Psychological treatment of social phobia

SOCIAL PHOBIA

Psychological treatment of social phobia

Summary of behavioural and cognitive treatments for social phobia Behavioural treatments • Exposure therapy • Social skills training (SST) • Applied relaxation

Adrian Wells Dean McMillan

Cognitive treatments • • • •

In the 1980s social phobia was the ‘neglected anxiety disorder’. The diagnosis was new, few treatments were available and the evidence base was sparse. Today, the situation is more encouraging: the clinician can select from a wide variety of treatments, many of which have been extensively evaluated. This contribution describes and reviews psychological interventions for social phobia, grouping them into the broad categories of behaviour therapy and cognitive therapy (Figure 1). Psychodynamic approaches are not considered here.

1

Behaviour therapy Exposure therapy Exposure therapy for social phobia is the same as that used in the treatment of other phobias. In this approach the therapist and patient generate a graded hierarchy of anxiety-provoking situations, and the patient is then encouraged to enter and remain in the feared situation until the level of anxiety decreases. The least anxiety-provoking situation is tackled first, followed by the next step in the hierarchy and so on. In behavioural terms, social anxiety is seen as a learned (conditioned) response, which can be unlearned through prolonged, repeated exposure. This strategy – according to behaviour theory – will lead to the habituation of the anxiety. The efficacy of this and a range of other treatments for social phobia has been examined in four meta-analyses: • Feske and Chambless, 1995 • Taylor, 1996 • Gould et al., 1997 • Fedoroff and Taylor, 2001. However, the conclusions about exposure therapy are mixed (Figure 2). Feske and Chambless (1995) calculated a range of effect sizes, including self-report measures of social anxiety, cognition and depression assessed at post-treatment and follow-up. The majority of the effect sizes were significantly different from zero, and of the significant ones many were large, which suggests that exposure is effective. The most recent review (Fedoroff and Taylor, 2001) calculated three effect sizes: observer-rated measures, self-report measures of anxiety at post-treatment and self-report measures of anxiety at follow-up. Only one of these (anxiety at follow-up) differed significantly from zero, which raises doubts about the efficacy of this technique. The clinical significance of the treatment effects is also unclear. Even if we select from the four meta-analyses the individual exposure studies that have the largest effect sizes, the degree of change on standardized social phobia measures might not indicate a clinically significant reduction in symptoms. Exposure therapy is a strikingly effective treatment for a range of specific phobias, but why are the findings for social phobia less compelling? Data from Feske and Chambless (1995) provide a possible clue. Although the majority of the effect sizes for exposure

Treatment considerations Therapeutic techniques, no matter how effective, will be of benefit only if patients present to services and remain in treatment. This requires interaction with other people – the very experience that sparks fear in social phobia. Reluctance to seek help can be a difficulty for both pharmacological and psychological interventions, but psychological techniques also require extensive face-to-face contact with a clinician and most involve exposure to other social situations. It is therefore important to forestall these difficulties. It helps, for example, to explain to the patient that the interaction with the clinician, anxiety-provoking though it may be, is an opportunity to practise the skills learned in therapy. The clinician must also provide a compelling rationale for any exposure work and then check that the patient understands it. Drop-out rates for psychological and pharmacological approaches are comparable (Fedoroff and Taylor, 2001), which suggests that the face-to-face nature of psychological treatments is not an inevitable problem.

Adrian Wells is Reader in Clinical Psychology at the University of Manchester, Manchester, UK, Professor of Clinical Psychology at the Norwegian University, Trondheim, Norway, and Consultant Clinical Psychologist in Manchester Mental Health Partnership. His research interests include: cognitive factors in the cause and maintenance of emotional disorders; cognitive theory; and cognitive therapy of anxiety disorders, including social phobia, generalized anxiety, obsessive– compulsive disorder and PTSD. He is a pioneer of metacognitive theory and treatment based on information-processing approaches to emotion. Dean McMillan is a trainee clinical psychologist at the University of Manchester, Manchester, UK. His research interests include cognitive theory and the treatment of emotional disorders.

PSYCHIATRY

Cognitive restructuring Combined cognitive–behavioural techniques Cognitive–behavioural group therapy (CBGT) Clark and Wells’s (1995) cognitive therapy

56

© 2004 The Medicine Publishing Company Ltd

SOCIAL PHOBIA

were significant, of the three reflecting changes in cognition, two did not reach statistical significance. This could indicate that exposure does not produce long-term cognitive change. It is possible that cognitive processes, such as the allocation of attentional resources, prevent the patient learning from exposure experiences (Clark and Wells, 1995). Another possibility is that people with social phobia use a range of unhelpful coping strategies – termed safety behaviours – that also prevent learning during exposure (Clark and Wells, 1995). Unless these maintaining mechanisms are addressed directly, cognitive change is unlikely to occur and the treatment is unlikely to be effective.

