehead ... eyebrows ... eyelids ... jaws ... tongue and throat ... lips ... your entire face ... Relax your neck ... shoulders ... arms ... hands ... and all the way out to your fingertips ... Breathe calmly and regularly with your stomach all the time ... Let the relaxation spread to your stomach ... waist and back ... Relax the lower part of your body, your behind ... thighs ... knees ... calves ... feet ... and all the way down to the tips of your toes ... Breathe calmly and regularly and feel how you relax more and more by each breath ... Take a deep breath and hold your breath a couple of seconds ... and let the air out slowly ... slowly ... Notice how you relax more and more.' (Ost, 1987, p. 409)

Cue-controlled Relaxation

The next step is cue-controlled relaxation. This is a brief (30s) relaxation exercise that can be used repeatedly throughout the day in all kinds of situations, including those in which patients feel anxious or panicky. Cuecontrolled relaxation focuses on breathing, and can incorporate slow diaphragmatic breathing (Ost, 1987). When used during the therapy session, cue-controlled relaxation starts by having patients relax using release-only relaxation. Then patients are instructed to imagine saying 'inhale' as they breathe in, and 'relax' as they breathe out. Patients are instructed to 'let go' of any muscle tension as they breathe out. Cuecontrolled relaxation is practiced several times in the therapy session and then practiced as homework. Patients should practice cue-controlled relaxation frequently and regularly. Coloured dots or other salient stimuli are used as cues or reminders to practice relaxation. To illustrate, a blue dot might be stuck on a patient's cell phone as a reminder to use the relaxation exercise. Cues are changed periodically (e.g., blue dots are replaced with red dots) to prevent habituation to the reminders.

376

ADJUNCTIVECOGNITIVE-BEHAVIOURAL INTERVENTIONS

Applied Relaxation Applied relaxation (Ost, 1987) builds on progressive muscle relaxation in various ways. Although it might appear that applied relaxation is a standalone treatment for panic disorder, it actually contains interoceptive and situational exposure, which are used as part of training to use relaxation methods under stressful circumstances. Accordingly, applied relaxation is more properly viewed as a set of interventions. The following is a summary of this multicomponent protocol, which has been shown to be effective in reducing panic disorder (Ost, 1988; Ost & Westling, 1995). Psychoeducation (1 Session)

Patients are instructed that the purpose of applied relaxation is to teach a coping skill to enable them to (1) recognize the early signs of arousal, and (2) relax rapidly in order to counteract, and eventually to abort panic attacks. The goal is to relax within 20-30s. Clark's (1986) model of panic is presented, and it is emphasized that arousal-related sensations are unpleasant but harmless. The treatment program is outlined, and patients are instructed in methods for monitoring symptoms, such as a panic diary. Progressive Muscle Relaxation: Tense-release and Release-only Relaxation (4-7 Sessions)

These exercises are the same as described earlier in this chapter, with the exception that no relaxation tapes are used. The various exercises (e.g., tense-release relaxation) are practiced two or three times per therapy session. Differential Relaxation (2 Sessions)

The primary aim of this exercise is to make relaxation increasingly more portable. A secondary aim is to teach the patient to perform activities with a minimum of muscle tension. The exercise starts with cue-controlled relaxation, relaxing from head to foot, while seated in a comfortable chair. Then patients are instructed to perform a series of exercises in which activities are performed in which only the essential muscles are tensed. Examples include: (1) opening one's eyes and looking around while relaxing all the muscles except those required to sit upright and look about; (2) lifting one arm, and then the other, lifting one foot, one leg and then the other, while relaxing all the muscles that are not needed for those activities; and (3) tensing the biceps while keeping the hands relaxed. Patients then practice relaxing under increasingly more active circumstances; e.g., walking

l

I

I

I

l

BREATHINGRETRAINING 377

about while relaxing all the muscles except those required for ambulation. For homework, the exercises are practiced in various settings four times per day for 5min each time. Rapid Relaxation (1-2 Sessions)

This exercise is essentially the same as cue-controlled relaxation. Rapid relaxation is used to teach patients to relax in everyday nonstressful situations, and to further reduce the time it takes to relax. The goal is to relax within 20-30s. Rapid relaxation consists of (1) taking 1-3 deep breaths and slowly exhaling, (2) thinking 'relax' before each exhalation, and (3) scanning one's body for tension and trying to relax as much as possible. The exercise is initially practiced in nonstressful situations, such as the therapist's office, and then practiced 15-20 times per day in everyday situations. Cues (e.g., coloured dots) are used to remind patients to practice. Application Training (2-3 Sessions)

Application training combines relaxation exercises with situational and interoceptive exposure; e.g., rapid relaxation for 30 s followed by 2 min of hyperventilation, and then by another 30 s of rapid relaxation. Relaxation is practiced before, during, and after situational exposure. As with all relaxation methods, the importance of continued practice is emphasized. Maintenance Program (1-2 Sessions)

Each patient is advised to develop the habit of scanning their body at least once a day, and to implement rapid relaxation if any tension is detected. Differential relaxation also should be practiced twice per week. Other maintenance strategies, applicable to all CBT2 interventions, are also used (see Chapter 15).

Other Relaxation Exercises In Situ Isometric Relaxation

Franklin (1986, 1996) developed a portable, rapid relaxation procedure based on isometric exercises. Known as in situ isometric relaxation, it is used whenever patients feel tense. It is unobtrusive because it involves tensing opposing muscles so patients can tense and release with no apparent movement. Franklin developed various forms of isometric relaxation, including an exercise for standing and one for sitting in a public place. Here is an example of the latter:

378

ADJUNCTIVE COGNITIVE-BEHAVIOURAL INTERVENTIONS

Take a small breath and hold it for up to 7 s. At the same time do the following: •

•

•

Cross your feet at the ankles and (1) press down with the upper leg while trying to lift the lower leg, (2) push the legs sideways in opposite directions while keeping them locked at the ankles, and now (3) combine these two leg movements. Then, tense your hand and arm muscles by doing one of the following: (1) place hands comfortably in the lap, palm against palm, and press down with the top hand while trying to lift the lower hand, (2) place hands under thighs and try to lift up, (3) place hands under the side of a chair and pull self down in the chair, (4) grasp hands behind the chair and try to pull them apart while simultaneously pushing them in against the back of the chair; (5) place hands behind head, interlock fingers and while pushing head backwards into hands try to pull hands apart (continue to interlock fingers). On releasing the tension, (1) breathe out slowly, and quietly say the word 'relax' to yourself after every third or fourth breath, (2) separate your legs and hands, moving them to a new position, (3) either close your eyes or let them rest passively on something such as a spot on the floor or wall.

Only one of the hand exercises is used at any one time, depending on which is most appropriate to the situation. The exercises are practiced several times in succession, as needed. The procedure initially takes 3-5 min to relax fully, and with practice relaxation can be achieved in under a minute. The advantages of these exercises are that they are quick, portable, and versatile. A disadvantage is that they are complicated, which may be offputting for some patients. Preliminary evidence suggests an in situ isometric relaxation is useful in reducing anxiety and panic (Franklin, 1989), although further evaluation is warranted.

Relaxation Cue Procedure A promising alternative to Franklin's isometric method is the relaxation cue procedure(Otto et al., 1996a). The latter takes only 10-15s to apply and is practiced a minimum of four times per day. It entails a brief tension induction followed by relaxation:

RELAXATIONTRAINING

379

Take a deep breath, wrinkle your face, and tense your shoulders by shrugging them up toward your ears. Hold all the tension for 5 s and then quickly it let go as you exhale.

Compared to Franklin's isometric relaxation, the relaxation cue procedure is simpler and easier to learn. Clinically, patients have no difficulty learning this procedure and find it effective in reducing tension. Unfortunately, there has yet to be any published research on its efficacy. Choosing Among Relaxation Procedures

An important issue in choosing among protocols is how much time-if any-to devote to relaxation training. The decision to include relaxation training should be based on the case formulation (Chapter 10), and by considering the indications and caveats for relaxation training described earlier in this chapter. If relaxation is to play a modest role in treatment, then the therapist might as well use the relaxation procedures described in the Albany protocol (Craske et al., 1994), along with the relaxation cue procedure (Otto et al., 1996a). If relaxation training is to be the emphasis of treatment-as may be the case if the therapist plans to use relaxation to simultaneously treat panic disorder and a comorbid pain disorderthen applied relaxation could be used, with tense-release relaxation omitted if necessary. Applied relaxation is useful because there is some evidence that panic attacks are more effectively treated by applied relaxation than by simply relying on progressive muscle relaxation (Chapter 6).

Troubleshooting •

Problem: 'Itching, twitching, and bitching' (Clarke & Jackson, 1983). Solution: When patients practice release-only relaxation they often become aware of itching sensations, muscle twitching, and intrusive thoughts or images (e.g., thoughts about the chores they have to do later that day). If patients feel itchy they shouldn't desperately try to ignore the itch; they should scratch. Patients should be encouraged to regard muscle twitching as a harmless expression of residual tension that will gradually abate as they become more proficient at relaxation. When intrusive thoughts or images occur, or if patients become aware of other distractions (e.g., muffled voices from the office next door), they should mentally acknowledge the distracters, and then gently

380

•

ADJUNCTIVE COGNITIVE-BEHAVIOURAL INTERVENTIONS

draw their attention back to the task of relaxing. It doesn't matter how often they become distracted. The important thing is that they repeatedly refocus their attention on the relaxation exercise (Ost, 1987). Problem: Some patients are embarrassed about performing the relaxation exercises, particularly exercises that require them to relax in public places (e.g., rapid relaxation). These patients worry that other people will notice them practicing the exercises. Solution: The therapist can point out that relaxation can reduce embarrassment (Ost, 1987). If embarrassment is a problem, ask the patient to practice the exercises in front of you, in order to check whether they are performing the exercises correctly and unobtrusively (Ost, 1987). Concerns about embarrassment also can be addressed by accompanying the patient to a public place (e.g., a marketplace) and then have the patient observe you perform one of the portable relaxation exercises, such as rapid relaxation. Then the patient and therapist can discuss whether the exercise attracted the attention of others.

Other Problems and Solutions Many of the other problems encountered in relaxation training are similar to those encountered in breathing retraining: (1) relaxation training becoming a safety behaviour, (2) difficulty finding time to practice, (3) difficulty concentrating on the exercise, (4) relaxation exercises becoming a source of frustration, and (5) relaxation-induced anxiety or panic. The solutions to these problems are the same as those for breathing retraining. If relaxation exercises become safety behaviours, then they should be faded out so that patients learn that arousal-related sensations are harmless. If patients have difficulty finding time to practice, then their daily schedules can be reviewed to find opportunities for practicing relaxation. If patients have difficulty concentrating on a relaxation exercise, then they could practice with the aid of a relaxation tape. If patients become frustrated because they are trying to perform the relaxation exercises perfectly, then they can be asked to expend less effort, and to perform the exercises 'imperfectly'. Relaxing is like getting to sleep at night; the harder one tries, the less one succeeds. Patients also can be given the instruction that deep relaxation will not be attained until they have practiced for a few weeks (Ost, 1987). Relaxation-induced anxiety or panic (Adler et al., 1987; Cohen et al., 1985; Heide & Borkovec, 1983, 1984) often arises from the catastrophic beliefs about relaxation sensations. Such sensations include warmth, numbness, falling or floating feelings, depersonalization, and slowed heart rate. The salience of these sensations is increased because many relaxation exercises are practiced in quiet places with limited distractions, and because the

VAGAL INNERVATIONTECHNIQUES 381

person's attention is directed inwards (Clark, 1988; Wells, 1990). Beliefs about relaxation sensations include misconceptions that the sensations will lead to death (e.g., 'My heart will slow down until it stops'), beliefs about insanity ('Unreal feelings will become so strong that I lose touch with reality'), and beliefs about loss of control ('If I "let go" of muscle tension then I will lose control of my behaviour'). Relaxation-induced anxiety and panic can be treated by simply providing corrective information about the harmlessness of the sensations; e.g., increased peripheral blood flow and reduction in muscle tension produces many of the sensations. The occurrence of the sensations can be reframed as signs of successful relaxation (Otto et al., 1996a). If problems persist then relaxation can be used as a form of interoceptive exposure (Chapter 13) to test the catastrophic beliefs and their noncatastrophic alternatives. Examples of the latter are: 'My heart slows down, but does not stop,' and 'I feel temporarily unreal but I can snap out of it at any time, no matter how strong the feelings of unreality'.

VAGAL INNERVATION TECHNIQUES There are several methods for rapidly stimulating the vagal receptors, thereby producing short-term reductions in heart rate. The methods have been successfully used in treating panic disorder (Sartory & Olajide, 1989) and in helping patients taper offbenzodiazepines (Elsesser et al., 1996: see Chapter 7). These vagal innervation techniques include: • • •

Massaging the carotid artery (i.e., rubbing one side of the neck). Pressing on one eye while exhaling. The Valsalva manoeuvre: deeply inhaling and then increasing the pressure in the chest by tensing the intercostal and abdominal muscles while pushing air against the closed glottis for 10-15 s (Mathias & Bannister, 1992; Wieling, 1992).

Of these methods, the Valsalva manoeuvre produces the most rapid reduction in heart rate (Bannister, 1983). Patients are trained to use vagal innervation techniques whenever they become aware of the first signs of panic (Sartory & Olajide, 1989). The techniques are easily learned and rapidly implemented, and can be effective for treating patients who catastrophically misinterpret cardiac sensations. By terminating tachycardia, these techniques provide patients with methods of controlling feared cardiac sensations, which may help them reappraise the sensations as harmless ('Tachycardia can't be dangerous if I can easily turn it off'). However, these techniques also can be misused as safety behaviours.

382

ADJUNCTIVE COGNITIVE-BEHAVIOURAL INTERVENTIONS

There are additional caveats in using the Valsalva manoeuvre. If patients apply this method with extreme effort, then fainting can occur, even in physically healthy people (Picornell-Darder et al., 1978). Case reports have documented other negative side effects of forcefully using this method, including spontaneous pneumothorax, hemorrhagic retinopathy, and orbital hematoma (Duane, 1973; Feldman et al., 1993; Katz & Carmody, 1985; Roberts & MacKay, 1987). Clearly, patients should be instructed not to vigorously apply the Valsalva manoeuvre. The procedure, however used, may be unsafe for patients with serious pulmonary, cardiac, retinal, or cerebrovascular disorders. Given these concerns, along with the potential for vagal innervation techniques to become safety behaviours, these procedures are not recommended for routine treatment of panic disorder. The techniques are sometimes used in conjunction with routine cardiologic care for treating panic patients with benign, recurrent tachycardia (e.g., benign paroxysmal atrial tachycardia). Once the patient learns (e.g., via cognitive restructuring) that the tachycardia does not lead to imminent catastrophe, vagal innervation techniques can be useful in reducing unpleasant but harmless tachycardic episodes.

INTERPERSONALLY FOCUSED INTERVENTIONS Assertiveness

Training

Assertiveness deficits merit treatment in their own right, and may be doubly important when these deficits exacerbate panic disorder and undermine its treatment. Assertiveness deficits can create problems in adherence to panic treatment, particularly when the patient has to request time off work or request significant others to assume child care responsibilities while the patient is attending CBT2 sessions or completing homework assignments. In these cases a course of assertiveness training can be helpful. Strategies for improving assertiveness have been well described by many authors, which the interested reader should consult (e.g., Jakubowski & Lange, 1978; Lange & Jakubowski, 1976).

Couples and Family Interventions On average, the dyadic relationships (marriages or other spousal relationships) of people with panic disorder tend to be no better or worse than those in the general population (Marks, 1987). Dissatisfaction with these

INTERPERSONALLY FOCUSEDINTERVENTIONS 383

relationships also tends to be a weak predictor of outcome of CBT2, although few studies have examined the effects of very severe dyadic or family discord on the treatment of panic disorder (Chapter 8). Extreme discord can occur for panickers as for other people, and may interfere with panic treatment. Sometimes, both panic disorder and relationship discord need to be addressed in treatment. The question arises as to the best way to treat these problems when they co-occur. Behavioural Couples Therapy

Patients with panic disorder and relationship discord can be treated with CBT2 interventions combined with behavioural marital therapy. The latter is also sometimes called behavioural couples therapy (BCT), which is a more appropriate descriptor given that the treatment was designed for improving dyadic relationships, regardless of whether couples are married. Based on social learning principles, BCT is a collection of strategies emphasizing communication training and interpersonal problem solving. Interventions include training in conflict resolution, negotiation training, and methods for increasing the exchange of positive interactions and decreasing aversive ones. Detailed descriptions of BCT are available elsewhere (Baucom & Epstein, 1990; Friedman, 1987; Jacobson & Margolin, 1979; Stuart, 1980). When panic disorder co-occurs with significant problems in the patient's dyadic relationship, both CBT2 and BCT can be implemented. BCT can be used to address general problems in relationship functioning (e.g., poor communication of needs and wants, poor problem solving skills, power imbalances), as well as any problems specific to panic disorder. As observed by Daiuto et al. (1998), problems specific to panic disorder include how the partners communicate about the disorder and how they communicate about the individual and lifestyle changes that may occur, or may need to occur, once panic disorder begins to remit. A primary goal of BCT for these couples is to facilitate a marital environment that reinforces the patient's moves towards autonomy. Thus, the skills that are taught, and the way that the spouses use them, must support the goal of eliminating the agoraphobic [and panic] symptoms. The application of BCT for this purpose may, at times, result in therapeutic recommendations for change that are initially quite disruptive and dissatisfying for the spouses. For example, despite mutual satisfaction with an arrangement in which the husband does all of the household shopping, the therapist might ask the spouses to problem solve about changing the division of labor so

T

384 ADJUNCTIVECOGNITIVE-BEHAVIOURAL INTERVENTIONS that the wife does not remain at home as much. As such, decreases in relationship satisfaction may occur as the partners are encouraged to make changes in the ways that they interact. Such negative reactions to the loss of old ways of relating should be viewed as progress and normalized, as treatment-induced dissatisfaction with the relationship is likely to decrease the spouses' motivation to work in a united way to eliminate the agoraphobia .... BCT can be used to offset some of the negative reactions spouses may have in response to the agoraphobic partner's improvement ... For example, the use of cognitive restructuring techniques could be useful in modifying husbands' negative attributions about their wives' increased autonomy .... If a permanent dissatisfaction arises on the part of one or both partners, this needs to be recognized and the goals of the therapy need to be re-evaluated. (Daiuto et al., 1998, pp. 680-681)

What are the best ways of combining CBT2 with BCT? Should they be implemented simultaneously or sequentially? Chernen and Friedman (1993) described two case studies in which agoraphobia and high levels of patient-reported marital conflict were successfully reduced by 10 sessions of situational exposure followed by 10 sessions of BCT. (The effects of treatments on panic attacks were not reported.) The treatment package was of little help for two other panic disordered patients who initially denied marital conflict but were subsequently found to have marital difficulties. Thus, a sequential treatment of exposure followed by BCT was useful only for the two patients who were able-for unknown reasonsto report marital problems at the beginning of treatment. Panic treatments (CBT2 or situational exposure) and BCT also can be implemented concurrently rather than sequentially. A simultaneous approach may prove most effective for addressing relationship issues that either interfere with, or are caused by, the patient's initial gains from treatment for panic disorder (Daiuto et al., 1998). This possibility merits investigation. Meanwhile, treatment may be best guided by a formulation (Chapter 10) of how the patient's symptoms are related to his or her relationship problems. The most distressed, poorly adjusted couples may be least likely to seek couples therapy. In such cases the panic disordered patient or the spouse may decline to participate in couples therapy. Daiuto et al. suggested that in these cases BCT could be implemented in an individual format, as described by Bennun (1985, 1997). Here, the patient would be treated for panic disorder, and also receive training in strategies for improving the dyadic relationship. BCT would be used to help the patient understand dysfunctional behaviour patterns in the relationship, and to help the patient identify strategies for altering these patterns. Cognitive restructuring of relationship-related dysfunctional beliefs also could be implemented.

INTERPERSONALLY FOCUSEDINTERVENTIONS 385

Integrated CBT2 and Family Therapy When relationship problems extend beyond the dyadic relationship to other family members, it can be useful to integrate panic treatments with some form of family therapy. To assess the role of family factors, and thereby decide whether to use family therapy methods, Lange and de Beurs (1992) recommended that the therapist gather data to answer the following questions.

•

•

•

Are the patient's symptoms reinforced by one or more family members? If so, then in conjunction with individual treatment of panic, the therapist may attempt to change the patterns of interaction between the patient and family members. This could be done by having treatment sessions with all the pertinent family members. Is the patient 'isolated within the family'? In other words, is the family unsupportive or a source of stress for the patient? If so, then this issue should be explored to identify how this problem might be solved. Do the patient's symptoms of panic disorder serve the interests or needs of the other family members? Lange and de Beurs suggest that a paradoxical intervention is one way to deal with this problem: The therapist might give indirect paradoxical suggestions [to the family] ... by stating: 'When one family member's symptoms are treated, the situation for the other can become so threatening that it might even be better not to change anything.' Parallel to this, the therapist might start with symptom-directed interventions, such as breathing retraining and exposure exercises, but he should express a pessimistic opinion about the chances of success. (Lange & de Beurs, 1992, p. 61)

Lange and de Beurs (1992) reported five case studies in which patients received some combination of CBT2 interventions (e.g., interoceptive exposure, situational exposure, and breathing retraining) in conjunction with various family therapy techniques (e.g., Boszormenyi-Nagi, 1987; Haley, 1980; Lange & van der Hart, 1983). An integrated package of CBT2 methods and family therapy interventions was developed for each patient, presumably on the basis of some sort of case formulation. Lange and de Beurs' case studies suggest that the combination of CBT2 and family therapy methods can successfully treat panic disorder, while also addressing family or relationship problems. The reader is referred to their article and supporting references (e.g., Lange & van der Hart, 1983) for details on how to implement the various family therapy interventions.

386

ADJUNCTIVECOGNITIVE-BEHAVIOURAL INTERVENTIONS

IMPROVING QUALITY OF LIFE Telch et al. (1995) claimed that CBT2 improved quality of life (QOL) of panic patients. However, the measures in their study assessed QOL primarily in terms of impairments in functioning. CBT2 can help patients get out of the house to get a job and shop for groceries, and cognitive interventions can be used to help patients deal with life stressors. But there is more to QOL than the absence of impairment. QOL refers to the person's evaluation of the degree to which their most important needs, goals, and wishes have been fulfilled (Frisch et al., 1992). Thus, QOL also entails life satisfaction and subjective well-being. The absence of impairment does not ensure that the patient will feel fulfilled with life, in the same way that the absence of anxiety or depression does not ensure the presence of happiness (Frisch et al., 1992). Accordingly, CBT2 for panic disorder may not be sufficient for improving QOL. The therapist may need to implement strategies from Frisch's (1998, 1999) cognitive-behavioural QOL therapy. This could be done at the end of CBT2, once panic disorder has abated. QOL therapy offers specific treatment strategies for each of 16 domains of QOL: health, self-esteem, goals and values, money, work, play, learning, creativity, helping, love, friends, children, relatives, home, neighborhood, and community. The therapy also offers several general strategies that can be used for boosting satisfaction in any of these domains. General strategies include: • • •

•

•

Changing the objective circumstances of a life domain; e.g., increasing one's circle of friends, or changing where one lives or works. Changing one's attitude about domains; e.g., learning to be satisfied with one's current job ('Are things really so bad at my job?'). Examining one's goals and standards for fulfillment. This involves setting realistic goals and lowering excessive standards; e.g., 'Do I have to have a six-figure salary in order to be happy?' Examining the importance placed on each domain for one's overall happiness. Priorities may be revised to emphasize those domains the person can change; e.g., if one's job is unsatisfying and other employment opportunities are limited, the person could increasingly cultivate other life domains, such as friends, family, and recreational pursuits. Cultivating interests in other areas not considered before (e.g., hobbies and other pleasant activities) even while continuing to work on improving other life domains.

Numerous other interventions can be used, as needed, to further improve QOL. Quality of relationships with one's children can be improved by

SUMMARY AND CONCLUSIONS 387

Parent Effectiveness Training (Gordon, 1975). Vocational training may be useful for patients who have long avoided work situations (as part of their agoraphobia) (Rapee & Barlow, 1991). Strategies for a healthy lifestyle can also be implemented (e.g., regular exercise, a balanced diet, sufficient sleep, reduction in smoking and alcohol consumption), along with specific stress-management strategies such as time management skills (e.g., Friedman, 1996). A useful effect of these strategies is that they can strengthen the patient's view of him- or herself as a person who is healthy and resilient, instead of someone at risk for dying, going crazy, or losing control.

SUMMARY AND CONCLUSIONS This chapter described _a number of adjunctive cognitive-behavioural interventions that can be used to supplement cognitive restructuring and exposure exercises in the treatment of panic disorder. Source references were provided for detailed accounts of these interventions. Breathing retraining, relaxation training, and the vagal innervation methods should be used sparingly and with caution, because these 'coping' strategies have the potential of becoming safety behaviours, thereby preventing catastrophic beliefs from disconfirmation. Other contraindications were also noted. A case formulation (Chapter 10) can guide the selection and timing of adjunctive interventions. One needs to be careful not to overwhelm the patient by implementing too many interventions at once. Quality of life therapy might be best implemented once the patient's panic attacks and agoraphobia have gone into remission. Quality of life issues may be most salient at that time, when the patient is no longer preoccupied with panic and avoidance.

STRATEGIES FOR IMPROVING TREATMENT ADHERENCE AND PREVENTING RELAPSE This chapter has two aims. First, to identify general obstacles to successful treatment and to propose some solutions. Second, to describe several strategies for increasing the odds that the beneficial effects of treatment will continue after therapy formally ends. Strategies include maintenance programs containing methods for preventing relapse.

OVERCOMING OBSTACLES TO SUCCESSFUL TREATMENT Previous chapters discussed troubleshooting strategies for specific interventions, such as methods for dealing with problems in implementing interoceptive exposure. In the present chapter we will discuss problems with treatment adherence that apply to most interventions and procedures used in second generation cognitive-behavioural therapy (CBT2). Poor adherence can be expressed in a variety of ways, including failure to complete homework assignments, consistently missing therapy appointments or arriving late for sessions, and failing to work with the therapist in exploring important issues (e.g., keeping the discussion vague or superficial, or repeatedly changing topics) (Schaap et al., 1993). Poor adherence can be due to a variety of factors, including insufficient motivation to complete treatment, personality pathology that prevents the patient from becoming an active collaborator in therapy, and treatmentrelated beliefs (e.g., 'Therapy is too difficult for me'). Some adherence problems are forms of avoidance behaviour. The first step toward overcoming these difficulties is to develop a working hypothesis about the causes. The case formulation approach (Chapter 10) can be used. The therapist can discuss the adherence problems with the patient, and collaboratively look for ways of overcoming them.

OVERCOMING OBSTACLES TO SUCCESSFUL TREATMENT 389

Facilitating Treatment Adherence Treatment adherence can be facilitated by targeting and restructuring therapy-harming beliefs such as 'I'll never overcome my problems, so what's the point in trying'. Adherence also can be facilitated by structuring homework assignments in ways that maximize the chances of successful completion, and by enhancing patient motivation to work at the assignments.

Structuring Assignments to Improve Adherence Homework assignments-such as exposure assignments and cognitive restructuring exercises-play a vital role in CBT2. There are numerous strategies for structuring homework assignments so that they are likely to be completed (e.g., Emmelkamp & Bouman, 1991; Persons, 1989). These include the following: •

• • •

• •

•

Pretherapy education can encourage patients to persist with treatment. To illustrate, Emmelkamp and Emmelkamp-Benner (1975) presented patients with a video in which three 'ex-patients' were interviewed about their experiences in treatment, which consisted of selfdirected situational exposure. Although the video did not increase the efficacy of treatment, it did reduce attrition. A combination of written handouts and educational videos may be even more effective. Collaboratively plan each homework assignment so patients assume at least some responsibility. Ensure the assignment is not too difficult. If necessary, break it up into manageable steps or subtasks. Ensure that patients understand the rationale for their homework assignments. Have them describe the rationale in their own words. If patients do not understand the rationale then they will wonder why they should bother with the tasks. Provide patients with a written copy of each assignment so it is not forgotten. Use rating forms so patients can record details of their assignments, such as the date and duration of the task, and ratings specific to the nature of the task (e.g., severity of anxiety, degree of relaxation). Ask patients to use weekly schedules or day-planners to list out all the tasks they will be performing during the following week; e.g., going to work, mowing the lawn, taking the children to football practice. The schedule is structured so that there is a specified time for the homework assignments. This strategy is particularly useful for patients who have difficulties with time management.

390

•

•

STRATEGIES FOR IMPROVING TREATMENTADHERENCE

If adherence is an ongoing problem, written contracts can be used. The contract specifies the nature of the assignment and outlines the responsibilities of the patient and the therapist; e.g., 'I, [name of patient] agree to practice [name of task] for [specified frequency and duration]. In return, I, [name of therapist] agree to review progress and to consult on ways of maximizing the effectiveness of the task.' This approach might seem unduly formal, although it can be quite useful when adherence is problem. Ask patients to make a commitment to their significant other(s) to perform the homework assignments.

Highlight Discrepancies between Current and Desired States

Motivation can be enhanced by discussing the discrepancy between the patient's current situation and long-term goals. This increases the patient's awareness of the costs of current maladaptive behaviours. One satisfactory way to develop discrepancies between current behavior and future goals is to inquire about what the person would most enjoy doing when unshackled from their anxiety disorder. When this image (or images) has been developed, it can be contrasted with the person's present state. The resulting discomfort can then be used to motivate the person to engage in therapeutic exercises. (Andrews et al., 1994, p. 211)

Use Reinforcers

Rewards are also good motivators that can help patients achieve their goals or subgoals. Rewards for subgoals are important because subgoals are more rapidly attained than ultimate goals, and so rewards are more frequently received. For rewards to be meaningful, patients should attribute progress to their own efforts rather than those of the therapist. To this end, remind patients that the therapist simply serves as consultant and facilitator; it is largely through the patient's efforts that the therapeutic exercises are completed. For some patients, the mere attainment of goals or subgoals is rewarding (i.e., intrinsic motivation). Other patients may benefit from some form of extrinsic motivation to lead them to complete the sometimes difficult work of therapy. For these patients, it can be fruitful to help them look for extrinsic rewards to encourage them to attain their goals. Sometimes, however, patients are reluctant to use rewards, and instead they attempt to use self-criticism and other punishments to motivate themselves. This strategy is typically self-defeating, contributing to dysphoria and

OVERCOMING OBSTACLESTO SUCCESSFUL TREATMENT 391

low self-esteem. Strategies for encouraging patients to use rewards are as follows. •

•

•

Review the efficacy of the patients' current strategies for motivating themselves. Have self-criticisms or other punishments helped overcome their problems? If not, then the therapist can suggest that perhaps it's time for a new strategy. Ask patients how they would motivate a good friend. Would rewards or criticisms be used? Most choose to use rewards. This can lead into a discussion of why rewards are used and why they are more effective than punishments. Review the motivators in the patients' daily lives, and identify which they prefer; negative reinforcement (avoidance or escape), punishment, or rewards. Most prefer to work for rewards. Rewards can be idiosyncratic and may not be immediately obvious to the therapist. To illustrate, Mr P., a severely agoraphobic man, collected Air Miles, which are points earned by purchasing goods at particular stores. The points can be redeemed for free airline tickets. Mr P. collected Air Miles with enjoyment. Ironically, he doubted he ever would redeem them; Mr P. was terrified of flying.

Even if patients can be persuaded to use rewards, they may need help coming up with specific reinforcers. Patients may be able to come up with a suitable reinforcer for attaining a long-term goal, but have trouble coming up with the more frequent reinforcers for attaining subgoals. A good reward is one that does not cause problems of its own. Eating junk food would not be a good reinforcer for someone who is morbidly obese. Buying clothing would not be a good reward for someone with financial problems. Effective rewards include self-praise. That is, patients can be encouraged to praise themselves for achieving their goals. Praise from the therapist is also usually a good motivator. Another simple but useful reward strategy is based on the Premack Principle. Naturally occurring, high-frequency behaviours are identified in the patient's daily life. For example, the patient might watch a favourite TV program each day. This activity is then made contingent on homework completion. If the homework assignment was to practice an interoceptive exposure exercise each day, the patient would watch the TV program only if the assignment was completed. If the goal is not attained, then the therapist and patient can conduct a 'post-mortem' in an effort to identify and overcome any obstacles. For example, perhaps the assignment was too difficult and needs to be broken up into subtasks.

392

STRATEGIES FORIMPROVINGTREATMENT ADHERENCE

Target Secondary Gains

Some patients are under reinforcement contingencies that reward them for not working at therapy, and punish them for overcoming their problems. These 'secondary gains' or 'reinforcement traps' can sometimes interfere with CBT2 (see Chapter 10). To circumvent these difficulties the therapist should explore, at the outset of treatment, what pleasant or aversive things might happen if the patient was free of panic disorder. This approach can alert the clinician to possible secondary gains. If you believe that secondary gains could undermine treatment, one approach is to ask the patient to make a list of all the ways a person could keep treatment from working, and then ask the patient to identify the particular method that they might use (McMullin, 1986). Then discuss why these might be used and review how they would keep the patient from reaching the therapeutic goals.

The patier,t and therapist can conduct a cost-benefit analysis in which the costs of overcoming panic disorder are weighed against the benefits. Costs include the effort required to complete treatment as well as the loss of secondary gains. If the costs exceed the benefits then solutions to this problem should be explored. This is illustrated in the following case. The case also shows that sometimes secondary gains do not become apparent to the patient and therapist until later in therapy. Mrs 0. formerly worked at a stressful, unsatisfying job. She then took sick leave as her panic disorder worsened. Over the course of CBT2 her panics were eliminated and her worry about panic substantially abated. As treatment focused on reducing her remaining agoraphobia, Mrs 0. became increasingly preoccupied with the prospect of returning to work. She also became more 'forgetful' about completing homework assignments. A review of the problem revealed that as she became less worried about panic attacks, her focus of apprehension shifted to her return to work. Homework assignments acutely reminded her of this. To avoid these unpleasant thoughts, Mrs 0. made efforts to avoid all reminders of work, including homework assignments. To overcome this problem, Mrs O.'s occupational prospects were discussed during the treatment sessions. Initially, this was quite anxietyevoking. Mrs 0. saw herself in a no-win situation; either she retained her panic disorder or returned to a highly stressful job. As the problem was explored, however, she came to see that there were other options she had not considered. She contacted an employment assistance officer in her company and eventually arranged to be transferred to another, less stressful position. Once progress had been made on her job situation, Mrs 0.

OVERCOMING OBSTACLES TO SUCCESSFUL TREATMENT 393

became more optimistic about returning to work and began to work harder at homework assignments for reducing her agoraphobia.

Strategies for Managing Difficult Patients When panic disorder is comorbid with a personality disorder, both disorders may merit treatment in their own right. Some personality disorders-such as paranoid and avoidant personality disorders-are particularly important to address because they can interfere with the treatment of panic disorder (Chapter 8). These comorbid conditions can be treated with CBT2 for panic disorder combined with cognitive-behavioural approaches to treating personality disorders (e.g., Beck et al., 1990; Schaap et al., 1993). The case formulation (Chapter 10) can provide guidance on how to implement these treatments. It is beyond the scope of this volume to describe the treatment of personality pathology. Instead, we will highlight some of the strategies for managing these difficult patients, which can be implemented during CBT2 for panic disorder. Patients with paranoid personality features tend to be suspicious and distrusting. They are reluctant to confide in, or become close to, others because they fear the information they share will be used against them (American Psychiatric Association, 1994). These features make it difficult to establish a collaborative therapeutic relationship, and also make it difficult to identify and restructure panic-related catastrophic beliefs. These patients should be handled with care. The therapist needs to be courteous, attentive, and tolerant, without submission or capitulation (Schaap et al., 1993). Patients with paranoid personalities expect people to exploit or harm them. When these patients have comorbid panic disorder they may harbour catastrophic beliefs about the social consequences of panic or anxiety. Some of these beliefs can be readily tested, particularly beliefs about the behaviours of others (e.g., 'People will laugh at me if I start to tremble'). It is more difficult to refute beliefs about the thoughts or attitudes of others (e.g., 'People will think I'm pathetic if I start to tremble'). Accordingly, behavioural experiments should test, as far as possible, objectively observable outcomes (e.g., behaviours of others). Another strategy is to explore the patient's beliefs about the importance of others (e.g., 'Suppose someone did see you tremble and thought you were pathetic, what would be so bad about that?'). Panic patients with avoidant personality disorders can be passive and pessimistic during treatment, which can impede the progress of CBT2 assignments that require some degree of autonomous functioning (e.g., traveling alone by bus to a shopping mall). The therapist needs to work

394

STRATEGIES FOR IMPROVINGTREATMENT ADHERENCE

at the patient's pace, while showing understanding and interest, and not raising expectations too high; 'the pessimistic client does not feel understood by an optimistic therapist' (Schaap et al., 1993, p. 66). Treatment goals should be broken into small steps, and extrinsic reinforcers can be used as motivators. Cognitive restructuring can be used to challenge pessimistic beliefs (e.g., Beck et al., 1979). Indirect methods can be used to motivate the patient. Sometimes, an approach is needed in which therapists remain empathic, friendly and understanding, and refrain from instilling hope and pushing the client. Instead, the therapist may become even more passive than the client and only indirectly suggest ways that change may be effected, even though the client is not yet ready. (Schaap et al., 1993, p. 66)

When the therapist becomes passive during treatment, the patient is placed in the role of making decisions about how to proceed. For example, decisions about which sorts of exposure assignments to undertake. This approach is best implemented after the therapist has developed a good understanding of the patient and the patient's problems. Armed with this knowledge, the therapist can predict how the patient will react to indirect treatment strategies. Some patients react by becoming more actively involved in therapy; others react by becoming frustrated and dropping out of treatment.

STRATEGIES FOR MAINTAINING TREATMENT GAINS Preparing for the End of Treatment A formal course of CBT2 typically consists of 8-16 sessions, depending on the protocol. The formal course of treatment is followed by a long-term maintenance phase in which the patient implements CBT2 skills, as needed, with or without booster therapy sessions. As the formal course of CBT2 draws to a close, there are several tasks the therapist needs to accomplish in order to increase the chances that treatment-related gains will be maintained, if not improved over time. •

Fade out the frequency of therapy sessions-e.g., from weekly to biweekly, then to monthly-so that patients become gradually more accustomed to implementing CBT2 strategies on their own, without regular contact with a therapist.

STRATEGIES FOR MAINTAINING TREATMENT GAINS 395

•

•

•

•

Review the gains made during treatment and the extent to which the therapy goals were attained. A graph of the treatment gains, such as a chart showing the reductions in panic frequency over time, can be used to illustrate treatment gains (Beck & Zebb, 1994), thereby encouraging patients to continue using the exercises learned in CBT2. This approach can be effective for patients who fail to recognize the gains made in therapy. Such patients often underestimate how severe their problems were at the beginning of treatment. Ensure that patients understand what they did to produce the treatment gains. Ask patients to produce a written description of what they learned during treatment so that the CBT2 exercises can be used again in the future, if needed. Identify and challenge maladaptive beliefs that thwart autonomous functioning after the end of treatment; e.g., 'I won't be able to cope without my therapist'. Review problems that remain to be worked on, such as residual levels of agoraphobic avoidance. Develop a written maintenance program. That is, a plan for maintaining or increasing treatment gains, and strategies for relapse prevention.

Developing

a Maintenance Program

A long-term maintenance program is set up in the last one or two sessions of the formal course of treatment. The program is intended to help patients function as their own therapists to continue to work on any remaining problems and to deal with any difficulties that may arise. There are several elements of maintenance program, as described below. Available evidence suggests that these are effective in sustaining treatmentrelated gains (e.g., Jansson et al., 1984; Ost, 1989). Plan for Ongoing Self-directed Treatment

Toward the end of the formal course of treatment, the patient is asked to write out short-term goals (e.g., those to be attained in the coming weeks or months) and longer-term goals. Self-directed treatment initially focuses on the short-term goals. The patient is asked to write out, in chronological order, a list of further exposure assignments and other CBT2 exercises to be completed over the next 2-3 months. A self-directed treatment plan is developed, such as a plan to complete at least two exposure assignments per week. Patients should continue using homework monitoring forms and any other relevant forms for charting progress. A patient with residual panic disorder who no longer has full-blown panic attacks might

396

STRATEGIES FOR IMPROVINGTREATMENT ADHERENCE

use a panic diary to record details of the sensations and thoughts occurring during limited symptom panic attacks. The use of rewards and other incentives are used to facilitate adherence. Patients are also asked to make a commitment to their significant others about the importance of continued practice. The patient and therapist also can sign a contract on their respective duties during the maintenance phase. The patient's responsibility is to continue to practice, as needed, the CBT2 exercises. The therapist's responsibility is to be available for periodic telephone consultations or booster sessions with the patient (Ost, 1989.). As part of this program the patient may need to further reduce their fear of arousal-related sensations. This may involve learning to accept that anxiety is a fact of life for everyone, without worrying that it will lead to panic. Patients with panic disorder in full or partial remission sometimes worry that periods of normal anxiety are precursors of panic attacks or harbingers of panic disorder relapse. Episodes of anxiety can be framed as naturally occurring behavioural experiments in which the patients allow themselves to experience anxiety without trying to suppress this emotion. This enables patients to discover the consequences of anxiety. Patients can be advised to use anxiety as a signal that they are perceiving something in their lives as threatening. Rather than ignore or dampen this signal, the patient can learn whether the threat is a realistic one that needs to be dealt with in some manner. By the end of treatment the strength of catastrophic beliefs may be substantially reduced, but not to the point that the patient confidently regards the beliefs as untrue. To identify these lingering doubts the therapist can ask the patient to generate a list of evidence for their catastrophic beliefs and against their noncatastrophic beliefs (Clark & Salkovskis, 1987). If the patient can generate such a list then that means the catastrophic beliefs have not been completely eradicated. If so then cognitive and exposure interventions can be set as tasks to be performed during the maintenance program. For example, the patient might be asked to conduct further tests of their beliefs.

Plan for Periodic Therapist Contact Booster sessions and telephone contacts with the therapist can be arranged as needed. These can be used as 'check-ups' scheduled at regular intervals after the end of the acute phase of treatment; e.g., 3, 6, and 12 months post-treatment. Boosters and telephone contacts are used to assess the maintenance of treatment gains, to reinforce the patient's efforts, to provide assistance in helping the patient deal with any new or enduring problems, and to develop plans for ongoing exposure or other therapeu-

STRATEGIES FOR MAINTAINING TREATMENT GAINS 397

tic exercises. Patients should write down what they learned from the consultations so that the information is not forgotten. Medication Discontinuation

Toward the end of 8-16 sessions of CBT2, many panic patients will have successfully discontinued whatever anti-panic medications they may have been taking. Some, however, will have only begun this process. Medication tapering can continue during the followup interval under the guidance of the prescribing physician. A tapering schedule can be implemented. Detailed schedules are described in Otto et al. (1996a). Periodic telephone consultations or booster sessions can be used as needed. Relapse Prevention

Even when panic disorder is successfully treated, symptoms of this disorder may reoccur at some point in the future. The important issue is not whether symptoms return, but how the patient deals with them. In order to successfully maintain treatment gains, it is important that a lapsedefined as a transient increase in symptoms of panic disorder-does not progress into a relapse of panic disorder. Accordingly, the maintenance program should provide the patients with relapse prevention strategies. These should be written down so that patients can consult them in the future, if necessary. The points to convey are as follows, which are supplemented by a written handout such as that shown in Table 15.1. •

•

•

A lapse is not a relapse. The occurrence of full or limited-symptom panic attacks or an increase in agoraphobia does not mean that the patient has lost all the gains that he or she made during treatment. It simply means that it is time to practice the exercises learned in therapy. By practicing these the patient can prevent a lapse from developing into a relapse. Lapses can be framed as opportunities for continuing to practice the skills that were learned during treatment. Patients are asked to write out a plan of what CBT2 strategies they would implement if they had another panic attack. As part of these exercises the patient can write out and challenge any catastrophic beliefs they have about the recurrence of panic (e.g., 'If I have another panic attack I'll be back at square 1; all my therapy gains will be lost'). It is important that patients realize that as a result of therapy they have acquired effective skills for dealing with lapses. To prepare to deal with lapses, the patient and therapist can compile a list of 'high-risk' situations. These are events or circumstances most

398

STRATEGIES FOR IMPROVING TREATMENTADHERENCE

Table 15.1 Guidelines for dealing with the reemergence of panic disorder (with or without agoraphobia)

If you begin to experience problems with panic, anxiety, or avoidance: l.

2. 3.

4. 5. 6.

Remind yourself that a setback is not a catastrophe. It is not a relapse, but a temporary failure to manage a situation that you have managed before. Analyze the situation. Try to learn why the setback occurred. Write out the frightening thoughts that occurred during any episodes of anxiety or panic, and write out what you fear will happen in situations you are avoiding. Practice the exercises you learned in therapy, such as the cognitive restructuring methods. Write out the evidence for and against your beliefs. Set up tests of your catastrophic beliefs; contrast them with noncatastrophic alternative explanations. Restrict the setback; don't let it spread to similar situations. If you are starting to avoid some situations, try not to avoid other situations. Return to situations you are starting to avoid. Do this as soon as possible. If the situation is too difficult to enter, try an easier situation and then gradually work up to more difficult situations. If this still doesn't work, call your therapist as soon as possible to discuss the problem or to arrange further therapy sessions.

Modified from Ost, L.-G. (1989). A maintenance program for behavioral treatment of anxiety disorders. Behaviour Research and Therapy, 27, 123-130.

•

likely to lead to the return of panic disorder. To identify these situations the therapist can review the case formulation, including circumstances in which the patient's problems initially developed. This provides clues about what might contribute to lapses. Common high-risk situations include stressful life events and experiences that lead the patient to believe that arousal-related sensations are dangerous. For example, at some point in the future one of the patient's relatives might suddenly die of a heart attack. This could lead the patient to believe that palpitations are dangerous, and thereby increase the chances of panicking when this sensation is experienced. By educating patients about the possibility of future catastrophic learning, they can be prepared to identify and challenge any emerging catastrophic beliefs. To inoculate against the future development of catastrophic beliefs, one should review body sensations that patients might not have experienced before (Clark & Salkovskis, 1987); e.g., 'Suppose that one day you noticed that you had difficulty breathing. What would you think was happening?' Socratic dialogue can be used to elicit catastrophic and noncatastrophic interpretations of the new sensations, and to discuss ways of testing the interpretations.

l

J

SUMMARY AND CONCLUSIONS 399

•

•

If a lapse does occur, then the patient should attempt to analyze the situation to identify any catastrophic beliefs. Once identified, the patient can use cognitive and exposure strategies learned in therapy to evaluate these beliefs in comparison to noncatastrophic alternatives. Patients are encouraged to restrict the lapse before it gets worse. For example, returning as soon as possible to the setting in which a panic attack occurred in order to limit the development of agoraphobic fear and avoidance. An exposure hierarchy can be implemented as necessary. If the patient is unable to manage the lapse, then the therapist should be contacted for one or more telephone contacts or booster sessions. The consultations should examine the cause of the lapse and thereby lead to a plan for correcting the problem. The consultation also can examine the meaning of the lapse to the patient. This involves identifying and challenging any catastrophic beliefs about the fact that a lapse occurred; e.g., 'Just when things were going so well I had another panic attack. That means that therapy didn't work-my panic disorder is coming back!'

Emergence of Other Problems

The concept of the general neurotic syndrome (Chapter 2) suggests that patients with panic disorder are at risk for developing other neurotic spectrum disorders, even if their panic disorder is successfully treated with CBT2. Thus, even when panic disorder is in remission, patients may develop other anxiety or mood disorders at some point in the future. This is not inevitable and it is not helpful to alarm the patient about this possibility. Instead, simply advise patients that other emotional problems might arise in the future, and if this happens then they should contact their therapist or primary care physician because there are effective cognitive-behavioural and other treatments for these problems.

