Preliminary evaluation of a broad-spectrum cognitive-behavioral group therapy for anxiety

https://doi.org/10.1016/j.jbtep.2004.07.002Get rights and content

Abstract

This study involved the development and outcome evaluation of a cognitive-behavioral group treatment program for anxiety disorder using a randomized controlled design. This treatment program deviated from traditional anxiety disorder treatment protocols in that anxiety diagnosis was de-emphasized and treatment focused on shared common features across the anxiety disorders. Twenty-three participants were recruited and randomly assigned to either immediate treatment or waitlist control conditions. Nine randomly assigned participants representing a range of anxiety diagnoses completed the 12-week group treatment, and were compared with 10 waitlist control participants who stayed in the study during the waitlist period. Participants in the two conditions were compared on change in diagnosis, diagnostic severity, self-report fear on ideographic measures, and self-report questionnaires of anxiety and state negative affect. Results were generally supportive of the efficacy of the treatment program. Compared to controls participants, those receiving treatment showed significantly greater improvement on diagnostic measures and ideographic fear-avoidance hierarchies. Data from self-report measures of anxiety and state negative affect were less supportive of the treatment efficacy. Implications for emerging conceptualizations of anxiety disorders, as well as implications for treatment and treatment dissemination, are discussed.

Introduction

Anxiety disorders represent a major national mental health problem, following substance use disorders as the second most prevalent category of psychological diagnoses in the United States, with a lifetime prevalence estimate of 24.9%. Furthermore, when limiting to 12-month prevalence, anxiety disorders surpass substance use disorders as the most prevalent class of diagnoses (Kessler et al., 1994). Prevalence estimates Canada (Offord et al., 1996), Australia (Henderson, Andrews, & Hall, 2000), and the Netherlands (Bijl, Ravelli, & Van Zessen, 1998) are generally similar to the estimates from the United States.

Not only are anxiety disorders prevalent, but they also yield a considerable economic impact. Greenberg et al. (1999) estimated the overall economic impact of anxiety disorders (in 1990 dollars) to be over $42 billion per annum in the United States, or an annual cost of over $1500 per individual with an anxiety disorder. Anxiety disorders are also associated with considerable impairment in quality of life (Hanson, 2002; Mendelowitz & Stein, 2000) and disability (Andrews, Issakidis, & Slade, 2000). Indeed, although anxiety disorders do not tend to be associated with as much individual disability as some other psychological disorders, when examined in relation to prevalence, the total societal disability impact of anxiety disorders exceeds that of schizophrenia or other “major mental illnesses” and is second only to major depression (Andrews et al., 2000).

Despite the foreboding prevalence and impact statistics, many highly effective treatments exist for the anxiety disorders. Meta-analyses of psychological and pharmacological treatments consistently reveal strong treatment effects of cognitive-behavioral treatments (CBT) and various pharmacological agents for panic disorder and agoraphobia (e.g., Bakker, van Balkom, Spinhoven, Blaauw, & van Dyck, 1998), social anxiety disorder (e.g., Fedoroff & Taylor, 2001), obsessive-compulsive disorder (OCD; Kobak, Greist, Jefferson, Katzelnick, & Henk, 1998), generalized anxiety disorder (GAD; Gould, Otto, Pollack, & Yap, 1997), and post-traumatic stress disorder (PTSD; Sherman, 1998). However, only approximately 20–30% of individuals with anxiety disorders actually receive such treatments (Young, Klap, Sherbourne, & Wells, 2001). Clearly, the problem lies not with our treatments, but with their delivery.

