Despite a great volume of literature on the subject, we were able to locate only a small number of RCTs for the antidepressant classes of interest. Notably absent were studies examining desipramine (a single out-patient study) and nortriptyline (no published RCTs found), cheap drugs with reputedly lower rates of ADRs, particularly of the anticholinergic variety. No supporting evidence was found. Such studies should be published, especially if ADRs are of interest. Nonetheless, we were able to retrieve information on nearly 3000 patients, which is a reasonably large sample to provide estimates.
Efficacy rates were statistically superior for venlafaxine in out-patients but not in-patients. There were fewer patients in the in-patient analysis; however, rates were similar among drug classes. More information about in-patients is required.
The range of success rates was wide for TCAs – from 22% to 90% (heterogeneity P<0.001). SSRIs also displayed significant heterogeneity, albeit somewhat less. On the other hand, venlafaxine rates were much more consistent (P<0.05). No reasons for these findings were found.
Subgroup analyses were performed on individual drugs (Table 3). For out-patients, efficacy rates were similar for amitriptyline and imipramine; SSRIs were consistent as a group. For in-patients, the efficacy rate of amitriptyline was somewhat higher (8.6% above the mean), and that of imi-pramine somewhat lower (14% below the mean), than the mean; however, both drugs had wide 95% CIs, which tended to be true for all analyses. Thus, it seems reasonable to combine results for the entire classes.