SUMMARY AND CONCLUSIONS When difficulties are encountered in successfully implementing CBT2, the therapist should develop a working hypothesis about the causes. This guides the therapist in selecting strategies for overcoming the problems. In this chapter we reviewed several common problems and solutions, including a number of strategies for facilitating treatment adherence. After the end of the formal course of CBT2, the patient moves into the

400

STRATEGIES FOR IMPROVING TREATMENTADHERENCE

maintenance phase of treatment. This involves largely self-directed treatment of any remaining problems, along with procedures for dealing with relapses. Although CBT2 tends to be a short-term treatment, occasional booster sessions and telephone contacts are useful. These brief, periodic consultations can help maintain gains over the long term.

Chapter 16

COGNITIVE-BEHAVIOURAL PROTOCOLS FOR SPECIAL •POPULATIONS

Most protocols for treating panic disorder have been developed for, and evaluated on, Caucasian adults, aged 18-60 years. Although previous studies have found that demographic variables fail to consistently predict treatment outcome (Chapter 8), these studies have not represented all demographic groups. Protocols of second generation cognitive-behaviour therapy (CBT2) may need to be modified to optimally treat particular demographically defined groups. Accordingly, the first aim of this chapter is to review the literature on behavioural and cognitive-behavioural therapies of the following demographically defined groups: children and adolescents, older adults, and culturally defined groups. The limited research will be examined to assess how CBT2 might need to be modified for treating panic disorder in these special populations. Special populations also can be specified in terms of clinically defined groups, such as patients with particular combinations of disorders. The second aim of this chapter is to review the CBT2 protocols developed for some of these groups. Recall from Chapter 8 that some types of comorbidity seem to have little prognostic significance in treating panic disorder. For example, the efficacy of CBT2 in reducing panic disorder does not seem to be impaired when comorbid generalized anxiety disorder is present. In these cases the clinician might begin by treating the most severe disorder and then treat the remaining disorder(s). Other combinations of comorbidity, however, may require special protocols for treating panic disorder. For example, the CBT2 protocol may need to be modified when panic disorder occurs in a person with schizophrenia. When panic disorder is comorbid with other disorders, it is also important to consider how the therapist can efficiently combine treatments for two different disorders, such as treatments for panic disorder and posttraumatic stress disorder. Such protocols will be reviewed in the latter part of this chapter.

402

COGNITIVE-BEHAVIOURAL PROTOCOLS FOR SPECIAL POPULATIONS

PANIC TREATMENTS FOR DEMOGRAPHICALLY DEFINED GROUPS Children and Adolescents Panic Attacks in Children Although panic disorder is most commonly found in young adults, it occurs in adolescents and occasionally is seen in children (American Psychiatric Association, 1998). It is not known, however, whether there is a lower limit on the age in which panic occurs. Histories provided by panic disordered adolescents and adults, and case studies of panic disordered children, suggest that panic attacks can occur in children as young as 5 years old (Dummit & Klein, 1994; Ollendick et al., 1994). Are panic attacks in children due to the same mechanisms as those in adults? It may be asked whether children have sufficient cognitive sophistication to make catastrophic misinterpretations of body sensations, as described in cognitive models of panic (Nelles & Barlow, 1988). Ollendick's (1995) account of the first unexpected panic attack in an 11 year-old boy suggests a catastrophic misinterpretation that was much the same as those occurring in adults. The boy was sitting at home working on his stamp collection, when: From nowhere, it felt like I couldn't breathe and that I was going to die .. . I was just sitting there and it happened ... I just couldn't breathe ... it reminded me of when my grandfather died ... mom said that he just stopped breathing and died ... I thought it was happening to me, too. (p. 521)

Mattis and Ollendick (1997) found that children aged 8-14 years were cognitively capable of making catastrophic misinterpretations of body sensations. It remains to be determined whether children under 8 years also have this capability. It may be that the cognitive models of panic (Chapter 3) are better able to account for the panic attacks in older children and adolescents, rather than in younger children. Ollendick (1997) speculated that most panic attacks in children are cued, not unexpected attacks, and that children tend to make noncatastrophic interpretations of body sensations. Ollendick further suggested that the most common triggers for cued panics include stressors, especially when the child has little family support or is separated from caregivers. The causes of panic attacks in young children have important implications for treatment. If the attacks arise from catastrophic misinterpreta-

PANIC TREATMENTSFOR DEMOGRAPHICALLYDEFINEDGROUPS 403

tions, then presumably some form of CBT2 would be helpful. On the other hand, if the panics in very young children (e.g., 5-year-olds) arise from other mechanisms, then different treatment may be necessary. Unfortunately, there have been few studies of the treatment of pediatric panic, and so the issue remains unresolved. Interventions for childhood panic may need to address associated disorders. Research suggests that panic disorder in children tends to co-occur with separation anxiety disorder, and with specific phobias such as claustrophobia (American Psychiatric Association, 1998). It is not known whether panic disorder is best reduced by concurrently treating the associated disorders (e.g., via situational exposure), or whether it is best to first treat one disorder (e.g., panic disorder) and then the others. In the following sections we will consider the assessment and treatment of panic disorder in older children (8-12 years) and adolescents.

Assessment Issues The American Academy of Child and Adolescent Psychiatry (1997) outlined a comprehensive set of guidelines for the assessment and treatment of anxiety disorders in children and adolescents. The reader should consult these guidelines for general information on the biopsychosocial assessment and treatment of panic and other anxiety disorders. Structured interviews have been developed for diagnosing panic disorder and other anxiety disorders in children and adolescents. A recent, psychometrically sound example is the Anxiety Disorders Interview Schedule for Children for DSM-IV (Silverman & Albano, 1996). This can be used to elicit diagnostic information directly from the child or adolescent, or from a parent or other caregiver. The most widely used method for assessing anxiety sensitivity in children and adolescents is the Childhood Anxiety Sensitivity Index (CASI: Silverman et al., 1991). This 18-item self-report questionnaire is a 'down-scaled' version of the most widely used adult measure of anxiety sensitivity, the Anxiety Sensitivity Index (Peterson & Reiss, 1992). The CASI was designed for use with school-aged children and adolescents (6-17 years) and has performed generally well on tests of reliability and validity (Silverman & Weems, 1999). However, researchers have debated whether CASI scores are valid for children under 11 years (Chorpita et al., 1996; Silverman & Weems, 1999). CASI scores should be interpreted with caution when obtained from children under 11 years. For older children, the CASI provides a valid indication of the severity of anxiety sensitivity, which can be used to estimate the strength of arousal beliefs (i.e., beliefs about the dangerousness of arousal-related sensations). Unfortunately,

404

COGNITIVE-BEHAVIOURALPROTOCOLSFOR SPECIALPOPULATIONS

direct measures for assessing arousal beliefs in children have yet to be developed. Treatment Protocols

According to the guidelines developed by the American Academy of Child and Adolescent Psychiatry (1997), clinicians treating panic disorder in children and adolescents should proceed as follows: •

• • •

Educate the patient, family, and school if appropriate about the nature of the disorder and about 'long-term management of this chronic relapsing illness' (p. 815). Enlist the family's support and cooperation with treatment. Consult with the family physician and appropriate school personnel when school functioning is affected by the disorder. Implement panic treatments such as pharmacotherapy or cognitivebehavioural therapy.

These guidelines are generally reasonable, particularly with regard to family involvement in treatment (e.g., the parents could assist in the implementation of situational exposure and interoceptive exposure assignments). However, some of the guidelines are questionable. It is not known whether panic disorder is a 'chronic relapsing illness' when treated early. It may be that CBT2, implemented soon after the onset of the disorder, can reduce the risk of relapse. Accordingly, it may be best to inform the family that relapse is possible, particularly during times of stress, but not inevitable. Also, it is unclear whether treatment outcome is optimized by CBT2, pharmacotherapy, or their combination. The literature on the treatment of adults (Chapter 7) supports the use of CBT2 alone rather than in combination with drugs. Further research is needed to see whether this conclusion applies to the treatment of children and adolescents. Most of the empirical literature on the treatment of panic disorder in youngsters consists of case studies and small uncontrolled trials, with patients consisting of adolescents. A handful of these studies have evaluated behavioural or cognitive-behavioural treatments. Barlow and Seidner (1983) reported a study of three panic disordered adolescents, aged 15-17 years. They were treated in a small group for 10 weekly sessions. The patients' mothers were involved as cotherapists. Therapy consisted of self-directed situational exposure and cognitive restructuring. Details of treatment were lacking in Barlow and Seidner's

l

1

PANIC TREATMENTS FOR DEMOGRAPHICALLYDEFINEDGROUPS 405

report, although it appears that the therapist attempted to persuade patients that panic symptoms were harmless. Two of the three patients improved over the course of treatment, with reductions in agoraphobic fear and avoidance (panic data were not reported). Gains were maintained at 6 month followup. The third patient failed to respond. Barlow and Seidner (1983) observed that treatment of these adolescents was more challenging than treatment of the typical panic disordered adult. Our adolescents ... would not tolerate anxiety during practice despite the detailed treatment rationale on the necessity to do this; concluding, contrary to discussions in group sessions, that the anxiety or panic did represent impending physical catastrophes of one sort or another. Inevitably, they would turn to their parents and ask for 'relief' from these sensations. In this respect our adolescents resembled adult agoraphobics who do not respond to treatment, possibly due to ... 'overvalued ideation' or an inability to comprehend that fears are basically unrealistic. The fact that two of our adolescents responded to treatment in spite of this thinking may be the strongest indication of the need for parental input into the treatment of adolescent agoraphobia. (p. 525)

Other studies of adolescent panickers have not reported these difficulties, suggesting that Barlow and Seidner's patients were particularly severe. Nevertheless, the latter study suggests that it can be useful to engage parents in treatment. Kolko (1984) described the treatment of a 16 year-old panic disordered girl who also had obsessive-compulsive symptoms. She was treated with three weekly sessions of paradoxical instruction to embellish anxiety and to try to induce the feared consequences of her arousal symptoms during prolonged, graded situational exposure. This treatment represents a procedure for disconfirming catastrophic beliefs (Chapter 13). Treatment reduced agoraphobic fear and avoidance, anxiety, and obsessivecompulsive symptoms, with gains maintained at 9 month followup. Outcome data for panic attacks were not reported. Most recently, Ollendick (1995, 1997) treated four adolescents, aged 13-17 years, with a version of cognitive-behavioural therapy. The elements of this multicomponent treatment are summarized in Table 16.1. Treatment duration varied across patients, depending on the amount of time required to achieve a satisfactory treatment response. Patients received 6-9 weekly 50-60min sessions, plus a mean of 3h of therapist- or parentassisted situational exposure between sessions. After treatment, panic attacks were eliminated, and anxiety sensitivity and trait anxiety were reduced to normal levels. Patients were also more confident

406

COGNITIVE-BEHAVIOURAL PROTOCOLS FOR SPECIAL POPULATIONS

Table 16.1

Protocol for treating panic disordered

Content

Session

1

• • •

2

•

3 4

• • • •

Between sessions 4 and 5 5 onwards

adolescents (Ollendick, 1995)

• • •

Information about the nature of panic and anxiety. Description of an interoceptive conditioning model of panic, based on Barlow (1988). (Barlow's model is quite similar to the cognitive models of panic.) Description of the methods used in the forthcoming sessions. Progressive muscle relaxation. As with all skills taught in this protocol, relaxation is practiced in session and as homework. Homework is reviewed in the following session. Breathing retraining. Cue-controlled relaxation and rapid relaxation. Self-instruction training, involving the development of positive self-statements for coping with anxiety and pamc. Planning of graded situational exposure, with instructions to use the relaxation and self-instruction techniques during exposure. Therapist-assisted and parent-assisted situational exposure, scheduled outside clinic sessions. Review of progress made during treatment and rehearsal of relaxation and self-instruction strategies. Problem-solving to address remaining difficulties.

Note: It is important to ensure that breathing retraining, relaxation training, and coping strategies are not misused as safety behaviours (see Chapter 14).

about coping with future panic attacks. Gains were maintained at 6 month followup. The case studies described in this section suggest that behavioural and cognitive-behavioural strategies, involving parents as cotherapists, are useful in treating panic disorder in adolescents. It remains to be seen whether treatment outcome can be further improved by using a full CBT2 package, including interoceptive exposure exercises. It is important to tailor treatment to the patient's developmental level (American Psychiatric Association, 1998; Ollendick, 1995). An explanation of panic disorder and a rationale for treatment should be presented in a way that can be comprehended by the child or adolescent. Explaining catastrophic misinterpretations as 'false alarms'-like the misfiring of a smoke detector or a car alarm-can be readily understood by most panic disordered youths.

PANIC TREATMENTSFOR DEMOGRAPHICALLYDEFINEDGROUPS 407

Self-helpMaterials Garland and Clark (1995) developed one of the few workbooks for helping school age children and adolescents overcome anxiety disorders. In their engaging, illustrated workbook, anxiety symptoms are described as 'worry dragons'. Panic attacks are described as 'false alarms' caused by these dragons. The goal of the treatment program is to impart skills for 'taming' worry dragons. In the sections of the book dealing with panic disorder, the child or adolescent is taught various coping strategies. The first is a form of re-attribution; labeling panic attacks as 'false alarms' caused by bothersome but harmless worry dragons. Other coping strategies include relaxation training, soothing imagery, situational exposure, self-instruction training (i.e., use of coping statements and other positiveself statements), and other simple forms of cognitive restructuring. The workbook is best suited as an adjunct to therapist-administered treatment, rather than as a stand-alone intervention.

Older Adults Epidemiologic studies suggest that panic disorder is less prevalent in later life than in mid-life (American Psychiatric Association, 1998). Even so, clinicians treating psychiatric problems in older adults will encounter cases of panic disorder, and therefore should be capable of making an accurate diagnosis and arranging for suitable treatment.

Assessment Issues There have been a number of documented cases of panic disorder arising -after age 60 (Hassan & Pollard, 1994; Raj et al., 1993). However, when an older adult presents with recent-onset panic symptoms it is important to determine whether the symptoms are (1) the result of a general medical condition (a medical mimic) or (2) the side effects of drugs used for treating general medical conditions (American Psychiatric Association, 1998). When assessing agoraphobia in the elderly, the clinician needs to consider the patient's realistic age-related fears. Problems with ambulation, frailty, and the associated risk of victimization can cause older adults to acquire realistic fears of venturing from home (Lydiard & Brawman-Mintzer, 1997). This should not be confused with agoraphobia.

TreatmentProtocols At the time of writing, there appear to be only two published studies of psychological treatment of panic disorder in the elderly. Both studies used

408

COGNITIVE-BEHAVIOURAL PROTOCOLS FOR SPECIAL POPULATIONS

CBT2. In an uncontrolled study, Swales et al. (1996) treated 20 communitydwelling older adults, aged 55-80 years (mean= 63 years). None of the patients were dementing or otherwise cognitively impaired. Patients with unstable pulmonary disorders were excluded in order 'to safeguard against the remote possibility that their condition might worsen when they were asked to hyperventilate ... during some therapy sessions' (p. 48). The authors did not report how many potential patients they excluded for this reason. Swales et al.'s treatment consisted of an·adaptation of Barlow and Cerny's (1988) protocol, administered in ten 90min one-to-one sessions over 12 weeks. Treatment involved psychoeducation, cognitive restructuring, breathing retraining, progressive muscle relaxation, and situational and interoceptive exposure. A quarter of the sample dropped out of treatment, some because they considered their panic disorder to be insufficiently severe to warrant therapy. Others dropped out because they were in greater need of treatment for comorbid depression. The proportion of dropouts would have been lower if the authors had been more selective in whom they offered treatment, screening out patients with mild panic disorder and patients requiring immediate treatment for depression or other problems. For patients completing CBT2, scores on measures of panic frequency, anxiety, and agoraphobia declined over the course of therapy, with improvements maintained at 3 month followup. Rathus and Sanderson (1996) reported the treatment of two panic disordered adults, aged 69 and 70 years. Treatment consisted of 15-19 weekly lh sessions involving psychoeducation, cognitive restructuring, breathing retraining, and situational exposure. Interoceptive exposure was also used for one patient. Over the course of treatment both patients reported reductions in panic frequency, agoraphobia, and associated symptoms, with gains maintained at 8-10 month followup. The studies by Rathus and Sanderson (1996) and Swales et al. (1996) suggest that CBT2 protocols developed for use with younger adults (18-60 years) can be successfully used with older adults. However, there are some special treatment considerations in the elderly. It is particularly important that the CBT2 therapist check with the patient's physician about whether there are any medical contraindications for using situational or interoceptive exposure. Cognitive restructuring can also pose a challenge when treating older persons. For one of their patients, Rathus and Sanderson had difficulty challenging the patient's catastrophic beliefs. The patient believed that because of her advanced age she was

PANIC TREATMENTSFOR DEMOGRAPHICALLY DEFINEDGROUPS 409

more frail than the typical panic patient. She believed that her arousalrelated sensations indicated a serious disease (e.g., heart disease), citing the diseases of her age cohorts as evidence. If her fears came true (e.g., collapsing during a panic attack) she also believed the consequences would be worsened by her age (e.g., she would be more likely to break a hip). The patient also believed that if she practiced cognitive restructuring (e.g., tried to ignore or discount catastrophic thoughts) then this might lead her to overlook a true sign of disease and thereby delay its detection and treatment. The therapist was able to correct the cognitive distortions associated with these concerns. The patient was led to recognize that she was overfocusing on sick age-cohorts and ignoring the vast majority of her social circle who were healthy for their age. To successfully challenge the patient's dysfunctional beliefs, 'the cognitive restructuring component of treatment required extensive drilling and repetition' (Rathus & Sanderson, 1996, p. 277). The authors further observed that: One of the most important challenges that arose during the course of treatment was not blithely accepting the clients' self-perceptions of frailty and increased vulnerability to risk while undergoing the protocol treatment. With a bit of common sense caution (e.g., medical check-up at treatment outset; not sending an elderly patient out to shovel an icy sidewalk as an interoceptive exposure exercise), we realized that our elderly patients could undergo the same treatment procedures as our other patients, and that their self-perceptions of frailty and vulnerability were examples of probability overestimation and catastrophic thinking-cognitions to be challenged like any others. Perhaps just as important, we realized that as therapists newly working with such clients, our own fears of age-related adverse consequences of this treatment were cognitive distortions to be challenged as well. (p. 279)

Culturally Defined Groups Cross-cultural Differences in Panic Disorder

Cultural factors may influence the way panic disorder is manifested. That is, the expression of panic may depend to some extent on culture-specific norms, values, and ways of labeling emotional states (Amering & Katschnig, 1990), and is possibly influenced by biological differences across cultural groups. Cultures differ in the reported frequency of symptoms occurring during panic attacks. Some symptoms (e.g., dyspnea, choking or smothering sensations, chest pain or discomfort, paresthesias) are more commonly reported in southern than northern

41 0

COGNITIVE-BEHAVIOURAL PROTOCOLSFOR SPECIALPOPULATIONS

European countries, whereas other symptoms (e.g., trembling and shaking) are more commonly reported in northern countries (Amering & Katschnig, 1990). In the US, panic disorder in African Americans, compared to Caucasians, is more likely to co-occur with hypertension and isolated sleep paralysis (American Psychiatric Association, 1998). Isolated sleep paralysis consists of an inability to speak or move, typically for a few seconds, as the person is falling asleep or awakening (Lishman, 1987). The implications of these comorbidities are still being investigated. Friedman et al. (1994) found that sleep paralysis had little relevance for planning or outcome of CBT2. Similarly, I have successfully treated panic in patients with isolated sleep paralysis, using Clark and Salkovskis' (1987) CBT2 protocol.

Assessment Issues A structured diagnostic interview, along with an understanding of the patient's cultural background, can be important for arriving at an accurate diagnosis. This is illustrated by a case described by Paradis et al. (1992). The patient was a 43 year-old Guyanian woman who was initially diagnosed by her psychiatrist-using an unstructured interview-as having schizophrenia with bizarre delusions. Paradis et al. (1992), who are experienced in the role of cultural factors in psychopathology (e.g., Friedman & Paradis, 1991; Friedman et al., 1994), subsequently re-assessed the patient with a structured diagnostic interview (the ADIS-R: see Chapter 9). Instead of diagnosing schizophrenia, the authors found that the patient had panic disorder. The initial misdiagnosis appeared to arise from the way the patient's cultural background influenced the way she described her panic attacks: e.g., 'I feel like I'm going to die ... I worry it's not natural ... not natural causes, evil like someone put a curse on me' (Paradis et al., 1992, p. 63). Such misdiagnoses are not rare. Paradis and colleagues found that panic disorder in minority patients was often misdiagnosed when US clinicians used unstructured interviews. Instead of identifying panic disorder, the clinicians tended to assign other diagnoses (e.g., psychotic disorders, mood disorders, adjustment reactions). Interviews using a structured diagnostic interview seemed to help overcome these errors. The authors concluded that 'although the ADIS-R does not focus on cultural issues, it may be an effective means for diagnosing panic disorder in a minority population because it focuses specifically on physical and cognitive symptoms pathognomonic of panic disorder' (p. 63). This conclusion should also apply to structured interviews for DSM-IV, such as the ADIS-IV and SCID-IV (these interviews are reviewed in Chapter 9).

PANIC TREATMENTSFOR DEMOGRAPHICALLYDEFINEDGROUPS 411

Treatment Issues and Protocols

CBT2 and situational exposure have been found to effectively treat panic disorder in many countries, including English-speaking countries (Chapters 5 and 6) and countries in which English is not the primary language, such as Germany, Hungary, and Japan (e.g., Kopp et al., 1986; Margraf & Schneider, 1991; Takahashi, 1993). However, there may be particular countries or cultural groups in which panic treatments need to be extensively modified in order to be effective. Knowledge of the patient's cultural (and socioeconomic) background, and conveying this understanding to the patient, can enhance patient satisfaction with treatment (Yamamoto et al., 1984) and thereby may improve treatment adherence and outcome. An understanding of the patient's cultural background can help the therapist understand how the patient developed catastrophic beliefs. This can help the therapist devise effective means of correcting these beliefs. To date, it appears that only two studies have been published on the role of cultural factors in the treatment of panic disorder. In a study of 15 Caucasian and 15 African American panickers, Friedman and Paradis (1991) found the African Americans were more likely to drop out of a treatment consisting of CBT2 with or without anti-panic medication. In a later study, Friedman et al. (1994) compared the treatment responses of 100 Caucasian and 43 African American panickers. As in the earlier study, patients were treated with CBT2 with or without drugs. The groups did not differ in treatment response or dropout rates. Friedman et al. argued that the difference across studies was due to differences in the treatment protocols. The 1991 findings led the authors to modify their treatment protocol in an attempt to reduce dropouts among African Americans. Treatment was modified in three main ways: • •

•

In the 1994 study, more emphasis was placed on educating and involving extended families in the treatment of African Americans. The authors changed the therapy format from individual (one-to-one) to group treatment; 'the addition of group therapy to the treatment protocol was especially beneficial to African American patients, most of whom had been unaware that other African Americans also suffered from panic disorder' (Friedman et al., 1994, p. 802). The lack of awareness might be partly a result of mass media descriptions of panic disorder, which typically portray panic sufferers as Caucasian. In their 1994 study, Friedman and colleagues referred patients to a wider array of social services, and provided patients with instruction in child management skills. The authors drew on these interventions

41 2

COGNITIVE-BEHAVIOURAL PROTOCOLSFOR SPECIALPOPULATIONS

because many of the African American patients were single, unemployed mothers. The interventions presumably eased some of the stressors in the patients' lives (e.g., financial stressors) and increased their social suppqrt. Note that in using these interventions, Friedman et al. were tailoring treatment to socioeconomic status rather than to cultural background. Nevertheless, it appears that tailoring the interventions to the patient's demographic characteristics was helpful in reducing treatment dropout.

Comment CBT2 protocols appear to need only slight modification for treating panic disordered adolescents and the elderly. The treatment of childhood panic is likely to require greater modification to existing protocols, in order to take the child's developmental level into consideration. If Ollendick (1997) is correct in suggesting that cued panic attacks are more common than unexpected panics in childhood panic disorder, then the protocols made need to be modified to focus more on cued panic. Treatment might focus more on graded situational exposure, and devote less time to interventions such as breathing retraining and interoceptive exposure. The small but growing literature on the use of panic treatments in culturally defined groups suggests that behavioural and cognitivebehavioural therapies can successfully reduce panic disorder in a variety of cultures. The treatments sometimes need to be modified to the specifics of the cultural groups. Understanding the patient's cultural and socioeconomic background is important for accurate diagnosis, treatment planning, and engaging the patient in treatment. Much more research needs to be done before we can move beyond these broad, tentative conclusions to more specific recommendations.

TREATING PANIC IN THE CONTEXT OF OTHER DISORDERS Panic disorder is frequently comorbid with other clinical conditions, and the severity of the latter often abates when panic disorder is successfully treated (Chapter 6). This suggests a simple sequential strategy for treating many comorbid presentations; use a protocol for treating the most severe disorder and then apply a protocol for treating whatever symptoms or disorders are left remaining. This approach is often useful, although in the absence of a case formulation it can lead to an incomplete or short-lived treatment response (Anderson et al., 1996).

TREATING PANIC IN THE CONTEXT OF OTHER DISORDERS 413

In recent years there has been growing interest in developing protocols that integrate the treatments of two or more disorders. In such protocols two disorders might be treated more or less at the same time (e.g., panic disorder and posttraumatic stress disorder, or panic and irritable bowel syndrome). In the following sections we will review these protocols. The protocols have been developed from adults treated in western countries. Accordingly, the treatment may need to be adapted for use in the demographic groups discussed in the first part of this chapter.

Posttraumatic Stress Disorder Panic attacks and panic disorder commonly occur in people with posttraumatic stress disorder (PTSD). In a study of 62 people seeking treatment for PTSD symptoms, Falsetti and Resnick (1997) found that 69% had panic attacks within the past few weeks. Many (72%) of the panickers thought they were going crazy or losing control during their attacks, and 38% thought they were having a heart attack. Thus, the feared consequences during these attacks were similar to those seen in panickers who do not suffer from PTSD. However, there also are some ways that panics in PTSD are different. Panic attacks sometimes arise from the catastrophic misinterpretation of the re-experienced symptoms of PTSD, such as misinterpretation of intrusive thoughts or flashbacks of the traumatic event(s) (Franklin, 1996; Zeanah, 1988). A PTSD victim, for example, might misinterpret the occurrence of vivid, intrusive images of the trauma as a sign that he or she is on the verge of some catastrophe (e.g., going crazy, dying, or losing control), thereby precipitating the vicious cycle of panic (Chapter 3). In a sample of 34 people with PTSD arising from motor vehicle accidents, we found that 29% had pariic disorder (Taylor et al., 1999). In each case PTSD and pariic disorder were precipitated by a motor vehicle accident. Anxiety sensitivity predicted the severity of PTSD symptoms, and treatment-related changes in anxiety sensitivity predicted changes in PTSD symptoms (Fedoroff et al., 2000). This suggests that treatments for reducing anxiety sensitivity (e.g., CBT2) may be useful in simultaneously treating panic and PTSD. However, panic treatments might not always be sufficient for treating PTSD (Elsenga & Emmelkamp, 1990; Saper & Brasfield, 1998). Saper and Brasfield described a case study showing that CBT2 reduced pariic disorder but not PTSD. The authors reduced the PTSD by means of an imaginal exposure treatment known as implosion therapy (Stampfl & Levis, 1967). The efficacy of treatments for concurrent panic disorder and PTSD might be improved by developing protocols that combine elements of

414

COGNITIVE-BEHAVIOURAL PROTOCOLS FOR SPECIAL POPULATIONS

empirically supported panic and PTSD treatments. Falsetti (1997a,b) has been developing and evaluating such a protocol, called Multiple-Channel Exposure Therapy (M-CET). This is a 12 week treatment conducted either in groups or individually, which can be used to treat PTSD arising from a variety of traumata. M-CET combines Craske et al.'s (1994) CBT2 for panic disorder with contemporary cognitive-behavioural therapies for PTSD (Kilpatrick et al., 1982; Resick & Schnicke, 1993). The protocol is still being refined (Sherry A. Falsetti, personal communication, 3 March 1998). The current version (Falsetti, 1997b) is summarized in Table 16.2. To make exposure to trauma-related stimuli more tolerable for patients with panic attacks and PTSD, M-CET begins by treating the panic attacks. After this, patients receive imaginal exposure to memories of the trauma, and in vivo exposure to harmless but fear evoking stimuli associated with the trauma (e.g., trauma-related places, objects, sounds). M-CET also includes cognitive restructuring to address distorted trauma-related beliefs. Clients are taught to look at the evidence for their beliefs and to identify when they are overestimating the risk of a negative outcome, catastrophizing, overgeneralizing, basing their thoughts on feelings instead of facts, and disregarding important aspects of situations. Cognitive skills are employed to assist participants in challenging distorted thinking and to fully process traumatic memories. Worksheets [address] disruptions [sic] in beliefs about safety, trust, esteem, and power/competency. (Falsetti, 1997a, p. 111)

The panic component of treatment includes breathing retraining, interoceptive exposure, and cognitive restructuring. Cued panics, unlike spontaneous panics, are regarded as conditioned responses to reminders of the trauma (Falsetti, 1997a). Exposure to panic symptoms 'serves not only to decrease fear of the symptoms themselves, but also provides exposure to the physiological component of fear associated with the trauma, and thereby weakens the association of physiological arousal and the traumatic memory' (Falsetti, 1997a, p. 112). Falsetti (1997a) summarized pilot data on 10 patients treated with M-CET. Of the 8 patients completing treatment, all were said to no longer experience panic attacks and all had a significant reduction in PTSD symptoms to the point that they no longer met DSM-IV criteria for PTSD. In summary, M-CET is a promising method for treating panic and PTSD, and merits further evaluation in controlled studies.

-i._l

Table 16.2 Multiple channel exposure therapy (M-CET): A protocol for treating posttraumatic stress disorder (PTSD) with comorbid panic attacks or panic disorder (Falsetti, 1997a,b) Session

1

Content

• •

2

• •

3,4

• •

5,6

• • •

7

• • • •

8

• • • •

9-11

•

12

• •

•

Education about the effects of traumatic events, PTSD symptoms, and panic attacks, based on Foa et al.'s (1989) information processing model of PTSD and Barlow's (1988) interoceptive conditioning model of panic. Homework: (1) monitoring PTSD symptoms and panic attacks, (2) identifying panic triggers, and (3) writing about the meaning of the traumatic event. The latter assignment is presumably used for the purposes of assessment and also may be used as imaginal exposure. Review of homework (reviewed in this and in all subsequent sessions). Assessment and education regarding the connection between hyperventilation, traumatic events, panic attacks. Breathing retraining practiced in session and set as homework. Cognitive restructuring for panic and PTSD. For example, restructuring catastrophic beliefs about arousal-related body sensations, and beliefs about external sources of danger and safety (e.g., 'I am never safe'). Cognitive restructuring exercises are also set as homework. Continued practice of cognitive restructuring and breathing retraining (in session and as homework). Interoceptive exposure (as per CBT2 for panic disorder) . In session 6 the patient is asked to write out the traumatic event (used as a form of imaginal exposure). Writing out the traumatic event and, for individual treatment, reading the event out loud. The event is not read out when M-CET is conducted as a group treatment, in order to avoid vicarious traumatization of other patients. Cognitive restructuring is used, as needed, to challenge maladaptive beliefs identified when the patient reads out the description of their traumatic event. Interoceptive exposure used as needed . Homework: practice skills learned so far and write out a description of the traumatic event again. Continue writing out the traumatic event. Continue challenging maladaptive beliefs . Continue interoceptive exposure, as needed . Review how the trauma has affected the person's beliefs about safety. Sessions focus primarily on graded in vivo exposure to traumarelated and panic-related stimuli (i.e., exposure to harmless but upsetting situations). Review of treatment progress and review of any difficulties encountered during therapy. Discussion of relapse prevention. Planning for continued self-directed treatment (e.g., in vivo exposure)

Note: It is important to ensure that breathing retraining, relaxation training, and coping strategies are not misused as safety behaviours (see Chapter 14).

416

COGNITIVE-BEHAVIOURAL PROTOCOLSFOR SPECIALPOPULATIONS

Schizophrenia Among people with schizophrenia, at least a third have panic attacks and 11-20% have concurrent panic disorder (Cutler, 1994; Hofmann, 1999). Compared to panic attacks in people without schizophrenia, the panics in schizophrenia sufferers are more likely to have psychotic features. For instance, one patient with schizophrenia and panic felt that her panic attacks were due to a device connected to her nervous system, which was controlled by 'the voices' (Cutler, 1994). Panic attacks and panic disorder are more frequent in people with schizophrenia compared to the general population. This raises the question of how panic and schizophrenia are related to one another. They may have etiologic factors in common (Hofmann, 1999) and clinical observations suggest that both may exacerbate one another (Kahn et al., 1987, 1988; Sandberg & Siris, 1987). Clinical reports also suggest that increases in the dose of anti-psychotic medications can sometimes exacerbate panic attacks in people with comorbid panic and schizophrenia (Argyle, 1990). This could occur if the medications induce arousal-related sensations that the person catastrophically misinterprets. Panic disorder is often ignored by clinicians treating schizophrenia (Argyle, 1990), particularly when clinicians regard schizophrenia as the central or most important of the patient's problems. This is an unfortunate oversight because panic disorder can cause much suffering in people with schizophrenia (Argyle, 1990). Several studies have evaluated panic treatments in patients concurrently receiving pharmacotherapy for schizophrenia. Benzodiazepines and imipramine have been found to reduce panic frequency in schizophrenia sufferers (e.g., Siris et al., 1989; Wolkowitz & Pikar, 1991). However, these drugs are associated with the same problems encountered when they are administered to non-schizophrenic panickers: benzodiazepine abuse and dependence, nonadherence with imipramine due to side effects, and panic relapse or exacerbation when the drugs are discontinued. Recently, Arlow et al. (1997) and Hofmann and colleagues (e.g., Bufka & Hofmann, 1999) adapted CBT2 to treat panic disorder in people with schizophrenia. The results are preliminary but offer some guidance about how standard CBT2 protocols (e.g., Craske et al., 1994) can be adapted for treating panic in schizophrenia. Arlow et al. (1997) conducted an open trial of group CBT2 administered as a component of a day program for 11 patients with comorbid panic

TREATINGPANICIN THECONTEXTOF OTHERDISORDERS417

disorder and schizophrenia. The CBT2 protocol consisted of 16 weekly sessions, beginning with 2-4 weeks of psychoeducation about panic disorder, including information about the role of catastrophic misinterpretations. Psychoeducation was continued until it was clear that patients fully understood the material. Handouts were used to ensure that patients understood the rationale for treatment and the treatment plan. Concurrent with the educational phase, patients were also instructed in breathing retraining, meditation, progressive muscle relaxation, thought stopping, positive self-talk, and positive visualization. Several weeks were spent practicing these methods, during sessions and as homework assignments. Then patients received 3-6 weeks of cognitive restructuring. They learned about faulty beliefs, and began to identify their own catastrophic cognitions. According to Arlow and coworkers, 'most patients enjoyed doing this work and completed it with little or no difficulty' (p. 146). The final 4-6 weeks were devoted to graded situational exposure. Throughout, the group leader followed the general approach of keeping sentences short, clear, and direct, repeating or reviewing important concepts often, and building the information base slowly. Consistent gentle encouragement was employed, with frequent acknowledgement of small gains. Many of the relaxation and visualization exercises were geared toward improving patients' self-esteem, and at least one complete session was devoted to a thorough discussion of self-concepts. However, every session focused to some degree on helping participants develop a more satisfying self-concept. (Arlow et al., 1997, p. 146)

Of 11 patients, 8 completed treatment. The other 3 were hospitalized during. the early weeks of CBT2. The a_uthors admitted all available patients into their study, including those who were becoming increasingly psychotic. This may account for why 3 patients (27% of their sample) required inpatient treatment for psychosis. After completing their study, Arlow and colleagues modified their protocol to exclude patients who are unable to focus on the methods used in CBT2 because of severe psychosis, hostility, or uncooperativeness. For the 8 patients completing CBT2, the effects of treatment were not as great as those seen in studies of non-psychotic panickers. Only 3 of the 8 completers reported a decline in panic frequency. The others reported no change. Contrast this with studies of non-psychotic panic patients, in which a mean of 88% are panic-free by the end of 8-16 weeks of treatment (Chambless & Gillis, 1994). Arlow et al. did not use interoceptive exposure, which may partly account for the modest treatment response. Also,

418

COGNITIVE-BEHAVIOURAL PROTOCOLSFOR SPECIALPOPULATIONS

formal thought disorder, delusions, and other features of schizophrenia may have interfered with CBT2. For example, behavioural experiments to prove the harmlessness of arousal-related sensations may be unconvincing for patients whose reasoning abilities are impaired, and may fail to persuade patients who hold delusional beliefs about the dangerousness of the sensations. Arlow et al. did not report whether their treatment reduced agoraphobia. This omission is unfortunate because these investigators emphasized situational exposure in their protocol, and their treatment strategy suggests that the reductions in agoraphobia may have exceeded the modest reductions in panic. We discovered participants in the CBT group who claimed to have had no panic symptoms during the past week but revealed on further questioning that they did not leave their homes except to attend the program .... We quickly learned to place a higher priority on having the patient engage in risk-taking behaviors, even if this led to panic symptoms, than on having the patient hand in a self-report that showed no panic attacks. We considered it a success, for example, if a member had a panic attack in a crowded place but used techniques such as taking time out, controlling breathing, and reality testing, and then carried on. We also learned to measure success in terms of the length of time it took for someone to recover and amount of disruption of plans. (Arlow et al., 1997, p. 148)

Hofmann and colleagues have also looked at ways of adapting CBT2 to treat panic disorder in people with schizophrenia. Although an outcome study has yet to be published, Bufka and Hofmann (1999) have used their clinical experiences to suggest guidelines for treating panic in people with schizophrenia. These authors observed that panic treatment may need to proceed more slowly when the patient has comorbid schizophrenia. The pace of therapy may need to be slowed because of the decreased concentration and other cognitive impairments associated with schizophrenia. Similar to the observations made by Arlow et al. (1997), Bufka and Hofmann (1999) suggested that some patients may be better served by shorter treatment sessions, and by covering less material in each session, so that only one new aspect of treatment is introduced each session. The number of sessions may need to be increased to cover all the material in, say, the Craske et al. (1994) protocol. Bufka and Hofmann further observed that many patients with panic and schizophrenia also have a history of traumatic experiences (e.g., sexual or physical assault) and so the therapist should be alert to the possibility that the panic attacks may be triggered by trauma-related stimuli. Falsetti's (1997a,b) M-CET for treating panic and PTSD might be helpful for these patients.

TREATING PANIC IN THE CONTEXT OF OTHER DISORDERS 419

Panickers with comorbid schizophrenia sometimes fear they will go crazy or lose control during their panic attacks (Argyle, 1990; Kahn et al., 1988; Sandberg & Siris, 1987). Hofmann (1999) pointed out that for people with schizophrenia, these fears are not always unrealistic, especially given the possibility that panic attacks may exacerbate psychotic symptoms. This raises the question of how these fears should be addressed in CBT2. If a patient's psychotic symptoms are clearly worsened by panic attacks, then it may not be fruitful (or accurate) to challenge beliefs that arousal-related sensations lead to insanity or loss of control. It may be more useful to simply use applied relaxation or breathing retraining to reduce the feared sensations. Fear of going crazy also can be addressed by helping patients understand the causes of prior exacerbations in schizophrenia. Arlow et al.'s (1997) therapists 'were frank about what triggered past hospitalizations, and, for the most part, patients were helped to. recognize that while anxiety may have been a contributing factor, a number of other problems (nonadherence with medications, family issues, separating from home) were central to their decompensations' (p. 146). To summarize, preliminary research suggests that CBT2 is moderately effective in treating panic disorder in people with schizophrenia.

Irritable Bowel Syndrome Recall from Chapter 2 that irritable bowel syndrome (IBS) is a functional gastrointestinal disorder characterized by abdominal pain and disturbed defecation (bloating, diarrhea, constipation, passage of mucous), occurring in the absence of identifiable gastrointestinal pathology. It commonly co-occurs with panic disorder, and so protocols for treating concurrent panic and IBS would be clinically useful. A promising start in this direction was reported by Eldridge et al. (1993), who described the treatment of a patient with IBS and panic disorder. The patient's panic attacks were often accompanied by diarrhea and fear of losing bowel control. As is often the case in comorbid panic and IBS, the patient's panic attacks appeared to arise from the catastrophic belief that various internal and external stimuli portend an impending, publicly observable loss of bowel control; 'the patient panicked about any sensation that occurred between her knees and her chest-but only in a situation where access to a bathroom might be restricted (socially or environmentally)' (Gloria D. Eldridge, personal communication, 2 June 1999). External stimuli associated with panic attacks included situations in which the patient was required to eat or drink in public,

420

COGNITIVE-BEHAVIOURAL PROTOCOLSFOR SPECIALPOPULATIONS

and situations in which it was difficult to gain rapid, unlimited access to a toilet. The patient reported having frequent 'uncontrollable' attacks of diarrhea, but was apparently always able to make it in time to a toilet. She attempted to avoid situations in which there might not be immediate access to a toilet, and situations in which she might feel embarrassed about the amount of time spent in the toilet. On days that the patient had to leave her house, she refrained from eating or drinking for the entire day, so as to minimize the risk of a bout of diarrhea. She had tried a number of different treatments for her problems, including anti-diarrhea medications. All met with limited success. Eldridge et al. (1993) adapted Barlow and Cerny's (1988) CBT2 protocol to develop a treatment package that treated the patient's IBS, panic attacks, and agoraphobia. The components of treatment were as follows: • •

•

Correcting misconceptions about normal bowel functioning. Bowel control training. This involved training the patient to discriminate among abdominal sensations (e.g., diaphragm movements vs. muscle contractions vs. intestinal distention), and interposing increasing delays between the onset of bowel sensations and going to the toilet. Graduated exposure to internal stimuli the patient misinterpreted as signaling impending loss of bowel control (e.g., intestinal distention). These interventions overlapped with bowel control training. Gloria D. Eldridge (personal communication, 2 June 1999), who was the patient's therapist, provided the following instructive account of the sorts of interoceptive exposure exercises that were used: We did a lot of assignments to show her that most sensations in that area had nothing to do with her bowels. She misinterpreted a lot of diaphragmatic sensations as bowel sensations because they happened in roughly the same bodily area. I had her doing a lot of breathing exercises, singing, laughing, etc. to discriminate the diffeFence between a bowel sensation and sensations that originated in the diaphragm. I also got her to do a lot of abdominal ... exercises [e.g., sit-ups] to show her the sensations of muscular contractions in that area. The intestinal distention part ... overlapped with interposing increasing delays between legitimate bowel sensations and going to the toilet. The patient would be asked to eat at a big meal and then refrain from running to the bathroom. We asked her to wait increasing amounts of time (carefully incremented) between the first bowel sensation and going to the toilet. This enabled the patient to become used to the fact that the first sensation isn't immediately and inevitably followed by catastrophic evacuation.

TREATING PANIC IN THE CONTEXT OF OTHER DISORDERS 421

•

•

Graduated situational exposure to external stimuli associated with fears of loss of bowel control (e.g., situations in which there was no immediate access to a toilet). Establishing regular eating patterns. This was implemented for two reasons: (1) to make the patient's bowel habits more regular, and (2) to discourage avoidance of food and liquids on days she journeyed from her home.

Treatment consisted of seven 90-min sessions over 15 weeks. The patient was accompanied to therapy sessions by her sister, who assisted with situational exposure. Thirteen weeks after these sessions the patient received six 1-h booster sessions spaced over 11 weeks. The booster sessions were added because the patient reported increasing anxiety and avoidance related to emolling in a community college program. The boosters focused on situations associated with the program, such as sitting in classes and eating in cafeterias. Panic frequency, anxiety, agoraphobia, and diarrhea frequency declined over the course of treatment, with gains maintained at 18 month followup. By that time the patient had completed the college program and had been employed for a year. Treatment improvement was also associated with changes in the patient's cognitions about diarrhea (Gloria D. Eldridge, personal communication, 27 May 1999). This is what would be expected if cognitions played a causal role in the patient's problems. Prior to treatment, each diarrhea episode was accompanied by catastrophic thoughts such as 'Oh my God, here it goes again; I'm having another attack. How do I get to the toilet? Something awful is going to happen'. After treatment, diarrhea episodes were less frequent, but still occasionally occurred. When they did occur, the patient had noncatastrophic thoughts, such as' Another little diarrhea episode. Everyone gets diarrhea. No big deal'. The results of this case study encourage controlled outcome studies of Eldridge et al.'s (1993) protocol for treating comorbid IBS and panic disorder. Dismantling studies also could be conducted in an attempt to identify the most effective components of treatment.

Non-cardiac Chest Pain Among patients attending cardiology clinics, it is not uncommon for some to have non~cardiacchest pain. That is, recurrent chest pain in the absence of cardiac disease (Katon, 1994). Many of these sufferers have panic disorder. Their panic attacks seem to arise from the catastrophic misinterpretation

------------------------~ 422

-

----

COGNITIVE-BEHAVIOURAL PROTOCOLSFOR SPECIALPOPULATIONS

of chest pain. In such patients, pain can occur during and between panic attacks. To illustrate, Mr Q. described having dull aching chest pain for hours on end. Panic attacks occurred only when he experienced intense stinging jabs of pain, which he misinterpreted as heart muscle spasms. In such patients, reassurance often does not change the patient's belief that pain is a symptom of heart disease (Lantinga et al., 1988). In fact, reassurance can further alarm patients, particularly if they receive vague or ambiguous medical advice that may be open to catastrophic misinterpretations (Klimes et al., 1990). A number of treatment packages have been developed for treating non-cardiac pain and associated symptoms such as panic attacks. These protocols can be used for treating patients with comorbid panic disorder and non-cardiac chest pain. Case reports and uncontrolled studies suggest that the following interventions are useful in reducing recurrent chest pain and associated panic attacks. •

•

•

•

EMG biofeedback applied to the chest muscles can be used to educate patients about the role of chest muscle tension in producing pain. This is likely to be especially useful for patients who have difficulty accepting the idea that their pain is 'simply' due to harmless tension (Eifert, 1992). Biofeedback might also be useful in reducing the frequency of chest pain, including pain occurring in between panic attacks. Increasing the patient's activity level also has been reportedly successful in reducing chest pain (Levenkron et al., 1985). Physical activity can serve as a behavioural experiment, challenging the patient's belief that chest pain is a symptom of heart disease. If pain does not increase with exertion then it is unlikely to be due to heart disease. This approach was used to treat Mr Q. He was asked to increase his activity level by participating in an aerobic exercise class three times per week, combined with prospective monitoring of chest pain, thoughts, emotions, and stressful events. Physical exercise decreased the pain, whereas perceived job demands increased the pain. This led Mr Q. to conclude that his pain was caused by worrisome thoughts about his job performance, not by heart disease. Removal of positive or negative reinforcement contingencies maintaining pain preoccupation and pain complaints (Levenkron et al., 1985). One can inform patients and their families that chest pain is not a legitimate excuse for avoiding household chores. Patients should be encouraged to complete their domestic and occupational responsibilities regardless of whether chest pain occurs. Relaxation training (e.g., progressive muscle relaxation or applied relaxation) to reduce episodes of tension-induced pain (Levenkron et al., 1985). As a caveat, note that if this intervention is used, then it is

TREATINGPANIC IN THE CONTEXTOF OTHERDISORDERS 423

•

•

important that patients understand that relaxation is intended simply to gain respite from pain, not to prevent some cardiac catastrophe (also see Chapter 14). Breathing retraining can be helpful, particularly for patients who are chronic hyperventilators (DeGuire et al., 1992; Hegel et al., 1989; Klimes et al., 1990). Such people tend to breathe with their chest muscles rather than their diaphragm, thereby causing chest pain (for details, see Chapter 4). As with relaxation training, it is important that patients understand that breathing retraining simply removes harmless sensations; it does not prevent any catastrophic event. Other methods used in CBT2 for panic disorder, such as cognitive restructuring, also can be helpful in convincing patients that their chest pain is harmless (Hegel et al., 1989; Klimes et al., 1990).

