Potential Predictors of Placebo Response: Lessons from a Large Database

Abstract

This report is a statistical assessment of four factors, each separated into two categories that might contribute to the high placebo response rates noted during major depressive disorder (MDD) drug development: mean age ( 23), and sex (% female and male). Data from 4,500 depressed patients who participated in 18 clinical trials comparing venlafaxine and placebo were pooled. Univariate χ2, multivariate analyses, generalized estimating equation, and mixed-model repeated-measures tests were used to analyze the impact of the four baseline parameters on remission rates in the placebo- and active-treated groups. Remission rates were significantly correlated with the 17-item Hamilton Rating Scale for Depression (HAM-D17) severity at baseline, the age and sex of the patient, and the MDD duration before treatment. The variables predicting remission outcome were significantly influenced by the statistical method used to analyze the data.