Thirty-two RCTs (5,562 patients) were included. There were 20 RCTs (3,844 patients) that compared venlafaxine with SSRIs, 9 RCTs (1,356 patients) that compared venlafaxine with TCAs, and 3 RCTs (418 patients) that compared venlafaxine with other drugs.
The average sample size was 179 patients (range: 28 to 382). Most trials used the last observation carried forward for the primary analysis. None of the RCTs indicated whether the concealment of allocation was conducted properly.
Efficacy.
Venlafaxine significantly reduced the depression score compared with all comparators combined and compared with SSRIs: the pooled effect sizes were -0.14 (29 studies, 95% CI: -0.22, -0.07) and -0.17 (95% CI: -0.27, -0.08), respectively. There was no statistically-significant difference between venlafaxine and either TCAs (pooled effect size -0.13, 95% CI: -0.33, +0.09) or other drugs (pooled effect size -0.09, 95% CI: -0.42, +0.23). The results appeared consistent across the SSRI studies but there were differences among the TCA studies.
Response rates.
Venlafaxine significantly increased the response rates compared with all comparators combined; the OR was 1.27 (95% CI: 1.07, 1.52), the risk difference was 0.05 (95% CI: 0.02, 0.09) and the NNT was 19 (95% CI: 11, 63)).
Remission rates (18 RCTs, including 16 RCTs that used SSRIs as the comparator).
Venlafaxine significantly increased the remission rates compared with all comparators combined; the OR was 1.36 (95% CI: 1.14, 1.61) and the NNT was 14 (95% CI: 9, 29).
None of the potentially explanatory factors were significantly predictive in the meta-regression analyses.
No evidence of publication bias was found by the meta-regression or visual inspection of the funnel plots.
Treatment discontinuation.
There was no statistically-significant difference in the discontinuation rates for venlafaxine; the overall risk difference was -0.004 (95% CI: -0.029, +0.020).