Studies using one or more of these methods have treated patients over 4-11 weekly lh sessions, implemented after the patient has had a thorough medical evaluation to rule out general medical conditions that might cause the symptoms. A sample protocol combining several of these interventions was developed by Klimes et al. (1990), and is summarized in Table 16.3. Results from several studies suggest that cognitive-behavioural interventions are generally useful in reducing chest pain, panic frequency, and pain-related avoidance (e.g., exercise avoidance), with gains maintained at followups of 4-12 months (DeGuire et al., 1992; Hegel et al., 1989; Klimes et al., 1990; Levenkron et al., 1985). Such treatments are generally acceptable to patients, even those who are skeptical about the role of psychological factors in causing their pain. Patients can be informed that their physical state could, if necessary, be medically re-evaluated at the end of cognitive-behavioural treatment (Klimes et al., 1990). According to Klimes and colleagues, Clinical observation of patients' positive reactions to an explanation of their complaints as compared with the previous reassurance suggests that the cognitive elements of therapy are indispensable. Patients' understanding of factors that can influence their pain and their realization that they can control their own symptoms is direct and convincing evidence to counter their fears of heart disease. Good initial response to our cognitive explanation appeared to predict therapeutic success. We believe that purely behavioural treatment (Lantinga et al., 1988) encouraging resumption of normal activities in the hope that attitudes and beliefs about illness will then change as a result ('If I can do these things, maybe I am healthy') is less effective. (p. 610)

424

COGNITIVE-BEHAVIOURAL PROTOCOLSFOR SPECIALPOPULATIONS

Table 16.3 Cognitive-behavioural protocol for treating non-cardiac chest pain in patients with or without panic disorder (Klimes et al., 1990) Session

1

2 3

4

5 onwards (up to a maximum of 11 sessions)

Content • •

Functional analysis of complaints. Explanation of why psychological treatment is being offered. • Description of treatment program. • Progressive muscle relaxation. This and other skills taught in this protocol are practiced in session and as homework. The latter is reviewed in the next session. • Hyperventilation (up to 2min) to demonstrate how easily 'real' sensations can be induced. • Breathing retraining. • Distraction training (focusing attention away from symptoms and from associated worries). • Self-monitoring homework: monitoring of chest pain, mood, and activity level in order to investigate the relationships among these variables. • Note that if patients are asked to distract themselves from chest pain, then they may record fewer episodes of pain on the homework monitoring form. • Continued practice of skills (relaxation, breathing retraining, and distraction). • Continued monitoring of chest pain, mood, and activity level. • Continued practice of skills. • Exposure to feared situations or activities (e.g., physical exercise). • Voluntary prevention of safety behaviours (e.g., refraining from repeated pulse taking and refraining from reassurance-seeking). • Pacing activities to reduce unrealistic demands. • Cognitive restructuring of remaining maladaptive beliefs about organic illness (i.e., restructuring of beliefs that were not corrected by information derived from monitoring or by the other interventions used in previous sessions). • Problem-solving for social problems.

Note: It is important to ensure that breathing retraining, relaxation training, and coping strategies are not misused as safety behaviours (see Chapter 14).

Comment In Chapter 6 we saw that comorbid conditions often abate when panic disorder is successfully treated. This suggests that a sequential treatment strategy is often useful; treat one disorder (e.g., panic disorder) and then

PANICAND PREGNANCY 425

treat the remaining symptoms or disorders (e.g., depression). This may not be the most effective or efficient way of treating all comorbid cases. A small but growing literature suggests alternative treatment strategies are useful, in which the CBT2 protocol is modified either (1) to treat panic attacks in the context of ongoing treatment for another disorder (e.g., treat panic in patients receiving drug treatment for schizophrenia) or (2) to initiate treatment for two disorders more or less simultaneously (e.g., concurrent treatment of panic disorder and irritable bowel syndrome). Only a handful of the possible comorbidities have been investigated. Empirically supported protocols have yet to be developed for treating panic disorder when it cooccurs with other disorders, such as personality disorders, hypochondriasis, or substance use disorders. The development and evaluation of protocols for the more common combinations of comorbidity is an important step in improving cognitive-behavioural treatment of panic disorder and co-occurring conditions. In the meantime, the therapist can use a case formulation to decide how to best treat comorbid disorders.

PANIC AND PREGNANCY Pregnancy is another clinical condition that can occur with panic disorder, and complicates the treatment of panic. Despite the fact that panic disorder most commonly occurs in women of child-bearing age, very little has been written on the treatment of panic in pregnant women. For pregnant panickers in their last trimester, the CBT2 therapist should be cautious about interventions requiring intense effort or those evoking intense anxiety or arousal-related sensations. Sometimes these activities may induce false labour. Thus, interoceptive exposure and situational exposure should be used with caution. The CBT2 therapist should consult with the patient's family physician or obstetrician about the medical advisability of CBT2. In some cases it may be worth postponing panic treatment until the pregnancy has come to term, particularly when the panic disorder is mild. However, as Robinson et al. (1992) pointed out, one needs to consider the effects of treatment-or no treatment-on the mother and child. These authors point out that is it not known how panic attacks and anxiety affect the fetus, the infant, and the mother-child relationship. Also unknown are the effects of anti-panic medication on the fetus and on breast-feeding infants. The use of a CBT2 package consisting of cognitive restructuring, applied relaxation, and breathing retraining may be the most benign, effective treatment for panic in pregnancy. If necessary, the full CBT2 package can be implemented post-partum.

426

COGNITIVE-BEHAVIOURAL PROTOCOLSFOR SPECIALPOPULATIONS

In what appears to be the only published study to investigate the treatment of pregnant panickers, Robinson et al. (1992) treated a 26-year-old woman 4 months before the scheduled birth of her baby. Treatment consisted of six weekly 1 h sessions, with a booster session 2 weeks after delivery of her child. Treatment consisted of psychoeducation, cognitive restructuring, and graded situational exposure. Interoceptive exposure was not used because of the authors' concerns about its effects on the fetus. Panic disorder declined with treatment, and treatment-related gains were maintained at 22 month followup. This finding supports the use of a partial CBT2 package in the treatment of panic in pregnancy. The efficacy of other partial packages (e.g., applied relaxation as the sole intervention) remains to be investigated.

SUMMARY AND CONCLUSIONS CBT2 protocols have been developed for treating panic disorder in a number of special populations. Although the results are promising, much remains to be learned, not only about the treatment of panic in the populations reviewed in this chapter, but also in other special populations, such as people who have panic disorder along with other psychiatric disorders (e.g., personality disorders) or general medical conditions. For many types of comorbidity, the clinician may best proceed by using a case formulation to guide treatment. Little is known about the treatment of rare or atypical manifestations of panic disorder. For the minority of panickers who suffer exclusively from nocturnal panic attacks, special CBT2 protocols may be required. These may incorporate sleep hygiene and methods for treating the patient's fear of sleep (Craske & Rowe, 1997). Similarly, special protocols may need to be developed for patients who have agoraphobia without a history of panic. Here, agoraphobia may be precipitated by recurrent paroxysmal bouts of symptoms such as vomiting, incontinence, or migraine headaches (Pollard et al., 1996). Effective treatment may need to directly treat the precipitants.

Chapter 17

OTHER PSYCHOLOGICAL TREATMENTS Throughout this book we have discussed strategies for improving the efficacy of behavioural and cognitive-behavioural treatments. These strategies may be of some help, but chances are that they will not help every patient in every circumstance. When patients fail to respond it may be that other psychological treatments can be usefully employed as adjunctive or alternative therapies. Many psychological therapies have been used to treat panic disorder, although only a fraction have been subject to any form of empirical evaluation. Of these we will review four therapies, selected because they are often used or seem promising for treating panic disorder: • • • •

psychodynamic psychotherapies hypnosis eye movement desensitization and reprocessing mindfulness meditation

This chapter will evaluate the efficacy of these interventions as standalone treatments, and consider whether they are worth integrating with second generation cognitive-behavioural therapy (CBT2) or with its components, such as situational exposure. Although the outcome literature from these studies was insufficient for inclusion in meta-analyses of panic treatments (Chapter 5), the literature does suggest some tentative conclusions.

PSYCHODYNAMIC PSYCHOTHERAPIES

Historical Perspectives Psychodynamic psychotherapies have long been used in an attempt to treat panic disorder. There have been several case reports of their successful use (see Milrod & Shear, 1991, for' a review), although it is unclear

428

OTHER PSYCHOLOGICAL TREATMENTS

whether the successes outnumber the failures. Psychodynamic therapists, like therapists of other persuasions, are reluctant to publish their failures. Moreover, in many case reports it is not mentioned whether the successful treatment consisted of psychodynamic therapy alone, or whether some form of situational exposure was also used. Even Freud believed situational exposure was necessary for treating clinically significant agoraphobia. Take the example of agoraphobia; there are two classes of it, one slight and the other severe. Patients belonging to the first indeed suffer from anxiety when they go about alone, but they have not yet given up going out alone on that account; the others protect themselves from the anxiety by altogether giving up going about alone. With these last one succeeds only when one can induce them through the influence of the analysis to behave like the first class, that is, to go about alone and to struggle with their anxiety while they make the attempt. One achieves, therefore, a considerable moderation of the phobia, and it is only when this has been attained by the physician's recommendation that the associations and memories come into the patient's mind enabling the phobia to be solved. (Freud, 1919/1950, p. 400)

Contemporary Psychodynamic

Psychotherapies

Wiborg and Dahl's Approach Recently, dynamic psychotherapies have been modified specially to reduce panic disorder, without necessarily using situational exposure exercises (Milrod et al., 1997; Wiborg & Dahl, 1996). Wiborg and Dahl reported that their brief psychodynamic psychotherapy (15 weekly sessions) plus clomipramine (9 months) was as effective as 9 months of clomipramine alone at post-treatment. At 9 month followup, the combined treatment was more effective in reducing relapse on measures of panic frequency, anticipatory anxiety, and agoraphobic avoidance. Wiborg and Dahl's dynamic treatment was commenced three weeks prior to clomipramine, in order to emphasize to patients the importance of psychotherapy. Patients with agoraphobia were not asked to expose themselves to feared situations. The dynamic treatment was based on concepts and techniques developed by other psychodynamic clinicians (Davanloo, 1978; Malan, 1979; Strupp & Binder, 1984), in addition to the authors' clinical experience. Unfortunately, the theory behind their treatment is vaguely described in Wiborg and Dahl's report. They hypothesized that people with panic disorder learn as children to be submissive to others. This originally serves a self-protective function, but later interferes with adult functioning and causes stress in interpersonal relationships. Pan-

PSYCHODYNAMIC PSYCHOTHERAPIES 429

ickers are also said to have 'the unconscious conflict between need for approval and fear of punishment for self-assertiveness' (Wiborg & Dahl, 1996, p. 691). This somehow contributes to panic attacks. The treatment based on this conceptualization seems to be a generic psychotherapy for interpersonal problems. The goal of the psychotherapy was to help the patient to develop insight into the origins and determinants of the dysfunctional patterns. This insight was needed to acquire more adaptive patterns of interpersonal relatedness. The therapist focused on the identification and working through of the dysfunctional patterns as they emerged in the past, the present, and the transference. The work included all the technical elements of dynamic psychotherapy: clarification, confrontation, and interpretation of the patient's resistance, defensive styles, and associated isolated affects. (Wiborg & Dahl, 1996, p. 691)

Wiborg and Dahl's (1996) study provides encouraging results that merit replication. It would be useful to see the treatment and its rationale spelled out in more detail. The publication of a treatment manual would help clinicians and clinical investigators better assess the usefulness of this therapy. Panic1ocused Psychodynamic Psychotherapy

Another new development in the treatment of panic disorder comes from the Cornell Panic-Anxiety Study Group (Busch et al., 1996; Milrod et al., 1997; Shear et al., 1993). On the basis of their clinical experiences, the psychodynamic literature, and research reports on panic disorder and related conditions, the Cornell group developed a psychodynamic theory of panic, and from this devised their Panic-focused Psychodynamic Psychotherapy (PFPP). The treatment was developed because the group believed that when dynamic therapy fails, it often does so because of an insufficient focus on panic symptoms (Busch et al., 1996). The Cornell group proposed that panic attacks arise from a number of factors, including catastrophic misinterpretations (Busch et al., 1996). People at risk for panic disorder are said to have (1) a neurophysiological vulnerability to panic attacks, as revealed in childhood by behavioural inhibition to novel stimuli and situations, and/ or (2) multiple experiences of 'developmental trauma'. These factors lead the person to become frightened of unfamiliar situations, and to become excessively dependent on the primary caregiver to provide a sense of safety. The caregiver is unable to always provide support, and so the child develops a fearful dependency on this person. This leads to the development of unconscious conflicts.

430

OTHER PSYCHOLOGICAL TREATMENTS

The child blames his or her inadequacies and fears on the parent, who is seen as unreliable and controlling and who may actually behave in an unreliable and controlling way. The child becomes angry at the parent's perceived or actual rejecting or frightening behavior but also fears loss because of the anger. A vicious cycle develops in which the child's anger threatens the needed tie to the parent and thereby increases fearful dependency, which promotes further frustration and rage at the parent. The ego's signal anxiety function-which generates smaller doses of anxiety to alert the ego to the presence of psychologically meaningful dangers and to act as a stimulus to mobilize defences ... -fails, leading to the onset of panic. This cycle recurs in adulthood, when threats to attachment can trigger intense feelings of abandonment, anger, and anxiety. (Busch et al., 1996, p. 75)

Thus, vulnerability to panic attacks is said to arise from conflicts about dependency (independence vs. reliance on others) and anger (expression vs. inhibition). The dependency conflict is said to express itself in a number of ways (Shear et al., 1993). Some panic-vulnerable people are sensitive to separation and overly reliant on others, while others are sensitive to suffocation and overly reliant on a sense of independence. In both instances, the self is represented in the unconscious as weak and ineffectual, and other people are represented as powerful. Avoidance of unfamiliar situations results in little opportunity to learn to accurately predict threats, and hampers the development of adaptive defenses and coping strategies. Instead, defenses remain immature and focused on the problem of maintaining a tolerable distance (not too close and not too far) from overly powerful others (Shear et al., 1993). The Cornell group proposed that other conflicts, such as sexual conflicts, also influence the vulnerability to panic attacks (Busch et al., 1996). The conflicts can activate conscious or unconscious fantasies of catastrophic danger, which can trigger panic attacks (Shear et al., 1993). In addition, the conflicts evoke aversive emotions, such as anxiety, anger, and guilt. The body sensations accompanying these emotions can become the focus of 'conscious as well as unconscious cognitive catastrophizing' (Shear et al., 1993, p. 862), thereby leading to panic attacks. The treatment derived from this theory-PFPP-focuses specifically on eliminating panic attacks. This is done by helping the patient understand and rectify the above-mentioned conflicts and maladaptive defences. This is achieved, in part, by examining the environmental and emotional precipitants of panic (Busch et al., 1996). PFPP is used in the short-term to reduce panic attacks (12-20 weeks of 1-2 sessions/week), and used in the longer-term to reduce the putative vulnerability to panic (9-12 months of 1-2 sessions/week). Note that this treatment is longer than the typical course of behavioural or cognitive-behavioural therapy (8-16 sessions).

PSYCHODYNAMIC PSYCHOTHERAPIES 431

PFPP is outlined in a manual by Milrod et al. (1997). PFPP can be used alone or in combination with other interventions, such as medication or CBT2. Uncontrolled case reports of the successful use of PFPP have been described (e.g., Busch et al., 1996; Milrod et al., 1997) but no controlled trials have been published. The availability of Milrod et al.'s treatment manual should be useful in testing their new therapy. Integrated Treatments It has often been suggested that treatment outcome may be enhanced by integrating behavioural or cognitive-behavioural therapies with other treatments, such as psychodynamic approaches (e.g., Goldfried et al., 1992; Scaturo, 1994; Wolfe, 1989). Combining behavioural or cognitive-behavioural therapies with contemporary dynamic treatments might improve the efficacy and durability of panic treatments. On the other hand, combination treatments might not be worth the extra effort. There have been few published studies of integrated treatments. Hoffart and Martinsen (1990) compared psychodynamic psychotherapy used either alone or in combination with cognitive-behavioural therapy for panic disorder. The dynamic treatment focused on conflicts thought to contribute to panic and anxiety (e.g., unresolved grief, repressed anger, feelings of isolation) and appears similar to the treatments developed by the Cornell group (Milrod et al., 1997) and by Wiborg and Dahl (1996). Hoffart and Martinsen's cognitive-behavioural treatment consisted of elements of CBT2; situational exposure and cognitive restructuring. After 11 weeks of inpatient treatment, patients in both conditions improved significantly on measures of anxiety, agoraphobia, and catastrophic beliefs (panic data were not reported). At 1 year followup the gains were maintained for the combined group, while patients treated with dynamic therapy alone tended to relapse. Although this finding supports the use of combined treatment, it remains to be seen whether this treatment is any better than CBT2 alone.

Comment Recent developments suggest that some forms of psychodynamic psychotherapy may be effective in treating panic disorder. The theories underpinning these treatments have not been empirically evaluated, and

432

OTHER PSYCHOLOGICAL TREATMENTS

so it is not known whether the theories can make clinically useful predictions for understanding and treating panic disorder. Uncontrolled case reports and a recent controlled study suggest that dynamic treatments may be useful, but further evaluation is required. It is possible that dynamic treatments could be used as adjuncts or alternatives to CBT2 or other empirically supported anti-panic treatments. However, this remains to be established. Clinicians should not assume that integrated treatments are necessarily better than CBT2 alone or exposure alone. The research on combining drug treatment with behavioural or cognitive-behavioural therapies indicates that adding drugs does not improve treatment efficacy (Chapter 7). The same may be true of adding dynamic therapies to either CBT2 or situational exposure.

HYPNOSIS Case Reports Several case reports suggest that hypnosis-used alone or in conjunction with other treatments-may be useful in eliminating panic attacks (e.g., Delmonte, 1995; Der & Lewington, 1990; Marriott, 1987; Matez, 1986; Stafrace, 1994; Wild, 1994). Some hypnotherapists have proposed that panic attacks arise from unresolved childhood problems (e.g., separation anxiety and childhood traumata) and have used hypnotic age regression to address these issues (Delmonte, 1995; Matez, 1986). Hypnosis has been used to 'install' positive self-statements (Marriott, 1987), to induce relaxation during situational exposure (Marriott, 1987), and to help patients vividly engage in imaginal exposure exercises (Der & Lewington, 1990; Wild, 1994). Hypnotic suggestions also have been used to remind patients to practice diaphragmatic breathing exercises, to use thought-stopping when fear-evoking thoughts occur, and to use coping statements (e.g., 'My heart is OK') during panic attacks (Stafrace, 1994). In a case study by Stafrace (1994), hypnosis was also used to 'facilitate' a behavioural experiment intended to teach the patient that missed heartbeats were 'caused' by focusing attention on his body. A trance was induced and the patient was instructed to concentrate on the sensation of his heart beating. The patient was given a hypnotic suggestion that he would occasionally notice a skipped beat. As intended, the manipulation of attention increased awareness of skipped beats. However, it is unclear how attention would cause arrhythmia. Focusing on one's body is more likely to simply increase the chances of detecting skipped beats (see

EYE MOVEMENT DESENSITIZATION AND REPROCESSING 433

Chapter 4). The experiment is a better illustration of the role of attention in the perception of body sensations. Although Stafrace thought that hypnosis facilitated this experiment, such experiments may work just as well without hypnosis (see Chapter 13). In contrast to the largely enthusiastic case reports of hypnotherapy for panic disorder, a controlled study by Marks et al. (1968) found that hypnosis was no better than systematic desensitization. In fact, there was a trend for systematic desensitization to be more effective. Hypnosis involved suggestions that patients would feel more relaxed while going out, and suggestions that their agoraphobia would gradually disappear. The finding that hypnosis tended to be less effective than systematic desensitization is notable, because desensitization tends to be weaker than situational exposure (Chapter 5). This suggests that hypnosis, as a stand-alone treatment, is weaker than situational exposure. An unanswered question, however, is whether or not it is useful to use hypnosis as an adjunct to either situational exposure or to the full CBT2 package.

Comment Little is known about the efficacy of hypnotherapy for panic disorder, either as a stand-alone treatment or as an adjunct to established treatments. Several case reports have described positive outcomes, although there is no telling how many unpublished case studies found hypnosis to be ineffective. It could be that hypnosis alone is simply a placebo. If this is the case then patients who respond to this treatment would be expected to relapse shortly afterwards. Controlled studies using long-term followup are needed to investigate this issue. Based on the current state of knowledge, the therapist would be better off using CBT2 or another empirically supported treatment as the mainstay for treating panic disorder. Hypnosis might be a used adjunct for patients who are sufficiently susceptible to hypnosis, although it remains to be shown that this improves treatment outcome.

EYE MOVEMENT DESENSITIZATION AND REPROCESSING One of the newest and most controversial developments in psychotherapy is eye movement desensitization and reprocessing (EMDR: Shapiro, 1995). The treatment was originally developed for reducing posttraumatic

434

OTHER PSYCHOLOGICAL TREATMENTS

stress disorder, but has since been applied to many other clinical conditions including panic disorder. EMDR has proved controversial for at least two reasons. First, Shapiro and others have claimed that it rapidly eliminates many clinical problems, sometimes within three 90min sessions. Thus, EMDR is claimed to work more rapidly than many other therapies. The second source of controversy is the ad hoc nature of the theory behind the treatment. The interested reader should consult a special issue of the Journal of Anxiety Disorders (1999, vol. 13, no. 1/2) for arguments on either side of these debates. The following sections will briefly describe the nature of EMDR, followed by a review of its efficacy in treating panic disorder.

Description and Rationale EMDR entails imaginal exposure to traumatic images while systematic saccadic eye movements are produced. Saccades are typically induced by tracking a therapist's finger as it is moved rapidly from side to side, about 12-14in (30-35cm) in front of the patient's face. While the patient tracks the therapist's finger, the patient is asked to visualize the traumatic memories, along with the body sensations and negative thoughts associated with the memory. After a series of about 24 eye movements, the patient is asked to report what comes to mind (e.g., new memories or sensations), and then notice those as the therapist waves his or her hand to induce another series of eye movements. The procedure continues in this fashion until the distress associated with the original memory is reduced. Coping statements also are introduced while the scene is being imagined. This is called 'cognitive interweave'. As a variation on this procedure, Shapiro (1995) suggested that eye movements in EMDR can be replaced with other forms of bilateral stimulation, such as tones delivered via headphones, oscillating between ears. Oscillatory hand taps also have been used, alternatively tapping the patient's left and right hands. Shapiro (1995) postulated that exposure to trauma produces neuronal changes and disruption of a physiological balance between excitatory and inhibitory systems in the brain, which prevents appropriate processing of traumatic memories. EMDR purportedly restores this balance and reverses the neural pathology, and in so doing, allows appropriate reprocessing and integration of the traumatic memories. Several researchers have raised doubts about this rationale, and alternative accounts of any treatment effects have been offered. Feske and Goldstein (1997) suggested

EYEMOVEMENTDESENSITIZATION AND REPROCESSING435

that exposure to fear-evoking stimuli is a core feature of EMDR. In other words, it may be that EMDR is effective in treating some clinical problems, but not for the reasons proposed by Shapiro (1995).

Efficacy in Treating Panic Disorder Over a dozen outcome studies have evaluated EMDR for posttraumatic stress disorder. In the aggregate, the evidence suggests that EMDR is more effective than placebo in treating this disorder (van Etten & Taylor, 1998). In comparison, only two studies have examined EMDR as a treatment for panic disorder, providing mixed support for its efficacy (Goldstein & Feske, 1994; Feske & Goldstein, 1997). In adapting EMDR for panic disorder, Goldstein and Feske conceptualized panic attacks as being similar to traumatic stressors. These investigators therefore used EMDR to reduce anxiety evoked by unpleasant memories of panic attacks, just as EMDR is used to reduce distress associated with memories of traumatic stressors. EMDR also was used to reduce fear of arousal-related sensations. Standard CBT2 methods were not used; e.g., there was no use of situational exposure homework assignments, breathing retraining, relaxation training, or interoceptive exposure. Very brief exposure to body sensations was induced only if patients failed to experience anxiety during treatment. This is different to the way that interoceptive exposure is used in CBT2 (see Chapter 13). Thus, in EMDR, interoceptive exposure was used only to elicit anxiety necessary to apply EMDR. Coping statements were used during some sets of eye movements, but these are a minor element of EMDR (Shapiro, 1995). In-their first study, Goldstein and Feske (1994) conducted an uncontrolled outcome trial in which 7 panic disordered patients received EMDR. Treatment consisted of a 60 min information gathering session followed by five 90min sessions of EMDR. There was a significant reduction in a range of symptoms, including panic frequency and anxiety. Anxiety sensitivity also declined over the course of treatment. In a second study, Feske and Goldstein (1997) randomly assigned 43 panic patients to either EMDR or to a placebo condition that contained all the elements of EMDR except for eye movements and other oscillatory stimulation (Feske & Goldstein, 1997; Ulrike Feske, personal communication, 14 June 1999). A wait-list control was also included. EMDR and placebo treatment consisted of five sessions; a 2h information gathering session followed by four 90min sessions. At the end of treatment, EMDR was

436

OTHER PSYCHOLOGICAL TREATMENTS

superior to the wait-list in terms of reductions in panic frequency, anxiety, agoraphobia, and anxiety sensitivity. EMDR was superior to the placebo control of two of five measures, with no difference in reductions in panic frequency. At 3 month followup, EMDR was no different from placebo on any outcome measure. Only 1 out of 15 patients completing EMDR (7%) achieved high end-state functioning (i.e., remission of panic disorder) at post-treatment. None achieved this level of functioning at followup. In comparison, estimates of high end-state functioning of cognitive-behavioural treatments for panic disorder range from 46 to 86% (Chambless & Gillis, 1993). Feske and Goldstein (1997) drew the following conclusions. The current study provides initial support for EMDR's efficacy in the treatment of panic disorder with agoraphobia. However, unless and until EMDR is shown to be as efficacious as more strongly evidence-based treatments such as exposure and cognitive-behavior therapy, we suggest that EMDR should not be the first-line treatment for this severe anxiety disorder. (p. 1034)

Comment The available research suggests that EMDR may be effective for panic disorder, although its efficacy may be largely due to nonspecific (placebo) effects. The evidence suggests that EMDR should not be used in place of CBT2 or other empirically supported treatments for panic disorder. It is possible that EMDR might be a useful adjunct to those therapies, but that remains to be demonstrated. EMDR may be better suited for treating posttraumatic stress disorder than panic disorder. When treating this combination of disorders, EMDR could be used to treat the symptoms of posttraumatic stress disorder, and CBT2 could be used to treat panic disorder.

MINDFULNESS MEDITATION Many different types of meditation have been used to help people overcome anxiety and related problems. Mindfulness meditation (MM) is among the most promising of these methods for treating panic disorder. MM is the major element of Kabat-Zinn's (1990) Stress Reduction and Relaxation Program. The program is generic; each patient completes the same program regardless of the reason for seeking treatment. Patients with a variety of disorders have been successfully treated with this approach,

MINDFULNESS MEDITATION 437

including pain disorder, hypertension, stress-related medical conditions (e.g., psoriasis), and anxiety disorders (Kabat-Zinn, 1990; Kabat-Zinn et al., 1985, 1992, 1998; Miller et al., 1995). When used to treat panic disorder, the rationale for using MM is broadly consistent with cognitive models of panic (Chapter 3) and is generally compatible with CBT2. The following sections will summarize MM, as used in Kabat-Zinn's program, and then review the research on applying this intervention to panic disorder.

Description and Rationale Mindfulness refers to nonjudgmental moment-to-moment awareness (Kabat-Zinn, 1990). MM consists of several exercises in which patients practice focusing their attention, in a sustained manner, on presentmoment experience. For example, focusing on one's breathing is one way of practicing mindfulness. Each time the person is distracted-e.g., by thoughts, memories, emotions, or body sensations-he or she simply acknowledges the distraction and then resumes focusing on breathing. A nonjudgmental stance is adopted, whereby the person attempts to simply observe the distractions, without evaluating them. Thus, in MM the patient attempts to 'let go' of distractions and consistently return to the task at hand (e.g., attending to one's breathing). During meditation we treat all our thoughts as if they are of equal value. We try to be aware of them when they come up and then we intentionally return our attention to the breath as the major focus of observation, regardless of the content of the thought! In other words, we intentionally practice letting go of each thought that attracts our attention, whether it seems important and insightful or unimportant and trivial. We just observe them as thoughts, as discrete events that appear in the field of our awareness. We are aware of them because they are there but we intentionally decline getting caught up in the content of the thoughts during meditation. (KabatZinn, 1990, p. 68, emphasis in original)

The purpose of MM is to cultivate stable, nonreactive present moment awareness. That is, to become more mindful in daily life-particularly in stressful situations-so one does not get carried away by ruminations about the past or worries about the future. In this way the person can live fully in the 'here and now', rather than operating in a 'mindless', automatic fashion. Other goals of MM are to increase self-understanding and to learn to achieve relaxation. Relaxation is not the main objective; it comes by itself with continued practice. Patients are advised to practice MM exercises during periods of low stress so they will be better able to

438

OTHER PSYCHOLOGICAL TREATMENTS

apply these exercises during high stress. There are many formal and informal MM exercises (see Kabat-Zinn, 1990). The following are some of the main exercises. •

•

•

•

Sitting meditation (10-20 min/ day). Sitting upright, with eyes closed, the aim of this exercise is to become aware of one's breathing; focusing on the sensations associated with breathing. The aim is to observe one's breathing without trying to control it; 'just let it happen and be aware of it, feeling how it feels, witnessing it as it flows in and out' (Kabat-Zinn, 1990, p. 22). The patient is asked to do this by focusing on the diaphragm rather than the nostrils, because focusing on the former seems to bring on a greater sense relaxation. Body scan (three 45min sessions/week, alternated with Mindful Hatha Yoga). In this exercise the patient lies back in a comfortable position, with eyes closed. The goal is not to fall asleep. The patient focuses on the rise and fall of the abdomen with each breath, and then follows a particular sequence in which he or she attends to various parts of the body. For example, spending some time paying attention to any sensations in the toes of the left foot, and then sequentially shifting attention to other parts of the body. Mindful Hatha Yoga (three 45min sessions/week). This consists of a series of gentle stretching and strengthening excises (e.g., rolling one's head in one direction and then the other). These are done very slowly, and should not be done to the point that pain occurs. The goal is to focus on the sensations of breathing and other sensations that arise during the exercises. Informal mindfulness exercises. There is no limit to the number of informal mindfulness exercises that can be devised. These often involve being mindful of routine activities. For example, in the walking meditation the person practices being mindful of walking, noticing the body sensations that arise as he or she paces back and forth. Eating meditation involves being mindful of the act of eating, noticing the taste of food and becoming aware of the sensations of mastication and swallowing.

Audiotapes are used to help the patient focus on the various exercises (e.g., the body scan). Also included is the practice of slow, diaphragmatic breathing, in which patients are instructed to relax their abdomens as much as they can. This is similar to the breathing retraining exercises used in CBT2 (Chapter 14). In Kabat-Zinn's stress reduction program, MM is practiced in 8 weekly 2hr classes, plus an all-day meditation retreat. There

MINDFULNESS MEDITATION 439

are about 30 patients in each class. Patients are also asked to complete homework assignn1ents for 45min/ day, 6 days/week, in which they practice the various MM exercises. Patients are counselled to get into the habit of practicing these exercises, and to continue to do so after the end of the 8 week program.

Efficacy in Treating Panic Disorder Kabat-Zinn et al. (1992) reported an uncontrolled trial of MM in treating 24 patients who had either panic disorder (n = 14) or generalized anxiety disorder (n = 10) as their primary (most severe) disorder. These patients completed the program in classes that included patients with other problems. Two patients dropped out, both of whom had generalized anxiety disorder as their main problem. In the course of the program, Kabat-Zinn and colleagues observed 'it can be a substantial challenge for patients with panic disorder to sit still for long periods of time, attempting to observe anxious thoughts and impulses as they arise and working with them mindfully rather than succumbing to impulses of reactivity and panic' (p. 942). Nevertheless, all the panic patients were able to complete the program. At the end of treatment, panic patients (as a group) had a significant reduction in panic frequency and in the severity of anxiety and agoraphobia, with gains maintained at 3 month followup. At followup, 91% of patients (20 of 22) reported that they continued to practice the MM exercises, typically 3-4 times per week. Medication status (taking vs. not taking psychotropic medication) and demographic variables failed to predict treatment outcome. Pretreatment expectancy for improvement also failed to predict outcome. Interestingly, self-reported amount of practice of the MM exercises also failed to predict treatment outcome. It could be that this was due to a restricted range in the amount of practice; if all patients completed most of their homework, then there would be insufficient variance for homework adherence to predict outcome. Miller et al. (1995) reported a 3 year followup of 18 of the 22 patients completing Kabat-Zinn et al.'s (1992) study. Gains tended to be maintained on measures of anxiety, agoraphobia, and panic frequency. A total of 89% (16 of 18) reported that they continued to use at least some of the MM techniques, and most (89%) thought the program had lasting value. Only 12% began some form of additional treatment during the followup interval.

440

OTHER PSYCHOLOGICAL TREATMENTS

Comment Preliminary results indicate that MM is a promising treatment for panic disorder. MM exercises are readily learned and patients appear willing and able to practice them on a long-term basis. The findings certainly encourage controlled trials of MM. In particular, it is of interest to determine how the efficacy of MM compares to CBT2, and whether outcome is improved by combining these interventions.

If controlled studies indicated that MM is effective in reducing panic disorder, then the issue would be to determine its mechanism of action. Panic patients could use attention-focusing exercises to distract themselves from the feared sensations. For example, using the eating meditation to distract oneself from feared palpitations. If this is the case then MM would reduce panic by helping patients avoid the body sensations that lead to panic attacks. 1 Although patients could use MM to better distract themselves from body sensations, fearful distraction is inconsistent with the practice of MM. Kabat-Zinn (1990) advises patients to observe 'distracting' stimuli in a non-judgmental fashion, not to avoid the stimuli. If the patient becomes frightened by body sensations, then he or she is encouraged to simply observe the fear and to dispassionately observe any associated thoughts (e.g., catastrophic misinterpretations). Thus, patients are instructed to distance themselves from fear-evoking sensations and catastrophic thoughts; observing the sensations and other contents of consciousness without trying to change them. This is similar to the distancing strategies used in cognitive restructuring (Chapter 12). Relaxation-induced anxiety is addressed in MM and CBT2 in similar ways; in both cases the patient is encouraged to nonjudgmentally observe the sensations of relaxation (e.g., slowed heartbeat), and thereby learn that relaxation is harmless. MM and CBT2 have several other things in common; both use some form of breathing retraining, both use exercises that induce relaxation, and both discourage patients from making catastrophic misinterpretations of body sensations. CBT2 changes interpretations of body sensations by structured exercises such as behavioural experiments. MM appears to change interpretations by encouraging patients to observe the sensations without judging them; e.g., to observe palpitations while suspending the judgment

1

The same problem might occur in transcendental meditation, which is quite different from MM. In transcendental meditation the patient focuses on a mantra-a word or phraserepeated silently to attain a meditative state. The mantra could be used to distract the person from feared sensations.

SUMMARY AND CONCLUSIONS 441

that these sensations are 'bad' or 'dangerous.' MM exercises may thereby allow patients to learn that arousal-related sensations are harmless. The meditative, cognitive, and cognitive-behavioral approaches share an emphasis on noting sensations and thoughts without viewing them as catastrophic and the use of stress-inducing situations as cues to engage in new behaviors. They also have in common the use of homework assignments to reinforce what was learned in the group sessions .... Patients who are able to identify anxious thoughts as thoughts, rather than as 'reality', report that this alone helps reduce their anxiety and increases their ability to encounter anxiety-producing situations more effectively. (Kabat-Zinn et al., 1992, pp. 941-942, emphasis in original)

In summary, when it comes to treating panic disorder, it may be that MM works in a similar way to CBT2; reducing hyperventilation (if this occurs), reducing arousal, and reducing the tendency to make catastrophic misinterpretations. Despite their similarities, there are a number of interventions and procedures that distinguish MM from CBT2 (Kabat-Zinn et al., 1992). The differences may have important implications for combining these treatments. In CBT2 an important goal is to teach the patient to evaluate whether their cognitions are realistic or unrealistic. In MM the patient is encouraged simply to observe the thoughts without judging them. If the clinician combines CBT2 and MM, it is important to ensure that these different approaches do not confuse patients about when they should or shouldn't evaluate and challenge their cognitions. One way of using MM within CBT2 is to explain to patients that MM exercises are intended to help them observe their catastrophic thoughts without reacting to them, in order to learn that the thoughts simply come and go, without any harmful consequences.

SUMMARY AND CONCLUSIONS Four approaches to panic disorder were reviewed in this chapterpsychodynamic psychotherapies, hypnosis, eye movement desensitization and reprocessing, and mindfulness meditation. None have been extensively evaluated as panic treatments, and none have been shown to outperform empirically supported treatments such as CBT2. In the routine treatment of panic disorder, CBT2 should be used as a first-line approach. If this treatment fails, the clinician should attempt to understand the reasons for treatment failure and try to implement a modified CBT2. This might include elements of the other therapies. If this fails then

442

OTHERPSYCHOLOGICAL TREATMENTS

the clinician might consider using one of the alternative approaches. For example, if a patient fails to respond to a standard course of CBT2, then perhaps a course of mindfulness meditation would be considered. Decisions to use the other therapies as adjuncts or alternatives to CBT2 need to be made on a case-by-case basis, because there are currently no empirical data to guide the clinician. Treatments other than those reviewed in this chapter also might prove useful as adjuncts or alternatives to CBT2. Interpersonal Psychotherapy (Klerman et al., 1984), for example, is a psychodynamic treatment that has proved useful in treating a number of disorders including depression and bulimia nervosa (Fairburn et al., 1995; Klerman & Weissman, 1993). This therapy might be usefully adapted to treat panic disorder as a stand-alone treatment. Procedural differences between interpersonal psychotherapy and CBT2 may make it difficult to combine the two (Wilson, 1996a). Regardless of how the therapist chooses to treat panic disorder, the treatment should be based on a sound rationale.

REFERENCES Abelson, J.L., & Curtis, G.C. (1993). Discontinuation of alprazolam after successful treatment of panic disorder: A naturalistic follow-up study. Journal of Anxiety Disorders, 7, 107-117. Adler, C.M., Craske, M.G., & Barlow, D.H. (1987). Relaxation-induced panic (RIP): When rest isn't peaceful. Integrative Psychiatry, 5, 94-100. Agras, W.S., Leitenberg, H., & Barlow, D.H. (1968). Social reinforcement in the modification of agoraphobia. Archives of General Psychiatry, 19, 423--427. Airola, P.O. (1977). Hypoglycemia: A better approach. Phoenix, AZ: Health Plus. Albus, M., Lecrubier, Y., Maier, W., Buller, R., Rosenberg, T., & Hippius, H. (1990). Drug treatment of panic disorder: Early response to treatment as a predictor of final outcome. Acta Psychiatrica Scandinavica, 82, 359-365. Alexander, P.E. (1991). Management of panic disorders. Journal of Psychoactive Drugs, 23, 329-333. Alford, B.A., & Beck, A.T. (1994). Cognitive therapy for delusions. Behaviour Research and Therapy, 32, 369-380. Allison, D.B., & Gorman, B.S. (1993). Calculating effect sizes for meta-analysis: The case of the single case. Behaviour Research and Therapy, 31, 621-631. American Academy of Child and Adolescent Psychiatry. (1997). Practice parameters for the assessment and treatment of children and adolescents with anxiety disorders. Journal of the American Academy of Child and Adolescent Psychiatry, 36, (suppl. 10), 69S-84S. American Psychiatric Association. (1980). Diagnostic and statistical manual of mental disorders (3rd ed.). Washington, DC: Author. American Psychiatric Association. (1987). Diagnostic and statistical manual of mental disorders (3rd ed., rev.). Washington, DC: Author. American Psychiatric Association. (1994). Diagnostic and statistical manual of mental disorders (4th ed.). Washington, DC: Author. American Psychiatric Association (1995). Practice guideline for psychiatric evaluation of adults. American Journal of Psychiatry, 152 (Nov. suppl.), 63-80. American Psychiatric Association. (1998). Practice guidelines for the treatment of patients with panic disorder. American Journal of Psychiatry, 155 (May suppl.), 1-34. Amering, M., & Katschnig, H. (1990). Panic attacks and panic disorder in crosscultural perspective. Psychiatric Annals, 20, 511-516. Amsterdam, E.A. (1990). Emotions, cardiac arrhythmias, and sudden death. In D.G. Byrne & R.H. Rosenman (Eds.), Anxiety and the heart (pp. 251-258). New York: Hemisphere. Anastasi, A. (1988). Psychological testing (6th ed.). New York: Macmillan. Andersch, S., Hanson, L., & Hallstrom, T. (1997). Panic disorder: A five-year follow-up study in 52 patients. European Journal of Psychiatry, 11, 145-155.

444

REFERENCES

Anderson, J.R. (1990). Cognitive psychology and its implications (3rd ed.). New York: Freeman. Anderson, K.W., Taylor, S., & McLean, P.D. (1996). Panic disorder associated with blood-injury reactivity: The necessity of establishing functional relationships among maladaptive behaviors. Behavior Therapy, 27, 463-472. Andrade, L., Eaton, W.W., & Chilcoat, H. (1994). Lifetime comorbidity of panic attacks and major depression in population-based study: Symptom profiles. British Journal of Psychiatry, 165, 363-369. Andrews, G., Crino, R., Hunt, C., Lampe, L., & Page, A. (1994). The treatment of anxiety disorders: Clinician's guide and patient manuals. Cambridge: Cambridge University Press. Andrews, G., & Moran, C. (1988). Exposure treatment of agoraphobia with panic attacks: Are drugs essential? In I. Hand & H.-U. Wittchen (Eds.), Panic and phobias II: Treatments and variables affecting course and outcome (pp. 89-99). Heidelberg: Springer-Verlag. Antony, M.M., Meadows, E.A., Brown, T.A., & Barlow, D.A. (1994). Cardiac awareness before and after cognitive-behavioural treatment for panic disorder. Journal of Anxiety Disorders, 8, 341-350. Antony, M.M., Roth, D.A., & Swinson, R.P. (1998). Symptom induction exercises in panic disorder with agoraphobia and non-anxious individuals. Unpublished data, Department of Psychiatry, McMaster University. Apfeldorf, W.J., Shear, M.K., Leon, A.C., & Portera, L. (1994). A brief screen for panic disorder. Journal of Anxiety Disorders, 8, 71-78. Arena, J.G., Blanchard, E.B., Andrasik, F., Cotch, B.A., & Myers, P.E. (1983). Reliability of psychophysiological assessment. Behaviour Research and Therapy, 21, 447-460. Argyle, N. (1988). The nature of cognitions in panic disorder. Behaviour Research and Therapy, 26, 261-264. Argyle, N. (1990). Panic attacks in chronic schizophrenia. British Journal of Psychiatry, 157, 430-433. Argyle, N., Deltito, J.,Allerup, P., Maier, W., Albus, M., Nutzinger, D., Rasmussen, S., Ayuso, J.L., & Bech, P. (1991). The Panic-Associated Symptom Scale: Measuring the severity of panic disorder. Acta Psychiatrica Scandinavica, 83, 20-26. Arlow, P.B., Moran, M.E., Bermanzohn, P.C., Stronger, R., & Siris, S.G. (1997). Cognitive-behavioral treatment of panic attacks in chronic schizophrenia. Journal of Psychotherapy Practice and Research, 6, 145-150. Arnow, B.A., Taylor, C.B., Agras, W.S., & Telch, M.J. (1985). Enhancing agoraphobia treatment outcome by changing couple communication patterns. Behavior Therapy, 16, 452-467. Arntz, A., & van den Hout, M. (1996). Psychological treatments of panic disorder without agoraphobia: Cognitive therapy versus applied relaxation. Behaviour Research and Therapy, 34, 113-121. Aronson, T.A., & Craig, T.J. (1986). Cocaine precipitation of panic disorder. American Journal of Psychiatry, 143, 643-645. Aronson, T.A., & Logue, C.M. (1987). On the longitudinal course of panic disorder: Developmental history and predictors of phobic complications. Comprehensive Psychiatry, 28, 344-355. Arrindel!, W.A. (1993). The fear of fear concept: Retest artefact and predictive power. Behaviour Research and Therapy, 31, 139-148. Arrindel!, W.A., Cox, B.J.,van der Ende, J., & Kwee, M.G.T. (1995). Phobic dimensions-II. Cross-national confirmation of the multidimensional structure underlying the Mobility Inventory (MI). Behaviour Research and Therapy, 33, 711-724.

REFERENCES445 Arrindel], W.A., Emmelkamp, P.M.G., & Sanderman, R. (1986). Marital quality and general life adjustment in relation to treatment outcome in agoraphobia. Advances in Behaviour Research and Therapy, 7, 139-185. Arrindell, W.A., Oei, T.P.S., Evans, L., & van der Ende, J. (1991a). Agoraphobic, animal, death-injury-illness and social stimuli clusters as major elements in a four-dimensional taxonomy of self-rated fears: First-order level confirmatory evidence from an Australian sample of anxiety disorder patients. Advances in Behaviour Research and Therapy, 13, 227-249. Arrindell, W.A., Pickersgill, M.J., Merckelbach, H., Ardon, A.M., & Cornet, F.C. (1991b). Phobic dimensions: III. Factor analytic approaches to the study of common phobic fears; an updated review of findings obtained with adult subjects. Advances in Behaviour Research and Therapy, 13, 73-130. Ascher, L.M. (1981). Employing paradoxical intention in the treatment of agoraphobia. Behaviour Research and Therapy, 19, 533-542. Ascher, L.M., Schotte, D.E., & Grayson, J.B. (1986). Enhancing effectiveness of paradoxical intention in treating travel restriction in agoraphobia. Behavior Therapy, 17, 124-130. Ashton, H. (1984). Benzodiazepine withdrawal: An unfinished story. British Medical Journal, 288, 1135-1140. Asmundson, G.J.G., Norton, G.R., Lanthier, N.J., & Cox, B.J. (1996). Fear of anxiety: Do current measures assess unique aspects of the construct? Personality and Individual Differences, 20, 607-612. Asmundson, G.J.G., Sandler, L.S., Wilson, K.G., & Walker, J.R. (1992). Selective attention toward physical threat in patients with panic disorder. Journal of Anxiety Disorders, 6, 295-303. Asmundson, G.J.G., & Taylor, S. (1996). Role of anxiety sensitivity in pain-related fear and avoidance. Journal of Behavioral Medicine, 19, 577-582. Asnis, J., Faisal, A., & Sanderson, W. (1999). Environmental factors in panic disorder. Journal of Clinical Psychiatry, 60, 264. AuBuchon, P.G., & Malatesta, V.J. (1998). Managing the therapeutic relationship in behavior therapy: The need for a case formulation. In M. Bruch & F.W. Bond (Eds.), Beyond diagnosis: Case formulation approaches in CBT (pp. 141-165). Chichester: Wiley. Bakker, A., van Balkom, A.J.L.M., Spinhoven, P., Blaauw, B., & van Dyck, R. (1998). Follow-up on the treatment of panic disorder with or without agoraphobia: A quantitative review. Journal of Nervous and Mental Disease, 186, 414--419. Ball, S.G., Otto, M.W., Pollack, M.H., & Rosenbaum, J.F. (1994). Predicting prospective episodes of depression in patients with panic disorder: A longitudinal study. Journal of Consulting and Clinical Psychology, 62, 359-365. Ball, S.G., Shekhar, A., Bedwell, C., & Clair, D. (1998, March). Basilar arterial responsivity to hyperventilation provocation in panic disorder: Implications for classification and treatment. Paper presented at the 18th annual meeting of the Anxiety Disorders Association of America, Boston, MA. Ballenger, J.C., Burrows, G.D., DuPont, R.L., Lesser, I.M., Noyes, R., Pecknold, J.C., Rifkin, A., & Swinson, R.P. (1988). Alprazolam in panic disorder and agoraphobia: Results from a multicenter trial. I. Efficacy in short-term treatment. Archives of General Psychiatry, 45, 413--422. Ballenger, J.C., & Pyer, A.J. (1996). Panic disorder and agoraphobia. In T.A. Widiger, A.J. Frances, H.A., Pincus, R. Ross, M.B. First, & W.W. Davis (Eds.), DSM-IV sourcebook, vol. 2 (pp. 411--471). Washington, DC: American Psychiatric Association.

