Twenty-four studies were included in the review. Sample sizes ranged from 15 to 189. Study quality results were not presented.
Panic disorder (sixteen studies): Combination of cognitive-behavioural therapy (CBT) and drug treatment was statistically superior to drug treatment alone on clinician rating (ES 0.39, 95% CI 0.09 to 0.69) and self-reporting (ES 0.61, 95% CI 0.22 to 0.99). This combination was also superior to CBT plus placebo using clinician rating (ES 0.23, 95% CI 0.05 to 0.41) and self-reporting (ES 0.43, 95% CI 0.21 to 0.66). All treatment modalities showed large pre-post effect sizes, with clinician-rated combined drug and cognitive-behavioural therapy showing the largest pre-post treatment difference (ES 2.07, 95% CI 1.77 to 2.38).
Social anxiety disorder (six studies): A statistically significant difference was found for the comparison between CBT plus drug treatment versus CBT plus placebo on clinician rating (ES 0.42, 95% CI 0.18 to 0.68). All treatment modalities showed large pre-post treatment effect sizes, with the largest for clinician ratings of drug treatments (ES 2.18, 95% CI 1.75 to 2.62).
Generalised anxiety disorder (two studies): There were insufficient data available to perform meta-analysis for this outcome. In general, CBT, or CBT in combination with drug treatment, was superior to drug treatment alone.
No statistically significant differences were found between drug and CBT in direct comparisons. The analysis of available follow-up studies did not show a clear long term effect of CBT. Mixed results were reported for other comparisons.