The effect sizes for this approach are at the lower end of the range of those for other active treatments (Figure 2). Two of the reviews (Fedoroff and Taylor, 2001; Taylor, 1996) calculated effect sizes for attention-placebo conditions, which are designed to control for non-specific factors such as therapist attention and treatment credibility. In these two reviews, although the effect sizes for SST were significantly larger than those for wait-list controls, they were not significantly larger than those for the placebo conditions. This suggests that any improvement in functioning achieved by SST might be due to these non-specific factors. A recent variant of this approach, social effectiveness training, combines SST with exposure and has already produced some promising results, including evidence that it outperforms a placebo condition. However, the combination of treatments makes it difficult to establish whether the same level of change would have been achieved using exposure alone. While SST is consistent with a behavioural explanation of social phobia, for cognitive theorists the strategy is controversial. Cogni-

Social skills training Social skills training (SST) is based on the assumption that people with social phobia either have a deficit in their social skills or are unable to use these skills because of their anxiety. The treatment uses behavioural techniques such as modelling, role-play and corrective feedback to improve the person’s social performance.

Summary of meta-analyses examining psychological treatments for social phobia Meta-analysis/treatment

Mean effect size1

No. of trials

Drop-out rate (%)

Feske and Chambless (1995)2 Exposure Cognitive–behavioural therapy

9 12

0.99 0.90

10.0 11.9

Taylor (1996)3 Wait-list control Placebo Exposure Social skills training Cognitive therapy Cognitive therapy + exposure

5 5 8 4 5 11

-0.12 0.48 0.82 0.65 0.63 1.06

5.7 14.7 16.4 16.6 12.2 18.0

Gould et al. (1997)4 Exposure Social skills training Cognitive restructuring Cognitive restructuring + exposure

9 3 4 8

0.89 0.60 0.60 0.80

12.6 3.7 14.5 10.3

Fedoroff and Taylor (2001)5 Wait-list control Attention-placebo Exposure Social skills training Applied relaxation Cognitive restructuring Cognitive restructuring + exposure

9 4 7 7 4 7 21

0.03 0.47 1.08 0.64 0.51 0.72 0.84

5.9 19.7 18.2 16.6 10.0 11.0 18.0

1

Based on post-treatment self-report measures of anxiety. The meta-analyses report additional categories of effect size, but this is the only type reported by all four Effect sizes were calculated using difference between pre- and post-treatment, and were weighted by sample size 3 Effect sizes were calculated using difference between pre- and post-treatment, and were not weighted by sample size 4 Effect sizes were calculated using difference between treatment and control conditions at post-treatment, and were not weighted by sample size 5 Effect sizes were calculated using difference between pre- and post-treatment, and were weighted by sample size 2

2

PSYCHIATRY

57

© 2004 The Medicine Publishing Company Ltd

SOCIAL PHOBIA

tive models (e.g. Clark and Wells, 1995) suggest that people with social phobia have rigid, extreme standards about social performance (e.g. ‘I must always appear calm during social interactions’; see also pages 51–5). One of the reasons the person experiences anxiety is their belief that they consistently fail to meet the demands they make of themselves. These standards remain intact because safety behaviours, which are used to reduce the risk of negative evaluation, prevent the person learning that catastrophic consequences do not follow if they fail in a social performance. SST, with its emphasis on appropriate social behaviour, might prevent the person being exposed to failures, and this experience may be necessary if these rigid beliefs are to change.

follow-up. Although this is the only psychological treatment to produce an effect size significantly larger than attention-placebo conditions (Taylor, 1996; Fedoroff and Taylor, 2001), none of the meta-analyses has found that the combined approach produces a significantly larger effect size than exposure alone. The addition of cognitive techniques to exposure does not seem to improve the effectiveness of treatment. One possible reason for this is the failure of these treatments to alter the key cognitive processes that maintain the difficulty. Of the cognitive mechanisms discussed by Clark and Wells (1995), few, if any, are addressed by the cognitive techniques used in the combined treatments. Furthermore, these treatments use a habituation rationale for exposure, whereas current cognitive approaches use exposure experiences to test and modify key cognitions. The combination of cognitive techniques and exposure might have had only a limited effect because neither strategy is configured to maximize cognitive change.