446

REFERENCES

Bandelow, B. (1995). Assessing the efficacy of treahnents for panic disorder and agoraphobia. IL The Panic and Agoraphobia Scale. International Clinical Psychopharmacology, 10, 73-81. Bandelow, B., Brunner, E., Broocks, A., Beinroth, D., Hajak, G., Pralle, L., & Ruther, E. (1998). The use of the Panic and Agoraphobia Scale in a clinical trial. Psychiatry Research, 77, 43-49. Bandelow, B., Hajak, G., Holzrichter, S., Kunert, H.J.,. & Ruther, E. (1995). Assessing the efficacy of treatments for panic disorder and agoraphobia. I. Methodological problems. International Clinical Psychopharmacology, 10, 8393. Bandura, A. (1986). Social foundations of thought and action. Englewood Cliffs, NJ: Prentice-Hall. Bandura, A. (1988). Self-efficacy conception of anxiety. Anxiety Research, 1, 77-98. Bannister, R. (1983). Autonomic failure. Oxford: Oxford University Press. Barbone, F., McMahon, A.D., Davey, P.G., Morris, A.D., Reid, LC., McDevitt, D.G., & McDonald, T.M. (1998). Association of road-traffic accidents with benzodiazepine use. Lancet, 352, 1331-1336. Barloon, T.J., & Noyes, R. (1997). Charles Darwin and panic disorder. Journal of the American Medical Association, 277, 138-141. Barlow, D.H. (1988). Anxiety and its disorders. New York: Guilford. Barlow, D.H. (1994). Effectiveness of behavior treatment for panic disorder with and without agoraphobia. In B.E. Wolfe & J.D. Maser (Eds.), Treatment of panic disorder: A consensus development conference (pp. 105-120). Washington, DC: American Psychiatric Press. Barlow, D.H. (1997). Cognitive-behavioral therapy for panic disorder: Current status. Journal of Clinical Psychiatry, 58 (suppl. 2), 32-36. Barlow, D.H. (1998, March). Results from the multi-center clinical trial on the treatment of panic disorder-cognitive behavior treatment vs. imipramine vs. their combination. Proceedings of the 18th annual meeting of the Anxiety Disorders Association of America. Barlow, D.H., Brown, T.A., Craske, M.G., Rapee, R.M., & Antony, M.M. (1991, November). Treatment of panic disorder: Follow-up and mechanisms of action. Paper presented at the 25th annual meeting of the Association for Advancement of Behavior Therapy, New York, NY. Barlow, D.H., & Cerny, J.A. (1988). Psychological treatment of panic. New York: Guilford. Barlow, D.H., & Craske, M.G. (1988). The phenomenology of panic. In S. Rachman & J.D. Maser (Eds.). Panic: Psychological perspectives. (pp. 11-35). Hillsdale, NJ: Erlbaum. Barlow, D.H., & Craske, M.G. (1994). Mastery of your anxiety and panic II: Client workbook. San Antonio, TX: Psychological Corporation. Barlow, D.H., Craske, M.G., Cerny, J.A., & Klosko, J.S. (1989). Behavioral treatment of panic disorder. Behavior Therapy, 20, 261-282. Barlow, D.H., O'Brien, G.T., & Last, CG. (1984). Couples treahnent of agoraphobia. Behavior Therapy, 15, 41-58. Barlow, D.H., O'Brien, G.T., Last, CG., & Holden, A.E. (1983). Couples treahnent of agoraphobia: Initial outcome. In K.D. Craig & R.J. McMahon (Eds.), Advances in clinical behavior therapy (pp. 99-126). New York: Brunner/Mazel. Barlow, D.H., & Seidner, A.L. (1983). Treahnent of adolescent agoraphobics: Effects on parent-adolescent relations. Behaviour Research and Therapy, 21, 519-526.

REFERENCES447

Barlow, D.H., Vermilyea, J.A., Blanchard, E.B., Vermilyea, B.B., DiNardo, P.A., & Cerny, J.A. (1985). The phenomenon of panic. Journal of Abnormal Psychology, 94, 320-328. Barlow, J.B., Bosman, C.K., Pocock, W.A., & Marchand, P. (1968). Late systolic murmurs and non-ejection (mid-late) systolic clicks: An analysis of 90 patients. British Heart Journal, 30, 203-218. Baron, RM., & Kenny, D.A. (1986). The moderator-mediator variable distinction in social psychological research: Conceptual, stategic, and statistical considerations. Journal of Personality and Social Psychology, 51, 1173-1182. Barsky, A.J., Barnett, M.C., & Cleary, P.D. (1994). Hypochondriasis and panic disorder: Boundary and overlap. Archives of General Psychiatry, 51, 918-925. Basoglu, M., Marks, J.M., Kilic;, C., Brewin, C.R., & Swinson, RP. (1994a). Alprazolam and exposure for panic disorder with agoraphobia: Attribution of improvement to medication predicts subsequent relapse. British Journal of Psychiatry, 164, 652-659. Basoglu, M., Marks, J.M., Kilic;, C., Swinson, RP., Noshirvani, H., Kuch, K., & O'Sullivan, G. (1994b). Relationship of panic, anticipatory anxiety, agoraphobia and global improvement in panic disorder with agoraphobia treated with alprazolam and exposure. British Journal of Psychiatry, 164, 647-652. Bass, C.M. (1990). Functional cardiorespiratory syndromes. In C.M. Bass (Ed.), Somatization: Psychical symptoms and psychological illness (pp. 171-206). Oxford: Blackwell. Bass, C.M., Gardner, W.N., & Jackson, G. (1994). Psychiatric and respiratory aspects of functional cardiovascular syndromes. In B.H. Timmons & R. Ley (Eds.), Behavioral and psychological approaches to breathing disorders (pp. 125-138). New York: Plenum. Battaglia, M., Bertella, S., Politi, E., Bernardeschi, L., Perna, G., Gabriele, A., & Bellodi, L. (1995). Age at onset of panic disorder: Influence of familial liability to the disease and of childhood separation anxiety disorder. American Journal of Psychiatry, 152, 1362-1364. Baucom, D.H., & Epstein, N. (1990). Cognitive behavioral marital therapy. New York: Brunner/ Maze 1. Beaconsfield, P., Ginsburg, J., & Rainsbury, R. (1972). Marijuana smoking: Cardiovascular effects in man and possible mechanisms. New England Journal of Medicine, 287, 209-212. Beck, A.T. (1976). Cognitive therapy and the emotional disorders. New York: International Universities Press. Beck, A.T. (1988). Cognitive approaches to panic disorder: Theory and therapy. In S. Rachman & J.D. Maser (Eds.), Panic: Psychological perspectives. (pp. 91-109). Hillsdale, NJ: Erlbaum. Beck, A.T., & Emery, G. (1985). Anxiety disorders and phobias: A cognitive perspective. New York: Basic Books. Beck, A.T., Freeman, A., & Associates. (1990). Cognitive therapy of personality disorders. New York: Guilford. Beck, A.T., Rush, A.J., Shaw, B.F., & Emery, G. (1979). Cognitive therapy of depression. New York: Guilford. Beck, A.T., Sokol, L., Clark, D.A., Berchick, R., & Wright, F. (1992). A crossover study of focused cognitive therapy for panic disorder. American Journal of Psychiatry, 149, 778-783. Beck, A.T., & Steer, RA. (1993). Manual for the Beck Anxiety Inventory. San Antonio, TX: Psychological Corporation.

448

REFERENCES

Beck, A.T., Steer, RA., Sanderson, W.C., & Skeie, T.M. (1991). Panic disorder and suicidal behavior: Discrepant findings in psychiatric outpatients. American Journal of Psychiatry, 148, 1195-1199. Beck, J.G., Stanley, M.A., Baldwin, L.E., Deagle, E.A., & Averill, P.M. (1994). Comparison of cognitive therapy and relaxation training for panic disorder. Journal of Consulting and Clinical Psychology, 62, 818-826. Beck, J.G., & Zebb, B.J. (1994). Behavioral assessment and treatment of panic disorder: Current status, future directions. Behavior Therapy, 25, 581-611. Beckfield, D.F. (1998). Master your panic and take back your life! (2nd ed.). San Luis Obispo, CA: Impact Publishers. Bennun, I. (1985). Behavioral marital therapy: An outcome evaluation of conjoint, group, and one spouse treatment. Scandinavian Journal of Behavior Therapy, 14, 157-168. Bennun, I. (1997). Relationship interventions with one partner. In W.K. Halford & H.J. Markman (Eds.), Clinical handbook of marriage and couples interventions (pp. 451-470). Chichester: Wiley. Bennun, I., & Schindler, L. (1988). Therapist and patient factors in the behavioural treatment of phobic patients. British Journal of Clinical Psychology, 27, 145-150. Berg, I., Marks, I., McGuire, R., & Lipsedge, M. (1972). School phobia and agoraphobia. Psychological Medicine, 4, 428-434. Bernstein, D.A., & Borkovec, T.D. (1973). Progressive relaxation training. Champaign, IL: Research Press. Bibb, J., & Chambless, D.L. (1986). Alcohol use and abuse among diagnosed agoraphobics. Behaviour Research and Therapy, 24, 49-58. Biederman, J., Rosenbaum, J.F., Hirshfeld, D.R., Faraone, S.J., Bolduc, E.A., Gersten, M., Meminger, S.R., Kagan, J., Snidman, N., & Reznick, S. (1990). Psychiatric correlates of behavioral inhibition in young children of parents with and without psychiatric disorders. Archives of General Psychiatry, 47, 21-26. Black, D.W., Monahan, P., Wesner, R., Gabel, J., & Bowers, W. (1996). The effect of fluvoxamine, cognitive therapy, and placebo on abnormal personality traits in 44 patients with panic disorder. Journal of Personality Disorders, 10, 185-194. Black, D.W., Wesner, R., Bowers, W., & Gabel, J. (1993). A comparison of fluvoxamine, cognitive therapy, and placebo in the treatment of panic disorder. Archives of General Psychiatry, 50, 44-50. Black, D.W., Wesner, R., Gabel, J.,Bowers, W., & Monahan, P. (1994). Predictors of short-term treatment response in 66 patients with panic disorder. Journal of Affective Disorders, 30, 233-241. Bland, K., & Hallam, RS. (1981). Relationship between response to graded exposure and marital satisfaction in agoraphobics. Behaviour Research and Therapy, 19, 335-338. Bonn, J.A., Readhead, C.P.A., & Timmons, B.H. (1984). Enhanced adaptive behavioural response in agoraphobic patients pretreated with breathing retraining. Lancet, ii, 665-669. Boszormeny-Nagi, I. (1987). Foundations of contextual therapy: Collected papers of Ivan Boszormeny-Nagi. New York: Brunner/Mazel. Bouchard, S., Gauthier, J.,Laberge, B., French, D., Pelletier, M.-H., & Godbout, C. (1996). Exposure versus cognitive restructuring in the treatment of panic disorder with agoraphobia. Behaviour Research and Therapy, 34, 213-224. Bouchard, S., Pelletier, M.-H., Gauthier, J., Cote, G., & Laberge, B. (1997). The assessment of panic using self-report: A comprehensive survey of validated instruments. Journal of Anxiety Disorders, 11, 89-111.

REFERENCES449

Bowen, R., South, M., Fischer, D., & Looman, T. (1994). Depression, mastery and number of group sessions attended predict outcome of patients with panic and agoraphobia in a behavioural/medication program. Canadian Journal of Psychiatry, 39, 283-288. Bowlby, J. (1990). Charles Darwin: A new life. New York: Norton. Boyd, J.H. (1986). Use of mental health services for the treatment of panic disorder. American Journal of Psychiatry, 143, 1569-1574. Boyd, J.H., & Crump, T. (1991). Westphal's agoraphobia. Journal of Anxiety Disorders, 5, 77-86. Boyer, W. (1995). Serotonin uptake inhibitors are superior to imipramine and alprazolam in alleviating panic attacks: A meta-analysis. International Clinical Psychopharmacology, 10, 45-49. Brandt, T. (1996). Physiological and psychophysiological vertigo. In R.W. Baloh & G.M. Halmagyi (Eds.), Disorders of the vestibular system (pp. 496-508). Oxford: Oxford University Press. Brehm, J.W. (1962). A theory of psychological reactance. New York: Academic. Breier, A., Charney, D.S., & Heninger, C.R. (1986). Agoraphobia with panic attacks: Development, diagnostic stability, and course of illness. Archives of General Psychiatry, 43, 1029-1036. Breslau, N., & Davis, G.C. (1993). Migraine, physical health and psychiatric disorder: A prospective epidemiologic study in young adults. Journal of Psychiatric Research, 27, 211-221. Brown, G.K., Beck, A.T., Newman, C.F., Beck, J.S., & Tran, G.Q. (1997). A comparison of focused and standard cognitive therapy for panic disorder. Journal of Anxiety Disorders, 11, 329-345. Brown, G.W., Harris, T.O., & Eales, M.J. (1993). Aetiology of anxiety and depressive disorders in an inner-city population. 2. Comorbidity and adversity. Psychological Medicine, 23, 155-165. Brown, T.A., Antony, M.M., & Barlow, D.H. (1995). Diagnostic comorbidity in panic disorder: Effect on treatment outcome and course of comorbid diagnoses following treatment. Journal of Consulting and Clinical Psychology, 63, 408-418. Brown, T.A., & Barlow, D.H. (1995). Long-term outcome in cognitive-behavioral treatment of panic disorder: Clinical predictors and alternative strategies for assessment. Journal of Consulting and Clinical Psychology, 63, 754-765. Bruce, T.J., Spiegel, D.A., Gregg, S.F., & Nuzzarello, A. (1995). Predictors of alprazolam discontinuation with and without cognitive behavior therapy in panic disorder. American Journal of Psychiatry, 152, 1156-1160. Bruce, T.J., Spiegel, D.A., & Hegel, M.T. (1999). Cognitive-behavioral therapy helps prevent relapse and recurrence of panic disorder following alprazolam discontinuation: A long-term follow-up of the Peoria and Dartmouth studies. Journal of Consulting and Clinical Psychology, 67, 151-156. Bruch, M. (1998). The UCL case formulation model: Clinical applications and procedures. In M. Bruch & F.W. Bond (Eds.), Beyond diagnosis: Case formulation approaches in CBT (pp. 19-41). Chichester: Wiley. Bruch, M., & Bond, F.W. (1998). Beyond diagnosis: Caseformulation approaches in CBT. Chichester: Wiley. Bruss, G.S., Gruenberg, A.M., Goldstein, R.D., & Barber, J.P. (1994). Hamilton Anxiety Rating Scale Interview Guide: Joint interview and test-retest methods for interrater reliability. Psychiatry Research, 53, 191-202. Bufka, L.F., & Hofmann, S.G. (1999). Modifying CBT to treat panic disorder in patients with schizophrenia. Cognitive and Behavioral Practice, 6, 10-15.

450

REFERENCES

Buller, R., Maier, W., Goldenberg, I.M., Lavori, P.W., & Benkert, 0. (1991). Chronology of panic and avoidance, age of onset in panic disorder, and prediction of treatment response. European Archives of Psychiatry, and Clinical Neuroscience, 240, 163-168. Burgess, LS., Jones, L.M., Robertson, S.A., Radcliffe, W.N., & Emerson, E. (1981). The degree of control exerted by phobic and non-phobic verbal stimuli over the recognition behaviour of phobic and non-phobic subjects. Behaviour Research and Therapy, 19, 233-243. Burns, D.D. (1981). Feeling good: The new mood therapy. New York: Signet. Burns, L.E., Thorpe, G.L., & Cavallaro, L.A. (1986). Agoraphobia eight years after behavioral treatment: A follow-up study with interview, self-report, and behavioral data. Behavior Therapy, 17, 580-591. Busch, F.N., Milrod, B., Cooper, A., & Shapiro, T. (1996). Psychodynamic approaches to panic disorder. Journal of Psychotherapy Practice and Research, 5, 73-83. Cameron, O.G., Kuttesch, D., McPhee, K., & Curtis, G.C. (1988). Menstrual fluctuation in the symptoms of panic anxiety. Journal of Affective Disorders, 15, 169-174. Canadian Pharmaceutical Association (1990). Compendium of pharmaceuticals and specialties (25th ed.). Ottawa: Author. Carey, K.B. (1994). Use of the structured clinical interview for DSM-III-R Personality Questionnaire in the presence of severe Axis I disorders: A cautionary note. Journal of Nervous and Mental Disease, 182, 669-671. Carr, R.E. (1999). Panic disorder and asthma. Journal of Asthma, 36, 143-152. Carr, R.E., Lehrer, P.M., Rausch, L.L., & Hochron, S.M. (1994). Anxiety sensitivity and panic attacks in an asthmatic population. Behaviour Research and Therapy, 32, 411-418. Carter, M.M., Turovsky, J., Sbrocco, T., Meadows, E.A., & Barlow, D.H. (1995). Patient dropout from a couples' group treatment for panic disorder with agoraphobia. Professional Psychology: Research and Practice, 26, 626-628. Casper, R.F., Graves, G.R., & Reid, R.L. (1987). Objective measurement of hot flushes associated with the premenstrual syndrome. Fertility and Sterility, 47, 341-344. Cattell, R.B. (1978). The scientific use of factor analysis in the behavioral and life sciences. New York: Plenum. Chadwick, P.D.J., & Lowe, C.F. (1994). A cognitive approach to measuring and modifying delusions. Behaviour Research and Therapy, 32, 355-367. Chambless, D.L. (1985). The relationship of severity of agoraphobia to associated psychopathology. Behaviour Research and Therapy, 23, 305-310. Chambless, D.L. (1988). Body sensations questionnaire. In M. Hersen & A.S. Bellack (Eds.), Dictionary of behavioral assessment techniques (pp. 85-86). New York: Pergamon. Chambless, D.L. (1990). Spacing of exposure sessions in treatment of agoraphobia and simple phobia. Behavior Therapy, 21, 217-229. Chambless, D.L., Caputo, G., Bright, P., & Gallagher, R. (1984). Assessment of fear of fear in agoraphobics: The Body Sensations Questionnaire and the Agoraphobic Cognitions Questionnaire. Journal of Consulting and Clinical Psychology, 52, 1090-1097. Chambless, D.L., Caputo, G., Jasin, S., Gracely, E.J., & Williams, C. (1985). The Mobility Inventory for agoraphobia. Behaviour Research and Therapy, 23, 35-44.

REFERENCES451 Chambless, D.L., Foa, E.B., Groves, G.A., & Goldstein, A.J. (1979). Brevital in flooding with agoraphobics. Behaviour Research and Therapy, 17, 243-251. Chambless, D.L., & Gillis, M.M. (1993). Cognitive therapy of anxiety disorders. Journal of Consulting and Clinical Psychology, 61, 248-260. Chambless, D.L., & Gillis, M.M. (1994). A review of psychosocial treatments for panic disorder. In B.E. Wolfe & J.D. Maser (Eds.), Treatment of panic disorder: A consensus development conference (pp. 149-173). Washington, DC: American Psychiatric Press. Chambless, D.L., & Goldstein, A.J. (1988). Fear of fear: A reply to Reiss. Behavior Therapy, 19, 85-88. Chambless, D.L., & Gracely, E.J. (1988). Predictors of outcome following in vivo exposure treatment of agoraphobia. In I. Hand & H.-U. Wittchen (Eds.), Panic and phobias II: Treatment and variables affecting outcome (pp. 209-220). New York: Springer-Verlag. Chambless, D.L., & Gracely, E.J. (1989). Fear of fear and the anxiety disorders. Cognitive Therapy and Research, 13, 9-20. Chambless, D.L., & Mason, J. (1986). Sex, sex-role stereotyping and agoraphobia. Behaviour Research and Therapy, 24, 231-235. Chambless, D.L., Renneberg, B., Goldstein, A., & Gracely, E.J. (1992). MCMIdiagnosed personality disorders among agoraphobic outpatients: Prevalence and relationship to severity and treatment outcome. Journal of Anxiety Disorders, 6, 193-211. Chaplin, E.W., & Levine, B.A. (1981). The effects of total exposure duration and interrupted versus continued exposure in flooding therapy. Behavior Therapy, 12, 360-368. Chernen, L., & Friedman, S. (1993). Treating the personality disordered agoraphobic patient with individual and marital therapy. Journal of Anxiety Disorders, 7, 163-177. Chobanian, A.V. (1982). Orthostatic hypotension. In H.R. Brunner & H. Gavras (Eds.), Clinical hypertension and hypotension (pp. 435-454). New York: Marcel Dekker. Chorpita, B.F., Albano, A.M., & Barlow, D.H. (1996). Child Anxiety Sensitivity Index: Considerations for children with anxiety disorders. Journal of Clinical Child Psychology, 25, 77-82. Clair, A.L., Oei, T.P.S., & Evans, L. (1992). Personality and treatment response in agoraphobia with panic attacks. Comprehensive Psychiatry, 5, 310-318. Clark, D.M. (1986). A cognitive approach to panic. Behaviour Research and Therapy, 24, 461-470. Clark, D.M. (1988). A cognitive model of panic attacks. In S. Rachman & J.D. Maser (Eds.). Panic: Psychological perspectives. (pp. 71-89). Hillsdale, NJ: Erlbaum. Clark, D.M. (1989). Anxiety states: Panic and generalized anxiety. In K. Hawton, P.M. Salkovskis, J. Kirk, & D.M. Clark (Eds.), Cognitive behaviour therapy for psychiatric problems: A practical guide (pp. 52-96). Oxford: Oxford University Press. Clark, D.M. (1993). Cognitive mediation of panic attacks induced by biological challenge tests. Advances in Behaviour Research and Therapy, 15, 75-84. Clark, D.M. (1996). Panic disorder: From theory to therapy. In P.M. Salkovskis (Ed.), Frontiers of cognitive therapy (pp. 318-344). New York: Guilford. Clark, D.M. (.1999). Anxiety disorders: Why they persist and how to treat them. Behaviour Research and Therapy, 37, S5-S27. Clark, D.M., & Ehlers, A. (1993). An overview of the cognitive theory and treatment of panic disorder. Applied & Preventive Psychology, 2, 131-139.

452

REFERENCES

Clark, D.M., & Salkovskis, P.M. (1987). Cognitive therapy for panic attacks: Therapist's manual. Department of Psychiatry, University of Oxford. Clark, D.M., Salkovskis, P.M., & Anastasiades, P. (1990, November). Cognitive mediation of lactate induced panic. Paper presented at the 24th meeting of the Association for Advancement of Behavior Therapy, San Francisco, CA. Clark, D.M., Salkovskis, P.M., & Chalkey, A.J. (1985). Respiratory control as a treatment for panic attacks. Journal of Behavior Therapy and Experimental Psychiatry, 16, 23-30. Clark, D.M., Salkovskis, P.M., Gelder, M., Koehler, C., Martin, M., Anastasiades, P., Hackmann, A., Middleton, H., & Jeavons, A. (1988). Tests of a cognitive theory of panic. In I. Hand & H.-U. Wittchen (Eds.), Panic and phobias 2: Treatments and variables affecting course and outcome (pp. 149-158). New York: Springer-Verlag. Clark, D.M., Salkovskis, P.M., Hackmann, A., Middleton, H., Anastasiades, P., & Gelder, M. (1994). A comparison of cognitive therapy, applied relaxation and imipramine in the treatment of panic disorder. British Journal of Psychiatry, 164, 759-769. Clark, D.M., Salkovskis, P.M., Hackmann, A., Wells, A., & Gelder, M. (1997a). Longterm outcome of brief cognitive therapy for panic disorder. Paper presented at the XXVII Congress of the European Association for Behavioural & Cognitive Therapies, Venice, Italy. Clark, D.M., Salkovskis, P.M., Hackmann, A., Wells, A., Ludgate, J., & Gelder, M. (1997b ). A controlled trial of brief cognitive therapy for panic disorder. Paper presented at the XXVII Congress of the European Association for Behavioural & Cognitive Therapies, Venice, Italy. Clark, D.M., Salkovskis, P.M., Hackmann, A., Wells, A., Ludgate, J., & Gelder, M. (1999). Brief cognitive therapy for panic disorder: A randomized controlled trial. Journal of Consulting and Clinical Psychology, 67, 583-589. Clark, D.M., Salkovskis, P.M., Ost, L.-G., Breitholtz, E., Koehler, K.A., Westling, B.E., Jeavons, A., & Gelder, M. (1997c). Misinterpretation of body sensations in panic disorder. Journal of Consulting and Clinical Psychology, 65, 203-213. Clarke, J.C., & Jackson, J.A. (1983). Hypnosis and behavior therapy. New York: Springer. Clifft, M.A. (1986). Writing about psychiatric patients: Guidelines for disguising case material. Bulletin of the Menninger Clinic, 50, 511-524. Clum, G.A. (1990). Coping with panic. Pacific Grove, CA: Brooks/Cole. Clum, G.A., Clum, G., & Surls, R. (1993). A meta-analysis of treatments for panic disorder. Journal of Consulting and Clinical Psychology, 61, 317-326. Clum, G.A., & Knowles, S.L. (1991). Why do some people with panic disorders become avoidant: A review. Clinical Psychology Review, 11, 295-313. Cobb, J.P., Mathews, A.M., Childs-Clarke, A., & Blowers, C.M. (1984). The spouse as co-therapist in the treatment of agoraphobia. British Journal of Psychiatry, 144, 282-287. Cohen, A.S., Barlow, D.H., & Blanchard, E.B. (1985). Psychophysiology of relaxation-associated panic attacks. Journal of Abnormal Psychology, 94, 96-101. Cohen, S.D., Monteiro, W., & Marks, I.M. (1984). Two-year follow-up of agoraphobics after exposure and imipramine. British Journal of Psychiatry, 144, 276-281. Cohn, T.E., & Lasley, D.J. (1990). Wallpaper illusion: Cause of disorientation and falls on escalators. Perception, 19, 573-580. Collins, R.E.C. (1969). A new escalator injury. Lancet, i, 1268.

REFERENCES453 Colp, R. (1977). To be an invalid: The illness of Charles Darwin. Chicago, IL: University of Chicago Press. Conway, A.V. (1994). Breathing and feeling. In B.H. Timmons & R. Ley (Eds.), Behavioral and psychological approaches to breathing disorders (pp. 243-251). New York: Plenum. Cook, B.L., Noyes, R., Garvey, M.J., Beach, V., Sobotka, J., & Chaudhry, D. (1990). Anxiety and the menstrual cycle in panic disorder. Journal of Affective Disorders, 19, 221-226. Cooper, M.L., Russel, M., Skinner, J.B., Prone, M.R., & Mudar, P. (1992). Stress and alcohol use: Moderating effects of gender, coping, and alcohol expectancies. Journal of Abnormal Psychology, 101, 139-152. Coryell, W., Noyes, R., & Clancy, J. (1982). Excess mortality in panic disorder. Archives of General Psychiatry, 39, 701-703. Coryell, W., Noyes, R., & House, D. (1986). Mortality among outpatients with anxiety disorders. American Journal of Psychiatry, 143, 508-510. Costa, P.T., & McCrae, R.R. (1992). Revised NEO Personality Inventory (NEO-PI-R) and NEO Five-Factor Inventory (NEO-FFI) professional manual. Odessa, FL: Psychological Assessment Resources. Costello, C.G. (1992). Problems in recent tests of two cognitive theories of panic. Behaviour Research and Therapy, 30, 1-5. Cote, G., Bouchard, S., Gauthier, J.C., & Laberge, B. (1994a, November). Reduced therapist contact cognitive behavioral treatment for panic disorder: Effectiveness at three-year follow-up. Paper presented at the 28th annual meeting of the Association for Advancement of Behavior Therapy, San Diego, CA. Cote, G., Gauthier, J.C., Laberge, B., Cormier, H.J., & Plamondon, J. (19946). Reduced therapist contact in the cognitive behavioral treatment of panic disorder. Behavior Therapy, 25, 123-145. Cottraux, J., Note, I.-D., Cungi, C., Legeron, P., Heim, F., Chneiweiss, L., Bernard, G., & Bouvard, M. (1995). A controlled study of cognitive behaviour therapy with buspirone or placebo in panic disorder with agoraphobia. British Journal of Psychiatry, 167, 635-641. Cox, B.J. (1996). The nature and assessment of catastrophic thoughts in panic disorder. Behaviour Research and Therapy, 34, 363-374. Cox, B.J., Borger, S.C., & Enns, M.W. (1999). Anxiety sensitivity and emotional disorders: Psychometric studies and their theoretical implications. In S. Taylor (Ed.), Anxiety sensitivity: Theory, research, and treatment of the fear of anxiety (pp. 115-148). Mahwah, NJ: Erlbaum. Cox, B.J., Cohen, E., Direnfeld, D.M., & Swinson, RP. (1996). Does the Beck Anxiety Inventory measure anything beyond panic attack symptoms? Behaviour Research and Therapy, 34, 949-954. Cox, B.J., Endler, N.S., Lee, P.S., & Swinson, RP. (1992a). A meta-analysis of treatments for panic disorder with agoraphobia: Imipramine, alprazolam, and in vivo exposure. Journal of Behavior Therapy and Experimental Psychiatry, 23, 175-182. Cox, B.J., Norton, C.R., & Swinson, RP. (19926). The Panic Attack Questionnaire, Revised. Toronto: Clarke Institute of Psychiatry. Cox, B.J., Norton, C.R., Swinson, RP., & Endler, N.S. (1990). Substance abuse and panic-related anxiety: A critical review. Behaviour Research and Therapy, 28, 385-393. Cox, B.J., Swinson, RP., Kuch, K., & Reichman, J.T. (1993). Dimensions of agoraphobia assessed by the Mobility Inventory. Behaviour Research and Therapy, 31, 427-431.

454

REFERENCES

Cox, B.J., Swinson, RP., & Shaw, B.F. (1991). Value of the Fear Questionnaire in differentiating agoraphobia and social phobia. British Journal of Psychiatry, 159, 842-845. Cox, B.J., & Taylor, S. (1999). Anxiety disorders: Panic and phobias. In T. Millon, P. Blaney, & R. Davis (Eds.), Oxford textbook of psychopathology (pp. 81-113). Oxford: Oxford University Press. Cox, D.J., Ballenger, J.C., Laraia, M., Hobbs, W.R., Peterson, G.A., & Hucek, A. (1988). Different rates of improvement of different symptoms in combined pharmacological and behavioral treatment of agoraphobia. Journal of Behavior Therapy and Experimental Psychiatry, 19, 119-126. Coyne, C. (1983). Muscle tension and its relation to symptoms in the premenstruum. Research in Nursing and Health, 6, 199-205. Craske, M.G. (1991). Phobic fear and panic attacks: The same emotional states triggered by different cues? Clinical Psychology Review, 11, 599-620. Craske, M.G., & Barlow, D.H. (1989). Nocturnal panic. Journal of Nervous and Mental Disease, 177, 160-167. Craske, M.G., & Barlow, D.H. (1993). Panic disorder and agoraphobia. In D.H. Barlow (Ed.), Clinical handbook of psychological disorders, 2nd ed. (pp. 1-47). New York: Guilford. Craske, M.G., & Barlow, D.H. (1994). Agoraphobia supplement to the MAP-II program: Client workbook. San Antonio, TX: Psychological Corporation. Craske, M.G., Brown, T.A., & Barlow, D.H. (1991). Behavioral treatment of panic disorder: A two-year follow-up. Behavior Therapy, 22, 289-304. Craske, M.G., Burton, T., & Barlow, D.H. (1989). Relationships among measures of communication, marital satisfaction and exposure during couples treatment of agoraphobia. Behaviour Research and Therapy, 27, 131-140. Craske, M.G., & Kreuger, M. (1990). The prevalence of nocturnal panic in a college population. Journal of Anxiety Disorders, 4, 125-139. Craske, M.G., Maidenberg, E., & Bystritsky, A. (1995). Brief cognitive-behavioral versus nondirective therapy for panic disorder. Journal of Behavior Therapy and Experimental Psychiatry, 26, 113-120. Craske, M.G., Meadows, E.A., & Barlow, D.H. (1994). Mastery of your anxiety and panic II and agoraphobia supplement: Therapist guide. San Antonio, TX: Psychological Corporation. Craske, M.G., Rachman, S.J., & Tallman, K. (1986). Mobility, cognitions, and panic. Journal of Psychopathology and Behavioral Assessment, 8, 199-210. Craske, M.G., Rowe, M. (1997). Nocturnal panic. Clinical Psychology: Science and Practice, 4, 153-174. Craske, M.G., Rowe, M., Lewin, M., & Noriega-Dimitri, R. (1997). Interoceptive exposure versus breathing retraining within cognitive-behavioural therapy for panic disorder with agoraphobia. British Journal of Clinical Psychology, 36, 85-99. Cross National Collaborative Panic Study. (1992). Drug treatment of panic disorder: Comparative efficacy of alprazolam, imipramine and placebo. British Journal of Psychiatry, 160, 191-202. Cutler, J. (1994). Panic attacks and schizophrenia: Assessment and treatment. Psychiatric Annals, 24, 473-476. Dager, S.R., Comess, K.A., Saal, A.K., & Dunner, D.L. (1986). Mitra! valve prolapse in a psychiatric setting: Diagnostic assessment, research and clinical implications. Integrative Psychiatry, 4, 211-213. Daiuto, A.A., Baucom, D.H., Epstein, N., & Dutton, S.S. (1998). The application of behavioral couples therapy to the assessment and treatment of agoraphobia: Implications of empirical research. Clinical Psychology Review, 18, 663-687.

REFERENCES455 Daley, D.C., & Marlatt, G.A. (1997). Managing your drug or alcohol problem. New York: Graywind. • Damas-Mora, J., Davies, L., Taylor, W., & Jenner, F.A. (1980). Menstrual respiratory changes and symptoms. British Journal of Psychiatry, 136, 492-497. Dattilio, F.M. (1987). The use of paradoxical intention in the treatment of panic attacks. Journal of Counseling and Development, 66, 102-103. Dattilio, F.M. (1994). Paradoxical intention as a proposed alternative in the treatment of panic disorder. Journal of Cognitive Psychotherapy, 8, 33-40. Davanloo, H. (1978). Basic principles and techniques in short-term dynamic psychotherapy. New York: Spectrum. de Beurs, E. (1993). The assessment and treatment of panic disorder and agoraphobia. Amsterdam: Thesis Publishers. de Beurs, E., Lange, A., Koele, P., & van Dyck, R. (1993). Frequency of panic as an outcome measure in agoraphobia research: Latent effects of exposure on panic. Journal of Anxiety Disorders, 7, 307-319. de Beurs, E., Lange, A., & van Dyck, R. (1992). Self-monitoring of panic attacks and retrospective estimates of panic: Discordant findings. Behaviour Research and Therapy, 30, 411-413. de Beurs, E., Lange, A., van Dyck, R., Blonk, R.W.B., & Koele, P. (1991). Behavioral assessment of avoidance in agoraphobia. Journal of Psychopathology and Behavioral Assessment, 13, 285-300. de Beurs, E., Lange, A., van Dyck, R., & Koele, P. (1995a). Respiratory training prior to exposure in vivo in the treatment of panic disorder with agoraphobia: Efficacy and predictors of outcome. Australian and New Zealand Journal of Psychiatry, 29, 104-113. de Beurs, E., van Balkom, A.J.L.M., Lange, A., Koele, P., & van Dyck, R. (1995b). Treatment of panic disorder with agoraphobia: Comparison of fluvoxamine, placebo, and psychological panic management combined with exposure and of exposure in vivo alone. American Journal of Psychiatry, 152, 683-691. DeGuire, S., Gevirtz, R., Kawahara, Y., & Maguire, W. (1992). Hyperventilation syndrome and the assessment of treatment for functional cardiac symptoms. American Journal of Cardiology, 70, 673-677. de Leon, A.C., & Cheng, T.O. (1986). History and physical examination. In J.W. Hurst (Ed.), The heart: Arteries and veins (6th ed.) (pp. 22-46). New York: McGraw-Hill. Delmonte, M.M. (1995). The use of hypnotic regression with panic disorder: A case report. Australian Journal of Clinical Hypnotherapy and Hypnosis, 16, 69-73. Deltito, J.A., Argyle, N., Buller, R., Nutzinger, D., Ottosson, J.-O., Brandon, S., Mellergard, M., & Shera, D. (1991). The sequence of improvement of the symptoms encountered in patients with panic disorder. Comprehensive Psychiatry, 32, 120-129. Der, D.-F., & Lewington, P. (1990). Rational self-directed hypnotherapy: A treatment for panic attacks. American Journal of Clinical Hypnosis, 32, 160-167. de Ruiter, C., Rijken, H., Garssen, B., & Kraaimaat, F. (1989). Breathing retraining, exposure and a combination of both, in the treatment of panic disorder with agoraphobia. Behaviour Research and Therapy, 27, 647-655. Devereux, R.B., Perloff, J.K., Reichek, N., & Josephson, M.E. (1976). Mitra! valve prolapse. Circulation, 54, 3-14. Dewey, D., & Hunsley, J. (1990). The effects of marital adjustment and spouse involvement on the behavioral treatment of agoraphobia: A meta-analytic review. Anxiety Research, 2, 69-83.

456

REFERENCES

Diaferia, G., Sciuto, G., Perna, G., Bernardeschi, L., Battaglia, M., Rusmini, S., & Bellodi, L. (1993). DSM-III-R personality disorders and panic disorder. Journal of Anxiety Disorders, 7, 153-161. DiNardo, P., Brown, T.A., & Barlow, D.H. (1994). Anxiety disorders interview schedule for DSM-IV. New York: Graywind. DiNardo, P., Moras, K., Barlow, D.H., Rapee, R.M., & Brown, T.A. (1993). Reliability of DSM-III-R anxiety disorder categories: Using the Anxiety Disorders Interview Schedule-Revised (ADIS-R). Archives of General Psychiatry, 50, 251-256. Dixon, J.C. (1960). Depersonalization phenomenon in a sample population of college students. British Journal of Psychiatry, 109, 371-375. Docherty, J.P., Fiester, S.J., & Shea, T. (1986). Syndrome diagnosis and personality disorder. In A.J. Frances & R.E. Hales (Eds.), Psychiatry update: American Psychiatric Association annual review, vol. 5 (pp. 315-355). Washington, DC: American Psychiatric Press. Dohrenwend, B.S., Krasnoff, L., Askenasy, A.R., & Dohrenwend, B.P. (1978). Exemplification of a method for scaling life events: The PERI Life Events Scale. Journal of Health and Social Behavior, 19, 205-229. Donnell, C.D., & McNally, R.J. (1990). Anxiety sensitivity and panic attacks in a nonclinical population. Behaviour Research and Therapy, 28, 83-85. Dreessen, L., Arntz, A., Luttels, C., & Sallaerts, S. (1994). Personality disorders do not influence the results of cognitive behavior therapies for anxiety disorders. Comprehensive Psychiatry, 35, 265-274. Drossman, D.A., Sandler, R.S., McKee, D., & Lovitz, H.A. (1982). Bowel patterns among subjects not seeking health care: Use of a questionnaire to identify a population with bowel dysfunction. Gastroenterology, 83, 529-534. Duane, T.D. (1973). Valsalva hemorrhagic retinopathy. American Journal of Ophthalmology, 75, 637--642. Dummit, E.S., & Klein, R.G. (1994). Panic disorder. In T.H. Ollendick, N.J. King, & W. Yule (Eds.), International handbook of phobic and anxiety disorders in children and adolescents (pp. 241-266). New York: Plenum. Dupont, R.L., & Pecknold, J. (1985, May). Alprazolam withdrawal in panic disorder patients. Paper presented at the 138th annual meeting of the American Psychiatric Association, Dallas, TX. Eaton, W.W., Dryman, A., & Weissman, M.M. (1991). Panic and phobia. In L.N. Robins & D.A. Reiger (Eds.), Psychiatric disorders in America: The epidemiologic catchment area study (pp. 155-179). New York: Free Press. Echeburua, E., De Corral, P., Bajos, E.G., & Borda, M. (1993). Interactions between self-exposure and alprazolam in the treatment of agoraphobia without current panic: An exploratory study. Behavioural and Cognitive Psychotherapy, 21, 219-238. Edelman, R.E., & Chambless, D.L. (1993). Compliance during sessions and homework in exposure-based treatment of agoraphobia. Behaviour Research and Therapy, 31, 767-773. Edlund, M.J., & Swann, A.C. (1987). The economic and social costs of panic disorders. Hospital and Community Psychiatry, 38, 1277-1280. Ehlers, A. (1993a). Interoception and panic disorder. Advances in Behaviour Research and Therapy, 15, 3-21. Ehlers, A. (i993b). Somatic symptoms and panic attacks: A retrospective study of learning experiences. Behaviour Research and Therapy, 31, 269-278.

REFERENCES457 Ehlers, A. (1995). A 1-year prospective study of panic attacks: Clinical course and factors associated with maintenance. Journal of Abnormal PsychologiJ, 104, 164-172. Ehlers, A., & Breuer, P. (1992). Increased cardiac awareness in panic disorder. Journal of Abnormal Psychology, 101, 371-382. Ehlers, A., & Breuer, P. (1996). How good are patients with panic disorder at perceiving their heartbeats? Biological Psychology, 42, 165-182. Ehlers, A., Margraf, J., Davies, S., & Roth, W.T. (1988). Selective processing of threat cues in subjects with panic attacks. Cognition and Emotion, 2, 201-219. Eifert, G.H. (1992). Cardiophobia: A paradigmatic behavioural model of heartfocused anxiety and non-angina! chest pain. Behaviour Research and Therapy, 30, 329-345. Eldridge, G.D., Walker, J.R., & Holborn, S.W. (1993). Cognitive-behavioral treatment for panic disorder with gastrointestinal symptoms: A case study. Journal of Behavior Therapy and Experimental Psychiatry, 24, 367-371. Ellis, A. (1962). Reason and emotion in psychotherapy. New York: Lyle-Stuart. Ellis, A. (1979). Discomfort anxiety: A new cognitive-behavioral construct, Part I. Rational Living, 14, 3-8. Ellis, A. (1980). Discomfort anxiety: A new cognitive-behavioral construct, Part IL Rational Living, 15, 25-30. Elsenga, S., & Emmelkamp, P.M.G. (1990). Behavioral treatment of an incestrelated trauma in an agoraphobic client. Journal of Anxiety Disorders, 4, 151162. Elsesser, K., Sartory, G., & Maurer, J. (1996). The efficacy of complaints management training in facilitating benzodiazepine withdrawal. Behaviour Research and Therapy, 34, 149-156. Emmelkamp, P.M.G. (1980). Agoraphobics' interpersonal problems: Their role in the effects of exposure in vivo therapy. Archives of General Psychiatry, 37, 1303-1306. Emmelkamp, P.M.G., & Bouman, T.K. (1991). Psychological approaches to the difficult patient. In J.R. Walker, G.R. Norton, & C.A. Ross (Eds.), Panic disorder and agoraphobia:A comprehensive guide for the practitioner (pp. 398-429). Pacific Grove, CA: Brooks/Cole. Emmelkamp, P.M.G., Brilman, E., Kuiper, H., & Mersch, P.P. (1986). The treatment of agoraphobia: A comparison of self-instructional training, rational emotive therapy, and exposure in vivo. Behavior Modification, 10, 37-53. Emmelkamp, P.M.G., & Emmelkamp-Benner, A. (1975). Effects of historically portrayed modeling and group treatment on self-observation: A comparison with agoraphobics. Behaviour Research and Therapy, 13, 135-139. Emmelkamp, P.M.G., & Kuipers, A. (1979). Agoraphobia: A follow-up study four years after treatment. British Journal of Psychiatry, 134, 352-355. Emmelkamp, P.M.G., Kuipers, A.CM., & Eggeraat, J.B. (1978). Cognitive modification versus prolonged exposure in vivo: A comparison with agoraphobics as subjects. Behaviour Research and Therapy, 16, 33-41. Emmelkamp, P.M.G., & Mersch, P.P. (1982). Cognition and exposure in vivo in the treatment of agoraphobia: Short-term and delayed effects. Cognitive Therapy and Research, 6, 77-88. Emmelkamp, P.M.G., & Ultee, K.A. (1974). A comparison of 'successive approximation' and 'self-observation' in the treatment of agoraphobia. Behavior Therapy, 5, 606-613.

458

REFERENCES

Emmelkamp, P.M.G., & van der Hout, A. (1983). Failure in treating agoraphobia. In E.B. Foa & P.M.G. Emmelkamp (Eds.), Failures in behavior therapy (pp. 58-81). New York: Wiley. Emmelkamp, P.M.G., van Dyck, R., Bitter, M., Heins, R., Onstein, E.J., & Eisen, B. (1992). Spouse-aided therapy with agoraphobics. British Journal of Psychiatry, 160, 51-56. Emmelkamp, P.M.G., & Wessels, H. (1975). Flooding in imagination versus flooding in vivo: A comparison with agoraphobics. Behaviour Research and Therapy, 13, 7-15. Evans, I., Holt, C., & Oei, T.P.S. (1991). Long term follow-up of agoraphobics treated by brief intensive group cognitive behavioural therapy. Australian and New Zealand Journal of Psychiatry, 25, 343-349. Eysel, U.T., & Burandt, U. (1984). Fluorescent tube light evokes flicker responses in visual neurons. Vision Research, 24, 943-948. Eysenck, H.J., & Eysenck, S.B.G. (1975). Eysenck Personality Questionnaire. Essex: Hodder Stoughton. Fahy, T.J., O'Rourke, D., Brophy, J., Schazmann, W., & Sciascia, S. (1992). The Galway study of panic disorder I: Clomipramine and lofepramine in DSM III-R panic disorder: A placebo controlled trial. Journal of Affective Disorders, 25, 63-76. Fairburn, C.G., Norman, P.A., Welch, S.L., O'Connor, M.E., Doll, H.A., & Peveler, R.C. (1995). A prospective study of outcome in bulimia nervosa and the longterm effects of three psychological treatments. Archives of General Psychiatry, 52, 304-312. Falsetti, S.A. (1997a). The decision-making process of choosing a treatment for patients with civilian trauma-related PTSD. Cognitive and Behavioral Practice, 4, 99-121. Falsetti, S.A. (1997b). Treatment of PTSD with comorbid panic attacks. National Center for PTSD Clinical Quarterly, 7, 46-48. Falsetti, S.A., & Resnick, H.S. (1997). Frequency and severity of panic attack symptoms in a treatment seeking sample of trauma victims. Journal of Traumatic Stress, 10, 683-689. Faravelli, C. (1985). Life events preceding the onset of panic disorder. Journal of Affective Disorders, 9, 103-105. Faravelli, C., & Albanesi, G. (1987). Agoraphobia with panic attacks: 1-year prospective follow-up. Comprehensive Psychiatry, 28, 481-487. Faravelli, C., Pallanti, S., Biondi, F., Paterniti, S., & Scarpato, M.A. (1992). Onset of panic disorder. American Journal of Psychiatry, 149, 827-828. Faravelli, C., Webb, T., Ambonetti, A., Tonnesu, F., & Sessarego, A. (1985). Prevalence of traumatic early life events in 31 agoraphobia patients with panic attacks. American Journal of Psychiatry, 142, 1493-1494. Fava, G.A., Grandi, S., Canestrari, R., Grasso, P., & Pesarin, F. (1991). Mechanisms of change of panic attacks with exposure treatment of agoraphobia. Journal of Affective Disorders, 22, 65-71. Fava, G.A., Kellner, R., Zielezny, M., & Grandi, S. (1988a). Hypochondriacal fears and beliefs in agoraphobia. Journal of Affective Disorders, 14, 239-244. Fava, G.A., Zielezny, M., Luria, E., & Canestrari, R. (1988b). Obsessivecompulsive symptoms in agoraphobia: Changes with treatment. Psychiatry Research, 23, 57-63. Fava, G.A.; Zielezny, M., Savron, G., & Grandi, S. (1995). Long-term effects of behavioural treatment for panic disorder with agoraphobia. British Journal of Psychiatry, 166, 87-92.

