Computer prediction of psychotropic drug assignment in state mental facilities: effect of eliminating alcoholics from the study sample.

A previous study on assignment of four major categories of psychotropic drugs to psychiatric inpatients generated the hypothesis that patients assigned minor tranquilizers consisted of two distinct sub-groups, one of which consisted of diagnosed alcoholics. The present study confirmed that hypothesis. More importantly, it found that minor tranquilizer assignment to non-alcoholic inpatients was a virtually random procedure. The general ability of predicting psychotropic drug assignment via a statistical model was nevertheless supported.