Applied relaxation Applied relaxation consists of teaching the patient to use progressive muscle relaxation, and ultimately to relax in response to a cue word. This is practised first in situations that are not anxiety-provoking and it is then used during anxiety-provoking social situations. Few studies have examined this treatment and only one meta-analysis (Fedoroff and Taylor, 2001) calculates an effect size for it. The effect size was significantly superior to that for waitlist control, but it was not different from attention-placebo. Any effectiveness of the treatment might be because of non-specific factors rather than a change in the maintaining psychological mechanisms in social phobia.

Cognitive–behavioural group therapy: of the combined approaches the most extensively evaluated is Heimberg’s cognitive–behavioural group therapy (CBGT). This intervention is delivered in a group format (six patients, two therapists) over 12 2½-hour sessions; the main components of this treatment are summarized in Figure 3. • Heimberg et al. (1998) provide one of the most sophisticated evaluations of this or any other treatment for social phobia. They used a randomized controlled trial (n=133) to compare four treatments: CBGT; the monoamine oxidase inhibitor (MAOI) antidepressant phenelzine sulphate; pill placebo; and an attention-placebo condition. Independent assessors, who were blind to treatment condition, classified a similar proportion of the active medication group (77%) and the CBGT group (75%) as improved, compared with 41% of the pill placebo group and 35% of the attention-placebo group. Although the performance of the two active

Cognitive therapy Cognitive restructuring A number of early studies examined the efficacy of cognitive restructuring when used alone in the treatment of social phobia. These studies used different types of treatment, including rational emotive therapy (RET) and self-instructional training (SIT). RET vigorously challenges and disputes underlying irrational beliefs, whereas SIT uses positive self-statements to help social performance. Although these techniques are more effective than a wait-list control, they do not appear to be more effective than attentionplacebo treatment (Taylor, 1996; Fedoroff and Taylor, 2001). Cognitive restructuring of this type may not be particularly effective because it is not combined with experiences that can unambiguously challenge beliefs. Furthermore, the prescriptive forms of restructuring used by these treatments might not modify the specific mechanisms that maintain the problem, such as negative beliefs about the consequences of failed performance and excessive self-consciousness.

Components of Heimberg’s cognitive–behavioural group therapy Education

Cognitive restructuring

Combined cognitive–behavioural treatments Many treatments use cognitive techniques alongside exposure, and researchers have sought to establish whether this combined cognitive–behavioural approach is superior to exposure alone. Feske and Chambless (1995) identified eight studies that directly compared the two approaches; in only two of these did the combination treatment outperform exposure. However, half of the studies had fewer than 12 participants in each condition, and so a lack of statistical power might account for these findings. The meta-analysis of Feske and Chambless (1995) was specifically designed to address this possibility. They pooled the studies to increase statistical power, but the conclusion was unchanged: effect sizes for the two treatments were broadly comparable at both post-treatment and

PSYCHIATRY

Exposure

Homework

Patients are given a description of the cognitive–behavioural model of social phobia and a rationale for the treatment Patients are taught to identify negative thoughts and are then taught how to challenge those thoughts by looking for thinking errors Repeated exposure to anxietyprovoking situations using in-session role-plays (exposure is seen as an opportunity to practise identifying and challenging automatic thoughts ) Patients are encouraged to use exposure and cognitive restructuring between sessions

3

58

© 2004 The Medicine Publishing Company Ltd

SOCIAL PHOBIA

treatments was similar, treatment response was slower for CBGT than phenelzine therapy, and phenelzine outperformed CBGT on some measures. • Liebowitz et al. (1999) conducted a 1-year follow-up study of those who responded to treatment in the Heimberg et al. (1998) study, and found a trend towards a higher rate of relapse in those patients treated with phenelzine than in those treated with CBGT. Although CBGT produces a slower rate of initial response, its treatment gains seem to be more durable than those of phenelzine. Follow-up studies of other samples treated with CBGT have found that the effects are maintained at 5 years. The attention to detail in the studies by Heimberg et al. (1998) and Liebowitz et al. (1999) is impressive. Along with a number of other well-conducted randomized trials, these provide a solid evidence base for the treatment strategy. CBGT is currently seen as the psychological treatment of choice for social phobia, but further improvements could be made. Heimberg et al. (1998), for example, conducted an intention-to-treat analysis and found that only 58% of patients responded to CBGT. There is also the question of whether the amount of change typically achieved by this treatment represents a clinically significant improvement. CBGT outperforms control conditions on a number of standardized measures, but the amount of change is often small, and it is unclear whether this level of change translates into an improvement in a person’s functioning.