Two issues are pertinent to the problem of treatment delivery: dissemination of effective treatments and accessibility to effective treatments. With regard to dissemination, Addis, Wade, and Hatgis (1999) suggest that concerns regarding training expenses and time demands are primary factors underlying poor dissemination of empirically supported treatments. Indeed, these concerns are not unwarranted, considering the financial and time costs of purchasing and studying treatment manuals and training materials, attending training seminars, and receiving supervision in the delivery of treatments for one, much less six or more anxiety diagnoses. In addition, there are only a limited number of clients that a trained CBT therapist can treat in any given period of time. One potential method to increase this number is by providing group CBT. However, this is not feasible in many settings. In areas of smaller population or greater therapist saturation, it is difficult to obtain sufficient numbers of patients with the same diagnosis, who present to clinic within a similar time-frame, with similar availabilities for scheduling to run one, much less several, CBT groups.

One treatment delivery model that holds potential for improving adoption and accessibility is a CBT for diagnostically heterogeneous anxiety groups. This treatment model posits that, from a clinical intervention viewpoint, the similarities among the anxiety diagnoses outweigh their differences. Indeed, this model holds closely to emerging models of anxiety that implicate a common core pathology across the different diagnoses (see Barlow, 2002; Craske, 1999). This construct bears close resemblance to the constructs of negative affectivity (Clark & Watson, 1991), neuroticism (e.g., Eysenck, 1967), and trait anxiety (Gray, 1982; Speilberger, 1985). Indeed, following reviews of the primary characteristics of these constructs, (Zinbarg and Barlow 1996; Barlow, 2000) concluded that the similarity and overlap among the definitions of these constructs is so pronounced that the terms likely reflect the same construct.

If the premise of a common core pathology were true, it should follow that treatments acting upon the core pathology should be effective regardless of the specific feared stimuli. As noted earlier, CBT approaches, which typically all share the same core therapeutic components—education, cognitive restructuring, and exposure to feared stimuli (Barlow & Lehman, 1996; Brown & Barlow, 1992)—have demonstrated considerable effectiveness in treating anxiety disorders. Some differences do exist among the common components, although these differences, however, are differences in content, not function. For example, the format for exposure depends on the nature of a client's fears, but the rationale for exposure, overcoming fears by facing them, does not vary across disorders. Indeed, Borkovec, Abel, and Newman (1995) and Brown, Antony, and Barlow (1995) have reported that following treatment for a primary anxiety diagnosis, un-targeted comorbid anxiety diagnoses typically abate.

In light of this conceptualization of clinical anxiety as sharing a common pathological entity, this study was conducted to evaluate the efficacy a single anxiety disorder treatment protocol designed for use with diagnostically heterogeneous groups of clients with a primary anxiety diagnosis. A waitlist control design was employed, as it affords the methodological advantages of no-treatment control condition, while overcoming the ethical problems surrounding withholding treatment (Kazdin, 1992). The treatment was a 12-week group CBT protocol (Norton & Hope, 2002) that de-emphasizes the differences between specific diagnoses or feared stimuli and highlights the similarities in the core features.

Three general sets of hypotheses were proposed. The first set of hypotheses were that participants in the treatment condition would demonstrate significant improvement on diagnostic indices, including (a) the proportion meeting criteria for a clinically severe anxiety diagnosis, and (b) Clinician Severity Ratings (see below) of the primary and all anxiety diagnoses. The second set of hypotheses were that participants in the treatment condition would show a significant decrease from pre-treatment to the end of treatment on measures of anxiety symptoms, the DASS-Anxiety and MASQ Somatic Anxiety subscales, and Fear Avoidance hierarchy ratings, whereas no significant change would be observed among participants in the waitlist control condition. Finally, it was hypothesized that during treatment, particularly during the second phase of treatment, participants in the treatment condition would show improvement on measures of the common core pathology whereas no change in on this construct was anticipated for waitlist participants. To assess change on this construct, state measures were required. In order to avoid over-aggrandizing any results from these data (i.e., implying that change is necessarily occurring on the trait dimension of negative affectivity) the term state negative affect will be used throughout this manuscript to refer to these observations.