REFERENCES459 Febbrano, G.A.R., Clum, G.A., Roodman, A.A., & Wright, J.H. (1999). The limits of bibliotherapy: A study of the differential effectiveness of self-administered interventions in individuals with panic attacks. Behavior Therapy, 30, 209-222. Fedoroff, I.C., Taylor, S., Asmundson, G.J.G., & Koch, W.J. (2000). Cognitive factors in traumatic stress reactions: Predicting PTSD symptoms from anxiety sensitivity and beliefs about harmful events. Behavioural and Cognitive Psychotherapy, 28, 5-15. Feldman, A.L., Sullivan, J.T., Passero, M.A., & Lewis, D.C. (1993). Pneumothorax in polysubstance-abusing marijuana and tobacco smokers: Three cases. Journal of Substance Abuse, 5, 183-186. Feske, U., & de Beurs, E. (1997). The Panic Appraisal Inventory: Psychometric properties. Behaviour Research and Therapy, 35, 875-882. Feske, U., & Goldstein, A.J. (1997). Eye movement desensitization and reprocessing treatment for panic disorder: A controlled outcome and partial dismantling study. Journal of Consulting and Clinical Psychology, 65, 1026-1035. Fewtrell, D., & O'Connor, K. (1995). Clinical phenomenology and cognitive psychology. London: Routledge. Fiengo, M.N., & Michelson, E.L. (1996). Evaluation of the patient with palpitations and non-life-threatening cardiac arrhythmias. In J.S. Alpert (Ed.), Cardiology for the primary care physician (pp. 59-65). St. Louis, MO: Mosby. Finlay-Jones, R., & Brown, G.W. (1981). Types of stressful life events and the onset of anxiety and depressive disorders. Psychological Medicine, 11, 803-815. First, M.B., Spitzer, R.L., Gibbon, M., & Williams, J.B.W. (1996). Structured Clinical Interview for DSM-IV Axis I-Patient edition. New York: Biometrics Research Department, New York State Psychiatric Institute. First, M.B., Spitzer, R.L., Gibbon, M., & Williams, J.B.W., Davies, M., Borus, J., Howes, M.J., Kane, J., Pope, H.G., & Rounsaville, B. (1995). The structured Clinical Interview for DSM-III-R personality disorders (SCID-II). Part II: Multisite test-retest reliability study. Journal of Personality Disorders, 9, 92-104. First, M.B., Spitzer, R.L., Gibbon, M., & Williams, J.B.W., & Lorna, B. (1994). Structured Clinical Interview for DSM-IV Axis II Personality Disorders (SCID-II) (Version 2.0). New York: Biometrics Research Department, New York State Psychiatric Institute. Fischer, J., & Corcoran, K. (1994). Measures for clinical practice: A sourcebook (2nd ed., vols. 1 & 2). New York: Free Press. Fischer, M., Hand, I., Angenendt, J., Buttner-Westphal, H., & Manecke, C. (1988). Failures in exposure treatment of agoraphobia: Evaluation and prediction. In I. Hand& H.-U. Wittchen (Eds.), Panic and phobias II: Treatment and variables affecting outcome (pp. 195-208). New York: Springer-Verlag. Fleet, R.P., & Beitman, B.D. (1998). Cardiovascular death from panic disorder and panic-like anxiety: A critical review of the literature. Journal of Psychosomatic Research, 44, 71-80. Flores, T.C., Cross, F.S., & Jones, R.D. (1981). Abnormal esophageal manometry in globus hystericus. Annals of Otology, Rhinology, and Laryngology, 90, 383-386. Foa, E.B., Jameson, J.S., Turner, R.M., & Payne, L.L. (1980). Massed vs. spaced exposure sessions in the treatment of agoraphobia. Behaviour Research and Therapy, 18, 333-338. Foa, E.B., & Kozak, M.J. (1986). Emotional processing of fear: Exposure to corrective information. Psychological Bulletin, 99, 20-35. Foa, E.B., Steketee, G., & Rothbaum, B.O. (1989). Behavioral/cognitive conceptualizations of post-traumatic stress disorder. Behavior Therapy, 20, 155-176.

460

REFERENCES

Folinsbee, L.J. (1995). Heat and air pollution. In M.L. Pollock & D.H. Schmidt (Eds.), Heart disease and rehabilitation (3rd ed.) (pp. 327-335). Champaign, IL: Human Kinetics. Fontaine, R., Chouinard, G., & Annable, L. (1984). Rebound anxiety in anxious patients after abrupt withdrawal of benzodiazepine treatment. American Journal of Psychiatry, 141, 848-852. Frances, A., Shear, M.K., & Fyer, M. (1988). Personality disorders and anxiety. In C. Last & M. Hersen (Eds.), Handbook of anxiety disorders (pp. 306-314). New York: Pergamon. Frank, J.D. (1973). Persuasion and healing (rev. ed.). Baltimore, MD: Johns Hopkins University Press. Frankl, V.E. (1975). Paradoxical intention and dereflection. Psychotherapy: Theory, Research, and Practice, 12, 226-227. Franklin, J.A. (1986). In-situ isometric relaxation training. Australian Psychologist, 21, 413--425. Franklin, J.A. (1989). A 6-year follow-up of the effectiveness of respiratory training, in-situ isometric relaxation, and cognitive modification in the treatment of panic disorder. Behavior Modification, 13, 139-167. Franklin, J.A. (1996). Overcoming panic: A complete nine-week home-based treatment program for panic disorder. Carlton, Australia: Australian Psychological Society. Freud, S. (1894/1949). The justification for detaching from neurasthenia a particular syndrome: The 'anxiety neurosis'. In E. Jones (Ed.), Collected papers of Sigmund Freud, vol. I. (pp. 76--106). London: Hogarth Press. Freud, S. (1895a/1949). A reply to criticisms on the anxiety neurosis. In E. Jones (Ed.), Collected papers of Sigmund Freud, vol. I. (pp. 107-127). London: Hogarth Press. Freud, S. (1895b/1949). Obsessions and phobias: Their psychical determinants and their ~tiology. In E. Jones (Ed.), Collected papers of Sigmund Freud, vol. I. (pp. 128-137). London: Hogarth Press. Freud, S. (1919 /1950). Turnings in the ways of psychoanalytic therapy. In E. Jones (Ed.), Collected papers of Sigmund Freud, vol. II (pp. 392--402). London: Hogarth Press. Fried, R. (1993). The psychology and physiology of breathing. New York: Plenum. Friedman, M. (1996). Type A behavior. New York: Plenum Press. Friedman, S. (1987). Technical considerations in the behavioral-marital treatment of agoraphobia. American Journal of Family Therapy, 15, 111-122. Friedman, S., & Paradis, C.M. (1991). African-American patients with panic disorder and agoraphobia. Journal of Anxiety Disorders, 5, 35--41. Friedman, S., Paradis, C.M., & Hatch, M. (1994). Characteristics of AfricanAmerican and White patients with panic disorder and agoraphobia. Hospital and Community Psychiatry, 45, 798-803. Frisch, M.B. (1998). Quality of life therapy and assessment in health care. Clinical Psychology: Science and Practice, 5, 19--40. Frisch, M.B. (1999). Quality of Life Therapy: Therapist Manual. Unpublished manuscript, Department of Psychology and Neuroscience, Baylor University. Frisch, M.B., Cornell, J., Villanueva, M., & Retzlaff, P.J. (1992). Clinical validation of the Quality of Life Inventory: A measure of life satisfaction for use in treatment planning and outcome assessment. Psychological Assessment, 4, 92-101. Furer, P., Walker, J.R., Chartier, M.J., & Stein, M.B. (1997). Hypochondriacal concerns and somatization in panic disorder. Depression and Anxiety, 6, 78-85.

REFERENCES461 Fux, M., Taub, M., & Zohar, J. (1993). Emergence of depressive symptoms during treatment for panic disorder with specific 5-hydroxytryptophan reuptake inhibitors. Acta Psychiatrica Scandinavica, 88, 235-237. Fyer, A.J., Liebowitz, M.R., Gorman, J.M., Campeas, R., Levin, A., Davis, S.O., Goetz, D., & Klein, D.F. (1987). Discontinuation of alprazolam in panic patients. American Journal of Psychiatry, 144, 303-308. Fyer, A.J., Liebowitz, M.R., Gorman, J.M., Davies, S.O., & Klein, D.F. (1983). Sodium lactate reinfusion of recovered lactate-vulnerable panic-disorder patients. Psychopharmacology Bulletin, 19, 576-577. Garland, E.J., & Clark, S.L. (1995). Taming worry dragons: A manual for children, parents, and other coaches. Vancouver, BC: British Columbia's Children's Hospital. Garssen, B., Buikhuisen, M., & van Dyck, R. (1996). Hyperventilation and panic attacks. American Journal of Psychiatry, 153, 513-518. Garssen, B., de Ruiter, C., & van Dyck, R. (1992). Breathing retraining: A rational placebo? Clinical Psychology Review, 12, 141-153. Garssen, B., de Ruiter, C., van Dyck, R., & Hornsveld, H. (1993). Will hyperventilation syndrome survive?: A rejoinder to Ley (1993) Clinical Psychology Review, 13, 683-690. Gelder, M.G., Bancroft, J.H., Gath, D.H., Johnston, D.W., Mathews, A.M., & Shaw, P.M. (1973). Specific and non-specific factors in behaviour therapy. British Journal of Psychiatry, 123, 445-462. George, D.T., Zerby, A., Noble, S., & Nutt, D.J. (1988). Panic attacks and alcohol withdrawal: Can subjects differentiate the symptoms? Biological Psychiatry, 24, 240-243. Gershon, M.D. (1998). The second brain. New York: Harper Collins. Gerz, H.O. (1967). The treatment of the phobic and obsessional patient using paradoxical intention. In V.E. Frankl (Ed.), Psychotherapy and existentialism: Selected papers on logotherapy (pp. 199-221). New York: Washington Square Press. Ghosh, A., & Marks, I.M. (1987). Self-treatment of agoraphobia by exposure. Behavior Therapy, 18, 3-16. Ghosh, A., Marks, I.M., & Carr, A.C. (1988). Therapist contact and outcome of selfexposure treatment for phobias: A controlled study. British Journal of Psychiatry, 152, 234--238. Gibbs, D.M. (1992). Hyperventilation-induced cerebral ischemia in panic disorder and effect of nimodipine. American Journal of Psychiatry, 149, 1589-1591. Gillis, M.M., Haaga, D.A.F., & Ford, G.T. (1995). Normative values for the Beck Anxiety Inventory, Fear Questionnaire, Penn State Worry Questionnaire, and Social Phobia and Anxiety Inventory. Psychological Assessment, 7, 450-455. Goldberg, R.J. (1988). Clinical presentations of panic-related disorders. Journal of Anxiety Disorders, 2, 61-75. Goldfried, M.R., Wiser, S.L., & Raue, P.J. (1992). On the movement toward psychotherapy integration: The case of panic disorder. Journal of Psychotherapy Practice and Research, l, 213-224. Goldstein, A.J., & Chambless, D.L. (1978). A reanalysis of agoraphobia. Behavior Therapy, 9, 47-59. Goldstein, A.J., & Feske, U. (1994). Eye movement desensitization and reprocessing for panic disorder: A case series. Journal of Anxiety Disorders, 8, 351-362. Golub, S. (1976). The magnitude of premenstrual anxiety and depression. Psychosomatic Medicine, 38, 4-12. Gordon, T. (1975). Parent effectiveness training. New York: New American Library.

462

REFERENCES

Gorman, J.M. (1994). New and experimental pharmacological treatments for panic disorder. In B.E. Wolfe & J.D. Maser (Eds.), Treatment of panic disorder: A consensus development conference (pp. 83-89). Washington, DC: American Psychiatric Press. Gorman, J.M., Fyer, M.R., Goetz, R.R., Askanazi, J., Liebowitz, M.R., Fyer, A.J., Kinney, J., & Klein, D.F. (1988a). Ventilatory physiology of patients with panic disorder. Archives of General Psychiatry, 45, 31-39. Gorman, J.M., Goetz, R.R., Fyer, M., King, D.L., Fyer, A.J., Liebowitz, M.R., & Klein, D.F. (1988b). The mitral valve prolapse-panic disorder connection. Psychosomatic Medicine, 50, 114--122. Gould, R.A., & Clum, G.A. (1995). Self-help plus minimal therapist contact in the treatment of panic disorder: Replication and extension. Behavior Therapy, 26, 533-546. Gould, R.A., Clum, G.A., & Shapiro, D. (1993). The use of bibliotherapy in the treatment of panic disorder: A preliminary investigation. Behavior Therapy, 24, 241-252. Gould, R.A., Otto, M.W., & Pollack, M.H. (1995). A meta-analysis of treatment outcome for panic disorder. Clinical Psychology Review, 15, 819-844. Gray, L.P. (1983). The relationship of the 'inferior constrictor swallow' and 'globus hystericus' or the hypopharyngeal syndrome. Journal of Laryngalogy and Otolagy, 97, 607-618. Green, M.A., & Curtis, G.C. (1988). Personality disorders in panic patients: Response to termination of anti panic medication. Journal of Personality Disorders, 2, 303-314. Greenberg, P.E., Sisitsky, T., Kessler, R.C., Finkelstein, S.N., Berndt, E.R., Davidson, J.R.T., Ballenger, J.C., & Fyer, A.J. (1999). The economic burden of anxiety disorders in the 1990s. Journal of Clinical Psychiatry, 60, 427-435. Greenberg, R.L. (1989). Panic disorder and agoraphobia. In J. Scott, J.M.G. Williams, & A.T. Beck (Eds.), Cognitive therapy in clinical practice (pp. 25-49). London: Routledge. Griez, E., Zandbergen, J., Lousberg, H., & van den Hout, M. (1988). Effects of low pulmonary CO 2 on panic anxiety. Comprehensive Psychiatry, 29, 490-497. Grossman, P. (1983). Respiration, stress and cardiovascular function. Psychophysiology, 20, 284--300. Groth-Mamat, G. (1990). Handbook of psychological assessment (2nd ed.). New York: Wiley. Grunhaus, L., Pantle, A.C., Brown, M.B., & Greden, J.F. (1994). Clinical characteristics of patients with concurrent major depressive disorder and panic disorder. American Journal of Psychiatry, 151, 541-546. Gustavson, B., Jansson, L., Jerremalm, A., & Ost, L.-G. (1985). Therapist behavior during exposure treatment of agoraphobia. Behavior Modification, 9, 491504. Guy, W. (1976). ECDEU assessment manual far psychopharmacolagy. US Department of Health, Education and Welfare, Publication ADM 76-338. Washington, DC: US Government Printing Office. Hafner, R.J. (1976). Fresh symptom emergence after intensive behaviour therapy. British Journal of Psychiatry, 129, 378-383. Hafner, Rl (1977). The husbands of agoraphobic women and their influence on treatment outcome. British Journal of Psychiatry, 131, 588-602. Hafner, R.J. (1981). Agoraphobia in men. Australian and New Zealand Journal of Psychiatry, 15, 243-249.

REFERENCES463

Hafner, R.J. (1983). Behavior therapy for agoraphobic men. Behaviour Research and Therapy, 21, 51-56. Hafner, R.J. (1984). Predicting the effects on husbands of behaviour therapy for wives' agoraphobia. Behaviour Research and Therapy, 22, 217-226. Hafner, R.J., & Marks, I.M. (1976). Exposure in vivo of agoraphobics: Contributions of diazepam, group exposure and anxiety evocation. Psychological Medicine, 6, 71-88. Hafner, R.J., & Ross, M.W. (1983). Predicting the outcome of behaviour therapy for agoraphobia. Behaviour Research and Therapy, 21, 375-382. Haley, J. (1980). Leaving home. New York: McGraw-Hill. Hallam, R.S. (1978). Agoraphobia: A critical review of the concept. British Journal of Psychiatry, 133, 314-319. Hamann, M.S., & Mavissakalian, M. (1988). Discrete dimensions in agoraphobia: A factor analytic study. British Journal of Clinical Psychology, 27, 137-144. Hamilton, M. (1969). Diagnosis and rating of anxiety. British Journal of Psychiatry, Special Publication, 3, 76-79. Hand, I., & Lamontagne, Y. (1976). The exacerbation of interpersonal problems after rapid phobia removal. Psychotherapy: Theory, Research and Practice, 13, 405-411. Harrington, P.H., Schmidt, N.B., & Telch, M.J. (1996). Prospective evaluation of panic potentiation following 35% CO 2 challenge in a nonclinical sample. American Journal of Psychiatry, 153, 823-825. Harvey, J.M., Richards, J.C., Dziadosz, T., & Swindell, A. (1993). Misinterpretation of ambiguous stimuli in panic disorder. Cognitive Therapy and Research, 17, 235-248. Hassan, R., & Pollard, C.A. (1994). Late-life-onset panic disorder: Clinical and demographic characteristics of a patient sample. Journal of Geriatric Psychiatry and Neurology, 7, 86-90. Hawthorn, J. (1995). Understanding and management of nausea and vomiting. Cambridge, MA: Blackwell Science. Hawton, K., Salkovskis, P.M., Kirk, J., & Clark D.M. (1989). Cognitive behaviour therapy for psychiatric problems: A practical guide. Oxford: Oxford University Press. Hayward, C., Taylor, C.B., Roth, W.T., King, R., & Agras, W.S. (1989). Plasma lipid levels in patients with panic disorder or agoraphobia. American Journal of Psychiatry, 146, 917-919. Hazell, J., & Wilkins, A.J. (1990). A contribution of fluorescent lighting to agoraphobia. Psychological Medicine, 20, 591-596. Hecker, J.E., Losee, M.C., Fritzler, B.K., & Fink, C.M. (1996). Self-directed versus therapist-directed cognitive behavioral treatment for panic disorder. Journal of Anxiety Disorders, 10, 253-265. Hecker, J.E., & Thorpe, G.K. (1992). Agoraphobia and panic: A guide to psychological treatment. Needham Heights, MA: Allyn and Bacon. Hegel, M.T., Abel, G.G., Etscheidt, M., Cohen-Cole, S., & Wilmer, C.I. (1989). Behavioral treatment of angina-like chest pain in patients with hyperventilation syndrome. Journal of Behavior Therapy and Experimental Psychiatry, 20, 31-39. Hegel, M.T., Ravaris, C.L., & Ahles, T.A. (1994). Combined cognitive-behavioral and time-limited alprazolam treatment of panic disorder. Behavior Therapy, 25, 183-195. Heide, F.J., & Borkovec, T.D. (1983). Relaxation-induced anxiety: Paradoxical anxiety enhancement due to relaxation training. Journal of Consulting and Clinical Psychology, 51, 171-182.

464

REFERENCES

Heide, F.J., & Borkovec, T.D. (1984). Relaxation-induced anxiety: Mechanisms and theoretical implications. Behaviour Research and Therapy, 22, 1-12. Heimberg, R.G. (1994). Cognitive assessment strategies and the measurement of outcome of treatment for social phobia. Behaviour Research and Therapy, 32, 269-280. Heller, R. (1984). Munch: His life and work. Chicago, IL: University of Chicago Press. Hibbert, G.A. (1984). Ideational components of anxiety: Their origin and content. British Journal of Psychiatry, 144, 618-624. Hibbert, G.A., & Chan, M. (1989). Respiratory control: Its contribution to the treatment of panic attacks. British Journal of Psychiatry, 154, 232-236. Hibbert, G.A., & Pilsbury, D. (1989). Hyperventilation: Is it a cause of panic attacks? British Journal of Psychiatry, 155, 805-809. Higgitt, A., Golombok, S., Fonagy, P., & Lader~ M. (1987). Group treatment of benzodiazepine dependence. British Journal of Addiction, 82, 517-532. Himadi, W.G., Cerny, J.A., Barlow, D.H., Cohen, S., & O'Brien, G.T. (1986). The relationship of marital adjustment to agoraphobia treatment outcome. Behaviour Research and Therapy, 24, 107-115. Hobbs, W.R., Wilkinson, C., Peterson, G., Laraia, M., Cox, D., & Ballenger, J. (1984). The psychophysiology of agoraphobia.Washington, DC: American Psychiatric Press. Hoffart, A. (1994). State and personality in agoraphobic patients. Journal of Personality Disorders, 8, 333-341. Hoffart, A. (1995). A comparison of cognitive and guided mastery therapy of agoraphobia. Behaviour Research and Therapy, 33, 423--434. Hoffart, A. (1997). Interpersonal problems among patients suffering from panic disorder with agoraphobia before and after treatment. British Journal of Medical Psychology, 70, 149-157. Hoffart, A. (1998). Cognitive and guided mastery therapy of agoraphobia: Longterm outcome and mechanisms of change. Cognitive Therapy and Research, 22, 195-207. Hoffart, A., Friis, S., & Martinsen, E. (1992). Assessment of fear of fear among agoraphobic patients: The Agoraphobic Cognitions Scales. Journal of Psychopathology and Behavioral Assessment, 14, 175-187. Hoffart, A., & Hedley, L.M. (1997). Personality traits among panic disorder with agoraphobia patients before and after symptom-focused treatment. Journal of Anxiety Disorders, 11, 77-87. Hoffart, A., & Martinsen, E. (1990). Exposure-based integrated vs. pure psychodynamic treatment of agoraphobic inpatients. Psychotherapy, 27, 210-218. Hoffart, A., & Martinsen, E. (1993). The effects of personality disorders and anxious-depressive comorbidity on outcome of patients with unipolar depression and with panic disorder and agoraphobia. Journal of Personality Disorders, 7, 304--311. Hofmann, S.G. (1999). The relationship between panic and schizophrenia. Depression and Anxiety, 9, 101-106. Hofmann, S.G., & Barlow, D.H. (1996). Ambulatory psychophysiological monitoring: A potentially useful tool when treating panic relapse. Cognitive and Behavioral Practice, 3, 53-61. Hofmann, S.G., Barlow, D.H., Papp, L.A., Detweiler, M.F., Ray, S.E., Shear, M.K., Woods, S.W., & Gorman, J.M. (1998a). Pretreatment attrition in a comparative outcome study on panic disorder. American Journal of Psychiatry, 155, 43--47. Hofmann, S.G., Shear, M.K., Barlow, D.H., Gorman, J.M., Hershberger, D., Patterson, M., & Woods, S.W. (1998b). Effects of panic disorder treatments on personality disorder characteristics. Depression and Anxiety, 8, 14--20.

REFERENCES465 Holden, A.E., & Barlow, D.H. (1986). Heart rate and heart rate variability recorded in vivo in agoraphobics and nonphobics. Behavior Therapy, 17, 2642. Holden, A.E., O'Brien, G.T., Barlow, D.H., Stetson, D., & Infantino, A. (1983). Selfhelp manual for agoraphobia: A preliminary report on effectiveness. Behavior Therapy, 14, 545-556. Hollon, S.D., & Kendall, P.C. (1980). Cognitive self-statements in depression: Development of an Automatic Thoughts Questionnaire. Cognitive Therapy and Research, 4, 383-395. Holmes, T.M., & Rahe, R.H. (1967). The Social Readjustment Rating Scale. Journal of Psychosomatic Research, 11, 213-218. Hope, D.A., Rapee, R.M., Heimberg, R.G., & Dombeck, M.J. (1990). Representations of the self in social phobia: Vulnerability to social threat. Cognitive Therapy and Research, 14, 177-189. Hornig, C.D., & McNally, R.J. (1995). Panic disorder and suicide attempt: A reanalysis of data from the Epidemiological Catchment Area study. British Journal of Psychiatry, 167, 76-79. Hunt, C., & Andrews, G. (1998). Long-term outcome of panic disorder and agoraphobia. Journal of Anxiety Disorders, 12, 395-406. Hunter, J.E., & Schmidt, F.L. (1990). Methods of meta-analysis: Correcting error and bias in researchfindings. Newbury Park, CA: Sage. Hyler, S.E. (1994). Personality Diagnostic Questionnaire (4th ed.) (PDQ-4). New York: New York State Psychiatric Institute. Hyler, S., Reider, R., Spitzer, R., & Williams, J.B.W. (1983). Personality Diagnostic Questionnaire (PDQ). New York: New York State Psychiatric Institute. Ito, L.M., Noshirvani, H., Basoglu, M., & Marks, I.M. (1996). Does exposure to internal cues enhance exposure to external cues in agoraphobia with panic? Psychotherapy and Psychosomatics, 65, 24-28. Jacob, R.G., Furman, J.M., & Balaban, C.D. (1996a). Psychiatric aspects of vestibular disorders. In R.W. Baloh & G.M. Halmagyi (Eds.), Disorders of the vestibular system (pp. 509-528). Oxford: Oxford University Press. Jacob, R.G., Furman, J.M., Durrant, J.D., & Turner, S.M. (1996b). Panic, agoraphobia, and vestibular dysfunction. American Journal of Psychiatry, 153, 503512. Jacob, R.G., & Lilienfeld, S.O. (1991). Panic disorder: Diagnosis, medical assessment, and psychological assessment. In J.R. Walker, G.R. Norton, & C.A. Ross (Eds.), Panic disorder and agoraphobia:A comprehensive guide for the practitioner (pp. 16-102). Pacific Grove, CA: Brooks/Cole. Jacobsberg, L., Perry, S., & Frances, A. (1995). Diagnostic agreement between the SCID-II Screening Questionnaire and the Personality Disorder Examination. Journal of Personality Assessment, 65, 428-433. Jacobson, N.S., & Margolin, G. (1979). Marital therapy: Strategies based on social learning and behavior exchange principles. New York: Brunner/Maze!. Jacobson, N.S., Wilson, L., & Tupper, C. (1988). The clinical significance of treatment gains resulting from exposure-based interventions for agoraphobia: A reanalysis of outcome data. Behavior Therapy, 19, 539-554. Jakubowski, P., & Lange, A.J. (1978). The assertive option. Champaign, IL: Research Press. • Jang, K.L., Stein, M.B., Taylor, S., & Livesley, W.J. (1999). Gender differences in the aetiology of anxiety sensitivity: A twin study. Journal of Gender Specific Medicine, 2, 39-44.

466

REFERENCES

Jannoun, L, Munby, M., Catalan, J., & Gelder, M. (1980). A home-based treatment program for agoraphobia: Replication and controlled evaluation. Behavior Therapy, 11, 294-305. .. Jansson, L, Jerremalm, A., & Ost, L-G. (1984). Maintenance procedures in the behavioral treatment of agoraphobia: A program and some data. Behavioural Psychotherapy, 12, 109-116. Jansson, L, & Ost, L-G. (1982). Behavioral treatments for agoraphobia: An evaluative review. Clinical Psychology Review, 2, 311-336. Jansson, L, Ost, L-G., & Jerremalm, A. (1987). Prognostic factors in the behavioral treatment of agoraphobia. Behavioural Psychotherapy, 15, 31-44. Johnston, D., & Gath, D. (1973). Arousal levels and attribution effects in diazepamassisted flooding. British Journal of Psychiatry, 123, 463-466. Johnston, M., Johnston, D.W., Wilkes, H., Burns, LE., & Thorpe, G.L (1984). Cumulative scales for the measurement of agoraphobia. British Journal of Clinical Psychology, 23, 133-143. Jorm, A.F. (1989). Modifiability of trait anxiety and neuroticism: A meta-analysis of the literature. Australian and New Zealand Journal of Psychiatry, 23, 21-29. Kabat-Zinn, J. (1990). Full catastrophe living. New York: Delta. Kabat-Zinn, J., Lipworth, L, & Burney, R. (1985). The clinical use of mindfulness meditation for the self-regulation of chronic pain. Journal of Behavioral Medicine, 8, 163-190. Kabat-Zinn, J., Massion, A.O., Kristeller, J., Peterson, LG., Fletcher, K.E., Pbert, L, Lenderking, W.R., & Santorelli, S.F. (1992). Effectiveness of a meditation-based stress reduction program in the treatment of anxiety disorders. American Journal of Psychiatry, 149, 936-943. Kabat-Zinn, J., Wheeler, E., Light, T., Skillings, A., Scharf, M.J., Cropley, T.G., Hosmer, D., & Bernhard, J.D. (1998). Influence of a mindfulness meditationbased stress reduction intervention on rates of skin clearing in patients with moderate to severe psoriasis undergoing phototherapy (UVB) and photochemotherapy (PUVA). Psychosomatic Medicine, 60, 625-632. Kagan, J.S., & Snidman, N. (1991). Temperamental factors in human development. American Psychologist, 46, 856-862. Kahn, J., Puertollano, M.A., & Klein, D.F. (1987). Schizophrenia, panic anxiety, and alprazolam [letter]. American Journal of Psychiatry, 144, 527-528. Kahn, J., Puertollano, M.A., & Klein, D.F. (1988). Adjunctive alprazolam for schizophrenia with panic anxiety: Clinical observation and pathogenic implications. American Journal of Psychiatry, 145, 742-744. Kahneman, D., Slovic, P., & Tversky, A. (1982). Judgment under uncertainty: Heuristics and biases. Cambridge: Cambridge University Press. Kamieniecki, G.W., Wade, T., & Tsourtos, G. (1997). Interpretive bias for benign sensations in panic disorder and agoraphobia. Journal of Anxiety Disorders, 11, 141-156. Karajgi, B., Rifkin, A., Doddi, S., & Kolli, R. (1990). The prevalence of anxiety disorders in patients with chronic obstructive pulmonary disease. American Journal of Psychiatry, 147, 200-201. Kaspi, S.P., Otto, M.W., Pollack, M.H., Eppinger, S., & Rosenbaum, J.P. (1994). Premenstrual exacerbation of symptoms in women with panic disorder. Journal of Anxiety Disorders, 8, 131-138. Katemdahl, D.A. (1993). Panic and prolapse: Meta-analysis. Journal of Nervous and Mental Disease, 181, 539-544. Katemdahl, D.A., & Realini, J.P. (1995). Where do panic attack sufferers seek care? Journal of Family Practice, 40, 237-243.

REFERENCES467 Katemdahl, D.A., & Realini, J.P. (1997). Quality of life and panic-related work disability in subjects with infrequent panic and panic disorder. Journal of Clinical Psychiatry, 58, 153-158. Katon, W. (1984). Panic disorder and somatization: Review of 55 cases. American Journal of Medicine, 77, 101-106. Katon, W. (1988). Panic disorder: The importance of phenomenology. Journal of Family Practice, 26, 23-24. Katon, W. (1994). Panic disorder in the medical setting. Washington, DC: National Institute of Mental Health. (NIH publication no. 94-3482.) Katon, W., Vitaliano, P.P., Russo, J., Jones, M., & Anderson, K. (1987). Panic disorder: Spectrum of severity and somatization. Journal of Nervous and Mental Disease, 175, 12-19. Katz, B., & Carmody, R. (1985). Subperiosteal orbital hematoma induced by the Valsalva maneuver. American Journal of Ophthalmology, 100, 617-618. Keijsers, G., Hoogduin, C.A.L., & Schaap, C. (1994). Prognostic factors in the behavioral treatment of panic disorder with and without agoraphobia. Behavior Therapy, 25, 689-708. Keijsers, G., Schaap, C., Hooghuin, C., & Lammers, M.W. (1995). Patient-therapist interaction as a predictor of outcome in the behavioural treatment of panic disorder. Behavior Modification, 19, 491-517. Keijsers, G., Schaap, C., Hooghuin, C., & Peters, W. (1991). The therapeutic relationship in the behavioural treatment of anxiety disorders. Behavioural Psychotherapy, 19, 359-367. Kellner, R. (1991). Psychosomatic syndromes and somatic symptoms. Washington, DC: American Psychiatric Press. Kelly, D., Mitchell-Heggs, N., & Sherman, D. (1971). Anxiety and the effects of sodium lactate assessed clinically and physiologically. British Journal of Psychiatry, 119, 129-141. Kendell, R.E. (1974). The stability of psychiatric diagnoses. British Journal of Psychiatry, 124, 352-356. Kendler, K.S., Neale, M.C., Kessler, R.C., Heath, A.C., & Eaves, L.J. (1992). Childhood and parental loss and adult psychopathology in women: A twin study perspective. Archives of General Psychiatry, 49, 109-116. Kendler, K.S., Neale, M.C., Kessler, R.C., Heath, A.C., & Eaves, L.J. (1993). Panic disorder in women: A population-based twin study. Psychological Medicine, 23, 397-406. Kendler, K.S., Walters, E.E., Neale, M.C., Kessler, R.C., Heath, A.C., & Eaves, L.J. (1995). The structure of the genetic and environmental risk factors for six major psychiatric disorders in women: Phobia, generalized anxiety disorder, panic disorder, bulimia, major depression and alcoholism. Archives of General Psychiatry, 52, 374-383. Kennedy, B.L., & Schwab, J.J. (1997). Utilization of medical specialists by anxiety disorder patients. Psychosomatics, 38, 109-112. Kern, J.M. (1983). Relationships between obtrusive laboratory and unobtrusive naturalistic behavioral fear assessments: Treated and untreated subjects. Behavioral Assessment, 6, 45-60. Kernberg, O.F., Selzer, M.A., Koenigsberg, H.W., Carr, A.C., & Appelbaum, A.H. (1989). Psychodynamic psychotherapy of borderline patients. New York: Basic Books. Kessler, R.C., McGonagle, K.A., Zhao, S., Nelson, C.B., Hughes, M., Eshleman, S., Wittchen, H.-U., & Kendler, K.S. (1994). Lifetime and 12-month prevalence of DSM-III-R psychiatric disorders in the United States. Archives of General Psychiatry, 51, 8-19.

468

REFERENCES

Kessler, R.C., Stang, P.E., Wittchen, H.-U., Ustun, T.B., Roy-Byrne, P.P., & Walters, E.E. (1998). Lifetime panic-depression comorbidity in the National Comorbidity Survey. Archives of General Psychiatry, 55, 801-808. Khawaja, N.G., & Oei, T.P.S. (1998). Catastrophic cognitions in panic disorder with and without agoraphobia. Clinical Psychology Review, 18, 341-365. Khawaja, N.G., Oei, T.P.S., & Baglioni, A.J. (1994). Modification of the Catastrophic Cognition Questionnaire (CCQ-M) for normals and patients: Exploratory and LISREL analyses. Journal of Psychopathology and Behavioral Assessment, 16, 325-342. Kilpatrick, D.G., Veronen, L.J., & Resick, P.A. (1982). Psychological sequelae to rape: Assessment and treatment strategies. In D.M. Dolays & R.L. Meredith (Eds.), Behavioral medicine: Assessment and treatment strategies (pp. 473-497). New York: Plenum. Kirk, J. (1989). Cognitive-behavioural assessment. In K. Hawton, P.M. Salkovskis, J. Kirk, & D.M. Clark (Eds.), Cognitive behaviour therapy for psychiatric problems: A practical guide (pp. 13-51). Oxford: Oxford University Press. Kirsch, I. (1999). How expectancies shape experience. Washington, DC: American Psychological Association Press. Klein, D.F. (1964). Delineation of two drug-responsive anxiety syndromes. Psychopharmacologia, 5, 397-408. Klein, D.F. (1980). Anxiety reconceptualized. Comprehensive Psychiatry, 21, 411-427. Klein, D.F. (1993). False suffocation alarms, spontaneous panics, and related conditions: An integrative hypothesis. Archives of General Psychiatry, 50, 306-317. Klein, D.F., & Fink, M. (1962). Psychiatric reaction patterns to imipramine. American Journal of Psychiatry, 119, 432-438. Klein, D.F., Ross, D.C., & Cohen, P. (1987). Panic and avoidance in agoraphobia: Application of path analysis to treatment studies. Archives of General Psychiatry, 44, 377-385. Klein, M.H., Benjamin, L.S., Rosenfeld, R., Treece, C., Husted, J., & Greist, J.H. (1993). The Wisconsin Personality Disorders Inventory, I: Development, reliability, and validity. Journal of Personality Disorders, 7, 285-303. Kleiner, L., Marshall, W.L., & Spevack, M. (1987). Training in problem-solving and exposure treatment for agoraphobics with panic attacks. Journal of Anxiety Disorders, 1, 219-238. Klerman, G.L., & Weissman, M.M. (1993). New applications of interpersonal psychotherapy. Washington, DC: American Psychiatric Association. Klerman, G.L., Weissman, M.M., Rounsaville, B.J., & Chevron, E.S. (1984). Interpersonal psychotherapy of depression. New York: Basic Books. Klimes, I., Mayou, R.A., Pearce, M.J.,.Coles, L., & Fagg, J.R. (1990). Psychological treatment for atypical non-cardiac chest pain: A controlled evaluation. Psychological Medicine, 20, 605-611. Kloska, J.S., Barlow, D.H., Tassinari, R., & Cerny, J.A. (1990). A comparison of alprazolam and behavior therapy in treatment of panic disorder. Journal of Consulting and Clinical Psychology, 58, 77-84. Kloska, J.S., Barlow, D.H., Tassinari, R., & Cerny, J.A. (1995). A comparison of alprazolam and behavior therapy in treatment of panic disorder: Correction. Journal of Consulting and Clinical Psychology, 63, 830. Knapp, T.J., & Schumacher, M.T. (1988). Westphal's 'Die agoraphobie' with commentary: The .beginnings of agoraphobia. New York: University Press of America. Kniker, W.T. (1988). Anaphylaxis in children and adults. In C.W. Bierman & D.S. Pearlman (Eds.), Allergic diseases from infancy to adulthood (pp. 667-677). Philadelphia, PA: Saunders.

REFERENCES469 Kobak, K.A., Taylor, L., Dotti, S.L., Greist, J.H., Jefferson, J.W., Burroughs, D., Mantle, J.M., Katzelnick, D.J., Norton, R., Henk, H.J., & Serlin, R.C. (1997). A computer-administered telephone interview to identify mental disorders. Journal of the American Medical Association, 278, 905-910. Kolko, D.J. (1984). Paradoxical instruction in the elimination of avoidance behavior in an agoraphobic girl. Journal of Behavior Therapy and Experimental Psychiatry, 15, 51-57. Kopp, M., Milhaly, K., Tringer, A., & Vadasz, P. (1986). Agoraphobics es panikneurotikus betegek legzesi keyelese. Ideggyogyaszati Szemle, 39, 185-196. Kraaier, V., van Huffelen, A.C., Wieneke, G.H., & Keppel Hesselink, J.M. (1991). Nimodipine tested in a human model of cerebral ischemia. European Journal of Clinical Pharmacology, 40, 17-21. Kuch, K., Cox, B.J., Woszczyna, C.B., Swinson, R.P., & Shulman, I. (1991). Chronic pain in panic disorder. Journal of Behavior Therapy and Experimental Psychiatry, 22, 255-259. Kuch, K., & Swinson, R.P. (1992). Agoraphobia: What Westphal really said. Canadian Journal of Psychiatry, 37, 133-136. Kushner, M.G., Mackenzie, T.B., Fiszdon, J., Valentiner, D.P., Foa, E., Anderson, N., & Wangensteen, D. (1996). The effects of alcohol consumption on laboratory-induced panic and state anxiety. Archives of General Psychiatry, 53, 264-270. Kwon, S.M., Evans, L., & Oei, T.P.S. (1990). Factor structure of the Mobility Inventory for agoraphobia: A validation study with Australian samples of agoraphobic patients. Journal of Psychopathology and Behavioral Assessment, 12, 365-374. Laberge, B., Gauthier, J.G., Cote, G., Plamondon, J., & Cormier, H.J. (1993). Cognitive-behavioral therapy of panic disorder with secondary major depression: A preliminary investigation. Journal of Consulting and Clinical Psychology, 61, 1028-1037. Lang, P.J. (1968). Fear reduction and fear behavior: Problems in treating a construct. In J.M. Shlien (Ed.), Research in psychotherapy, vol. 3 (pp. 90-102). Washington, DC: American Psychological Association. Lang, P.J., & Lazovik, A.D. (1963). Experimental desensitization of a phobia. Journal of Abnormal and Social Psychology, 66, 519-525. Lange, A., & de Beurs, E. (1992). Multidimensional family treatment of agoraphobia. Journal of Family Psychotherapy, 3, 45-72. Lange, A., & van der Hart, 0. (1983). Directive family therapy. New York: Brunner/ Maze!. Lange, A., & van Dyck, R. (1992). The function of agoraphobia in the marital relationship. Acta Psychiatrica Scandinavica, 85, 89-93. Lange, A.J., & Jakubowski, P. (1976). Responsible assertive behavior. Champaign, IL: Research Press. Lantinga, L.J., Sprafkin, R.P., McCroskery, J.H., Baker, M.T., Warner, R.A., & Hill, N.E. (1988). One-year psychosocial follow-up of patients with chest pain and angiographically normal coronary arteries. American Journal of Cardiology, 62, 209-213. Last, C.G., Barlow, D.H., & O'Brien, G.T. (1984). Precipitants of agoraphobia: Role of stressful life events. Psychological Reports, 54, 567-570. Lazarus, R.S. (1966). Psychological stress and the coping process. New York: McGrawHill. Lazarus, RS. (1991). Cognition and motivation in emotion. American Psychologist, 46, 352-367.

470

REFERENCES

Lazarus, R.S., & Folkman, S. (1984). Stress, appraisal, and coping. New York: Springer. Lazarus, R.S., & Folkman, S. (1989). Manual of the Hassles and Uplifts Scales. Palo Alto, CA: Consulting Psychologists Press. Lehman, C.L., Brown, T.A., & Barlow, D.H. (1998). Effects of cognitive-behavioral treatment for panic disorder with agoraphobia on concurrent alcohol abuse. Behavior Therapy, 29, 423-433. Lelliott, P., Marks, I., McNamee, G., & Tobefia, A. (1989). Onset of panic disorder with agoraphobia: Toward an integrated model. Archives of General Psychiatry, 46, 1000-1004. Lelliott, P., Marks, I.M., Monteiro, W.O., Tsakiris, F., & Noshirvani, H. (1987). Agoraphobics 5 years after imipramine and exposure: Outcome and predictors. Journal of Nervous and Mental Disease, 175, 599-605. Leon, A.C., Portera, L., & Weissman, M.M. (1995). The social costs of anxiety disorders. British Journal of Psychiatry, 166 (suppl. 27), 19-22. Levenkron, J.C., Goldstein, M.G., Adamides, 0., & Greenland, P. (1985). Chronic chest pain with normal coronary arteries: A behavioral approach to rehabilitation. Journal of Cardiopulmonary Rehabilitation, 5, 475-479. Levenson, R.W., Oyama, O.N., & Meek, P.S. (1987). Greater reinforcement from alcohol for those at risk: Parental risk, personality risk, and sex. Journal of Abnormal Psychology, 96, 242-253. Ley, R. (1991). The efficacy of breathing retraining and the centrality of hyperventilation in panic disorder: A reinterpretation of experimental findings. Behaviour Research and Therapy, 29, 301-304. Ley, R. (1993). Breathing retraining in the treatment of hyperventilatory complaints and panic disorder: A reply to Garssen, de Ruiter, and van Dyck. Clinical Psychology Review, 13, 393-408. Lidren, D.M., Watkins, P.L., Gould, R.A., Clum, G.A.,Asterino, M., & Tulloch, H.L. (1994). A comparison of bibliotherapy and group therapy in the treatment of panic disorder. Journal of Consulting and Clinical Psychology, 62, 865-869. Liebowitz, M., Fyer, A.J., Gorman, J.M., Campeas, R., Levin, A., Davies, S.R., Goetz, D., & Klein, D.F. (1986). Alprazolam in the treatment of panic disorders. Journal of Clinical Psychopharmacology, 6, 13-20. Liebowitz, M., Fyer, A.J., Gorman, J.M., Dillon, D., Appleby, LL., Levy, G., Anderson, S., Levitt, M., Palij, M., Davies, S.O., & Klein, D.F. (1984). Lactate provocation of panic attacks, I: Clinical and behavioral findings. Archives of General Psychiatry, 41, 764-770. Light, R.J., & Pillemer, D.B. (1984). Summing up: The science of reviewing research. Cambridge, MA: Harvard University Press. Linehan, M.M. (1993). Cognitive-behavioral treatment of borderline personality disorder. New York: Guilford. Lishman, W.A. (1987). Organic psychiatry (2nd ed.). Oxford: Blackwell Scientific Publications. Loerch, B., Graf-Morgenstern, M., Hautzinger, M., Schlegel, S., Hain, C., Sandmann, J., & Benkert, 0. (1999). Randomised placebo-controlled trial of moclobernide, cognitive-behavioural therapy and the combination in panic disorder with agoraphobia. British Journal of Psychiatry, 174, 205-212. Loranger, A.W., Sartorius, N., Andreoli, A., Berger, P., Buchleim, P., Channabasavanna, S.M., Coid, B., Dahl, A., Diekstra, R.F.W., Feguson, B., Jacobsberg, L.B., Mombour, W., Pull, C., Ono, Y., & Regier, D. (1994). The International Personality Disorder Examination: The World Health Organization/ Alcohol, Drug

REFERENCES471 Abuse, and Mental Health, Administration international pilot study of personality disorders. Archives of General Psychiatry, 51, 215-224. Louie, A.K., Lannon, R.A., Ritzick, E.A., Browne, D., Lewis, T.B., & Jones, R. (1996). Clinical features of cocaine-induced panic. Biological Psychiatry, 40, 938-940. Lum, L.C. (1981). Hyperventilation and anxiety state. Journal of the Royal Society of Medicine, 74, 1-4. Lum, L.C. (1994). Hyperventilation syndromes: Physiological considerations in clinical management. In B.H. Timmons & R. Ley (Eds.), Behavioral and psychological approaches to breathing disorders (pp. 113-123). New York: Plenum. Lydiard, R.B., & Brawman-Mintzer, 0. (1997). Panic disorder across the life span: A differential diagnostic approach to treatment resistance. Bulletin of the Menninger Clinic, 61 (2, suppl. A), A66-A94. Lydiard, R.B., Fossey, M.D., & Ballenger, J.C. (1991). Irritable bowel syndrome in patients with panic disorder. American Journal of Psychiatry, 148, 1614. Lydiard, R.B., Greenwald, S., Weissman, M.M., Johrison, J.,Drossman, D.A., & Ballenger, J.C. (1994). Panic disorder and gastrointestinal symptoms: Findings from the NIMH Epidemiologic Catchment Area project. American Journal of Psychiatry, 151, 64-70. Lydiard, R.B., Laraia, M.T., Ballenger, J.C., & Howell, E.F. (1987). Emergence of depressive symptoms in patients receiving alprazolam for panic disorder. American Journal of Psychiatry, 144, 664-665. Mackay, W., & Liddell, A. (1986). An investigation into the matching of specific agoraphobic characteristics with specific types of treatment. Behaviour Research and Therapy, 24, 361-364. Maddock, R.J., Carter, C.S., Blacker, K.H., Beitman, B.D., Krishnan, K.R.R., Jefferson, J.W., Lewis, C.P., & Liebowitz, M.R. (1993). Relationship of past depressive episodes to symptom severity and treatment response in panic disorder with agoraphobia. Journal of Clinical Psychiatry, 54, 88-95. Maffei, C., Fossati, A., Agostoni, I., Barraco, A., Bagnato, M., Donati, D., Namia, C., Novella, L., & Petrachi, M. (1997). Interrater reliability and internal consistency of the Structured Clinical Interview for DSM-IV Axis II personality disorders (SCID-II), version 2.0. Journal of Personality Disorders, 11, 279284. Magarian, G.J. (1982). Hyperventilation syndromes: Infrequently recognized common expressions of anxiety and stress. Medicine, 61, 219-236. Maier, W., & Buller, R. (1988). One-year follow-up of panic disorder: Outcome and prognostic factors. European Archives of Psychiatry and Neurological Science, 238, 105-109. Maier, W., Buller, R., Philipp, M., & Heuser, I. (1988). The Hamilton Anxiety Scale: Reliability, validity, and sensitivity to change in anxiety and depressive disorders. Journal of Affective Disorders, 14, 61-68. Maier, W., Roth, M., Argyle, N., Buller, R., Lavori, P., Brandon, S., & Benkert, 0. (1991). Avoidance behavior: A predictor of the efficacy of pharmacotherapy in panic disorder? European Archives of Psychiatry and Neurological Science, 241, 151-158. Malan, D. (1979). Individual psychotherapy and the science of psychodynamics. London: Butterworths. Maller, R.G., & Reiss, S. (1992). Anxiety sensitivity in 1984 and panic attacks in 1987. Journal of Anxiety Disorders, 6, 241-247. Manning, A.P., Thompson, W.G., Heaton, K.W., & Morris, A. (1978). Toward a positive diagnosis of the irritable bowel. British Medical Journal, 2, 653-654.

