To carry out a metaanalysis of antidepressant studies to calculate the effect sizes for antidepressant effect in depressive disorder.
A metaanalysis of all antidepressant studies that included an active comparison drug as well as placebo was used to calculate the effect size. Articles were selected from a MEDLINE search for the period January 1966 to June 1995. Forty-nine studies were included in the metaanalysis.
The effect sizes for antidepressant treatment are moderately larger than for placebo. A larger effect size was observed in studies where objective diagnostic criteria for depression were used.
We conclude that the superior efficacy of antidepressants over placebo can be demonstrated.