Prediction of response to chlordiazepoxide and placebo in anxious outpatients: an attempt at replication.

The present study seeks to determine the extent to which a set of non-specific factors can stably predict response to either chlordiazepoxide (CDZ) or placebo (PBO) and the extent to which such prediction is specific to or distinctive for each treatment agent. For this purpose data were assembled for 447 primarily anxious neurotic outpatients treated with either CDZ or PBO in 4 week double blind drug trials performed over the past ten years and divided into two comparable subsamples for each treatment agent. A series of analyses revealed modest replicability but considerable drug specificity for a Global improvement measure. Replicability was considerable higher for a patient measure of Symptom Change, but its specificity to treatment agent was considerably less.