Assessing the comparative effectiveness of antidepressant therapies: a prospective clinical practice study

Although efficacy studies suggest equal potency among antidepressant treatments, their effectiveness in clinical practice appears more variable, particularly in that the newer antidepressants may be less effective in either more severe depression or the melancholic subtype of depression. We pursue some factors that may impact the effectiveness of antidepressant treatments in a clinical sample.

A sample of 182 patients with DSM-IV major depressive disorder was assessed at baseline and 12 months later to establish treatments provided, identify patients who had recovered from the index episode, and quantify likely treatment determinants. Four systems for distinguishing patients with melancholic and non-melancholic depression were examined to assess for differential effects of the antidepressant strategies across those subtypes.

Multimodal therapy (commonly, psychotherapy combined with an antidepressant drug) and patients' frequent attribution of recovery to spontaneous improvement made for difficulty in disentangling recovery determinants. After excluding a spontaneous improvement component, electroconvulsive therapy (ECT) and the irreversible monoamine oxidase inhibitors (MAOIs) appeared to be the most effective therapies across the sample, while the reversible inhibitor of monoamine oxidase-A (RIMA) appeared to be the least effective. The distinct gradient of suggested effectiveness of various strategies appeared to be contributed to principally by the varied effectiveness of alternate treatments across the melancholic subtype, whereby ECT, tricyclic antidepressants, and MAOIs were the most effective, and the selective serotonin reuptake inhibitors (SSRIs), RIMAs, and antipsychotic drugs were much less effective. For the nonmelancholic disorders, the effectiveness of SSRIs appeared to be comparable with that of older antidepressants.

Although most patients received a physical treatment, they commonly judged psychotherapy and spontaneous improvement to be influential in their recovery. Reasons for such attributions are worthy of clarifying studies. Despite patients' concerns about the side effects and stigma of ECT as well as the side effects associated with the older antidepressants, these therapies were rated as more helpful by patients-and were more strongly associated with recovery-than the newer antidepressant drugs. Such overall results are compatible with an earlier study undertaken by us involving an independent sample and retrospective data. The overall gradient is clarified by studying depressive subtypes, allowing an important conclusion. Although the newer and older antidepressant drugs may be of similar effectiveness in nonmelancholic depression, the newer agents appear comparatively inferior for the treatment of melancholia, findings that have clinical implications and perhaps inform us about the pathogenesis of melancholia.