Effectiveness of cognitive behavioural group therapy for social phobia in a community clinic: A benchmarking study

Abstract

This study used a benchmarking strategy to investigate the transportability of cognitive behavioural group therapy (CBGT) for social phobia to a community mental health clinic. The influence of common exclusion criteria on effect sizes was also examined. Patients (N=153) attended seven 4-h sessions of CBGT, which resulted in significant reductions in symptoms of social anxiety and depression. Effect sizes compared favourably to previous efficacy and effectiveness studies of both group and individual treatment. More than half of the treatment completers achieved reliable change. One-third achieved clinically significant change (CSC) on the Social Phobia Scale and the Beck Depression Inventory, but fewer achieved CSC on the Social Interaction and Anxiety Scale. Restricting the sample on the basis of depression symptoms, age above or below 50 years, comorbidity, alcohol problems, or student status did not significantly moderate effect sizes for social anxiety. It is concluded that CBGT is effective within community mental health clinics.

Introduction

Social phobia is characterised by a fear of negative evaluation within social or performance situations, where the individual is under scrutiny and may be embarrassed (American Psychiatric Association, 1994). Epidemiological studies have shown that social phobia is common, unremitting without treatment, and debilitating (e.g. Andrews, Henderson, & Hall, 2001). Moreover, only a minority of individuals with social phobia seek treatment, and few seek an empirically supported treatment (Andrews, Issakidis, & Carter, 2001). Cognitive behavioural therapy (CBT) has been shown to be efficacious in treating social phobia in numerous randomised controlled trials (RCTs; Butler, Chapman, Forman, & Beck, 2006). Efficacy studies prioritise internal validity and thus typically adhere to strict inclusion and exclusion criteria, are well controlled, use well-trained and regularly supervised therapists, and monitor adherence to manualised protocols. However, dissemination of empirically supported treatments has probably been impeded by the perception that efficacious treatments will be less effective with more severe and complex cases seen within mental health clinics (Barlow, Levitt, & Bufka, 1999).

Benchmarking strategies have been used to compare outcomes from efficacy studies with naturalistic settings for depression (Merrill, Tolbert, & Wade, 2003), panic disorder (Wade, Treat, & Stuart, 1998), obsessive–compulsive disorder (Franklin, Abramowitz, Kozak, Levitt, & Foa, 2000), with heterogeneous groups (McEvoy & Nathan, 2007), and in private practice (e.g. Gaston, Abbott, Rapee, & Neary, 2006). However, few studies have examined the transportability of CBT for social phobia. Haug et al. (2000) found significant improvement following exposure therapy within a general practice setting. However, results were not directly compared with previous benchmarks and the study had a number of RCT qualities, such as excluding those with comorbid Axis I diagnoses and partial recruitment via newspaper advertisements. Lincoln et al. (2003) evaluated the generalisability of cognitive behavioural individual therapy (CBIT) for social phobia to naturalistic clinical settings by benchmarking outcomes to meta-analyses and previous effectiveness and efficacy studies. These researchers found effect sizes of a similar magnitude, thus concluding that CBIT can be transported to the field of mental health. Lincoln et al. (2003) also found that applying exclusion criteria commonly used in efficacy studies did not improve treatment outcome. Similarly, Gaston et al. (2006) found that cognitive behavioural group therapy (CBGT) for social phobia was transportable to private practice.

CBGT is more effective than placebo controls (Taylor, 1996), and as efficacious as CBIT (Fedoroff & Taylor, 2001). Moreover, CBGT is considered to be the treatment of choice for social phobia, because CBIT is less time- and cost-effective (Heimberg, 2001). Additional therapeutic benefits to group treatment include normalisation of symptoms, learning from others’ experiences, a social context within which to practice skills, and social pressure to comply with treatment and complete homework. In contrast, potential logistical and therapeutic problems with CBGT have been postulated, such as patients having to wait longer as groups are arranged, less flexibility with regard to scheduling sessions, and less individualised treatment strategies (Scholing & Emmelkamp, 1993; Stangier, Heidenreich, Peitz, Lauterbach, & Clark, 2003). In order to assess the potential impact of these factors, Stangier et al. (2003) conducted an RCT to compare the effectiveness of CBIT to CBGT for social phobia. Patients allocated to CBIT achieved larger effect sizes on measures of social phobia, but were comparable to those receiving CBGT on measures of mood and general symptomatology. Attrition was also comparable, suggesting that the reduced flexibility of groups was not detrimental to attendance.

Few studies have evaluated the transportability of CBGT for social phobia to community mental health clinics (CMHCs), where patients are typically referred by medical practitioners and are highly comorbid, and treatment programmes are less controlled than efficacy studies. Therefore, the first aim of this study was to benchmark treatment outcomes from CBGT in a community clinic to previous efficacy and effectiveness studies of CBGT and CBIT for social phobia. It may be expected that CBGT in real-world settings will be less effective than in research settings. The second aim was to partially replicate and extend Lincoln et al.'s (2003) study by examining the effect of moderate to severe depression, age above 50, comorbid anxiety or affective disorders, or student samples, on outcomes. In addition, the effect of potentially harmful levels of alcohol consumption was examined.