A Typological Model for Estimation of Drug and Placebo Effects in Depression

One reason for failure to find specific treatment effects in drug versus placebo trials with patients who have depression is an insufficient period of observation. Also, differentiating between early fleeting response and maintained response has been shown relevant to detecting specific drug action. A model in which various types of drug and placebo response are specifically stipulated and which takes advantage of a two-stage experimental design is proposed. Substantive findings when patients who have major depression with atypical features are studied are: 1) about 6% have only a fleeting response to placebo and no response to drug; 2) by week-10 about 28% of the population will respond even if no medication is given; and 3) specific response to imipramine (21%) can be determined by week-4 but specific response to fluoxetine (20%) cannot be determined until later (week-10 in this study).