Section snippets

Participants

Twenty-three individuals (60.9% women) meeting DSM-IV criteria for an anxiety disorder were recruited for participation via advertisements and articles in local and neighborhood newspapers, referrals from health and mental health professions, and public service announcements. The following criteria were established for inclusion in the study: (a) age 19 or older; (b) primary DSM-IV diagnosis of an anxiety disorder; (c) adequate proficiency in English; (d) willingness to accept random

Results

All measures were assessed for multivariate outliers, univariate outliers, and distribution normality. No multivariate outliers were identified within the data at any assessment timepoint. Two univariate outliers were noted in the post-treatment and were Windsorized, replacing the outlying data with non-outlying values while retaining the sequential order among the outliers (Hoaglin, Mosteller, & Tukey, 1983). Finally, the data were examined for skew, defined a priori as >±0.8, but no variable

Discussion

The results of this study were partially, albeit not completely, supportive of the efficacy of a broad-spectrum CBT protocol in treating a diagnostically heterogeneous sample of individuals with anxiety disorders together in a group format. Six of nine participants receiving the treatment no longer met criteria for a clinically significant anxiety disorder diagnosis at post-treatment, while none of the waitlist control participants’ symptoms spontaneously remitted to sub-clinical levels.

Acknowledgements

This research was conducted as part of the first author's Ph.D. dissertation at the University of Nebraska-Lincoln. P.J. Norton thanks his committee members (D.A. Hope, R. Deinstbeier, M. Scheel, and W. Spaulding) for their helpful comments. The authors would also like to thank T.J. De Coteau, L.S. Ham, S.A. Hayes, E. Kimberly, C. Kraus, and E.L. Moore, who served as therapists and diagnosticians for this study.

References (48)

  • D.H. Barlow et al.

    Advances in the psychosocial treatment of anxiety disordersimplications for national health care

    Archives of General Psychiatry

    (1996)
  • R.V. Bijl et al.

    Prevalence of psychiatric disorders in the general populationresults of the Netherlands Mental Health Survey and Incidence Study (NEMESIS)

    Social Psychiatry and Psychiatric Epidemiology

    (1998)
  • T.D. Borkovec et al.

    Effects of psychotherapy on comorbid conditions in generalized anxiety disorder

    Journal of Consulting and Clinical Psychology

    (1995)
  • T.A. Brown et al.

    Comorbidity among anxiety disordersimplications for treatment and DSM-IV

    Journal of Consulting and Clinical Psychology

    (1992)
  • T.A. Brown et al.

    Anxiety disorders interview schedule for DSM-IV (Adult Version)

    (1994)
  • T.A. Brown et al.

    Diagnostic comorbidity in panic disordereffect on treatment outcome and course of comorbid diagnoses following treatment

    Journal of Consulting and Clinical Psychology

    (1995)
  • T.A. Brown et al.

    Reliability of DSM-IV anxiety and mood disordersimplications for the classification of emotional disorders

    Journal of Abnormal Psychology

    (2001)
  • L.A. Clark et al.

    Tripartite model of anxiety and depressionpsychometric evidence and taxonomic implications

    Journal of Abnormal Psychology

    (1991)
  • M.G. Craske

    Anxiety disorderspsychological approaches to theory and treatment

    (1999)
  • H.J. Eysenck

    The dynamics of anxiety and hysteriaan experimental application of modern learning theory to psychiatry

    (1957)
  • I.C. Fedoroff et al.

    Psychological and pharmacological treatments of social phobiaa meta-analysis

    Journal of Clinical Psychopharmacology

    (2001)
  • H. Friedman

    Simplified determinations of statistical power, magnitude of effect and research sample sizes

    Educational & Psychological Measurement

    (1982)
  • J.A. Gray

    The neurobiology of anxiety

    (1982)
  • P.E. Greenberg et al.

    The economic burden of anxiety disorders in the 1990s

    Journal of Clinical Psychiatry

    (1999)
  • Cited by (110)

    • Transdiagnostic Approaches to Mental Disorders

      2022, Comprehensive Clinical Psychology, Second Edition
    View all citing articles on Scopus
    View full text