472

REFERENCES

Mannuzza, S., Fyer, A.J., Klein, D.F., & Endicott, J. (1986). Schedule for Affective Disorders-Lifetime, Anxiety (SADS-LA): Rationale and conceptual development. Journal of Psychiatric Research, 20, 317-325. Mannuzza, S., Fyer, A.J., Liebowitz, M.R., & Klein, D.F. (1990). Delineating the boundaries of social phobia: Its relationship to panic disorder and agoraphobia. Journal of Anxiety Disorders, 4, 42-59. Marchand, A., Goyer, L.R., Dupuis, G., & Mainguy, N. (1998). Personality disorders and the outcome of cognitive-behavioural treatment of panic disorder with agoraphobia. Canadian Journal of Behavioural Science, 30, 14-23. Margolin, G., Michelli, J., & Jacobson, N. (1988). Assessment of marital dysfunction. In A.S. Bellack & M. Hersen (Eds.), Behavioral assessment: A practical handbook (3rd ed.) (pp. 441-489). New York: Pergamon. Margraf, J., & Ehlers, A. (1988). Etiological models of panic-Psychophysiological and cognitive aspects. In R. Baker (Ed.), Panic disorder: Theory, research and therapy (pp. 205-231). Chichester, UK: Wiley. Margraf, J., Ehlers, A., & Roth, W.T. (1988). Mitra! valve prolapse and panic disorder: A review of their relationship. Psychosomatic Medicine, 50, 93-113. Margraf, J., & Schneider, S. (1991, November). Outcome and active ingredients of cognitive-behavioural treatments for panic disorder. Paper presented at the 25th annual meeting of the Association for Advancement of Behavior Therapy, New York, NY. Margraf, J., Taylor, C.B., Ehlers, A., Roth, W.T., & Agras, W.T. (1987). Panic attacks in the natural environment. Journal of Nervous and Mental Disease, 175, 558-565. Markowitz, J.S., Weissman, M.M., Ouellette, R., Lish, J.D., & Klerman, G.L. (1989). Quality of life in panic disorder. Archives of General Psychiatry, 46, 984-992. Marks, I.M. (1971). Phobic disorders four years after treatment: A prospective follow-up. British Journal of Psychiatry, 118, 683-688. Marks, I.M. (1978). Living with fear. New York: McGraw-Hill. Marks, I.M. (1981). Cure and care of neuroses: Theory and practice of behavioural psychotherapy. New York: Wiley. Marks, I.M. (1987). Fears, phobias, and rituals. New York: Oxford University Press. Marks, I.M., & Bebbington, P. (1976). Space phobia: Syndrome or agoraphobic variant? British Medical Journal, 2, 345-347. Marks, I.M., Gelder, M.G., & Edwards, J.G. (1968). Hypnosis v desensitization for phobias: A controlled prospective trial. British Journal of Psychiatry, 114, 1263-1274. Marks, I.M., Gray, S., Cohen, D., Hill, R., Mawson, D., Ramm, E., & Stem, R.S. (1983). Imipramine and brief therapist-aided exposure in agoraphobics having self-exposure homework. Archives of General Psychiatry, 40, 153-162. Marks, I.M., & Mathews, A.M. (1979). Brief standard self-rating for phobic patients. Behaviour Research and Therapy, 17, 263-267. Marks, I.M., Shaw, S., & Parkin, R. (1998). Computer-aided treatments of mental health problems. Clinical Psychology: Science and Practice, 5, 151-170. Marks, I.M., Swinson, R.P., Basoglu, M., Kuch, K., Noshirvani, H., O'Sullivan, G., Lelliott, P.T., Kirby, M., McNamee, G., Sengun, S., & Wickwire, K. (1993a). Alprazolam and exposure alone and combined in panic disorder with agoraphobia: A controlled study in London and Toronto. British Journal of Psychiatry, 162, 776-787. Marks, I.M.; Swinson, R.P., Basoglu, M., Noshirvani, H., Kuch, K., O'Sullivan, G., & Lelliott, P.T. (1993b). Reply to comment on the London/Toronto study. British Journal of Psychiatry, 162, 790-794.

REFERENCES473 Marlatt, G.A., & Gordon, J.R. (1985). Relapse prevention: Maintenance strategies in the treatment of addictive behaviors. New York: Guilford. Marriott, J.A. (1987). Brief intervention in acute panic anxiety-case review. Australian Journal of Clinical Hypnotherapy and Hypnosis, 8, 93-95. Marriott, P., Judd, F., Jefferys, D., & Burrows, G. (1989). Panic and phobic disorders. Part I: Problems associated with drug therapy. Current Therapeutics, April, 107-121. Marshall, W.L. (1985). The effects of variable exposure in flooding therapy. Behavior Therapy, 16, 117-135. Martin, M. (1985). Neuroticism as predisposition toward depression: A cognitive mechanism. Personality and Individual Differences, 6, 353-365. Martin, R.L., Cloninger, C.R., Guze, S.B., & Clayton, P.J. (1985). Mortality in a follow-up of 500 psychiatric outpatients. Archives of General Psychiatry, 42, 47-71. Matez, A. (1986). The rapid treatment of fear, panic and phobia disorders using hypnoanalysis ... with illustrative case history summaries. Journal of the American Academy of Medical Hypnoanalysts, 1, 68-87. Mathew, R.J., & Wilson, W.H. (1990). Anxiety and cerebral blood flow. American Journal of Psychiatry, 147, 838--849. Mathews, A.M., Gelder, M.G., & Johnston, D.W. (1981). Agoraphobia: Nature and treatment. New York: Guilford. Mathews, A.M., Johnston, D.W., Lancashire, M., Munby, M., Shaw, P.M., & Gelder, M.G. (1976). Imaginal flooding and exposure to real phobic situations: Treatment outcome with agoraphobic patients. British Journal of Psychiatry, 129, 362-371. Mathews, A.M., Johnston, D.W., Shaw, P.M., & Gelder, M.G. (1974). Process variables and the prediction of outcome in behaviour therapy. British Journal of Psychiatry, 125, 256-264. Mathews, A.M., Teasdale, J., Munby, M., Johnston, D., & Shaw, P. (1977). Ahomebased treatment program for agoraphobia. Behavior Therapy, 8, 915-924. Mathias, C.J., & Bannister, R. (1992). Investigation of autonomic disorders. In R. Bannister & C.J. Mathias (Eds.), Autonomic failure: A textbook of clinical disorders of the autonomic nervous system (pp. 255-290). Oxford: Oxford University Press. Matthews, G., Quinn, C.E.J., & Mitchell, K.J. (1998). Rock music, task-induced stress and simulated driving performance. In G.B. Grayson (Ed.), Behavioural research and road safety, VIII (pp. 155-167). Crowthome, UK: TRL. Mattick, RP., Andrews, G., Hadzi-Pavlovic, D., & Christensen, H. (1990). Treatment of panic and agoraphobia: An integrative review. Journal of Nervous and Mental Disease, 178, 567-576. Mattis, S.G., & Ollendick, T.H. (1997). Children's cognitive responses to the somatic symptoms of panic. Journal of Abnormal Child Psychology, 25, 47-57. Mavissakalian, M. (1985). Male and female agoraphobia: Are they different? Behaviour Research and Therapy, 23, 469-471. Mavissakalian, M. (1986). The Fear Questionnaire: A validity study. Behaviour Research and Therapy, 24, 83-85. Mavissakalian, M. (1990). The relationship between panic disorder/ agoraphobia and personality disorders. Psychiatric Clinics of North America, 13, 661-684. Mavissakalian, M. (1993). Combined behavioral therapy and pharmacotherapy of agoraphobia. Journal of Psychiatric Research, 27 (suppl. 1), 179-191. Mavissakalian, M., & Hamann, M.S. (1987). DSM-III personality disorder in agoraphobia: II. Changes with treatment. Comprehensive Psychiatry, 28, 356361.

47 4

REFERENCES

Mavissakalian, M., & Hamann, M.S. (1992). DSM-III personality characteristics of panic disorder with agoraphobia patients in stable remission. Comprehensive Psychiatry, 33, 305-309. Mavissakalian, M., & Michelson, L. (1986a). Agoraphobia: Relative and combined effectiveness of therapist-assisted in vivo exposure and imipramine. Journal of Clinical Psychiatry, 47, 117-122. Mavissakalian, M., & Michelson, L. (1986b). Two-year follow-up of exposure and imipramine treatment of agoraphobia. American Journal of Psychiatry, 143, 1106-1112. Mavissakalian, M., Michelson, L., & Dealy, R.S. (1983). Pharmacological treatment of agoraphobia: Imipramine versus imipramine with programmed practice. British Journal of Psychiatry, 143, 348-355. Mavissakalian, M.R., & Pere!, J.M. (1989). Imipramine dose-response relationship in panic disorder with agoraphobia. Archives of General Psychiatry, 46, 127-131. Mavissakalian, M.R., & Pere!, J.M. (1992). Clinical experiments in maintenance and discontinuation of imipramine therapy in panic disorder with agoraphobia. Archives of General Psychiatry, 49, 318-323. McCabe, B., & Tsuang, M.T. (1982). Dietary considerations in MAO inhibitor regimens. Journal of Clinical Psychiatry, 43, 178-181. McDonald, R., Sartory, G., Grey, S.J., Cobb, J., Stern, R., & Marks, I. (1979). The effects of self-exposure instructions on agoraphobic outpatients. Behaviour Research and Therapy, 17, 83-85. McLean, P.D., Woody, S., Taylor, S., & Koch, W.J. (1998). Comorbid panic disorder and major depression: Implications for cognitive-behavioral therapy. Journal of Consulting and Clinical Psychology, 66, 240-247. McMullin, R.E. (1986). Handbook of cognitive therapy techniques. New York: Norton. McNally, R.J. (1994). Panic disorder: A critical analysis. New York: Guilford. McNally, R.J., Amir, N., Louro, C.E., Lukach, B.M., Riemann, B.C., & Calamari,J.E. (1994). Cognitive processing of idiographic emotional information in panic disorder. Behaviour Research and Therapy, 32, 119-122. McNally, R.J., & Foa, E.B. (1987). Cognition and agoraphobia: Bias in the interpretation of threat. Cognitive Therapy and Research, 11, 567-581. McNally, R.J., Hornig, C.D., Hoffman, E.C., & Han, E.M. (1999). Anxiety sensitivity and cognitive biases for threat. Behavior Therapy, 30, 51-61. McNally, R.J., Hornig, C.D., Otto, M.W., & Pollack, M.H. (1997). Selective encoding of threat in panic disorder: Application of a dual priming paradigm. Behaviour Research and Therapy, 35, 543-549. McNally, R.J., & Louro, C.E. (1992). Fear of flying in agoraphobia and simple phobia: Distinguishing features. Journal of Anxiety Disorders, 6, 319-324. McNally, R.J., Riemann, B.C., & Kim, E. (1990). Selective processing of threat cues in panic disorder. Behaviour Research and Therapy, 28, 407--412. McNamee, G., O'Sullivan, G., Lelliott, P., & Marks, I. (1989). Telephone-guided treatment for housebound agoraphobics with panic disorder: Exposure vs. relaxation. Behavior Therapy, 20, 491--497. McPherson, F.M., Brougham, L., & McLaren, L. (1980). Maintenance of improvements in agoraphobic patients treated by behavioral methods in a four-year follow-up. Behaviour Research and Therapy, 18, 150-152. Meehl, P.E. (1960). The cognitive activity of the clinician. American Psychologist, 15, 19-27. • Meichenbaum, D.H. (1975). Self instructional methods. In F.H. Kanfer & A.P. Goldstein (Eds.), Helping people change (pp. 357-391). New York: Pergamon.

REFERENCES 475

Mellman, T.A., Leverich, G.S., Hauser, P., Kramlinger, K., Post, RM., & Uhde, T.W. (1992). Axis II pathology in panic and affective disorders: Relationship to diagnosis, course of illness, and treatment response. Journal of Personality Disorders, 6, 53-63. Mellman, T.A., & Uhde, T.W. (1986). Withdrawal syndrome with gradual tapering of alprazolam. American Journal of Psychiatry, 143, 1464-1466. Mellman, T.A., & Uhde, T.W. (1989). Sleep panic attacks: New clinical findings and theoretical implications. American Journal of Psychiatry, 146, 1204-1207. Michelson, L. (1986). Treatment consonance and response profiles in agoraphobia: The role of individual differences in cognitive, behavioral and physiological treatments. Behaviour Research and Therapy, 24, 263-275. Michelson, L., & Ascher, L.M. (1984). Paradoxical intention in the treatment of agoraphobia and other anxiety disorders. Journal of Behavior Therapy and Experimental Psychiatry, 15, 215-220. Michelson, L.K., June, K., Vives, A., Testa, S., & Marchione, N. (1998). The role of trauma and dissociation in cognitive-behavioral psychotherapy outcome and maintenance for panic disorder with agoraphobia. Behaviour Research and Therapy, 36, 1011-1050. Michelson, L.K., Marchione, K.E., Greenwald, M., Glanz, L., Testa, S., & Marchione, N.J. (1990). Panic disorder: Cognitive-behavioral treatment. Behaviour Research and Therapy, 28, 141-151. Michelson, L.K., Marchione, K.E., Greenwald, M., Testa, S., & Marchione, N.J. (1996). A comparative outcome and follow-up investigation of panic disorder with agoraphobia: The relative and combined efficacy of cognitive therapy, relaxation training, and therapist-assisted exposure. Journal of Anxiety Disorders, 10, 297-330. Michelson, L.K., Mavissakalian, M., & Marchione, K. (1985). Cognitive and behavioral treatments of agoraphobia: Clinical, behavioral, and psychophysiological outcomes. Journal of Consulting and Clinical Psychology, 53, 913-925. Michelson, L.K., Mavissakalian, M., & Marchione, K. (1988). Cognitive, behavioral, and psychophysiological treatments of agoraphobia. Behavior Therapy, 19, 97-120. Michelson, L.K., Mavissakalian, M., & Meminger, S. (1983). Prognostic utility of locus of control in treatment of agoraphobia. Behaviour Research and Therapy, 21, 309-313. Miller, J.J., Fletcher, K., & Kabat-Zinn, J. (1995). Three-year follow-up and clinical implications of a mindfulness meditation-based stress reduction intervention in the treatment of anxiety disorders. General Hospital Psychiatry, 17, 192-200. Miller, P.P., Brown, T.A., DiNardo, P., & Barlow, D.H. (1994). The experimental induction of depersonalization and derealization in panic disorder and nonanxious subjects. Behaviour Research and Therapy, 32, 511-519. Millon, T. (1981). Disorders of personality. DSM-III: Axis II. New York: Wiley. Millon, T. (1987). Manual for the Millon Clinical Multiaxial Inventory II (MCMI II). Minneapolis, MN: National Computer Systems. Millon, T. (1994). Millon Clinical Multiaxial Inventory III (MCMI III) manual. Minneapolis, MN: National Computer Systems. Milrod, B.L., Busch, F.N., Cooper, A.M., & Shapiro, T. (1997). Manual of panicfocused psychodynamic psychotherapy. Washington, DC: American Psychiatric Association. Milrod, B., & Shear, M.K. (1991). Dynamic treatment of panic disorder: A review. Journal of Nervous and Mental Disease, 179, 741-743.

476

REFERENCES

Milton, F., & Hafner, J. (1979). The outcome of behavior therapy for agoraphobia in relation to marital adjustment. Archives of General Psychiatry, 36, 807-811. Mineka, S., & Hendersen, R.W. (1985). Controllability and predictability in acquired motivation. Annual Review of Psychology, 36, 495-529. Mizes, J.S., & Crawford, J. (1986). Normative values on the Marks and Mathews Fear Questionnaire: A comparison as a function of age and sex. Journal of Psychopathology and Behavioral Assessment, 8, 253-262. Monteiro, W., Marks, I.M., & Ramm, E. (1985). Marital adjustment and treatment in agoraphobia. British Journal of Psychiatry, 146, 383-390. • Moran, C. (1986). Depersonalization associated with marijuana use. British Journal of Medical Psychology, 59, 187-196. Moras, K., DiNardo, P.A., & Barlow, D.H. (1992). Distinguishing anxiety and depression: Reexamination of the reconstructed Hamilton scales. Psychological Assessment, 4, 224--227. Morey, L.C. (1991). Personality Assessment Inventory. Odessa, FL: Psychological Assessment Resources. Morris, A., Baker, B., Devins, G.M., & Shapiro, C.M. (1997). Prevalence of panic disorder in cardiac outpatients. Canadian Journal of Psychiatry, 42, 185-190. Mullaney, J.A., & Trippett, C.J. (1979). Alcohol dependence and phobias: Clinical description and relevance. British Journal of Psychiatry, 135, 565-573. Munby, M., & Johnston, D.W. (1980). Agoraphobia: The long-term follow-up of behavioural treatment. British Journal of Psychiatry, 137, 418-427. Munjack, D.J., Crocker, B., Cabe, D., Brown, R., Usigli, R., Zulueta, A., McManus, M., McDowell, D., Palmer, R., & Leonard, M. (1989). Alprazolam, propranolol and placebo in the treatment of panic disorder and agoraphobia with panic attacks. Journal of Clinical Psychopharmacology, 9, 22-27. Murphy, M.T., Michelson, L.K., Marchione, K., Marchione, N., & Testa, S. (1998). The role of self-direct in vivo exposure in combination with cognitive therapy, relaxation training, or therapist-assisted exposure in the treatment of panic disorder with agoraphobia. Journal of Anxiety Disorders, 12, 117-138. Murphy, S., Owen, R., & Tyrer, P. (1989). Comparative assessment of efficacy and withdrawal symptoms after 6 and 12 weeks treatment with diazepam or buspirone. British Journal of Psychiatry, 154, 529-534. Murrey, G.J., Bolen, J., Miller, N., Simensted, K., Robbins, M., & Truskowski, F. (1993). History of childhood sexual abuse in women with depressive and anxiety disorders: A comparative study. Journal of Sex Education and Therapy, 19, 13-19. Muskin, P., & Fyer, A.J. (1981). Treatment of panic disorder. Journal of Clinical Psychopharmacology, 1, 81-90. Nagler, R., & Spiro, H.M. (1961). Serial esophageal motility studies in asymptomatic young subjects. Gastroenterology, 41, 371-379. Nagy, L.M., Krystal, J.H., Charney, D.S., Merikangas, K.R., & Woods, S.W. (1993). Long-term outcome of panic disorder after short-term imipramine and behavioral group treatment: 2.9-year naturalistic follow-up study. Journal of Clinical Psychopharmacology, 13, 16-24. Nagy, L.M., Krystal, J.H., Woods, S.W., & Charney, D.S. (1989). Clinical and medication outcome after short-term alprazolam and behavioral group treatment of panic disorders: 2.5-year naturalistic follow-up study. Archives of General Psychiatry, 46, 993-999. Naifeh, K.H. (1994). Basic anatomy and physiology of the respiratory system and the autonomic nervous system. In B.H. Timmons & R. Ley (Eds.), Behavioral and psychological approaches to breathing disorders (pp. 17-45). New York: Plenum.

REFERENCES477 Nathan, R.G., Robinson, D., Cherek, D.R., Sebastian, C.S., & Hack, M. (1986). Alternative treatments for withdrawing long-term benzodiazepine users: A pilot study. International Journal of the Addictions, 21, 195-211. Neeleman, J. (1992). The therapeutic potential of positive reframing in panic. European Psychiatry, 7, 135-139. Nelles, W.B., & Barlow, D.H. (1988). Do children panic? Clinical Psychology Review, 8, 359-372. Neron, S., Lacroix, D., & Chaput, Y. (1995). Group vs individual cognitive behaviour therapy in panic disorder: An open clinical trial with a six month followup. Canadian Journal of Behavioural Science, 27, 379-392. Newman, M.G., Kenardy, J., Herman, S., & Taylor, C.B. (1996). The use of handheld computers as an adjunct to cognitive-behavior therapy. Computers in Human Behavior, 12, 135-143. Newman, M.G., Kenardy, J., Herman, S., & Taylor, C.B. (1997). Comparison of palmtop-computer-assisted brief cognitive-behavioral treatment to cognitivebehavioral treatment for panic disorder. Journal of Consulting and Clinical Psychology, 65, 178-183. Nietzel, M.T., Bernstein, D.A., & Russell, R.L. (1988). Assessment of anxiety and fear. In A.S. Bellack & M. Hersen (Eds.), Behavioral assessment (3rd ed.) (pp. 280-312). New York: Pergamon. Nisbett, R., & Ross, L. (1980). Human inference. Englewood Cliffs, NJ: Prentice-Hall. Norton, G.R. (1988). Panic attack questionnaire. In M. Hersen & A. Bellack (Eds.), Dictionary of behavioral assessment techniques (pp. 332-334). New York: Pergamon. Norton, G.R., Cox, B.J., & Malan, J. (1992). Nonclinical panickers: A critical review. Clinical Psychology Review, 12, 121-139. Norton, G.R., Malan, J., Cairns, S.L., Wozney, K.A., & Broughton, R. (1989). Factors influencing drinking behavior in alcoholic panickers and nonpanickers. Behaviour Research and Therapy, 27, 167-171. Noyes, R., Anderson, D.J., Ghoneim, M.N., & Hinricks, J.V. (1984). Diazepam and propranolol in panic disorder and agoraphobia. Archives of General Psychiatry, 41, 287-292. Noyes, R., Clancy, J., Woodman, C., Holt, C.S., Suelzer, M., Christiansen, J., & Anderson, D.J. (1993). Environmental factors related to the outcome of panic disorder. Journal of Nervous and Mental Disease, 181, 529-538. Noyes, R., Cook, B., Garvey, M., & Summers, R. (1990a). Reduction of gastrointestinal symptoms with treatment for panic disorder. Psychosomatics, 31, 75-79. Noyes, R., DuPont, R.L., Pecknold, J.C., Rifkin, A., Rubin, R.T., Swinson, R.P., Ballenger, J.C., & Burrows, G.D. (1988). Alprazolam in panic disorder and agoraphobia: Results from a multicenter trial. II. Patient acceptance, side effects, and safety. Archives of General Psychiatry, 45, 423-428. Noyes, R., Garvey, M.J., & Cook, B.L. (1989a). Follow-up study of patients with panic disorder and agoraphobia with panic attacks treated with tricyclic antidepressants. Journal of Affective Disorders, 16, 249-257. Noyes, R., Garvey, M.J., Cook, B.L., & Samuelson, L. (1989b). Problems with tricyclic antidepressant use in patients with panic disorder or agoraphobia: Results of a naturalistic follow-up study. Journal of Clinical Psychiatry, 50, 163-169. Noyes, R., Garvey, M.J., Cook, B.L., & Suelzer, M. (1991). Controlled discontinuation of benzodiazepine for patients with panic disorder. American Journal of Psychiatry, 148, 517-523. Noyes, R., Kathol, R.G., Fisher, M.M., Phillips, B.M., Suelzer, M.T., & Woodman, C.L. (1994). Psychiatric comorbidity among patients with hypochondriasis. General Hospital Psychiatry, 16, 78-87.

478

REFERENCES

Noyes, R., Reich, J., Christiansen, J., Suelzer, M., Pfohl, B., & Coryell, W.A. (1990b ). Outcome of panic disorder: Relationship to diagnostic subtypes and comorbidity. Archives of General Psychiatry, 47, 809-818. Nutzinger, D.O., & Zapotoczky, H.G. (1985). The influence of depression on the outcome of cardiac phobia (panic disorder). Psychopathology, 18, 155-162. Oatley, K., & Hodgson, D. (1987). Influence of husbands on the outcome of their agoraphobic wive's therapy. British Journal of Psychiatry, 150, 380-386. Obsessive Compulsive Cognitions Working Group (1997). Cognitive assessment of obsessive-compulsive disorder. Behaviour Research and Therapy, 35, 667-681. Oehrberg, S., Christiansen, P.E., & Behnke, K. (1995). Paroxetine in the treatment of panic disorder: A randomized double-blind placebo-controlled study. British Journal of Psychiatry, 167, 374-379. Oei, T.P.S., Llamas, M., & Evans, L. (1997). Does concurrent drug intake affect the long-term outcome of group cognitive behaviour therapy in panic disorder with or without agoraphobia? Behaviour Research and Therapy, 36, 851857. Oei, T.P.S., Moylan, A., & Evans, L. (1991). Validity and clinical utility of the Fear Questionnaire for anxiety-disorder patients. Psychological Assessment, 3, 391-397. Oei, T.P.S., Wanstall, K., & Evans, L. (1990). Sex differences in agoraphobia. Journal of Anxiety Disorders, 4, 317-324. Ogles, B.M., Lambert, M.J., Weight, D.G., & Payne, I.R. (1990). Agoraphobia outcome measurement: A review and meta-analysis. Psychological Assessment, 2, 317-325. Ollendick, T.H. (1995). Cognitive behavioral treatment of panic disorder and agoraphobia in adolescents: A multiple baseline design analysis. Behavior Therapy, 26, 517-531. Ollendick, T.H. (1997, August). Panic disorder in children and adolescents: New developments, new directions. Paper presented at the annual meeting of the American Psychological Association, Chicago, IL. Ollendick, T.H., Mattis, S.G., & King, N.J. (1994). Panic in children and adolescents: A review. Journal of Child Psychology and Psychiatry, 35, 113-134. Onyett, S.R., & Turpin, G. (1988). Benzodiazepine withdrawal in primary care: A comparison of behavioural group training and individual sessions. Behavioural Psychotherapy, 16, 297-312. O'Rourke, D., Fahy, T.J., Brophy, J.,& Prescott, P. (1996). The Galway study of panic .. disorder III. Outcome at 5 to 6 years. British Journal of Psychiatry, 168, 462-469. Ost, L.-G. (1987). Applied relaxation: Description of a coping technique and review .. of controlled studies. Behaviour Research and Therapy, 25, 397-409. Ost, L.-G. (1988). Applied relaxation in the treatment of panic disorder. Behaviour Research and Therapy, 26, 13-22. Ost, L.-G. (1989). A maintenance program for behavioral treatment of anxiety .. disorders. Behaviour Research and Therapy, 27, 123-130. Ost, L.-G., Jerremalm, A., & Jansson, L. (1984). Individual response patterns and the effects of different behavioral methods in the treatment of agoraphobia. Behaviour Research and Therapy, 22, 697-707. Ost, L.-G., & Westling, B.E. (1995). Applied relaxation vs cognitive behavior therapy in the treatment of panic disorder. Behaviour Research and Therapy, 33, 145-158. Ost, L.-G., Westling, B.E., & Hellstrom, K. (1993). Applied relaxation, exposure in vivo and cognitive methods in the treatment of panic disorder with agoraphobia. Behaviour Research and Therapy, 31, 383-394.

REFERENCES479

Ottaviani, R., & Beck, A.T. (1987). Cognitive aspects of panic disorders. Journal of Anxiety Disorders, 1, 15-28. Otto, M.W., Jones, J.C., Craske, M.G., & Barlow, D.H. (1996a). Stopping anxiety medication: Panic control therapy for benzodiazepine discontinuation. San Antonio, TX: Psychological Corporation. Otto, M.W., Pollack, M.H., & Barlow, D.H. (1995). Stopping anxiety medication: A workbook for patients wanting to discontinue benzodiazepine treatment. New York: Graywind. Otto, M.W., Pollack, M.H., Penava, S.J., & Zucker, B.G. (1999). Group cognitivebehavior therapy for patients failing to respond to pharmacotherapy for panic disorder: A clinical case series. Behaviour Research and Therapy, 37, 763-770. Otto, M.W., Pollack, M.H., & Sabatino, S.A. (1996b). Maintenance of remission following cognitive behavior therapy for panic disorder: Possible deleterious effects of concurrent medication treatment. Behavior Therapy, 27, 473-482. Otto, M.W., Pollack, M.H., Sachs, G.S., Reiter, S.R., Meltzer-Brody, S., & Rosenbaum, J.F. (1993). Discontinuation of benzodiazepine treatment: Efficacy of cognitive-behavior therapy for patients with panic disorder. American Journal of Psychiatry, 150, 1485-1490. Otto, M.W., & Reilly-Harrington, N.A. (1999). The impact of treatment on anxiety sensitivity. In S. Taylor (Ed.), Anxiety sensitivity: Theory, research, and treatment of the fear of anxiety (pp. 321-336). Mahwah, NJ: Erlbaum. Overholser, J.C. (1993a). Elements of the Socratic method: I. Systematic questioning. Psychotherapy, 30, 67-74. Overholser, J.C. (1993b). Elements of the Socratic method: II. Inductive reasoning. Psychotherapy, 30, 75-85. Overholser, J.C. (1994). Elements of the Socratic method: III. Universal definitions. Psychotherapy, 31, 286-293. Overholser, J.C. (1995). Elements of the Socratic method: IV. Disavowal of knowledge. Psychotherapy, 32, 283-292. Overholser, J.C. (1996). Elements of the Socratic method: V. Self-improvement. Psychotherapy, 33, 549-559. Paradis, C.M., Friedman, S., Lazar, R.M., Grubea, J., & Kesselman, M. (1992). Use of a structured interview to diagnose anxiety disorders in a minority population. Hospital and Community Psychiatry, 43, 61-64. Pecknold, J.C., Luthe, L., Munjack, D., & Alexander, P. (1994). A double-blind, placebo-controlled, multicenter study with alprazolam and extended-release alprazolam in the treatment of panic disorder. Journal of Clinical Psychopharmacology, 14, 314--321. Pecknold, J.C., Swinson, R.P., Kuch, K., & Lewis, C.P. (1988). Alprazolam in panic disorder and agoraphobia: Results from a multicenter trial: III. Discontinuation effects. Archives of General Psychiatry, 45, 429-436. Penava, S.J., Otto, M.W., Maki, K.M., & Pollack, M.H. (1998). Rate of improvement during cognitive-behavioral group treatment of panic disorder. Behaviour Research and Therapy, 36, 665-673. Pennebaker, J.W. (1982). The psychology of physical symptoms. New York: SpringerVerlag. Pennebaker, J.W., Burnam, M.A., Schaeffer, M.A., & Harper, D.C. (1977). Lack of control as a determinant of perceived physical sensations. Journal of Personality and Social Psychology, 35, 167-174. Perna, G., Bertani, A., Politim, E., Colombo, G., & Bellodi, L. (1997). Asthma and panic attacks. Biological Psychiatry, 42, 625-630.

480

REFERENCES

Persons, J.B. (1989). Cognitive therapy in practice: A caseformulation approach. New York: Norton. Persons, J.B. (1991). Psychotherapy outcome studies do not accurately represent current models of psychotherapy. American Psychologist, 46, 99-106. Persons, J.B., & Bertagnolli, A. (1999). Inter-rater reliability of cognitive-behavioral case formulations of depression: A replication. Cognitive Therapy and Research, 23, 271-283. Persons, J.B., Mooney, K.A., & Padesky, C.A. (1995). Interrater reliability of cognitive-behavioral case formulations. Cognitive Therapy and Research, 19, 21-34. Persons, J.B., & Tompkins, M.A. (1997). Cognitive-behavioral case formulation. In T.D. Eells (Ed.), Handbook of psychotherapy case formulation (pp. 314-339). New York: Guilford. Peterson, R.A., & Reiss, S. (1987). Test Manual for the Anxiety Sensitivity Index. Orland Park, IL: International Diagnostic Systems. Peterson, R.A., & Reiss, S. (1992). Anxiety Sensitivity Index Manual (2nd ed.). Worthington, OH: International Diagnostic Systems. Petursson, H., & Lader, M.H. (1981). Withdrawal from long-term benzodiazepine treatment. British Medical Journal, 283, 643-645. Pfohl, B., Blum, N., Zimmerman, M., & Stangl, D. (1989). Structured Interview for DSM-III-R Personality Disorders, SIDP-R. Iowa City, IA: Department of Psychiatry, University of Iowa. Picornell-Darder, I., Carrasco, J.L., Rostain, J.C., & Naquet, R. (1978). A study on the Valsalva manoeuvre in young healthy subjects. Electroencephalography and Clinical Neurophysiology, 45, 648-654. Pohl, R., Yergani, V., Balon, R., & Lycaki, H. (1988). The jitteriness syndrome in panic disorder patients treated with antidepressants. Journal of Clinical Psychiatry, 49, 100-104. Pollack, M.H., Kradin, R., Otto, M.W., Worthington, J., Gould, R., Sabatino, S.A., & Rosenbaum, J.F. (1996). Prevalence of panic in patients referred for pulmonary function testing at a major medical center. American Journal of Psychiatry, 153, 110-113. Pollack, M.H., Otto, M.W., Kaspi, S.P., Hammerness, P.G., & Rosenbaum, J.F. (1994a). Cognitive behavior therapy for treatment-refractory panic disorder. Journal of Clinical Psychiatry, 55, 200-205. Pollack, M.H., Otto, M.W., Rosenbaum, J.F., & Sachs, G.S. (1992). Personality disorders in patients with panic disorder: Association with childhood anxiety disorders, early trauma, comorbidity and chronicity. Comprehensive Psychiatry, 33, 78-83. Pollack, M.H., Otto, M.W., Rosenbaum, J.F., Sachs, G.S., O'Neil, C., Asher, R., & Meltzer-Brody, S. (1990). Longitudinal course of panic disorder: Findings from the Massachusetts General Hospital Naturalistic Study. Journal of Clinical Psychiatry, 51 (12, suppl. A), 12-16. Pollack, M.H., Otto, M.W., Sachs, G.S., Leon, A., Shear, M.K., Deltito, J.A., Keller, M.B., & Rosenbaum, J.F. (1994b). Anxiety psychopathology predictive of outcome in patients with panic disorder and depression treated with imipramine, alprazolam and placebo. Journal of Affective Disorders, 30, 273-281. Pollack, M.H., Otto, M.W., Tesar, G.E., Cohen, L.S., Meltzer-Brody, S., & Rosenbaum, J.F. (1993). Long-term outcome after acute treatment with alprazolam or clonazepam for panic disorder. Journal of Clinical Psychopharmacology, 13, 257-263. Pollack, M.H., & Smaller, J.W. (1995). The longitudinal course and outcome of panic disorder. Psychiatric Clinics of North America, 18, 785-801.

REFERENCES481 Pollard, C.A., & Lewis, L.M. (1989). Managing panic attacks in emergency patients. Journal of Emergency Medicine, 7, 547-552. Pollard, C.A., Obermeier, H.J., & Cox, G.L. (1987). Inpatient treatment of complicated agoraphobia and panic disorder. Hospital and Community Psychiatry, 38, 951-958. Pollard, C.A., Tait, RC., Meldrum, D., Dubinsky, I.H., & Gall, J.S. (1996). Agoraphobia without panic: Case illustrations of an overlooked syndrome. Journal of Nervous and Mental Disease, 184, 61-62. Pollard, H.J., & Pollard, C.A. (1993). Follow-up study of an inpatient program for complicated agoraphobia and panic disorder: A preliminary report. Anxiety Disorders Practice Journal, l, 37-40. Porzelius, J., Vest, M., & Nochomovitz, M. (1992). Respiratory function, cognitions, and panic in chronic obstructive pulmonary patients. Behaviour Research and Therapy, 30, 75-77. Pribor, E.F., & Dinwiddie, S.H. (1992). Psychiatric correlates of incest in childhood. American Journal of Psychiatry, 149, 52-56. Pyke, RE., & Kraus, M. (1988). Alprazolam in the treatment of panic attack patients with and without major depression. Journal of Clinical Psychiatry, 49, 66-68. Quitkin, F.M., Rifkin, A., Kaplan, J., & Klein, D.F. (1972). Phobic anxiety syndrome complicated by drug dependence and addiction: A treatable form of drug abuse. Archives of General Psychiatry, 27, 159-162. Rachman, S. (1984). Agoraphobia: A safety-signal perspective. Behaviour Research and Therapy, 22, 59-70. Rachman, S. (1985). Foreword. In I.D. Turkat (Ed.), Behavioral case formulation (p. ix). New York: Plenum. Rachman, S., & Bichard, S. (1988). The overprediction of fear. Clinical Psychology Review, 8, 303-312. Rachman, S., & Taylor, S. (1993). Analyses of claustrophobia. Journal of Anxiety Disorders, 7, 281-291. Radomsky, A.S., Rachman, S., Teachman, B.A., & Freeman, W.S. (1998). Why do episodes of panic stop? Journal of Anxiety Disorders, 12, 263-270. Raj, A., & Sheehan, D.V. (1987). Medical evaluation of panic attacks. Journal of Clinical Psychiatry, 48, 309-313. Raj, B.A., Corvea, M.H., & Dagon, E.M. (1993). The clinical characteristics of panic disorder in the elderly: A retrospective study. Journal of Clinical Psychiatry, 54, 150-155. Rapee, RM. (1985). A case of panic disorder treated with breathing retraining. Journal of Behavior Therapy and Experimental Psychiatry, 16, 63-65. Rapee, RM. (1986). Differential response to hyperventilation in panic disorder and generalized anxiety disorder. Journal of Abnormal Psychology, 95, 24c--28. Rapee, RM., & Barlow, D.H. (1991). The cognitive-behay:ioral treatment of panic attacks and agoraphobic avoidance. In J.R Wallser;G.R Norton, & C.A. Ross (Eds.), Panic disorder and agoraphobia: A comprehensive guide for the practitioner (pp. 252-305). Pacific Grove, CA: Brooks/C6le. Rapee, RM., Craske, M.G., & Barlow, D.W.(1990). Subject-described features of panic attacks using self-monitoring. Jof!n!!_lof Anxiety Disorders, 4, 171-181. Rapee, RM., Craske, M.G., Brown, T.1y, & Barlow, D.H. (1996). Measurement of perceived control over anxiety-relat~d events. Behavior Therapy, 27, 279-293. Rapp, M.S., Thomas, M.R, & Reyes, J;{C. (1983). Mega-doses of behaviour therapy for treatment-resistant agoraphobfcs. Canadian Journal of Psychiatry, 28, 105108. • I

482

/ __-

REFERENCES

Rathus, J.H., & Sanderson, W.C. (1996). Cognitive behavioral treatment of panic disorder in elderly adults: Two case studies. Journal of Cognitive Psychotherapy, 10, 271-280. Rathus, J.H., Sanderson, W.C., Miller, A.L., & Wetzler, S. (1995). Impact of personality functioning on cognitive behavioral treatment of panic disorder: Preliminary report. Journal of Personality Disorders, 9, 160-168. Ravaris, C.L., Friedman, M.J., Hauri, P.J., & McHugo, G.J. (1991). A controlled study of alprazolam and propranolol in panic-disordered and agoraphobic outpatients. Journal of Clinical Psychopharmacology, 11, 344-350. Rees, C.S., Richards, J.C., & Smith, L.M. (1998). Medical utilisation and costs in panic disorder: A comparison with social phobia. Journal of Anxiety Disorders, 12, 421--435. Reich, J.1--1. (1988). DSM-III personality disorders and the outcome of treated panic disorder. American Journal of Psychiatry, 145, 1149-1152. Reich, J.H., Noyes, R., Coryell, W., & O'Gorman, T. (1986). The effect of state anxiety on personality measurement. American Journal of Psychiatry, 143, 760-763. Reich, J.H., Noyes, R., Hirschfeld, R., Coryell, W., & O'Gorman, T. (1987). State and personality in depressed and panic patients. American Journal of Psychiatry, 144, 181-187. Reidenberg, M.M., & Lowenthal, D.T. (1968). Adverse non-drug reactions. New England Journal of Medicine, 279, 678-679. Reiss, S. (1991). Expectancy theory of fear, anxiety, and panic. Clinical Psychology Review, 11, 141-153. Reiss, S. (1999). The sensitivity theory of aberrant motivation. In S. Taylor (Ed.), Anxiety sensitivity: Theory, research, and treatment of the fear of anxiety (pp. 35-58). Mahwah, NJ:Erlbaum. Reiss, S., & McNally, R.J. (1985). The expectancy model of fear. In S. Reiss & R.R. Bootzin (Eds.), Theoretical issues in Behavior therapy (pp. 107-121). New \'.ork: Academic Press. / Reiss, S., Peterson, R.A., Gursky, M., & McNally, R.J. (1986). Anxiety sensitivity, anxiety frequency, and the prediction of fearfulness. Behaviour Research and Therapy, 24, 1-8. Rescorla, R.A. (1988). Pavlovian conditioning: It's not what you think it is. American Psychologist, 43, 151-160. Resick, P.A., & Schnicke, M.K. (1993). Cognitive processing therapy for rape victims: A treatment manual. Newbury Park, CA: Sage. Rice, D.P., & Miller, L.S. (1993). The economic burden of mental disorders. Advances in Health Economics and H':!fli:hServices Research, 14, 37-53. Rickels, K. (1990a). Discontinuation,,studies with alprazolam. Journal of Psychiatric Research, 24 (suppl. 1), 57-58/ Rickels, K. (1990b). Buspir~ in clinical practice. Journal of Clinical Psychiatry, 51 (suppl. 9), 51-54. Rickels, K., Schweizer .E., Csanalosi, I., Case, G., & Chung, H. (1988). Longterm treatment o anxiety and risk of withdrawal: Prospective comparison of chlorazepate )Uid buspirone. Archives of General Psychiatry, 45, 444--450. Rifkin, A., Klefu, D.F., Dillon, D., & Levitt, M. (1981). Blockade by imipramine or desipra:mfue of panic induced by sodium lactate. American Journal of Psychiatry, 6-677 . ., Kraaimaat, F., de Ruiter, C., & Garssen, B. (1992). A follow-up study on short-term treatment of agoraphobia. Behaviour Research and Therapy, 30, 63-66.

REFERENCES483 Riley, W.T., McCormick, M.G.F., Simon, E.M., Stack, K., Pushkin, Y., Overstreet, M.M., Carmona, J.J., & Magakian, C. (1995). Effects of alprazolam dose on the induction and habituation processes during behavioral panic induction treatment. Journal of Anxiety Disorders, 9, 217-227. Riskind, J.H., Beck, A.T., Brown, G., & Steer, RA. (1987). Taking the measure of anxiety and depression: Validity of the reconstructed Hamilton scales. Journal of Nervous and Mental Disease, 175, 474-479. Roberts, D.K., & MacKay, K.A. (1987). Microhemorrhagic maculopathy associated with aerobic exercise. Journal of the American Optometric Association, 58, 415-418. Roberts, W.W. (1960). Normal and abnormal depersonalization. Journal of Mental Science, 106, 478-493. Robins, L.N., Wing, J., Wittchen, H.-U., Helzer, J.E., Babor, T.F., Burke, J., Farmer, A., Jablensky, A., Pickens, R., Regier, D.A., Sartorius, N., & Towle, L.H. (1989). The Composite International Diagnostic Interview: An epidemiological instrument suitable for use in conjunction with different diagnostic systems and different culturEJ$.Archives of General Psychiatry, 45, 1069-1077. Robinson, L., w.a·ker, J.R., & Anderson, D. (1992). Cognitive-behavioural treatment of panic disor er during pregnancy and lactation. Canadian Journal of Psychiatry, 37, 623-62 . Roelofs, S. (198 . Hyperventilation, anxiety, craving for alcohol: A subacute alcohol withdrfwal syndrome. Alcohol, 2, 501-505. Rosenbaum, J.F.,i Biederman, J., Gersten, M., Hirshfeld, D.R., Meminger, S.R., Herman, J.B., ]l(agan, J., Reznick, S., & Snidman, N. (1988). Behavioral inhibition in childrep of parents with panic disorder and agoraphobia: A controlled study. Archive!; of General Psychiatry, 40, 1065-1069. Rosenbaum, J.H, & Pollack, M.H. (1991). Anxiety. In N. Cassem (Ed.), The Massachusetts General Handbook of General Hospital Psychiatry (pp. 159-190). St. Louis, MO: Mosby. Rosenberg, R., Bech, P., Mellergard, M., & Ottoson, J.-O. (1991a). Secondary depression in panic disorder: An indication of severity with a weak effect on outcome in alprazolam and imipramine treatment. Acta Psychiatrica Scandinavica, (suppl. 365), 39-45. Rosenberg, R., Bech, P., Mellergard, M., & Ottoson, J.-O. (1991b). Alprazolam, imipramine and placebo treatment of panic disorder: Predicting therapeutic response. Acta Psychiatrica Scandinavica, (suppl. 365), 46-52. Rosenthal, R. (1991). Meta-analytic procedures for social research (rev. ed.). Newbury Park, CA: Sage. Rowe, D.C. (1994). The limits offamily influence: Genes, experience, and Behavior. New York: Guilford. Roy-Byrne, P.P., Geraci, M., & Uhde, T. (1986). Life events and the onset of panic disorder. American Journal of Psychiatry, 143, 1424-1427. Roy-Byrne, P.P., Sullivan, M.D., Cowley, D.S., & Ries, R.K. (1993). Adjunctive treatment of benzodiazepine discontinuation syndromes: A review. Journal of Psychiatric Research, 27 (suppl. 1), 143-153. Rubinstein, B., & Erlandson, S.I. (1991). A stomatognathic analysis of patients with disabling tinnitus and craniomandibular disorders (CMD). British Journal of Audiology, 25, 77-84. Rudd, M.D., Dahm, P.F., & Rajab, M.H. (1993). Diagnostic comorbidity in persons with suicidal ideation and behavior. American Journal of Psychiatry, 150, 928-934. Salkovskis, P.M. (1989). Somatic problems. In K. Hawton, P.M. Salkovskis, J. Kirk, & D.M. Clark (Eds.), Cognitive behaviour therapy for psychiatric problems: A practical guide (pp. 235-276). Oxford: Oxford University Press.

484

REFERENCES

Salkovskis, P.M. (1991). The importance of behaviour in the maintenance of panic and anxiety: A cognitive account. Behavioural Psychotherapy, 19, 6-19. Salkovskis, P.M. (1999). Understanding and treating obsessive-compulsive disorder. Behaviour Research and Therapy, 37, S29-S52. Salkovskis, P.M., & Clark, D.M. (1986). Cognitive and physiological processes in the maintenance and treatment of panic attacks. In I. Hand & H.-U. Wittchen (Eds.), Panic and phobias (pp. 90-103). New York: Springer-Verlag. Salkovskis, P.M., & Clark, D.M. (1991). Cognitive therapy for panic attacks. Journal of Cognitive Psychotherapy, 5, 215-226. Salkovskis, P.M., Clark, D.M., & Gelder, M.G. (1996). Cognition-behaviour links in the persistence of panic. Behaviour Research and Therapy, 34, 453-458. Salkovskis, P.M., Clark, D.M., & Hackmann, A. (1991). Treatment of panic attacks using cognitive therapy without exposure or breathing retraining. Behaviour Research and Therapy, 29, 161-166. Salkovskis, P.M., Clark, D.M., Hackmann, A., Wells, A., & Gelder, M. (1997, September). Panic disorder with severe agoraphobia: Comparing exposure with a cognitive or behavioural emphasis. Paper presented at the XXVII Congress of the European Association for Behavioural & Cognitive Therapies, Venice, Italy. Salkovskis, P.M., Clark, D.M., Hackmann, A., Wells, A., & Gelder, M. (1999). An experimental investigation of the role of safety behaviours in the maintenance of panic disorder with agoraphobia. Behaviour Research and Therapy, 37, 559-574. Salkovskis, P.M., Jones, D.R.O., & Clark, D.M. (1986a). Respiratory control in the treatment of panic attacks: Replication and extension with concurrent measurement of behaviour and pCO 2 . British Journal of Psychiatry, 148, 526-532. Salkovskis, P.M., Warwick, H.M.C., & Clark, D.M. (1986b). A demonstration of acute hyperventilation during naturally occurring panic attacks. Behaviour Research and Therapy, 24, 91-94. Sanchez-Craig, M., Cappell, H., Busto, U., & Kay, G. (1987). Cognitive-behavioural treatment of benzodiazepine dependence: A comparison of gradual versus abrupt cessation of drug intake. British Journal of Addiction, 82, 1317-1327. Sandberg, L., & Siris, S.G. (1987). 'Panic disorder' in schizophrenia. Journal of Nervous and Mental Disease, 175, 627-628. Sanderson, W.C. (1987). The influence of perceived control on panic attacks induced via inhalation of 5% CO 2 enriched air. Unpublished doctoral dissertation, State University of New York at Albany. Sanderson, W.C., Rapee, RM., & Barlow, D.H. (1989). The influence of an illusion of control on panic attacks induced via inhalation of 5.5% carbon dioxideenriched air. Archives of General Psychiatry, 46,157-162. Sanderson, W.C., & Wetzler, S. (1993). Observations on the cognitive behavioral treatment of panic disorder: Impact of benzodiazepines. Psychotherapy, 30, 125-132. Sanderson, W.C., Wetzler, S., & Anis, G.M. (1994). Alprazolam blockade of COr provoked panic in patients with panic disorder. American Journal of Psychiatry, 151, 1220-1222. Saper, Z., & Brasfield, C.R. (1998). Two-phase treatment of panic disorder and posttraumatic stress disorder with associated personality features resulting from childhood abuse: Case study. Journal of Behavior Therapy and Experimental Psychiatry, 29, 171-178. Sarason, LG., Johnson, J.H., & Siegel, J.M. (1978). Assessing the impact of life changes: Development of the Life Experiences Survey. Journal of Consulting and Clinical Psychology, 46, 932-946.

