Early nonresponse to fluoxetine as a predictor of poor 8-week outcome

Andrew A. Nierenberg, Nancy E. McLean, Jonathan E. Alpert, John J. Worthington, Jerrold F. Rosenbaum, Maurizio Fava

Research output: Contribution to journalArticlepeer-review

187 Scopus citations


Objective: The purpose of this study was to quantify the proportion of patients who show no response to a fixed dose of fluoxetine after 2, 4, and 6 weeks of treatment and then respond by week 8. Method: In an open trial, 143 outpatients who met DSM-III-R criteria for major depressive disorder were treated with a regimen of fluoxetine, 20 mg/day. The authors analyzed the proportion of patients who had less than a 20% decrease from baseline in their scores on the Hamilton Rating Scale for Depression after 2, 4, and 6 weeks and who went on to have a 50% or greater reduction by week 8. A last- observation-carried-forward strategy was used to calculate conditional probabilities of 8-week response. Kaplan-Meier survival analysis was used to estimate probabilities of response at week 8 given degrees of response at week 2. Results: Eighty-two subjects (57.3%) who started the trial responded by week 8. Of those subjects who showed no improvement at weeks 2, 4, and 6, the proportions of responders at week 8 were 36.4%, 18.9%, and 6.5%, respectively. The Kaplan-Meier estimate of 8-week response given nonresponse at week 2 was 0.45. Conclusions: The proportion of patients with no response to antidepressant treatment by 4 or 6 weeks who responded by week 8 was substantially less than that for subjects who had at least a partial response. Nonresponse as early as week 2 predicted 8-week outcome.

Original languageEnglish (US)
Pages (from-to)1500-1503
Number of pages4
JournalAmerican Journal of Psychiatry
Issue number10
StatePublished - Oct 1995
Externally publishedYes

ASJC Scopus subject areas

  • Psychiatry and Mental health


Dive into the research topics of 'Early nonresponse to fluoxetine as a predictor of poor 8-week outcome'. Together they form a unique fingerprint.

Cite this