Predicting mental health service use: Diagnosis based systems and alternatives

Abstract

One of the main features of the modern British NHS is that health should be both purchased and provided in an efficient and cost effective manner. This is particularly relevant in the resource intensive area of mental health care. The achievement of such an aim entails the measurement of service inputs and patient outputs. Knowledge of the casemix (the grouping of 'similar' patients) is vital. This paper reviews the efficiency of the most prominent description of casemix, diagnosis-related groups (DRGs), in predicting levels of mental health service utilization. The results demonstrate that the application of DRGs to psychiatry is imperfect. Refinements to DRGs and alternative systems are also reviewed. The main finding is that diagnosis is a poor indicator of the level of services that patients require. More satisfactory predictors include social functioning, patient characteristics, and severity of illness.