Do antidepressants really work? A clinicians' guide to evaluating the evidence.

Although antidepressants represent the cornerstone of medical management of major depressive disorder, several widely publicized recent developments have called into question the safety and effectiveness of the antidepressant medications. This article reviews the methods used to conduct studies of antidepressant efficacy, with particular focus on the research conducted by the pharmaceutical industry. It is concluded that the specific efficacy of antidepressant medications in contemporary, industry-sponsored, randomized, controlled trials is modest compared with that of medications in double-blind, placebo trials. Sources of bias and artifact that detract from these studies' validity and limit their interpretability are reviewed. It is also argued that these studies--which are primarily conducted to obtain regulatory approval, to introduce new medications, or to showcase particular advantages of newer drugs after regulatory approval--form an inadequate basis for an evidence-based medicine assessment of antidepressant effectiveness.