Patient preference as a moderator of outcome for chronic forms of major depressive disorder treated with nefazodone, cognitive behavioral analysis system of psychotherapy, or their combination.

Little is known about moderators of response to psychotherapy, medication, and combined treatment for chronic forms of major depressive disorder (MDD). We hypothesized that patient preference at baseline would interact with treatment group to differentially affect treatment outcome. We report outcomes for 429 patients who participated in a randomized multicenter trial of nefazodone, Cognitive Behavioral Analysis System of Psychotherapy (CBASP), or combination therapy for chronic forms of MDD (DSM-IV criteria) and who indicated their preference for type of treatment at study entry. The primary outcome measures were total scores on the 24-item Hamilton Rating Scale for Depression (HAM-D-24) and categorical definitions of remission or partial response. The patients were recruited between June 1996 and December 1997. There was an interactive effect of preference and treatment group on outcome. The treatment effect varied as a function of preference, and was particularly apparent for patients who initially expressed preference for one of the monotherapies. Patients who preferred medication had a higher remission rate (45.5%) and lower mean HAM-D-24 score (11.6) at study exit if they received medication than if they received psychotherapy (remission rate, 22.2%; mean HAM-D-24 score, 21.0). Patients who preferred psychotherapy had a higher remission rate (50.0%) and lower mean HAM-D-24 score (12.1) if they received psychotherapy than if they received medication (remission rate 7.7%, mean HAM-D-24 score 18.3). Nevertheless, treatment preference was not associated with risk of dropout from the study. These results suggest that patient preference is a potent moderator of treatment response for patients with chronic forms of MDD; however, relatively low proportions of the patient sample preferred one of the monotherapies, participants were not blinded to treatment assignment, and there was no placebo group.