Quitkin F M, Rabkin J G, Ross D, McGrath P J
Arch Gen Psychiatry. 1984 Mar;41(3):238-45. doi: 10.1001/archpsyc.1984.01790140028003.
Data from three six-week placebo-controlled randomized antidepressant trials were pooled to test the hypothesis that a four-week trial is insufficient to reach a determination of drug failure in depressed patients. We compared global clinical ratings at weekly intervals for patients receiving drug and patients receiving placebo and calculated the proportion of patients whose clinical status changed over time. We predicted, and found, that a significant proportion of patients who showed no clear-cut response at four weeks would show much improvement at six weeks in drug but not placebo conditions. Baseline Research Diagnostic Criteria diagnosis, baseline illness severity, and drug dose adjustments after four weeks did not predict either late clinical improvement or relapse between four and six weeks. Additional placebo-controlled studies are needed to replicate our findings concerning the advantage of extending trials to five or six weeks in samples of patients of various depressive subtypes.
