When should a trial of fluoxetine for major depression be declared failed?

Objective: Although the newer antidepressants are widely used, little is known about how long it takes to see their full effect. The authors sought to determine how many weeks a fluoxetine trial with no improvement should continue before treatment is changed. Method: The data involved 840 patients in a 12-week open trial of fluoxetine, 20 mg/day, followed by a blinded, placebo-controlled discontinuation study. Outcomes at 4, 6, 8, 10, and 12 weeks were classified as nonresponse, partial response, response, and remission and were based on Hamilton Depression Rating Scale scores. The rate of remission at week 12 was calculated for each group of patients without remission at the earlier time points. The time to relapse during weeks 13-26 of the discontinuation study was examined in patients taking placebo and fluoxetine in relation to status at week 6. Results: Patients unimproved at week 6 had a remission rate at week 12 of 31%-41%. For patients with remission at week 12, level of improvement at week 6 did not affect prognosis in weeks 13-26. Of the unimproved patients at week 8, 23% had remissions by week 12. The week 12 remission rate for unimproved patients at week 4 was clearly high enough to justify continued treatment; the rate for unimproved patients at week 10 was too low. Conclusions: These data suggest that nonresponse to fluoxetine should not be declared until 8 weeks of treatment have elapsed. Practitioners should discuss trial length with patients at the beginning of treatment.