A computerized mental patient classification system.

Abstract

This paper presents the detailed logic and flow of a computerized actuarial report system based on the patient's demographic and clinical information routinely collected in a state mental health information system. Probability statements regarding the patient's diagnostic classifications, likelihood of harm to self and others, probable psychotropic drug prescription, likelihood of unauthorized absence, and prediction of length of stay are presented in the report. The multivariate linear discriminant function and linear regression are used in the calculation of the posterior probabilities and the predicted value of length of stay.