REFERENCES485

Sartory, G., Master, D., & Rachman, S. (1989). Safety-signal therapy in agoraphobics: A preliminary test. Behaviour Research and Therapy, 27, 205--209. Sartory, G., & Olajide, D. (1989). Vagal innervation techniques in the treatment of panic disorder. Behaviour Research and Therapy, 26, 431--434. Saunders, B.E., Villeponteaux, L.A., Lipovsky, J.A., Kilpatrick, D.G., & Veronen, L.J. (1992). Child sexual assault as a risk factor for mental disorders among women. Journal of Interpersonal Violence, 7, 189-204. Scaturo, D.J. (1994). Integrative psychotherapy for panic disorder and agoraphobia in clinical practice. Journal of Psychotherapy Integration, 4, 253--272. Schaap, C., Bennun, I., Schindler, L., & Hoogduin, K. (1993). The therapeutic relationship in behavioural psychotherapy. Chichester, UK: Wiley. Schatz, I.J. (1986). Orthostatic hypotension. Philadelphia, PA: F. A. Davis. Scheibe, G., & Albus, M. (1996). Predictors of outcome in panic disorder: A 5-year prospective follow-up study. Journal of Affective Disorders, 41, 111-116. Schmidt, N.B. (1999). Prospective evaluations of anxiety sensitivity. In S. Taylor (Ed.), Anxiety sensitivity: Theory, research, and treatment of the fear of anxiety (pp. 217-235). Mahwah, NJ: Erlbaum. Schmidt, N.B., Lerew, D.R., & Jackson, R.J. (1997a). The role of anxiety sensitivity in the pathogenesis of panic: Prospective evaluation of spontaneous panic attacks during acute stress. Journal of Abnormal Psychology, 106, 355-364. Schmidt, N.B., Lerew, D.R., & Trakowski, J.H. (1997b). Body vigilance in panic disorder: Evaluating attention to bodily perturbations. Journal of Consulting and Clinical Psychology, 65, 214-220. Schmidt, N.B., & Telch, M.J. (1997). Non-psychiatric medical comorbidity, health perceptions and treatment outcome in patients with panic disorder. Health Psychology, 16, 114-122. Schmidt, N.B., & Trakowski, J.H. (in press). Attentional focus and fearful responding in patients with panic disorder during a 35% CO 2 challenge. Behavior Therapy. Schmidt, N.B., Trakowski, J.H., & Staab, J.P. (1997c). Extinction of panicogenic effects of a 35% CO 2 challenge in patients with panic disorder. Journal of Abnormal Psychology, 106, 630-638. Schmidt, N.B., & Woolaway-Bickel, K. (2000). The effects of treatment compliance on outcome in cognitive behavioral therapy for panic disorder: Quality versus quantity. Journal of Consulting and Clinical Psychology, 68, 13-18. Schmidt-Traub, S., & Bamler, K.-J. (1997). The psychoimmunological association of panic disorder and allergic reaction. British Journal of Clinical Psychology, 36, 51-62. Schneck, J.M. (1967). Psychogenic contraction of visual fields. Psychosomatics, 8, 208-209. Schulte, D. (1996). Tailor-made and standardized therapy: Complementary tasks in behavior therapy. A contrarian view. Journal of Behavior Therapy and Experimental Psychiatry, 27, 119-126. Schulte, D., Kunzel, R., Pepping, G., & Schulte-Bahrenberg, T. (1992). Tailor-made versus standardized therapy of phobic patients. Advances in Behaviour Research and Therapy, 14, 67-92. Schweizer, E.E., Rickels, K., Case, W.G., & Greenblatt, D.J. (1990). Long-term use of benzodiazepines. II: Effects of gradual taper. Archives of General Psychiatry, 47, 908-915. Scupi, B.S., Maser, J.D., & Uhde, T.W. (1992). The National Institute of Mental Health Panic Questionnaire: An instrument for assessing clinical characteristics of panic disorder. Journal of Nervous and Mental Disease, 180, 566-572.

486

REFERENCES

Sedman, G. (1970). Theories of depersonalization: A re-appraisal. British Journal of Psychiatry, 117, 1-14. • Segal, Z.V., & Swallow, S.R. (1994). Cognitive assessment of unipolar depression: Measuring products, processes and structures. Behaviour Research and Therapy, 32, 147-158. Seligman, M.E.P., & Johnston, J.C. (1973). A cognitive theory of avoidance learning. In F.J. McGuigan & D.B. Lumsden (Eds.), Contemporary approaches to learning and conditioning (pp. 69-145). Washington, DC: Winston. Seuss, W.M., Alexander, A.B., Smith, D.D., Sweeney, H.W., & Marion, R.J. (1980). The effects of psychological stress on respiration: A preliminary study of anxiety and hyperventilation. Psychophysiology, 17, 535-540. Shapiro, F. (1995). Eye movement desensitization and reprocessing: Basic principles, protocols, and procedures. New York: Guilford. Sharp, D.M., Power, K.G., Simpson, R.J., Swanson, V., Moodie, E., Anstee, J.A., & Ashford, J.J. (1996). Fluvoxamine, placebo, and cognitive behaviour therapy used alone and in combination in the treatment of panic disorder and agoraphobia. Journal of Anxiety Disorders, 10, 219-242. Shaw, S.C., Marks, I.M., & Toole, S. (1999). Lessons from pilot tests of computer self help for agora/claustrophobia and panic. Unpublished manuscript, Institute of Psychiatry, London. Shea, S.C. (1988). Psychiatric interviewing: The art of understanding. Philadelphia, PA: Saunders. Shear, M.K., Ball, G., Fitzpatrick, M., Josephson, S., Klosko, J., & Frances, A. (1991a). Cognitive-behavioral therapy for panic: An open study. Journal of Nervous and Mental Disease, 179, 468-472. Shear, M.K., Brown, T.A., Barlow, D.H., Money, R., Sholomskas, D.E., Woods, S.W., Gorman, J.M., & Papp, L.A. (1997). Multicenter collaborative panic disorder severity scale. American Journal of Psychiatry, 154, 1571-1575. Shear, M.K., Cooper, A.M., Klerman, G.L., Busch, F.N., & Shapiro, T. (1993). A psychodynamic model of panic disorder. American Journal of Psychiatry, 150, 859-866. Shear, M.K., Pyer, A.J., Ball, G., Josephson, S., Fitzpatrick, M., Gitlin, B., Frances, A., Gorman, J., Liebowitz, M., & Klein, D.F. (1991b). Vulnerability to sodium lactate in panic disorder patients given cognitive-behavioral therapy. American Journal of Psychiatry, 148, 795-797. Shear, M.K., Kligfield, P., Harshfield, G., Devereux, R.B., Polan, J.J., Mann, J.J., Pickering, T., & Frances, A.J. (1987). Cardiac rate and rhythm in panic patients. American Journal of Psychiatry, 144, 633-637. Shear, M.K., & Maser, J.D. (1994). Standardized assessment for panic disorder research: A conference report. Archives of General Psychiatry, 51, 346-354. Shear, M.K., Pilkonis, P.A., Cloitre, M., & Leon, A.C. (1994). Cognitive behavioral treatment compared with nonprescriptive treatment of panic disorder. Archives of General Psychiatry, 51, 395-401. Sheehan, D.V. (1982). Current perspectives in the treatment of panic and phobic disorders. Drug Therapeutics, 12, 179-193. Sheehan, D.V. (1983). The anxiety disease. New York: Scribners. Sheehan, D.V. (1986). Tricyclic antidepressants in the treatment of panic and anxiety disorders. Psychosomatics, 27, 10-16. Sheehan, D.V., Ballenger, J., & Jacobsen, G. (1980). Treatment of endogenous anxiety with phobic, hysterical and hypochondriacal symptoms. Archives of General Psychiatry, 37, 51-59.

REFERENCES487 Sheehan, D.V., Raj, A.B., Harnett-Sheehan, K., Soto, S., & Knapp, E. (1993). The relative efficacy of high-dose buspirone and alprazolam in the treatment of panic disorder: A double-blind placebo-controlled study. Acta Psychiatrica Scandinavica, 88, 1-11. Sheehan, D.V., Raj, A.B., Sheehan, K.H., & Soto, S. (1990). Is buspirone effective for panic disorder? Journal of Clinical Psychopharmacology, 10, 3-11. Shulman, I.D., Cox, B.J., Swinson, R.P., Kuch, K., & Reichman, J.T. (1994). Precipitating events, locations and reactions associated with initial unexpected panic attacks. Behaviour Research and Therapy, 32, 17-20. Silove, D., Harris, M., Morgan, A., Boyce, P., Manicavasagar, V., Hadzi-Pavlovic, D., & Wilhelm, K. (1995). Is early separation anxiety a specific precursor of panic disorder-agoraphobia? A community study. Psychological Medicine, 25, 405411. Silove, D., Manicavasagar, V., Curtis, J., & Blaszczynski, A. (1996). Is early separation anxiety a risk factor for adult panic disorder? A critical review. Comprehensive Psychiatry, 37, 167-179. Silove, D., Manicavasagar, V., O'Connell, D., & Blaszczynski, A. (1993). Reported early separation anxiety symptoms in patients with panic and generalised anxiety disorders. Australian and New Zealand Journal of Psychiatry, 27, 489-494. Silverman, W.K., & Albano, A.M. (1996). The Anxiety Disorders Interview Schedule for Children-IV (Child and Parent Versions). San Antonio, TX: Psychological Corporation. Silverman, W.K., Fleisig, W., Rabian, B., & Peterson, R.A. (1991). Childhood Anxiety Sensitivity Index. Journal of Clinical Child Psychology, 20, 162-168. Silverman, W.K., & Weems, C.F. (1999). Anxiety sensitivity in children. In S. Taylor (Ed.), Anxiety sensitivity: Theory, research, and treatment of the fear of anxiety (pp. 239-268). Mahwah, NJ: Erlbaum. Siris, S.G., Aaronson, A., & Sellew, A.P. (1989). Imipramine-responsive panic like symptomatology in schizophrenia. Biological Psychiatry, 25, 485-488. Snaith, R., Baugh, S., Clayden, A., & Sipple, M. (1982). The Clinical Anxiety Scale: An instrument derived from the Hamilton Anxiety Scale. British Journal of Psychiatry, 141, 518-523. Sobell, L.C., & Sobell, M.B. (1992). Timeline followback: A technique for assessing self-reported alcohol consumption. In R.Z. Litten & J. Allen (Eds.), Measuring alcohol consumption: Psychosocial and biological methods (pp. 41-72). New Jersey: Humana Press. Sobell, L.C., Toneatto, T., & Sobell, M.B. (1994). Behavioral assessment and treatment planning for alcohol, tobacco, and other drug problems: Current status with an emphasis on clinical applications. Behavior Therapy, 25, 533580. Sochting, I., Taylor, S., Freeman, W., de Koning, E., Segerstrom, S., & Thordarson, D. (1998). In vivo exposure for panic disorder and agoraphobia: Does a cognitive rationale enhance treatment efficacy? In E. Sanavio (Ed.), Behavioural and cognitive therapy today: Essays in honor of Hans J. Eysenck (pp. 293-302). Amsterdam: Elsevier. Sokol, L., Beck, A.T., Greenberg, R.L., Wright, F.D., & Berchick, R.J. (1989). Cognitive therapy for panic disorder: A nonpharmacological alternative. Journal of Nervous and Mental Disease, 177, 711-716. Spiegel, D.A., & Bruce, T.J. (1997). Benzodiazepines and exposure-based cognitive behavior therapies for panic disorder: Conclusions from combined treatment trials. American Journal of Psychiatry, 154, 773-781.

488

REFERENCES

Spiegel, D.A., Bruce, T.J., Gregg, S.F., & Nuzzarello, A. (1994). Does cognitive behavior therapy assist slow-taper alprazolam discontinuation in panic disorder? American Journal of Psychiatry, 151, 876-881. Spiegel, D.A., Roth, M., Weissman, M., Lavori, P., Gorman, J., Rush, J., & Ballenger, J. (1993). Comment on the London/Toronto study of alprazolam and exposure in panic disorder with agoraphobia. British Journal of Psychiatry, 162, 788-789. Spielberger, C.D. (1983). Manual for the State-Trait Anxiety Inventory (Form Y). Palo Alto, CA: Consulting Psychologists Press. Spitzer, R.L., Williams, J.B.W., Gibbon, M., & First, M.B. (1990). User's guide for the Structured Clinical Interview for DSM-III-R. Washington, DC: American Psychiatric Association Press. Spitzer, R.L., Williams, J.B.W., Gibbon, M., & First, M.B. (1992). The Structured Clinical Interview for DSM-III-R (SCID). I: History, rationale, and description. Archives of General Psychiatry, 49, 624-629. Stafrace, S. (1994). Hypnosis in the treatment of panic disorder with agoraphobia. Australian Journal of Clinical Hypnosis, 22, 73-86. Stampfl, T.G., & Levis, D.J. (1967). Essentials of implosive therapy: A learning theory based psychodynamic behavioral therapy. Journal of Abnormal Psychology, 72, 496-503. Starcevic, V. (1992). Comorbidity models of panic disorder/agoraphobia and personality disturbance. Journal of Personality Disorders, 6, 213-225. Steer, RA., & Beck, A.T. (1996). Generalized anxiety and panic disorders: Response to Cox, Cohen, Direnfeld, and Swinson (1996). Behaviour Research and Therapy, 34, 955-957. Stein, D.J., Hollander, E., & Skoder, A.E. (1993). Anxiety disorders and personality disorders: A review. Journal of Personality Disorders, 7, 87-104. Stein, M.B., & Rapee, RM. (1999). Biological aspects of anxiety sensitivity: Is it all in the head? In S. Taylor (Ed.), Anxiety sensitivity: Theory, research, and treatment of the fear of anxiety (pp. 199-215). Mahwah, NJ: Erlbaum. Stein, M.B., Schmidt, P.J., Rubinow, D.R., & Uhde, T.W. (1989). Panic disorder and the menstrual cycle: Panic disorder patients, healthy control subjects, and patients with premenstrual syndrome. American Journal of Psychiatry, 146, 1299-1303. Steketee, G. (1993). Treatment of obsessive compulsive disorder. New York: Guilford. Stern, R., & Marks, I. (1973). Brief and prolonged flooding: A comparison in agoraphobic patients. Archives of General Psychiatry, 28, 270-276. Stewart, S.H., Samoluk, S.B., & MacDonald, A.B. (1999). Anxiety sensitivity and substance use and abuse. In S. Taylor (Ed.), Anxiety sensitivity: Theory, research, and treatment of the fear of anxiety (pp. 287-319). Mahwah, NJ: Erlbaum. Stewart, S.H., Taylor, S., Jang, K.L., Cox, B.J., Watt, M.C., Fedoroff, LC., & Borger, S.C. (2000). Causal modeling of relations among learning history, anxiety sensitivity, and panic attacks. Behaviour Research and Therapy. Stewart, W.F., Linet, M.S., & Celentano, D.D. (1989). Migraine headaches and panic attacks. Psychosomatic Medicine, 51, 559-569. Stockwell, T., Smail, P., Hodgson, R., & Canter, S. (1984). Alcohol dependence and phobic anxiety states, II: A retrospective study. British Journal of Psychiatry, 144, 58-63. Strupp, H.H., & Binder, J.L. (1984). Psychotherapy in a new key: A guide to time-limited dynamic psychotherapy. New York: Basic Books. Stuart, RB. (1980). Helping couples change: A social learning approach to marital therapy. New York: Guilford.

REFERENCES489 Suinn, R., & Richardson, F. (1971). Anxiety management training: A nonspecific behavior therapy program for anxiety control. Behavior Therapy, 2, 498-510. Svebak, S., Dalen, K., & Storfjell, 0. (1981). The psychological significance of taskinduced tonic changes in somatic and autonomic activity. Psychophysiology, 17, 403-409. Swales, P.J., Solfvin, J.F., & Sheikh, J.L (1996). Cognitive-behavioral therapy in older panic disorder patients. American Journal of Geriatric Psychiatry, 4, 46-60. Swinson, R.P., Cox, B.J., Shulman, LD., Kuch, K., & Woszczyna, C.B. (1992a). Medication use and the assessment of agoraphobic avoidance. Behaviour Research and Therapy, 30, 563-568. Swinson, R.P., Fergus, K.D., Cox, B.J., & Wickwire, K. (1995). Efficacy of telephoneadministered behavioral therapy for panic disorder with agoraphobia. Behaviour Research and Therapy, 33, 465-469. Swinson, R.P., Soulios, C., Cox, B.J., & Kuch, K. (1992b). Brief treatment of emergency room patients with panic attacks. American Journal of Psychiatry, 149, 944-946. Takahashi, T. (1993). A persuasion therapy for panic disorder in old Japanese medical literature. Comprehensive Psychiatry, 34, 31-35. Taylor, C.B., & Arnow, B. (1988). The nature and treatment of anxiety disorders. New York: Free Press. Taylor, C.B., Fried, L., & Kenardy, J. (1990). The use of a real-time computer diary for data acquisition and processing. Behaviour Research and Therapy, 28, 93-97. Taylor, L (1984). Self-exposure instructions by telephone with a severe agoraphobic: A case study. Behavioural Psychotherapy, 12, 68-72. Taylor, LL. (1985). The reactive effect of self-monitoring of target activities on agoraphobics: A pilot study. Scandinavian Journal of Behaviour Therapy, 14, 17-22. Taylor, S. (1995). Anxiety sensitivity: Theoretical perspectives and recent findings. Behaviour Research and Therapy, 33, 243-258. Taylor, S. (1998). The hierarchic structure of fears. Behaviour Research and Therapy, 36, 205-214. Taylor, S. (1999). Anxiety sensitivity: Theory, research, and treatment of the fear of anxiety. Mahwah, NJ: Erlbaum. Taylor, S., & Cox, B.J. (1998a). Anxiety sensitivity: Multiple dimensions and hierarchic structure. Behaviour Research and Therapy, 36, 37-51. Taylor, S., & Cox, B.J. (1998b). An expanded Anxiety Sensitivity Index: Evidence for a hierarchic structure in a clinical sample. Journal of Anxiety Disorders, 12, 463-484. Taylor, S., & Fedoroff, LC. (1999). The expectancy theory of fear, anxiety, and panic: A conceptual and empirical analysis. In S. Taylor (Ed.), Anxiety sensitivity: Theory, research, and treatment of the fear of anxiety (pp. 17-33). Mahwah, NJ: Erlbaum. Taylor, S., Fedoroff, I., & Koch, W.J. (1999). Posttraumatic stress disorder due to motor vehicle accidents: Patterns and predictors of response to cognitivebehaviour therapy. In E.J. Hickling & E.B. Blanchard (Eds.), International handbook of road traffic accidents and psychological trauma: Theory, treatment and the law (pp. 353-374). Amsterdam: Elsevier. Taylor, S., Koch, W.J., & McNally, R.J. (1992). How does anxiety sensitivity vary across the anxiety disorders? Journal of Anxiety Disorders, 7, 249-259. Taylor, S., & Livesley, W.J. (1995). The influence of personality on the clinical course of neurosis. Current Opinion in Psychiatry, 8, 93-97.

490

REFERENCES

Taylor, S., & Rachman, S. (1994). Stimulus estimation and the overprediction of fear. British Journal of Clinical Psychology, 33, 173-181. Taylor, S., Thordarson, D., & Sochting, D. (in press). Assessment, treatment planning, and outcome evaluation for obsessive-compulsive disorder. In D.H. Barlow & M.M. Antony (Eds.), Handbook of assessment, treatment planning, and outcome evaluation: Empirically supported strategies for psychological disorders. New York: Guilford. Taylor, S., Woody, S., Koch, W.J., McLean, P., & Anderson, K. (1996). Suffocation false-alarms and efficacy of cognitive-behavioural therapy for panic disorder. Behavior Therapy, 27, 115-126. Teasdale, J.D., Walsh, P.A., Lancashire, M., & Mathews, A.M. (1977). Group exposure for agoraphobics: A replication study. British Journal of Psychiatry, 130, 186-193. Telch, M.J. (1987). The Panic Appraisal Inventory. Unpublished scale, Department of Psychology, University of Texas at Austin. Telch, M.J. (1996, August). Singular and combined efficacy of in vivo exposure and CBT in the treatment of panic disorder with agoraphobia. Paper presented at the 26th International Congress of Psychology, Montreal, Quebec. Telch, M.J. (1997, June). Identification of risk factors in panic attacks: Preliminary results of a five-year prospective study. Paper presented at the annual convention of the Canadian Psychological Association, Toronto, Ontario. Telch, M.J., Agras, W.S., Taylor, C.B., Roth, W.T., & Gallen, C.C. (1985). Combined pharmacological and behavioral treatment for agoraphobia. Behaviour Research and Therapy, 23, 325-335. Telch, M.J., Brouillard, M., Telch, C.F., Agras, S., & Taylor, C.B. (1989). The role of cognitive appraisal in panic related avoidance. Behaviour Research and Therapy, 27, 373-385. Telch, M.J., & Harrington, P.J. (1992, November). The role of anxiety sensitivity and expectedness of arousal in mediating affective response of 35% carbon dioxide. Paper presented at the 26th annual meeting of the Association for Advancement of Behavior Therapy, Boston, MA. Telch, M.J., & Lucas, R.A. (1994). Combined pharmacological and psychological treatment of panic disorder: Current status and future directions. In B.E. Wolfe & J.D. Maser (Eds.), Treatment of panic disorder: A consensus development conference (pp. 177-197). Washington, DC: American Psychiatric Press. Telch, M.J., Lucas, J.A., Schmidt, N.B., Hanna, H.H., Jaimez, T.L., Lucas, R.A. (1993). Group cognitive-behavioral treatment of panic disorder. Behaviour Research and Therapy, 31, 279-287. Telch, M.J., Schmidt, N.B., Jaimez, L., Jacquin, K.M., & Harrington, P.J. (1995). Impact of cognitive-behavioral treatment on quality of life in panic disorder patients. Journal of Consulting and Clinical Psychology, 63, 823--830. Thomas-Peter, B.A., Jones, R.B., Sinnott, A., & Fordham, A.S. (1983). Prediction of outcome in the treatment of agoraphobia. Behavioural Psychotherapy, 11, 320-328. Thompson, A.H., Bland, R.C., & Orn, H.T. (1989). Relationship and chronology of depression, agoraphobia, and panic disorder in the general population. Journal of Nervous and Mental Disease, 177, 456--463. Thompson, W.G. (1984). The irritable bowel. Gut, 25, 305-320. Thompson! W.G., & Heaton, K.W. (1980). Functional bowel disorders in apparently healthy people. Gastroenterology, 79, 283-288. Thorpe, G.L., Hecker, J.E., Cavallaro, L.A., & Kulberg, G.E. (1987). Insight versus rehearsal in cognitive-behavior therapy: A crossover study with sixteen phobics. Behavioural Psychotherapy, 15, 319-336.

REFERENCES491 Timmons, B.H. (1994). Breathing-related issues in therapy. In B.H. Timmons & R. Ley (Eds.), Behavioral and psychological approaches to breathing disorders (pp. 261-292). New York: Plenum. Tobef\a, A., Sanchez, R., Pose, R., Masana, J., & Del Campo, A.M. (1990). Brief treatment with alprazolam and behavioral guidance in panic disorder. Anxiety Research, 3, 163-174. Trueman, D. (1984). Depersonalization in a non-clinical population. Journal of Psychology, 116, 107-112. Trull, T.J., & Hillerbrand, E. (1990). Psychometric properties and factor structure of the Fear Questionnaire Phobia Subscale items in two normative samples. Journal of Psychopathology and Behavioral Assessment, 12, 285-297. Trull, T.J., Nietzel, M.T., & Main, A. (1988). The use of meta-analysis to assess the clinical significance of behavior therapy for agoraphobia. Behavior Therapy, 19, 527-538. Tsao, J.C.I., Lewin, M.R., & Craske, M.G. (1998). The effects of cognitive-behavior therapy for panic disorder on comorbid conditions. Journal of Anxiety Disorders, 12, 357-371. Turk, D.C., & Salovey, P. (1986). Clinical information processing: Bias inoculation. In R. Ingram (Ed.), Information processing approaches to clinical psychology (pp. 305-323). San Diego, CA: Academic. Turkat, I.D. (1985). Behavioral caseformulation. New York: Plenum. Turkat, I.D. (1990). The personality disorders: A psychological approach to clinical management. New York: Pergamon. Turner, S.M., Beidel, D.C., & Costello, A. (1987). Psychopathology in the offspring of anxiety disorder patients. Journal of Consulting and Clinical Psychology, 55, 229-235. Turner, S.M., Beidel, D.C., & Wolff, P.L. (1996). Is behavioral inhibition related to the anxiety disorders? Clinical Psychology Review, 16, 157-172. Tweed, L.J., Schoenbach, J.J., George, L.K., Blazer, D.G. (1989). The effects of childhood parental death and divorce on six-month history of anxiety disorders. British Journal of Psychiatry, 154, 823-828. Tyrer, P. (1985). Neurosis divisible? Lancet, i, 685-688. Tyrer, P. (1988). Synthesis and prospectus. In P. Tyrer (Ed.), Personality disorders: Diagnosis, management and course (pp. 137-139). London: Wright. Tyrer, P. (1989). Classification of neurosis. Chichester, UK: Wiley. Tyrer, P., Murphy, S., Oates, G., & Kingdon, D. (1985). Psychological treatment for benzodiazepine dependence. Lancet, i, 1042-1043. Tyrer, P., Murphy, S., & Riley, P. (1990). The Benzodiazepine Withdrawal Symptom Questionnaire. Journal of Affective Disorders, 19, 53-61. Tyrer, P., Owen, R., & Dawling, S. (1983). Gradual withdrawal of diazepam after long-term therapy. Lancet, i, 1402-1406. Tyrer, P., Seivewright, N., Ferguson, B., & Tyrer, J. (1992). The general neurotic syndrome: A coaxial diagnosis of anxiety, depression and personality disorder. Acta Psychiatrica Scandinavica, 85, 201-206. Uhde, T.W., Boulenger, J.P., Roy-Byrne, P.P., Geraci, M.P., Vittone, B.J., & Post, R.M. (1985). Longitudinal course of panic disorder: Clinical and biological considerations. Progress in Neuro-Psychopharmacology and Biological Psychiatry, 9, 39-51. van Bakom, A.J.L.M., Bakker, A., Spinhoven, P., Blaauw, B.M., Smeenk, S., & Ruesink, B. (1997). A meta-analysis of the treatment of panic disorder with or without agoraphobia: A comparison of psychopharmacological, cognitivebehavioral, and combination treatments. Journal of Nervous and Mental Disease, 185, 510-516.

492

REFERENCES

van Balkom, A.J.L.M., de Beurs, E., Koele, P., Lange, A., & van Dyck, R. (1996). Long-term benzodiazepine use is associated with smaller treatment gain in panic disorder with agoraphobia. Journal of Nervous and Mental Disease, 184, 133-135. van Balkom, A.J.L.M., Nauta, M.C.E., & Bakkev, A. (1995). Meta-analysis on the treatment of panic disorder with agoraphobia. Review and re-examination. Clinical Psychology and Psychotherapy, 2, 1-14. van den Hout, M., Arntz, A., & Hoekstra, R. (1994). Exposure reduced agoraphobia but not panic, and cognitive therapy reduced panic but not agoraphobia. Behaviour Research and Therapy, 32, 447-451. van Etten, M., & Taylor, S. (1998). Comparative efficacy of treatments for posttraumatic stress disorder: A meta-analysis. Clinical Psychology and Psychotherapy, 5, 126-145. van Valkenburg, C., Akiskal, H.S., Pazantian, J., & Rosenthal, T. (1984). Anxious depressions: Clinical, family history, and naturalistic outcome-comparisons with panic and major affective disorders. Journal of Affective Disorders, 6, 67-82. van Zuuren, F.J. (1988). The Fear Questionnaire: Some data on validity, reliability and layout. British Journal of Psychiatry, 153, 659-662. Victor, M. (1973). The role of hypomagnesemia and respiratory alkalosis in the genesis of alcohol withdrawal symptoms. Annuals of the New York Academy of Science, 215, 235-248. von Korff, M., & Eaton, W.W. (1989). Epidemiologic findings on panic. In R. Baker (Ed.), Panic disorder: Theory, research, and therapy (pp. 35-50). New York: Wiley. Wade, S.L., Monroe, S.M., & Michelson, L.K. (1993). Chronic life stress and treatment outcome in agoraphobia with panic attacks. American Journal of Psychiatry, 150, 1494-1495. Wade, W.A., Treat, T.A., & Stuart, G.L. (1998). Transporting an empirically supported treatment for panic disorder to a service clinic setting: A benchmarking study. Journal of Consulting and Clinical Psychology, 66, 231-239. Walker, E.A., Katon, W.J., Hansom, J., Harrop-Griffiths, J., Holm, L., Jones, M.L., Hickok, L., & Jemelka, R.P. (1992). Medical and psychiatric symptoms in women with childhood sexual abuse. Psychosomatic Medicine, 54, 658-664. Waller, N.G., Putnam, F.W., & Carlson, E.B. (1996). Types of dissociation and dissociative types: A taxometric analysis of dissociative experiences. Psychological Methods, 1, 300-321. Waller, N.G., & Ross, C.A. (1997). The prevalence and biometric structure of pathological dissociation in the general population: Taxometric and behavior genetic findings. Journal of Abnormal Psychology, 106, 499-510. Wardle, J., Hayward, P., Higgitt, A., Stab!, M., Blizard, R., & Gray, J. (1994). Effects of concurrent diazepam treatment on the outcome of exposure therapy in agoraphobia. Behaviour Research and Therapy, 32, 203-215. Warwick, H.M.C., & Salkovskis, P.M. (1985). Reassurance. British Medical Journal, 290, 1028. Warwick, H.M.C., & Salkovskis, P.M. (1990). Hypochondriasis. Behaviour Research and Therapy, 28, 105-117. Watson, J.P., Mullet, G.E., & Pillay, H. (1973). The effects of prolonged group exposure to phobic situations upon agoraphobic patients treated in groups. Behaviour Research and Therapy, 11, 531-545. Watt, M.C., Stewart, S.H., & Cox, B.J. (1998). A retrospective study of the learning history origins of anxiety sensitivity. Behaviour Research and Therapy, 36, 505525.

REFERENCES493 Watts, F.N., Powell, G.E., & Austin, S.V. (1977). The modification of abnormal beliefs. British Journal of Medical Psychology, 46, 359-363. Watts, F.N., & Wilkins, A.J. (1989). The role of provocative visual stimuli in agoraphobia. Psychological Medicine, 19, 875-885. Wegner, D.M. (1994). Ironic processes of mental control. Psychological Review, 101, 34-52. Weissman, A.N. (1980). Assessing depressogenic attitudes: A validation study. Paper presented at the 51st annual meeting of the Eastern Psychological Association, Hartford, CT. _ Weissman, M.M. (1991). Panic disorder: Impact on quality of life. Journal of Clinical Psychiatry, 52, 6-9. Weissman, M.M., Bland, RC., Canino, G.J., Faravelli, C., Greenwald, S., Hwu, H.-G., Joyce, P.R., Karam, E.G., Lee, C.-K., Lellouch, J., Lepine, J.-P., Newman, S.C., Oakley-Browne, M.A., Rubio-Stipec, M., Wells, J.E., Wickramaratne, P.J., Wittchen, H.-U., & Yeh, E.-K. (1997). The cross-national epidemiology of panic disorder. Archives of General Psychiatry, 54, 305-309. Weissman, M.M., Broadhead, W.E., Olfson, M., Sheehan, D.V., Hoven, C., Conolly, P., Fireman, B.H., Farber, L., Blacklow, R.S., Higgins, S., & Leon, A.C. (1998). A diagnostic aid for detecting (DSM-IV) mental disorders in primary care. General Hospital Psychiatry, 20, 1-11. Weissman, M.M., Klerman, G.L., Markowitz, J.S., & Ouellette, R. (1989). Suicidal ideation and suicide attempts in panic disorder and attacks. New England Journal of Medicine, 321, 1209-1214. Weissman, M.M., Markowitz, J.S., Ouellette, R., Greenwald, S., & Kahn, J.P. (1990). Panic disorder and cardiovascular/ cerebrovascular problems: Results from a community survey. American Journal of Psychiatry, 147, 1504-1508. Welkowitz, L.A., Papp, L.A., Cloitre, M., Liebowitz, M.R., Martin, L.Y., & Gorman, J.M. (1991). Cognitive-behavior therapy for panic disorder delivered by psychopharmacologically oriented clinicians. Journal of Nervous and Mental Disease, 179, 473-477. Wells, A. (1990). Panic disorder in association with relaxation induced anxiety: An attentional training approach to treatment. Behavior Therapy, 21, 273-280. Wells, A. (1997). Cognitive therapy of anxiety disorders: A practice manual and conceptual guide. Chichester: Wiley. Westphal, C.F.O. (1871). Die Agoraphobia, eine neuropathische Erscheinung. Archiv far Psychiatrie und Nervenkrankheiten, 3, 138-161 & 219-221. Wiborg, I.M., & Dahl, A.A. (1996). Does brief dynamic psychotherapy reduce the relapse rate of panic disorder? Archives of General Psychiatry, 53, 689-694. Wieling, W. (1992). Non-invasive continuous recording of heart rate and blood pressure in the evaluation of neurocardiovascular control. In R. Bannister & C.J. Mathias (Eds.), Autonomic failure: A textbook of clinical disorders of the autonomic nervous system (pp. 291-311). Oxford: Oxford University Press. Wild, A.J. (1994). Hypnosis as an adjunct in the treatment of panic disorder. Australian Journal of Clinical and Experimental Hypnosis, 22, 109-117. Wilkins, A., Nimmo-Smith, I., Tait, A., McManus, C., Della Sala, S., Tilley, A., Arnold, K., Barrie, M., & Scott, S. (1984). A neurological basis for visual discomfort. Brain, 107, 989-1017. Wilkinson, G., Balestrieri, M., Ruggeri, M., & Bellantuono, C. (1991). Metaanalysis of double-blind placebo-controlled trials of antidepressants and benzodiazepines for patients with panic disorders. Psychological Medicine, 21, 991-998.

494

REFERENCES

Williams, J.B.W., Gibbon, M., First, M.B., Spitzer, R.L., Davis, M., Barus, J., Howes, M.J., Kane, J., Pope, H.G., Rounsaville, B., Wittchen, H.-U. (1992). The Structured Clinical Interview for DSM-III-R (SCID): IL Multisite test-retest reliability. Archives of General Psychiatry, 49, 630-636. Williams, J.M.G., Watts, F.N., MacLeod, C., & Mathews, A. (1997). Cognitive psychology and emotional disorders (2nd ed.). Chichester: Wiley. Williams, K.E., & Chambless, D.L. (1990). The relationship between therapist characteristics and outcome of in vivo exposure treatment for agoraphobia. Behavior Therapy, 21, 111-116. Williams, L.J. (1989). Foveal load affects the functional field of view. Human Performance, 2, 1-28. Williams, L.J. (1995). Peripheral target recognition and visual field narrowing in aviators and nonaviators. International Journal of Aviation Psychology, 5, 215-232. Williams, S.L. (1990). Guided mastery treatment of agoraphobia: Beyond stimulus exposure. Progress in Behavior Modification, 26, 89-121. Williams, S.L., & Falbo, J. (1996). Cognitive and performance-based treatments for panic attacks in people with varying degrees of agoraphobic disability. Behaviour Research and Therapy, 34, 253-264. Williams, S.L., & Zane, G. (1989). Guided mastery and stimulus exposure treatments for severe performance anxiety in agoraphobics. Behaviour Research and Therapy, 27, 237-245. Wilson, G.T. (1996a). Treatment of bulimia nervosa: When CBT fails. Behaviour Research and Therapy, 34, 197-212. Wilson, G.T. (1996b). Manual-based treatments: The clinical application of research findings. Behaviour Research and Therapy, 34, 295-314. Wilson, G.T. (1997). Treatment manuals in clinical practice. Behaviour Research and Therapy, 35, 205-210. Wilson, R.R. (1996). Don't panic: Taking control of anxiety attacks (rev. ed.). New York: Harper Perennial. Wittchen, H.-U. (1996). Critical issues in the evaluation of comorbidity of psychiatric disorders. British Journal of Psychiatry, 168 (suppl.), 9-16. Wolf, F.M. (1986). Meta-analysis: Quantitative methods for research synthesis. Beverly Hills, CA: Sage. Wolfe, B.E. (1989). Phobias, panic and psychotherapy integration. Journal of Integrative and Eclectic Psychotherapy, 8, 264-276. Wolfe, B.E., & Maser, J.D. (1994). Treatment of panic disorder: A consensus development conference. Washington, DC: American Psychiatric Press. Wolkowitz, O.M., & Pikar, D. (1991). Benzodiazepines in the treatment of schizophrenia: A review and reappraisal. American Journal of Psychiatry, 148, 714-726. Wolpe, J., & Lazarus, A.A. (1966). Behaviour therapy techniques: A guide to the treatment of neuroses. Oxford: Pergamon. Woodman, C.L., Noyes, R., Ballenger, J.C., Lydiard, R.B., Sievers, G., & Mihalko, D. (1994). Predictors of response to alprazolam and placebo in patients with panic disorder. Journal of Affective Disorders, 30, 5-13. Woods, S.W., Charney, D.S., Loke, J ., Goodman, W.K., Redmond, D.E., & Heninger, G.R. (1986). Carbon dioxide sensitivity in panic anxiety. Archives of General Psychiatry, 43, 900-910. Woody, S.; McLean, P., Taylor, S., & Koch, W.J. (1999). Treatment of major depression in the context of panic disorder. Journal of Affective Disorders, 53, 163174.

REFERENCES495 Yamamoto, J., Acosta, F.X., Evans, L.A., & Skilbeck, W.M. (1984). Orienting therapists about patients' needs to increase patient satisfaction. American Journal of Psychiatry, 141, 274-277. Yardley, L. (1994). Vertigo and dizziness. London: Routledge. Yardley, L., Luxon, L., Lear, S., Britton, J., & Bird, J. (1994). Vestibular and posturographic test results in people with symptoms of panic and agoraphobia. Journal of Audiological Medicine, 3, 48-69. Yellow lees, P.M., Alpers, J.H., Bowden, J.J., Bryant, G.D., & Ruffin, R.E. (1987). Psychiatric morbidity in patients with chronic airflow obstruction. Medical Journal of Australia, 146, 305-307. Yellowlees, P.M., Haynes, S., Potts, N., & Ruffin, R.E. (1988). Psychiatric morbidity in patients with life-threatening asthma: Initial report of a controlled study. Medical Journal of Australia, 149, 246-249. Zandbergen, J., Bright, M., Pols, H., Fernandez, I., de Loof, C., & Grez, E. (1991). Higher lifetime prevalence of respiratory diseases in panic disorder? American Journal of Psychiatry, 148, 1583-1585. Zane, G., & Williams, S.L. (1993). Performance-related anxiety in agoraphobia: Treatment procedures and cognitive mechanisms of change. Behavior Therapy, 24, 625-643. Zeanah, C.H. (1988). Atypical panic attacks and lack of resolution of mourning. General Hospital Psychiatry, 10, 373-377. Zillmann, D. (1983). Transfer of excitation in emotional behavior. In J.T. Cacioppo & R.E. Petty (Eds.), Social psychophysiology (pp. 215-240). New York: Guilford. Zimmerman, M. (1994). Diagnosing personality disorders: A review of issues and research methods. Archives of General Psychiatry, 51, 225-245. Zinbarg, R.E., & Barlow, D.H. (1996). The structure of anxiety and the anxiety disorders: A hierarchical model. Journal of Abnormal Psychology, 105, 181-193. Zitrin, C.M., Klein, D.F., & Woerner, M.G. (1980). Treatment of agoraphobia with group exposure in vivo and imipramine. Archives of General Psychiatry, 37, 63-72. Zitrin, C.M., Klein, D.F., Woerner, M.G., & Ross, D.C. (1983). Treatment of phobias, I. Comparison of imipramine hydrochloride and placebo. Archives of General Psychiatry, 40, 125-138. Zlotnick, C., Zakriski, A., Shea, M.T., Begin, A., Pearlstein, T., Simpson, E., & Costello, E. (1996). The long-term sequelae of childhood sexual abuse: Support for a complex PTSD. Jounial of Traumatic Stress, 9, 195-205.