of reduction are substantially larger than those reported for earlier psychological treatments. Clark et al. (2003) have examined the effectiveness of this new type of cognitive therapy in a randomized controlled trial. Sixty patients were randomly assigned to one of three treatments: cognitive treatment; the selective serotonin reuptake inhibitor (SSRI) fluoxetine combined with exposure; or placebo with exposure. At post-treatment, cognitive therapy was significantly superior to the other two conditions on social phobia measures, and this was maintained at 1-year follow-up. This is a particularly important finding because SSRIs are the pharmacological treatment of choice for social phobia (Fedoroff and Taylor, 2001; see also pages 61–2). The symptom reduction in the cognitive treatment group was

Components of Clark and Wells’s (1995) treatment1 Education Modifying self-focused attention

Safety behaviours

Clark and Wells’s treatment Clark and Wells (1995) argue that the reason earlier treatments for social phobia have had only a limited effect is that they fail to modify the key cognitive processes that maintain the difficulty. Their cognitive model of social phobia makes explicit predictions about these mechanisms, and this model informs their treatment. The main components of this treatment, each one designed to alter one or more of these mechanisms, are summarized in Figure 4. A central prediction is that when people with social phobia enter a social situation their attention shifts to a detailed monitoring of themselves (see also pages 51–5). This increases their awareness of anxiety symptoms and reduces their processing of information about the social situation. A further consequence is that the person uses the anxiety information to construct a distorted impression of how he or she appears to others, which often takes the form of a visual image. For example, one patient reported feeling hot during social encounters and believed, inaccurately, that she looked ‘as red as a beetroot’. Treatment techniques include using behavioural experiments to illustrate the effect of altering attentional focus in social interactions, and the use of video feedback to correct the distorted self-image. As indicated in Figure 4, there are a number of additional components of the treatment, including modifying safety behaviours and modifying pre- and post-event processing; these are described in more detail in Wells (1997). The first published study to assess the treatment package (Wells and Papageorgiou, 2001) used a single-case series design (n=6), and delivered an abbreviated form of the treatment (mean number of sessions, 5.5; range 4–8). The authors used a number of standardized measures to examine change, such as the Fear of Negative Evaluation (FNE) scale and the Social Avoidance and Distress (SAD) scale. The mean reduction on the FNE was 13.8 points, which translates into a 57.1% improvement. On the SAD the mean reduction was 12.8, a 62.4% improvement. These levels

PSYCHIATRY

Verbal reattribution

Video feedback

Modifying pre- and postevent processing Modifying conditional assumptions and beliefs

1

The therapist describes the model of social phobia to the patient Behavioural experiments are used to reduce self-consciousness and to illustrate the effect of switching from self-focused to externally focused attention During social interactions the patient practises increasing and decreasing safety behaviours to challenge beliefs about the consequences of not performing these behaviours The patient is taught to identify and challenge idiosyncratic negative thoughts related to social interactions The patient watches a video of themselves during a social encounter and their distorted selfimage is contrasted with the true observable self Patients practise disengaging from worry before and after social interactions Assumption and beliefs about social performance (e.g. ‘I must always speak fluently or people won’t take me seriously’) are challenged using standard verbal restructuring techniques. Behavioural experiments in which the patient purposefully fails during social performances (e.g. purposefully hesitating when speaking) are used to illustrate the inaccuracy of these beliefs

See Wells (1997) for a detailed description of this treatment protocol

4

59

© 2004 The Medicine Publishing Company Ltd

SOCIAL PHOBIA

comparable to that achieved in the Wells and Papageorgiou (2001) study, which provides further evidence that the newer cognitive treatment may be more effective than earlier psychological treatments.

Pharmacological treatment of social phobia Carlos Blanco

Conclusion

Carolina García Of the wide variety of treatments available for social phobia, CBGT has previously been the psychological treatment of choice. Although this treatment is effective, further improvements could be made. The newer cognitive treatment of Clark and Wells (1995), which directly targets the psychological mechanisms that maintain social phobia, appears to produce greater levels of symptom reduction than earlier psychological treatments, including CBGT. It also appears to outperform SSRIs (Clark et al., 2003), the firstline pharmacological treatment for social phobia (Fedoroff and Taylor, 2001). Further comparative treatment studies are needed ‹ to substantiate this.

