Background

Social phobia (SP), or social anxiety disorder, is a prevalent and disabling disorder. The growing evidence of the disorder's neurobiological basis has stimulated an increased interest in the use of medication in its treatment.

Objectives

To assess the effects of pharmacotherapy for social phobia, and to determine whether particular classes of medication are more effective and/or acceptable than others in its treatment.

Search methods

We searched the Cochrane Depression, Anxiety & Neurosis Group (CCDAN) specialised register, the Cochrane Central Register of Controlled Trials (The Cochrane Library issue 1, 2004), MEDLINE (1966 to 2003), PsycINFO (1966 to 2003), and reference lists of retrieved articles. We also requested published and unpublished RCTs from SP researchers and pharmaceutical companies.

Selection criteria

All placebo‐controlled randomised trials of the pharmacotherapy of SP were considered for the review.

Data collection and analysis

Two raters independently collated trial data, and assessed trial quality. Investigators were contacted to obtain missing data. Summary statistics were stratified by medication group (SSRIs ‐ selective serotonin reuptake inhibitors; MAOIs ‐ Monoamine oxidase inhibitors; moclobemide and brofaromine). Dichotomous and continuous measures were calculated using a random effects model, heterogeneity was assessed, and subgroup/sensitivity analyses were undertaken.

Main results

37 RCTs of a range of medications were included in the analysis (5264 participants), of which 23 were short‐term (14 weeks or less). A funnel plot provided evidence of publication bias.

Twenty‐six trials demonstrated short‐term superiority in treatment response of all medication groups over placebo (N = 3696; relative risk of non‐response (RR‐non) = 0.64; 95% CI = 0.57, 0.73). However, the SSRIs were significantly more effective than both moclobemide (Qb = 38.61; p < 0.00001), and, and to a lesser extent, brofaromine (Qb = 2.87; p = 0.09).

Sixteen comparisons of symptom severity showed a statistically significant difference between medication and placebo (weighed mean difference = ‐18, 95%CI = ‐25.17, ‐10.83). This effect was once again most evident for the SSRIs. Medication also reduced SP symptom clusters, comorbid depressive symptoms, and associated disability. The value of long‐term medication treatment in treatment response was demonstrated by 4 maintenance (RR‐non = 0.62; 95% CI = 0.50, 0.77) and 4 relapse prevention (RR of relapse = 0.33; 95% CI = 0.22, 0.49) studies. Two performance anxiety RCTs reported mixed results.

Authors' conclusions

Medication appears effective in treating SP over the short term (particularly amongst the SSRIs), and the long term. Nevertheless, the possibility of publication bias has to be acknowledged. Additional issues for future research include the use of medication in children and adolescents with SP, SP with comorbid psychiatric disorders, and performance anxiety.