AUTHOR INDEX

Abelson, J. L., 174 Adler, C. M., 380 Agras, W. S., 161, 248 Airola, P. 0., 80 Albanesi, G., 189, 190-3, 195, 198-9, 207 Albano, A. M., 403 Albus, M., 191-3, 195,211 Alexander, P. E., 155, 162, 318 Alford, B. A., 361-2 Allison, D. B., 84 American Academy of Child and Adolescent Psychiatry, 403-4 American Psychiatric Association, xiii, xiv, 5, 11-15, 17-18, 22-5, 30, 81, 83, 135-6, 155, 166, 204, 220-2, 232,393,402-4,406-7,410 Amering, M., 409-10 Amsterdam, E. A., 75 Anastasi, A., 217 Andersch, S., 134 Anderson, J. R., 309 Anderson, K. W., 255, 268, 412 Andrade, L., 24 Andrews, G., 135, 143-5, 153, 331, 335, 340, 368, 390 Antony, M. M., 114, 250-2, 340 Apfeldorf, W. J., 220 Arena, J. G., 254 Argyle, N., 53, 230-1, 416-9 Arlow, P. B., 416, 418-9 Arnow, B., 138, 234 Arntz, A., 126, 128-30, 138 Aronson, T. A., 12, 181, 184 Arrindel!, W. A., 16, 34, 142, 149, 187, 208,240 Ascher, L. M., 359 Ashton, H., 171 Asmundson, G. J. G., 26, 61, 240

Asnis, J., 70 AuBuchon, P. G., 278-9 Bakker, A., 109 Ball, S. G., 23, 72 Ballenger, J. C., 5, 12, 15, 27, 138 Bamler, K.-J., 78-9 Bandelow, B., 229-31, 233 Bandura, A., 53, 124 Bannister, R., 63, 381 Barbone, F., 160 Barloon, T. J., 33 Barlow, D. H., xv, 5-6, 9, 16, 21, 34, 37, 47, 55, 69, 83, 108, 114, 123, 126-7, 129, 133, 135, 137, 141, 149-50, 152-3, 161-2, 166, 177, 181-3, 187, 190-1, 204, 208, 210, 231, 234-5, 237-8, 254, 262, 285, 289, 290-1, 300, 302, 313, 336, 340, 342, 350, 353,367-9,372,387,402,404-5, 408 Barlow, J. B., 75 Baron, R. M., 112 Barsky, A. J., 25 Basoglu, M., 158, 167-8, 177, 189-92, 195, 211, 357 Bass, C. M., 7, 42-3, 71, 76 Battaglia, M., 22 Baucom, D. H., 383 Beaconsfield, P., 13 Bebbington, P., 18 Beck, A. T., xv, 7, 24, 35-6, 46-7, 83, 95, 116, 119, 135, 149, 178, 181-2, 185, 187, 197, 201, 219, 232-3, 271, 308, 310,322,361-2,393-5 Beck, J. G., 126-7, 142, 237-8 Beckfield, D. F., 302 Beitman, B. D., 5 Bennun, I., 210, 384

498

AUTHOR INDEX

Berg, I., 22 Bernstein, D. A., 373 Bertagnolli, A., 265 Bibb, J., 28 Bichard, S., 53 Biederman, J., 21 Binder, J. L., 428 Black, D. W., 133--4, 138, 146, 148, 151, 185-7, 189, 190-2, 195-6, 198,296 Bland, K., 149, 208 Bond, F. W., 262, 279 Bonn, J. A., 121, 340 Borkovec, T. D., 373, 380 Boszormeny-Nagi, I., 385 Bouchard, S., 112, 127, 219, 240, 242 Bouman, T. K., 63, 389 Bowen, R., 189-90, 192, 195, 210 Bowlby, J., 32-3, 73 Boyd, J. H., 4, 15 Boyer, W., 95, 97-100, 109 Brandt, T., 67 Brasfield, C. R., 413 Brawman-Mintzer, 0., 407 Brehm, J. W., 361 Breier, A., 5, 11, 17, 22--4, 79 Breslau, N., 26 Breuer, P., 62 Brown, G. K., 119-20 Brown, G. W., 11, 35 Brown, T. A., 114, 135-6, 140-1, 166, 169,177, 183--4, 190-1, 193, 195-6, 242 Bruce, T. J., 41, 160, 174-6, 203 Bruch, M., 262, 279, 282 Bruss, G. S., 233 Bufka, L. F., 416, 418 Buller, R., 189, 190, 192, 195 Burandt, U., 63 Burgess, I. S., 61 Burns, D. D., 243, 310 Burns, L. E., 135, 142, 144 Busch, F. N., 429-30 Cameron, 0. G., 79 Canadian Pharmaceutical Association, 113, 155 Carey, K. B., 224 Carmody, R., 382 Carr, R. E., 63, 73-4 Carter, M. M., 181-3, 187 Casper, R. F., 79

Cattell, R. B., 16, 34 Cerny, J. A., xv, 83, 231, 336, 350, 353, 408,420 Chadwick, P. D. J., 361 Chambless, D. L., 28, 41, 95-6, 100, 133--4, 165-6, 180-1, 185, 189-90, 192-5, 199, 200-2, 204, 208-11, 226, 230, 240, 255, 332, 354, 417, 436 Chan, M., 121-2 Chaplin, E. W., 354 Cheng, T. 0., 74 Chernen, L., 384 Chobanian, A. V., 75 Chorpita, B. F., 403 Clair, A. L., 190, 194, 198-9 Clark, D. M., xv, 36-7, 41, 43-7, 49, 50, 56, 58, 65, 71, 72, so,83, 95, 11214, 118, 120-3, 126,128,130, 1323, 135, 138, 142, 178, 190, 192-3, 195-6, 202, 209, 223, 234, 241-2, 252,261,268,273,285,287,288, 293-7, 304, 306, 308, 314-7, 319, 324,326,329, 333,334,336,33940, 346,367-8,381,396,398,410 Clark, S. L., 407 Clarke, J. C., 379 Clifft, M.A., xv Clum, G. A., 52, 103, 106, 234, 247, 301-2 Cobb, J. P., 149-50, 208 Cohen, A. S., 380 Cohen, S. D., 162 Cohn, T. E., 64 Collins, R. E. C., 64 Colp. R., 32 Conway, A. V., 72 Cook, B. L., 79 Cooper, M. L., 28 Corcoran, K., 254-5 Coryell, W., 5 Costa, P. T., 255 Costello, C. G., 240 cote,G., 135, 297 Cottraux, J., 165 Cox, B. J., 8, 16, 28, 38, 47-8, 102, 106, 225, 227, 233, 241 Cox, D. J., 186 Coyne, C., 79 Craig, T. J., 12, 178 Craske, M. G., xiii, 5-6, 9, 16-17, 37,

AUTHOR INDEX 499 46, 74, 83, 117, 128-30, 135, 138, 142, 149, 152, 166, 178, 181-3, 185, 187, 189, 190, 192-3, 195-6, 208, 226, 234-5,240,261-2,268, 283, 285, 287-92, 300, 302, 308, 313, 319, 326, 328, 330, 336, 340, 342, 353, 357, 359, 367, 371, 373, 379, 414, 416, 418, 426 Crawford, J., 225 Cross National Collaborative Panic Study, 138, 185-6 Crump, T., 15 Curtis, G. C., 174, 189-90, 199-200 Cutler, J., 416 Dager, S. R., 75 Dahl, A. A., 428-9, 431 Daiuto, A. A., 92-4, 208, 383-4 Daley, D. C., 144 Damas-Mora, J., 79 Dattilio, F. M., 359-60 Davanloo, H., 428-9 Davis, G. C., 26 de Beurs, E., 127, 139, 163, 168, 179, 183-6, 189, 191-2, 208-10, 238, 246, 248, 385 de Leon, A. C., 74 de Ruiter, C., 121, 127 DeGuire, S., 423 Delmonte, M. M., 432 Deltito, J. A., 138 Der, D.-F., 432 Devereux, R. B., 75 Dewey, D., 92-4 Diaferia, G., 30 DiNardo, P., 223 Dinwiddie, S. H., 11-12 Dixon, J. C., 66 Docherty, J.P., 20 Dohrenwend, B. S., 256 Donnell, C. D., 42 Dreessen, L., 199 Drossman, D. A., 27 Duane, T. D., 382 Dummit, E. S., 402 Dupont, R. L., 134 Eaton, W. W., 5, 9 Echeburua, E., 159 Edelman, R. E., 209-10 Edlund, M. J.,4

Ehlers, A., 37, 40, 41-2, 46, 61-2, 80, 190, 192-5, 203, 206 Eifert, G. H., 422 Eldridge, G.D., 419-21 Ellis, A., 45, 178 Elsenga, S., 413 Elsesser, K., 172-3, 381 Emery, G., 35, 119, 271, 308, 310, 322 Emmelkamp, P. M. G., 63, 108, 135-6, 149-50, 166, 179-80, 185-7, 18992, 195, 198, 204, 208-10, 332, 354, 389, 413 Emmelkamp-Benner, A., 209, 389 Epstein, N., 383 Erlandson, S. I., 68 Evans, I., 135, 142, 144, 189-90, 193, 195-6, 198, 208 Eysel, U. T., 63 Eysenck, H. J.,255 Eysenck, S. B. G., 255 Fahy, T. J., 155 Fairburn, C. G., 442 Falbo, J.,126-7, 129, 192 Falsetti, S. A., 413-5, 418 Faravelli, C., 11, 189-93, 195, 198-9, 207 Fava, G. A., 138, 143, 166, 189-90, 192, 194,199 Febbrano, G. A. R., 302 Fedoroff, I. C., 43, 413 Feldman, A. L., 382 Feske, U., 246, 434-6 Fewtrell, D., 8, 66, 68-9 Fiengo, M. N., 7 Fink, M., 81 Finlay-Jones, R., 11 First, M. B., 223-5 Fischer, J.,254-5 Fischer, M., 189-90, 192, 195 Fleet, R. P., 5 Flores, T. C., 77 Foa, E. B., 46, 112, 138, 339, 354 Folinsbee, L. J.,63 Folkman, S., 257 Fontaine, R., 171 Frances, A., 20 Frank, J.D., 115 Frankl, V. E., 359 Franklin, J.A., 135, 234, 302, 336-7, 340,358,377-9,413

500

AUTHOR INDEX

Freud, S., 8, 11, 13, 15,428 Fried, R., 71 Friedman, M., 387 Friedman, S., 383-4, 410-11 Frisch, M. B., 386 Furer, P., 25 Fux, M., 184 Fyer, A. J., 5, 12, 15, 27, 114, 134, 138 Garland, E. J., 407 Garssen, B., 69-70, 122 Gath, D., 156-7 Gelder, M. G., 332 George, D. T., 29 Gershon, M. D., 78 Gerz, H. 0., 360 Ghosh, A., 123, 300-1, 303-4 Gibbs, D. M., 72 Gillis, M. M., 95-6, 100, 133-4, 165, 180, 225, 232, 417, 436 Goldberg, R. J., 221 Goldfried, M. R., 431 Goldstein, A. J., 42, 240, 434-6 Golub, S., 79 Gordon, J. R., 144 Gordon, T., 387 Gorman, B. S., 84 Gorman, J.M., 70, 75, 97 Gould, R. A., 104, 107, 302 Gracely, E. J., 166, 189-90, 192-5, 202, 204,208-9,240 Gray, L. P., 77 Green, M. A., 189-90, 199, 200 Greenberg, P. E., 4 Greenberg, R. L., 242, 345 Gregg, S. F., 175 Griez, E., 69 Grossman, P., 70 Groth-Marnat, G., 217 Grunhaus, L., 24 Gustavson, B., 210 Guy, W., 218, 233 Hafner, R. J., 136, 142, 150, 156-7, 181, 187,189,192,198,208,354 Haley, J., 385 Hallam, R. S., 55, 149, 208 Hamann, M. S., 16, 31, 146, 148-9, 151, 199, 20d Hamilton, M., 233 Hand, I., 150-1

Harrington, P.H., 40-1 Harvey, J.M., 46 Hassan, R., 407 Hawthorn, J., 77 Hawton, K., 308 Hayward, C., 4 Hazell, J., 63-5 Heaton, K. W., 27 Hecker, J. E., 234, 301 Hedley, L. M., 145-6, 148, 191, 199, 200 Hegel, M. T., 138, 174, 176, 423 Heide, F. J ., 380 Heimberg, R. G., 255 Heller, R., 29 Hendersen, R. W., 35 Hibbert, G. A., 46, 69, 121-2 Higgitt, A., 171 Hillerbrand, E., 225 Himadi, W. G., 150, 186-7, 204, 208 Hobbs, W. R., 64 Hodgson, D., 187-8 Hoffart, A., 81, 112, 127-30, 135, 145-6, 148, 187, 191, 195, 199, 200, 242, 431 Hofmann, S. G., 69, 84, 146, 199, 200, 254, 416, 418-9 Holden, A. E., 53,254,301 Hollon, S. D., 255 Holmes, T. M., 256 Hope, D. A., 61 Hornig, C. D., 24 Hunsley, J., 92-4 Hunt, C., 135, 145, 153 Hunter, J. E., 87 Hyler, S., 145, 224 Ito, L. M., 127--8 Jackson, J. A., 379 Jacob, R. G., 18, 68, 221 Jacobsberg, L., 224 Jacobson, N. S., 89-92, 383 Jakubowski, P., 382 Jang, K. L., 41 Jannoun, L., 123, 187 Jansson, L., 192, 195, 208, 354, 395 Johnston, D., 135-6, 156-7, 193 Johnston, J. C., 52 Johnston, M., 16 Jorm, A. F., 232

AUTHOR INDEX 501

Kabat-Zinn, J., 436--41 Kagan, J. S., 56 Kahn, J., 416, 419 Kahneman, D., 265 Kamieniecki, G. W., 46 Karajgi, B., 73 Kaspi, S. P., 79-80 Katerndahl, 0. A., 4, 75 Katon, W., 11, 17, 26, 70, 74-5, 83, 219, 221,421 Katschnig, H., 409-10 Katz, B., 382 Keijsers, G., 149, 185-7, 192, 195, 199, 202, 208-11 Kellne1~ R., 7, 60, 77-8 Kelly, 0., 114 Kendall, P. C., 255 Kendell, R. E., 32 Kendler, K. S., 11, 34 Kennedy, B. L., 4 Kenny, 0. A., 112 Kern, J. M., 249 Kernberg, 0. F., 280 Kessler, R. C., 5, 23 Khawaja, N. G., 48, 243 Kilpatrick, 0. G., 414 Kirby, M., 158 Kirk, J., 238, 258 Kirsch, I., 59, 115 Klein, 0. F., 22, 81, 138 Klein, M. H., 70, 146, 224 Klein, R. G., 402 Kleiner, L., 142, 149, 182--4, 187 Klerman, G. L., 442 Klimes, I., 422-4 Klosko, J. S., 132, 181, 183--4, 186-7 Knapp, T. J., 15 Kniker, W. T., 78 Knowles, S. L., 52 Kobak, K. A., 220 Koch, W. J., 39--40 Kolko, 0. J., 142,405 Kopp, M., 70, 411 Kozak, M. J., 339 Kraaier, V., 72 Kraus, M., 195 Kreuger, M., 46 Kuch, K., 15, 26, 158 Kuipers, A., 135-6, 190, 194-5, 198 Kushner, M. G., 28 Kwon, S. M., 16,226

Laberge, B., 138, 186, 195-6 Lade1~ M. H., 171 Lamontagne, Y., 150-1 Lang, P. J., 248, 266 Lange, A., 150, 382, 385 Lantinga, L. J., 422-3 Lasley, 0. J., 64 Last, C. G., 12 Lazarus, A. A., 374 Lazarus, R. S., 113, 257 Lazovik, A. 0., 248 Lehman, C. L., 143 Lelliott, P., 11-12, 53, 158, 162, 195, 198-2, 208, 210 Leon, A. C., 4, 28 Levenkron, J. C., 422-3 Levenson, R. W., 28 Levine, B. A., 354 Levis, 0. J., 413 Lewington, P., 432 Lewis, L. M., 297-8 Ley R., 121-2 Liddell, A., 266 Lidren, 0. M., 88, 302 Liebowitz, M., 114, 138, 186, 189-90, 192-3, 195 Light, R. J., 87 Lilienfeld, S. 0., 221 Linehan, M. M., 207, 280 Lishman, W. A., 410 Livesley, W. J., 30-1 Loerch, B., 166 Logue, C. M., 181, 184 Loranger, A. W., 224 Louie, A. K., 27 Louro, C. E., 17 Lowe, C. F., 361 Lowenthal, D. T., 60 Lucas, R. A., 161 Lum, L. C., 67, 71-3 Lydiard, R. B., 27, 184, 407 MacKay, K. A., 382 Mackay, W., 266 Maddock, R. J., 186 Maffei, C., 224 Magarian, G. J., 73 Maier, W., 189, 192, 195, 233 Malan, D., 428 Malatesta, V. J., 278-9 Maller, R. G., 38, 40, 42

502

AUTHOR INDEX

Manning, A. P., 27 Mannuzza, S., 18, 223 Marchand, A., 192, 199 Margolin, G., 254, 383 Margraf, J., 37, 80, 127,238,411 Markowitz, J. S., 4 Marks, I. M., 18, 50, 53, 82, 88, 123, 135--6, 138, 141-2, 155-61, 167, 179, 182-4, 186-7, 192, 209, 225, 300-4,332, 354,382,433 Marlatt, G. A., 144 Marriott, J. A., 432 Marriott, P., 155 Marshall, W. L., 354 Martin, M., 35 Martin, R. L., 5 Martinsen, E., 199, 200, 242, 431 Maser, J. D., 81, 155, 217, 233, 237, 260 Mason, J., 181, 189 Matez, A., 432 Mathew, R. J., 63, 71 Mathews, A. M., 82, 88, 123-4, 141, 191, 193, 195, 198, 208-9, 211, 225, 248,300-2,332 Mathias, C. J., 381 Matthews, G., 63 Mattick, R. P., 100-1, 106, 177 Mattis, S. G., 402 Maurer, J., 172 Mavissakalian, M., 16, 30-1, 81, 123, 134, 144-5, 148-9, 151, 161-2, 181, 186, 189-92, 195,199,200,225 McCabe, B., 221 McCrae, R. R., 255 McDonald, R., 117 McLean, P. D., 24, 152, 181-5, 187, 196-7 McMullin, R. E., 303,322,335,392 McNally, R. J., xiii, 17, 24, 35, 37-40, 42, 46, 61, 112, 114 McNamee, G., 158, 179, 300-1 McPherson, F. M., 135-6 Meadows, E. A., 290 Meehl, P. E., 265 Meichenbaum, D. H., 178 Mellman, T. A., 9, 171, 199 Mersch, P. P., 108, 149 Michelson, E. L., 7, 11 Michelson, L., 81, 123, 127, 142-3, 149, 161-2, 181-4, 186, 189-93, 195, 202,207,210,266,359

Miller, J. J., 437, 439 Miller, L. S., 4 Miller, P. P., 66 Millon, T., 30, 146, 224 Milrod, B., 427-9, 431 Milton, F., 150, 208 Mineka, S., 35 Mizes, J. S., 225 Monteiro, W., 149 Moran, C., 66, 144 Moras, K., 233 Morey, L. C., 224 Morris, A., 74 Mullaney, J. A., 28 Munby, M., 135-6, 193 Munjack, D. J., 182, 186, 191, 195 Murphy, M. T., 171 Murrey, G. J., 11-12 Muskin, P., 138 Nagler, R., 77 Nagy, L. M., 134, 192 Naifeh, K. H., 70 Nathan, R. G., 171 Neeleman, J., 359 Nelles, W. B., 402 Neron, S., 88 Newman, M. G., 303-4 Nietzel, M. T., 249 Nisbett, R., 265 Norton, G. R., 9, 28, 227 Noshirvani, H., 158 Noyes, R., 25, 27, 33, 135, 145, 147, 185-7, 189-93, 195, 198-9 Nutzinger, D. 0., 195 Nuzzarello, A., 175 O'Connor, K., 8, 66, 68-9 O'Rourke, D., 190-1, 193, 195, 199 O'Sullivan, G., 158 Oatley, K., 187, 198 Obsessive Compulsive Cognitions Working Group, 255 Oehrberg, S., 164 Oei, T. P. S., 5, 48, 166, 225 Ogles, B. M., 225 Olajide, D., 381 Ollendick, T. H., 402, 405-6, 412 Onyett, S. R., 171 Ost, L.-G., 112, 126-30, 142, 150, 266, 354,372-3,375--6,380,395--6,398

AUTHOR INDEX 503

Ottaviani, R., 46 Otto, M. W., 63, 67, 112-4, 136, 166, 174-6, 189, 191-2, 203, 241, 261, 291, 327, 332, 340, 350-1, 359, 364, 368-71, 378-9, 381, 397 Overholser, J.C., 309, 311-2 Paradis, C. M., 410-1 Pecknold, J., 134-5, 138, 171, 186 Penava, S. J., 137-8, 229 Pennebaker, J. W., 59-61 Pere!, J. M., 134, 181, 186 Perna, G., 73 Persons, J. B., 219, 262, 265, 267-71, 273,277, 279-83, 389 Peterson, R. A., 39, 220, 241, 403 Petursson, H., 171 Pfohl, B., 224 Picornell-Darder, I., 388 Pikar, D., 416 Pillemer, D. B., 87 Pilsbury, D., 69 Pohl, R., 80, 162 Pollack, M. H., 73, 134-6, 166, 187, 191-5, 197,199,203 Pollard, C. A., 297-8, 305-6, 407, 426 Pollard, H. J., 306 Porzelius, J., 73 Pribor, E. F., 11-12 Pyke, R. E., 195 Quitkin, F. M., 28, 143 Rachman, S., 16, 53, 268 Radomsky, A. S., 8-9 Rahe, R. H., 256 Raj, A., 221 Raj, B. A., 407 Rapee, R. M., 6, 41, 120, 230, 234-6, 238, 246, 252, 368-9, 387 Rapp, M. S., 167 Rathus, J. H., 146-7, 199, 408-9 Ravaris, C. L., 138 Realini, J. P., 4 Rees, C. S., 4 Reich, J. H., 30, 145, 147, 185, 199,200 Reidenberg, M. M., 60 Reilly-Harrington, N. A., 112-14, 241 Reiss, S., 37-40, 42, 403 Rescorla, R. A., 82 Resick, P. A., 414

Resnick, H. S., 413 Rice, D. P., 4 Richardson, F., 172 Rickels, K., 135, 166, 171 Rifkin, A., 114 Rijken, H., 130, 189-91, 193, 195, 203 Riley, W. T., 159, 168 Riskind, J. H., 233 Roberts, D. K., 382 Roberts, W. W., 66 Robins, L. N., 223 Robinson, L., 425-6 Roelofs, S., 29 Rosenbaum, J. F., 21, 197 Rosenberg, R., 189-91, 195 Rosenthal, R., 87 Ross, C. A., 66 Ross, L., 265 Ross, M. W., 192, 198, 208 Rowe, D. C., 35 Rowe, M., 9, 426 Roy-Byrne, P. P., 171, 256 Rubinstein, B., 68 Rudd, M. D., 24 Salkovskis, P. M., xv, 43, 47, 50-1, 64, 69, 71-2, 80, 83, 95, 118, 120-4, 138, 178, 249, 261, 268, 285, 287-8, 293,295, 303,306,308,314-6,319, 324,329,333,339-40, 361,367-8, 396, 398, 410 Salovey, P., 265-6 Sanchez-Craig, M., 171 Sandberg, L., 416, 419 Sanderson, W. C., 46, 70, 114, 155, 408-9 Saper, Z., 413 Sarason, I. G., 256 Sartory, G., 168, 172, 184, 381 Saunders, B. E., 11 Scaturo, D. J., 431 Schaap, C., 115, 388, 393-4 Schatz, I. J., 75-6 Scheibe, G., 191-3, 195 Schindler, L., 210 Schmidt, F. L., 87 Schmidt, N. B., 38, 40-1, 61, 112, 114, 197, 210, 247, 252 Schmidt-Traub, S., 78-9 Schneck, J. M., 67 Schneider, S., 127, 411

504

AUTHOR INDEX

Schnicke, M. K., 414 Schulte, D., 219, 267 Schumacher, M. T., 15 Schwab, J. J., 4 Schweizer, E. E., 171 Scupi, B. S., 227-8 Sedman, G., 66 Segal, Z. V., 255 Segun, S., 158 Seidner, A. L., 404-5 Seligman, M. E. P., 53 Seuss, W. M., 72 Shapiro, F., 433-5 Sharp, D. M., 134, 138, 164 Shaw, S. C., 304 Shea. S. C.,217 Shear, M. K., 46, 75, 114--8, 183-4, 187, 217, 228-9, 233, 237, 260, 427, 429-30 Sheehan, D. V., 135, 138, 142, 166, 186, 191, 218, 221, 234 Shulman, I. D., 11, 27 Silove, D., 22 Silverman, W. K., 403 Siris, S. G., 416,419 Smaller, J. W., 135 Snaith, R., 233 Snidman, N., 56 Sobell, L. C., 254-5 Sobell, M. B., 254-5 Sochting, I., 112, 138 Sokol, L., 135, 138, 196, 199 Spiegel, D. A., 135, 159-60, 174, 174-6 Spielberger, C. D., 39, 232 Spiro, H. M., 77 Spitzer, R. L., 223-4 Stafrace, S., 432-3 Stampfl, T. G., 413 Starcevic, V., 20 Steer, R. A., 232-3 Stein, D. J., 30 Stein, M. B., 11, 41, 79, 252 Steketee, G., 345 Stern, R., 192,209,332,354 Stewart, S. H., 29 Stewart, W. F., 26, 41-2 Stockwell, T., 28 Strupp, H. H., 428 Stuart, R. B., 383 Suinn, R., 172 Svebak, S., 70

Swales, P. J., 408 Swallow, S. R., 255 Swann, A. C., 4 Swinson, R. P., 15, 158, 226, 251, 298, 300 Takahasi, T., 411 Taylor, C. B., 234 Taylor, I., 238, 299 Taylor, S., 8, 16, 21, 26, 28, 30-1, 34-5, 38-43,53-4, 86,153,202, 240-1, 254, 271, 295, 413, 435 Teasdale, J. D., 137 Telch, M. J., 40-1, 112, 126-30, 152-3, 161,186, 197,245-6,386 Thomas-Peter, B. A., 193, 198 Thompson, A. H., 23 Thompson, W. G., 27, 78 Thorpe, G. K., 234 Thorpe, G. L., 240 Timmons, B. H., 369 Tobefla, A., 186-7, 193, 195 Tompkins, M. A., 262, 265, 268-70, 273, 277, 283 Trakowski, J. H., 247 Trippett, C. J., 28 Trueman, D., 66 Trull, T. J., 88-9, 91, 225 Tsao, J.C. I., 140-1, 194 Tsuang, M. T., 221 Turk, D. C., 265-6 Turkat, I. D., 262, 280 Turner, S. M., 21 Turpin, G., 171 Tweed, L. J., 11 Tyrer, P., 20, 32-4, 171 Uhde, T. W., 5, 9, 11, 171 Ultee, K. A., 354 van Balkom, A. J. L. M., 105, 107-9, 168, 177 van den Hout, M., 117, 126, 128-30, 138 van der Hart, 0., 385 van der Hout, A., 166, 179-80, 185-6, 189-92, 195, 204, 208, 210 van Dyck, R., 150 van Etten, M., 86, 435 van Valkenburg, C., 195 van Zuuren, F. J., 225

AUTHOR INDEX 505 Victor, M., 29 von Korff, M., 9 Wade, S. L., 207 Wade, W. A., 142, 152-3, 184, 187, 189-90, 192, 194 Walker, E. A., 11 Waller, N. G., 66 Wardle, J., 156-7, 160, 183 Warwick, H. M. C., 43, 361 Watson, J. P., 195, 332 Watt, M. C., 42 Watts, F. N., 64-5, 70, 361 Weems, C. F., 403 Wegner, D. M., 333, 337 Weissman, A. N., 255 Weissman, M. M., 4-5, 24, 220, 442 Welkowitz, L. A., 184, 186 Wells, A., 124, 273, 285, 293-5, 308, 316, 327, 330, 334, 340, 372, 381 Wessels, H., 108, 209, 332, 354 Westling, B. E., 112, 126, 128-30, 202, 376 Westphal, C. F. 0., 14-15 Wetzler, S., 155 Wiborg, I. M., 428-9, 431 Wickwire, K., 158 Wieting, W., 381 Wild, A. J., 432 Wilkins, A., 63-5, 70 Wilkinson, G., 95, 97-8 Williams, J. B. W., 223

Williams, J. M. G., 35 Williams, K. E., 210 Williams, L. J., 67 Williams, S. L., 124-7, 129, 192, 344, 356 Wilson, G. T., 264-8, 442 Wilson, R. R., xiii, 302 Wilson, W. H., 71 Wittchen, H.-U., 219 Wolf, F. M., 81 Wolfe, B. E., 87, 155, 431 Wolkowitz, 0. M., 416 Wolpe, J., 374 Woodman, C. L., 186, 189-93, 195 Woods, S. W. Woody, S., 152, 197 Woolaway-Bickel, K., 210 Yamamoto, J., 411 Yardley, L., 18, 68-9 Yellowlees, P. M., 73 Zandbergen, J., 73 Zane, G., 124-5 Zapotoczky, H. G., 195 Zeanah, C. H., 413 Zebb, B. J., 395 Zillmann, D., 59 Zimmerman, M., 30, 224 Zinbarg, R. E., 21, 34 Zitrin, C. M., 161, 182-3, 187, 195 Zlotnick, C., 12

SUBJECT INDEX

Abbreviated treatment protocols, 296-7 Adherence to treatment common problems and solutions, 388-94 predictor of treatment outcome, 210, 213 Adolescents with panic disorder, 402-7 African Americans with panic disorder, 410-12 Age of onset of panic disorder predictor of treatment outcome, 181-2, 190-1, 212 typical ages of onset, 5 Agoraphobia assessment, 225-32 description and diagnosis, 14-18 relationship to panic disorder, 15 predictor of treatment outcome, 183, 192-3, 212 subtle form, 16-17 theories, 51-6 treatment, 81-3, 351-7 Agoraphobic Cognitions Questionnaire, 240-1 Agoraphobic Cognitions Scale, 242 Albany protocol, 285, 287-93 Alcohol-use disorders relationship to panic disorder, 28-30 treatment, 143-4 Allergic reactions, 1, 78-9 Alprazolam efficacy, 95-99, 101-9, 132, 138 interactions with other treatments, 157-61, 167-70 rebound panic, 135 side-effects, 113, 186 tapering, 174-6 Ambulatory monitoring, 253-4

American Academy of Child and Adolescent Psychiatry, 403,404 Anticipatory anxiety assessment, 228-31 treatment, 359 Antidepressant medications; see specific drugs (e.g., imipramine) Anxiety assessment, 232-3 effects of panic treatments on, 94, 101-3, 105 predictor of treatment outcome, 183, 193,212 Anxiety Control Questionnaire, 246-7 Anxiety Disorders Interview Schedule for Children for DSM-IV, 403 Anxiety Disorders Interview Schedule for DSM-IV, 223, 410 Anxiety management training, 172-3 Anxiety sensitivity assessment; see Anxiety Sensitivity Index defined, 37-8 effects of panic treatments on, 112-14, 168-70 origins, 41-3 predictor of treatment outcome, 185, 201-6, 212 role in anxiety disorders, 38-41, 54 see also Beliefs Anxiety Sensitivity Index, 220, 241, 245 Applied relaxation, 112, 126-31, 371-2, 376-7 Arrhythmia, 7, 75, 221, 432-3 Assertiveness effect of panic treatments on, 149 predictor of treatment outcome, 179, 212

508

SUBJECTINDEX

treatment of assertiveness deficits, 382 Attention biases, 61 internal versus external focus, 59-61 use of attentional shifts in treatment, 317 see also Interoceptive acuity Attitudes about treatment predictor of treatment outcome, 180, 185-6, 209, 212 strategies for modifying; see Socialization strategies Attributions for improvement in drug treatments, 167-8 Attrition comparison across treatments, 93-4, 96-8, 102-5, 107, 109-10, 173 predictors of, 180-8, 212-3 reducing attrition, 160-1, 389 Autonomic reactivity, 62-3 Avoidance; see Agoraphobia; Safety behaviours; Safety signals Avoidant personality disorder comorbidity with panic disorder, 30-1,32 effects of panic treatments on, 145, 147-8 predictor of treatment outcome, 200-1,212 'Bad days', causes of, 329 Balance control; see Vertigo Beck Anxiety Inventory, 232-3 Behavioural Avoidance Tests, 248-50 Behavioural Couples Therapy; see Couples therapy Behavioural experiments, 339-65 Behavioural inhibition, 21, 56, 429 Beliefs arousal-related, 37-8 assessment, 239-45 links with body sensations, 47-9 role in case formulation, 271-2 see also Overvalued ideation; Anxiety sensitivity Benzodiazepines efficacy, 95-9, 101-10, 135, 416 interactions with other treatments, 156-61

side-effects, 113 tapering, 171-6 withdrawal symptoms, 170-1 see also Alprazolam; Diazepam Bibliotherapy, 300-3 Blood-injury phobia, 141-2, 225 Body sensations induction of; see Interoceptive exposure interoceptive acuity and, 61-2 links with catastrophic beliefs, 47-9 natural causes of, 62-80 role of attention, 59-61 Body Sensations Interpretations Questionnaire, 241-2, 245 Body Sensations Questionnaire, 240-1 Body vigilance, 61, 80, 114, 239 Body Vigilance Scale, 247 Booster sessions, 154, 174, 293-4, 296, 396-7, 399-400 Borderline personality disorder, 21, 146-7, 199,201,280 Breadth of therapeutic effects of panic treatments, 140-52 Breathing, via chest versus abdomen, 70-1 Breathing retraining as a rational placebo, 121-3 efficacy, 120-1 indications and caveats, 366-8 protocols,368-70 troubleshooting, 370-1 Brief treatments; see Abbreviated treatment protocols Buspirone, 165-6 Caffeine, 13, 29, 37, 41, 45, 63, 66, 75, 77, 218, 318, 329 Cannabis; see Marijuana Carbon dioxide-induced panic, 46-7, 62, 114 Cardiac disorders, 74-5, 221-2, 382 Case formulations components, 269 criticisms of, 264-8 testing the formulation, 282-3 utility of, 262-4 versus manual-driven treatment, 266-7 Catastrophic Cognitions Questionnaire, 243

SUBJECTINDEX 509

Cerebral vasoconstriction causes of, 71-2 see also Hyperventilation Certainty, demand for, 328 Chest pain causes, 70-1, 74, 76, 221 description, 7, 26 treatment of non-cardiac chest pain, 123,421-4 Childhood Anxiety Sensitivity Index, 403 Childhood panic disorder, 402-7 Choking sensations, 7, 71, 77 Clark's cognitive model of panic description, 43-6 empirical support, 46-7 initial versus recurrent panic attacks, 47 relationship to anxiety sensitivity, 43 role of non-arousal sensations, 49-50 Clinical reasoning, 265-6 Clinical service settings, 152-3 Clomipramine, 97, 142, 428 Clonazepam,95, 97, 174-5 Cocaine, 12, 13, 27 Cognitive assessment, 239-47 Cognitive specificity hypothesis, 35 Cognitive-behaviour therapy commonly used package, 131 definitions of first and second therapies, xv, 82-3 Combining drug and psychological treatments efficacy, 156-66 mechanisms, 167-70 Comorbidity defined, 20 effects of panic treatments on, 140-9 models of, 20-1, 31-6 predictor of treatment outcome, 183-5, 194-201 prevalence in panic disorder, 22-3, 25-7, 28, 30 Complaints management training, 172-3 Computerized treatment, 303-5 Conditioning; see Operant conditioning Congruency hypothesis, 47-9 Control group confound, 86

Coping statements problems with, 330-1 use of, 359-60 Coping Strategies Questionnaire, 247 Cost-benefit analysis, 392 Cost-effectiveness of treatments, 104, 107 Couples therapy Behavioural Couples Therapy, 383-4 contribution to outcome of panic treatment, 92-5 see also Spouse-assisted treatment Cultural factors, 409-12 Darwin, Charles, 32-3, 73 Demographic variables as predictors of treatment outcome, 179, 181, 189-90, 212 Dependent personality disorder comorbidity with panic disorder, 30-1 effects of panic treatments on, 146-7, 149 predictor of treatment outcome, 200 Depersonalization and derealization categorical (taxonic) classification, 66 description, 7-8 causes of, 64, 66, 251 Depression comorbidity with panic disorder, 23-4 effects of panic treatments on, 89, 96, 101-3, 105,133,140,152 predictor of treatment outcome, 183-5, 194-7, 212 therapy for, 184-5, 197 worsening during panic treatment, 184-5 Devil's Advocate method, 328-9 Diagnostic evaluations screening methods, 219-20 structured interviews, 222-5 usefulness and limitations of, 219 Diazepam efficacy, 95, 97-8, 106 interactions with other treatments, 156-7, 160 side effects, 186 tapering, 170-3 Dieting, 80

510

SUBJECTINDEX

Difficulty level of interventions, 263, 348,353 Dismantling studies, 125-31 Distancing strategies, 288, 322-3, 440 Distraction, therapeutic use of, 172, 176,310,333, 336 Dizziness causes of, 67-9 description, 7-8 Downward arrow method, 243-5 Dropout; see Attrition Duration of panic disorder as a predictor of treatment outcome, 179, 181-2, 190-1,212 Effect size, 85-6 Emergency room treatment, 297-9 Empirical disputes, 324 End-state functioning, 135 Epilepsy; see Seizures Esophageal dysfunction, 76-7 Evolutionary models, 55 Expectancies, 51-3, 59 Experiential methods, 316-18 Exposure and response prevention, 345 Exposure probes, 247-53 Eye Movement Desensitization and Reprocessing, 433-6 Family therapy, 385 Fear Fighter, 304 Fear of fear; see Anxiety sensitivity Fear Questionnaire, 225-6 Fluorescent lighting, 63--4 Fluoxetine, 113, 142, 184 Fluvoxamine, 97, 99, 133--4, 136, 138, 142,146,151, 163-4 Focal versus non-focal interventions, 118-20 Followup studies of treatment efficacy, 109, 134--6 Fresh symptom emergence, 136 Freud, Sigmund, 8, 12, 13, 15, 428 Gastric motility dysfunction, 77-8 Gender gender _differences in disorder prevalence, 5 predictor of treatment outcome, 179, 181, 189-90,212

General medical conditions mimics of, or contributors to panic, 58,220-2 sources of body sensations, 68, 71-2, 73--4, 75, 78-9 General neurotic syndrome, 31-3 Generalized anxiety disorder comorbidity with panic disorder, 22-3 effects of panic treatments on, 140-1 predictor of treatment outcome, 194-5 Genetics, 34-5, 41-2 Global severity of panic disorder, as a predictor of treatment outcome, 182, 191-2, 212 Goal setting, 321-2 Guided mastery treatment, 124-30, 146 Hamilton Anxiety Rating Scale, 233 Headache,26, 64-5,221,347,426 Hierarchic structure of psychopathology, 34-5 High end-state functioning; see Endstate functioning History, personal and family, 218, 255-6 Histrionic personality disorder comorbidity with panic disorder, 30-1 effects of panic treatments on, 145-8 predictor of treatment outcome, 199 Homework adherence problems and solutions, 238, 389-93 descriptions, 279, 281-2, 287-8 predictor of treatment outcome, 210 see also descriptions of specific interventions Hostility, as a predictor of treatment outcome, 198,212 Hyperventilation acute, 71, 72 chronic, 72-3 effects of, 71-2 thoracic breathing, 70-1 environmental triggers, 70 use in assessment, 250-3 use in treatment, 340-2, 347 see also Breathing retraining

SUBJECTINDEX

Hypnosis, 432-3 Hypochondriasis, 25, 32, 50, 211, 361, 425 Hypoglycemia, 8, 14, 80, 221, 273 Imagery methods, 331-5 Imipramine, 81, 95, 97-110, 113, 132-3, 138,142, 144-6, 161-3,416 Inductive reasoning, 309-10 Inpatient treatment, 305-6 Interoceptive acuity effects of panic treatments on, 114 predictor of treatment outcome, 185, 206, 212 role in panic disorder, 61-2 Interoceptive exposure combined with other interventions, 357-60 contraindications, 346 efficacy, 126-31 examples of exercises, 340-4 exposure probes for assessment purposes, 250-3 protocols, 346-51 rationale, 339 troubleshooting, 362-5 Interpersonally-focused interventions assertiveness training, 382 couples and family interventions, 382-5 Interpretive bias, 46 Irritable bowel syndrome comorbidity with panic disorder, 27 definition, 27 treatment, 419-21 Isolated sleep paralysis, 9, 410 Lactate-induced panic, 46, 62, 114 Life Experiences Survey, 256-7 Lighting conditions, 63--4 Living circumstances, 256 London/Toronto study, 157-9 Long-term outcome, 109, 134-6 Lorazepam,97, 103,105,286 Maintaining treatment gains, 394-9 Major depression; see Depression Marijuana, 12, 13, 27, 66 Marital (dyadic) satisfaction effect of panic treatment on, 149-51

511

predictor of treatment outcome, 181, 187-8, 190,208,212 quality of marital relationships in panic disorder, 382-3 see also Couples therapy Medical conditions; see General medical conditions Medical evaluation, 220-2 Medication discontinuation, 134-5, 170-6 Menstrual cycle, 79-80 Meta-analysis basic concepts, 84-7 strengths and limitations, 87 Mindfulness Meditation description and rationale, 436-9 efficacy in treating panic disorder, 439 exercises, 438 integration with other treatments, 440-1 Mitra! valve prolapse, 75 Mobility Inventory, 226-7 Moclobemide, 166 Monoamine oxidase inhibitors (MAOis), 71, 103-6, 221 see also Moclobemide; Phenelzine Mortality associated with panic disorder, 4-5 Motivation for treatment predictor of treatment outcome, 185-6, 209, 213 strategies for improving motivation, 388-94 Multiple-Channel Exposure Therapy, 413-15 Munch, Edvard, 28-9 Myocardial infarction, 221 Neuroticism assessment, 255 cognitive factors, 35 defined, 34, 35 effect of panic treatments on, 144-5 predictor of treatment outcome, 198, 212 see also General Neurotic Syndrome NIMH Panic Questionnaire, 227-8 Nocturnal panic, 9, 45-6, 315-6, 426 Non-cardiac chest pain; see chest pain Nonprescriptive treatment, 115-18

512

SUBJECTINDEX

Nonspecific factors in panic treatments, 115-18 Obsessive-compulsive symptoms comorbidity with panic disorder, 22-3 effects of panic treatments on, 142-3 Obstacles to successful treatment common problems and solutions, 388-94 relapse prevention, 397-9 use of case formulation to identify and overcome obstacles, 269, 281-2 Older adults with panic disorder, 407-9 Operant conditioning, 51-2, 273 Orthostatic hypotension, 75-6 Overestimation of probability or cost of aversive events, 52-3, 326-8 Overprediction bias, 52-3 Overvalued ideation defined, 204 predictor of treatment outcome, 204-6 treating, 331, 360-2 Oxford protocol, 293-5 Pacing of treatment, 262-3 Pain, 26-7, 372 see also Chest pain; Headache Paired-words task, 46, 317 Palmtop computer-assisted treatment, 303-4 Palpitations, 7, 75 Panic and Agoraphobia Scale, 229-30 Panic Appraisal Inventory, 245-6 Panic-associated Symptom Scale, 230-1 Panic Attack Questionnaire, 227 Panic attack records description, 233-7 nonadherence problems and solutions, 237 Panic attacks definition and symptoms, 5-9 DSM-IV typology of panic, 9-10 limited symptom attacks, 5 nocturnal panic, 9, 45-6, 315-6, 426 predictor of treatment outcome, 182, 192, 212

prospective versus retrospective assessment, 238-9 treatment-related increases, 139 Panic Belief Questionnaire, 242 Panic Disorder Severity Scale, 228-9 Panic disorder diagnosis, 5 differential diagnosis, 13-14 precipitating factors, 11-13 relationship to agoraphobia, 15 Panic-focused Psychodynamic Psychotherapy, 429-31 Paradoxical intention, 359---60 Paranoid personality disorder effects of panic treatments on, 146-8 predictor of treatment outcome, 185, 199, 201 treatment issues, 393 Paresthesias, 7, 63 Paroxetine, 97, 164-5 Paroxysmal atrial tachycardia, 221-2, 382 Patterns of symptom change, 137-40 Perpetuating factors, 275-6 Personality disorders assessment, 224-5 comorbidity with panic disorder, 30-1 effect of panic treatments on, 145-9 predictors of outcome of panic treatment, 185, 198-201 treatment implications, 280, 393-4

see also specificpersonality disorders Phenelzine, 95, 142 Pheochromocytoma, 13,221 Physical fitness, 80 Physiological assessment; see Psychophysiological monitoring Pill phobia, 83-4 Placebo as an explanation of breathing retraining, 121-3 nonspecific factors in panic treatments, 115-15 Posttraumatic stress disorder, 413-15 Practice Guidelines for Psychiatric Evaluation of Adults, 220-1 Precipitating factors, 11-13, 269, 274-5 Predisposing factors, 40-1, 269,274 Pregnancy, 347, 425-6 Premenstrual symptoms, 79-80

SUBJECTINDEX

PRN (as needed) medications, 8, 169, 286,363 Problem context, 269, 272-3 Problem list, 269-71 Programmed practice, 123 Prospective monitoring assessing panic attacks, 233-7 homework monitoring, 238 monitoring symptoms other than panic attacks, 237-8 versus retrospective monitoring, 238 Protective factors, 269, 276-7 Psychodynamic psychotherapy contemporary treatments, 428-31 historical perspectives, 427-8 integration with other treatments, 431-2 Psychoeducation, 313-21 Psychophysiological monitoring, 253-4 Pulmonary diseases, 73-4 Puppy, reincarnation as, 328 Quality of Life Therapy, 386-7 Questions, types used in therapy, 310-12 Rate of symptom change across treatments, 137-40 as a predictor of treatment outcome, 187, 211, 213 Reinforcers, role in treatment, 390-1 Relapse prevention, 397-9 Relaxation training applied relaxation, 376-7 choosing among relaxation procedures, 379 indications and caveats, 371-3 in situ isometric relaxation, 377-8 progressive muscle relaxation, 373-5 relaxation cue procedure, 378-9 troubleshooting, 379-81 Relaxation, as a source of panic or anxiety, 61, 360, 380-1, 440 Respiration; see Hyperventilation Respiratory diseases, 73-4 Safety behaviours, 50-1, 53, 123, 124-5, 176, 231-2, 236, 246, 247, 249-50, 292,293,294,296, 330-1,

513

336-7,339,344-5,356-8,363,364, 367,370, 372,380-2 Safety signals, 53-4, 226-7, 292, 294, 316, 319, 344-5, 348-9, 352, 357, 358,363 Schedule for Affective Disorders-Lifetime, Anxiety, 223 Schizophrenia, 416-19 Seasonality and panic disorder, 12 Secondary gains, 258, 281-2, 392-3 Seizures, 13, 64, 65-6, 171, 220-1, 222, 252,347,352 Selective serotonin reuptake inhibitors, 97, 100, 103-5, 110, 163-5 see also Clomipramine; Fluoxetine; Fluvoxamine; Paroxetine; Zimeldine Separation anxiety, 22, 55-6, 403 Side effects drugs, 80, 83-4, 110, 113, 161-2, 163, 169, 186, 213 psychological treatments, 186-7, 213, 313, 382 Situational exposure contraindications, 352 exposure probes for assessment purposes, 248-50 protocols, 352-7 rationale, 339, 351-2 troubleshooting, 362-5 Sleep paralysis (see Isolated sleep paralysis) Social phobia comorbidity with panic disorder, 22-3 effects of panic treatments on, 140-2 predictors of treatment outcome, 194-5 Socialization strategies, 314-21 Socratic dialogue, 308-12 Space phobia, 18 Specific phobias comorbidity with panic disorder, 22-3 effects of panic treatments on, 140-2 Spouse (partner) assisted treatment, 92-5, 187-8,208 State-dependent learning, 167 State-Trait Anxiety Inventory, 232 Stress Reduction and Relaxation

514

SUBJECTINDEX

Program; see Mindfulness Meditation Stressful life events assessment, 256-7 predictor of treatment outcome, 187, 206-7,212 relationship to panic onset, 11-12 Structured Clinical Interviews for DSM-IV, 223-4 Structured interviews, 222-5 Style of therapy and structure of sessions, 288-9 Subjective Units of Distress, 248 Substance-use disorders, 27 see also Alcohol-use disorders; Cocaine; Marijuana Suicidal ideation and behaviour, 24-5 Suitability for treatment, 285-6 Systematic desensitization, 331-2, 433 Systematic questioning, 310-12 Telephone-administered treatment, 299-300 Testing the case formulation, 282-3 Therapeutic relationship, 278-81 Therapist characteristics predictor of treatment outcome, 180, 186, 210-1, 212 tailoring to particular patients, 278-81 Therapist training and evaluation, 306-7 Therapist-assisted exposure, 356-7 Thought stopping, 336-7 Timeline Follow-Back Procedure, 255-6

Timing and targets of treatment, 263-4 Transfer of excitation, 59 Trazodone, 95 Treatment packages (cognitivebehavioural therapy), 125-32 Treatment plan, 269, 277-81 Treatment obstacles, 281-2 Treatment refusal, 179-80, 212-13 Trembling, 7, 38, 43, 63, 80, 251, 309, 332,343, 344-5,358,360 Tricyclic antidepressants; see Imipramine Vagal innervation techniques, 381-2 Valsalva manoeuvre, 172-3, 176, 381-2 Verbal disputations description of methods, 322-9 troubleshooting, 329-31 Vertigo, 67-9 Vestibular disorders, 67-9 Vicious cycle of panic, 43-51 Virtual reality therapy, 303 Visual patterns, 64-6 Visual perception changes in the visual field, 67 lighting conditions, 63-64 role of hyperventilation, 67 Vomiting, 45, 48, 76-8, 171, 221, 341, 347,426 Westphal, Carl F. 0., 14-5 Wilkins' grid, 64-6 Working hypothesis, 269, 273-7 Zimeldine, 97

The Wiley Series in

CLINICAL PSYCHOLOGY Paul Chadwick, Max Birchwood and Peter T,·ower Peter Sturmey Frank Tallis

Cognitive Therapy for Delusions, Voices and Paranoia Functional Analysis in Clinical Psychology Obsessive Compulsive Disorder: A Cognitive and Neuropsychological Perspective

David Fowler, Philippa Garety and Elizabeth Kuipers

Cognitive Behaviour Therapy for Psychosis: Theory and Practice

Robert S. P Jones, Peter G. Walsh and Peter Stunney

Stereotyped Movement Disorders

D. Colin Drummond, Stephen T. Tiffany, Steven Glautier and Bob Remington (Editors) Carlo Perris, Willem A. Arrindel/ and Martin Eisemann (Editors) Chris Barker, Nancy Pis/rang and Robert Elliott Graham C. L. Davey and Frank Tallis (Editors) Paul Dickens

Addictive Behaviour: Cue Exposure Theory and Practice

Parenting and Psychopathology

Research Methods in Clinical and Counselling Psychology Worrying: Perspectives on Theory, Assessment and Treatment Quality and Excellence in Human Services

Edgar Miller and Robin Morris

The Psychology of Dementia

Ronald Blackburn

The Psychology of Criminal Conduct: Theory, Research and Practice

516

THE WILEY SERIESIN CLINICAL PSYCHOLOGY

Ian H. Gotlib and Constance L. Hammen

Psychological Aspects of Depression: Toward a Cognitive-Interpersonal Integration

Max Birchwood and Nicholas Tarrier (Editors)

Innovations in the Psychological Management of Schizophrenia: Assessment, Treatment and Services

Robert

J. Edelmann

Alastair Agar (Editor)

Bob Remington (Editor)

Colin A. Espie David Peck and C. M. Shapiro (Editors) Roger Baker (Editor) Friedrich Fosterling Anthony Lavender and Frank Holloway (Editors) John Clements

Anxiety; Theory, Research and Intervention in Clinical and Health Psychology Microcomputers and Clinical Psychology: Issues, Applications and Future Developments The Challenge of Severe Mental Handicap: A Behaviour Analytic Approach The Psychological Treatment of Insomnia Measuring Human Problems: A Practical Guide

Panic Disorder: Theory, Research and Therapy Attribution

Theory in Clinical Psychology

Community Care in Practice: Services for the Continuing Care Client Severe Leaming Disability and Psychological Handicap

Wiley

Series

in

Clinical

Psychology

UnderstandingandTreatingPanicDisorder Cognitive-Behavioural Approaches Steven Taylor, Department of Psychiatry, University of British Columbia, Canada This book is a comprehensive text and clinician'sguidewhich integratestheory,empirical findings,andtreatmentguidelines, to providea frameworkfor understanding andtreatingboth routineandcomplexcasesof panicdisorder. Thefirst Partof the bookcoversthetheoreticalfoundationsof cognitive-behavioural treatment {CBT)for panicdisorder{with or without agoraphobia), and the relevantempiricalfindings. Othertreatments for panicdisorder,suchaspharmacotherapies, arealsoreviewed, asa guideto selectingthe most appropriatetreatment.Importantclinicaloutcomes,such as treatment dropout,response, andrelapse,arealsodiscussed. The secondPartof the bookdescribesthe clinicalprotocolsand proceduresfor cognitivebehavioural assessment andtreatment.Theauthoremphasizes a caseformulationapproachto treatmentandincludestreatmentprotocolsfor uncomplicated casesaswell as guidelinesand strategiesfor dealingwith moredifficultcases.Thelatterincludecasesof panicdisorderthat havefailedto respondto conventional CBTapproaches, andcasesin whichpanicdisorderis comorbidwith otherdisorders(e.g.,schizophrenia, posttraumatic stressdisorder).Protocols are alsodescribed for implementing CBTin specificsettings{e.g.,emergency rooms,ruralsettings), specificpopulations (e.g.,children,adolescents, the elderly),andparticularculturalmilieux(i.e., culture-specific aspectsof treatment). Traineesand practitioners in clinicalpsychology, psychiatry,nursingandothermentalhealth disciplineswill welcomethis comprehensive and evidence-based guide to conceptsand treatmentof panicdisorder.

"StevenTaylor ...in thismasterfulbookhaswrittentheclearest, mostdetailedaccountavailable on contemporary CBTmethodsfor treatingthissyndrome ...his excellentbookshouldbe read by all mentalhealthprofessionals whotreatpanicdisorder.It providesthetoolsoneneedsto help thesepatientsin the fastest,most effectiveway,and it documentsthe sciencethat undergirds themethods."Professor RichardMcNally,HarvardUniversity,USA "Theclearestandmostcomprehensive textnowavailable oncognitive-behavioural approaches to panic.It is remarkable notonlyfor thebreadthanddepthof theoretical, empiricalandclinical coverage, butalsofor thewaytheseareintegrated. Noonewhois seriousaboutunderstanding andhelpingpeoplewith panicshouldbe withoutit." ProfessorPaulSalkovskis, Universityof OxfordDepartment of Psychiatry, UK This book appears in The Wiley Series in Clinical Psychology

ISBN 0-471-49067-9

JOHN WILEY & SONS, LTD Chichester• New York• Weinheim • Brisbane • Singapore • Toronto