Thirty RCTs (37 comparisons; 3,178 participants) were included in the review. The quality of the included studies was reported briefly; 14 studies reported independent allocation to treatment, 17 reported blinded outcome assessment, and 21 studies used intention-to-treat analysis. There was no evidence of publication bias.
The difference in effect between pharmacological and psychological therapy was small and not statistically significant (d=-0.07, 95% CI -0.15 to 0.01), with a low level of statistical heterogeneity (Ι²=21.31%). Drop-out rates were lower in psychological compared to pharmacological therapy, although there was moderate statistical heterogeneity (OR 0.66, 95%CI 0.47 to 0.92, Ι²=69.46%).
In patients with dysthymia, pharmacotherapy was significantly more effective than psychological therapy (d=-0.28, 95% CI -0.47 to -0.10, six comparisons, Ι²=26.36%), whereas there was no difference in effect between the two therapies in patients with major depression (d=-0.02, 95%CI -0.10 to 0.06, 31 comparisons, Ι²=4.22%).
Treatment with SSRIs was significantly more effective than psychological therapy, (d=-0.20, 95% CI -0.31 to -0.10, 15 comparisons, Ι²=0%). The effect of treatment with other antidepressants was similar to that of psychological therapy. Results of other subgroup analyses were reported.
Pre-treatment level of depression did not affect the results.
Studies which used intention-to-treat analyses (27 comparisons) had a smaller effect size than those based only on completers (10 comparisons), although neither was statistically significant.