Michael R Liebowitz

Social phobia is characterized by a fear of negative evaluation in social or performance situations and a strong tendency for sufferers to avoid feared social interactions or situations (see also pages 51–5). Over the last few years, several studies have documented the high prevalence and clinical significance of social phobia. At same time, a substantial body of literature has developed that shows that social phobia is quite responsive to psychopharmacological treatment yet approaches in this area are still under-utilized. This contribution reviews medications currently used in the treatment of social phobia (with a particular focus on those studied in placebo-controlled trials) (see Figure 1), and gives guidelines for future directions in research.

REFERENCES Clark D M, Ehlers A, McManus F et al. Cognitive therapy versus fluoxetine in generalized social phobia: a randomized placebo-controlled trial. J Consult Clin Psychol 2003; 71: 1058–67. Clark D M, Wells A. A cognitive model of social phobia. In: Heimberg R G, Liebowitz M R, Hope D A, Schneier F R, eds. Social Phobia: Diagnosis, Assessment, and Treatment. New York: Guilford, 1995. Fedoroff I C, Taylor S. Psychological and pharmacological treatments of social phobia: a meta-analysis. J Clin Psychopharmacol 2001; 21: 311–24. Feske U, Chambless, D L. Cognitive-behavioural versus exposure-only treatment for social phobia: a meta-analysis. Behav Ther 1995; 26: 695–720. Gould R A, Buckminster S, Pollack M H, Otto M W, Yap L. Cognitivebehavioral and pharmacological treatment for social phobia: a metaanalysis. Clinical Psychology: Science and Practice 1997; 4: 291–306. Heimberg R G, Liebowitz M R, Hope D A. Cognitive-behavioral group therapy versus phenelzine therapy for social phobia: 12-week outcome. Arch Gen Psychiatry 1998; 55: 1133–41. Liebowitz M R, Heimberg R G, Schneier F R et al. Cognitive-behavioral group therapy versus phenelzine in social phobia: long-term outcome. Depression and Anxiety 1999; 10: 89–98. Taylor S. Meta-analysis of cognitive-behavioral treatments for social phobia. J Behav Ther Exper Psychiatry 1996; 27: 1–9. Wells A, Papageorgiou C. Brief cognitive therapy for social phobia: a case series. Behav Res Ther 2001; 39: 713–20.

Monoamine oxidase inhibitors Phenelzine Traditionally, phenelzine was considered the best established treatment of social phobia. Four double-blind placebo-controlled trials have studied the efficacy of phenelzine in social phobia. • In the first study (Liebowitz et al., 1988; 1992), 85 patients meeting DSM-III criteria for social phobia were randomized to phenelzine, atenolol or placebo for 8 weeks. At the end of the study, phenelzine was significantly superior to both atenolol and placebo, but there were no significant differences between those two groups. • In the second study, Gelernter et al. (1991) randomized 60 patients meeting DSM-III criteria for social phobia to one of four groups for 12 weeks: phenelzine, alprazolam, placebo and cognitive–behavioural group therapy (CBGT). Phenelzine and alprazolam were superior to placebo on the Sheehan Disability Scale

Carlos Blanco is Assistant Professor of Clinical Psychiatry at the Department of Psychiatry of Columbia College of Physicians and Surgeons and the New York State Psychiatric Institute, New York, USA. His research interests include social phobia and post-traumatic stress disorder.

FURTHER READING Heimberg R G, Liebowitz M R, Hope D A, Schneier F R, eds. Social Phobia: Diagnosis, Assessment, and Treatment. New York: Guilford, 1995. (A comprehensive review of social phobia theory and research.) Wells A. Cognitive Therapy of Anxiety Disorders: A Practice Manual and Conceptual Guide. Chichester: Wiley, 1997. (Chapter 7 provides the detailed treatment protocol based on the Clark and Wells (1995) cognitive approach to social phobia.)

PSYCHIATRY

Carolina García is a graduate student at the New York University School of Social Work, New York, USA. Her research interests include social phobia. Michael R Liebowitz is Professor of Clinical Psychiatry, Columbia University, New York, USA, and Director of the Anxiety Disorders Clinic at the New York State Psychiatric Institute. His research interests include social phobia and obsessive–compulsive disorder

60

© 2004 The Medicine Publishing Company Ltd