Predicting resource utilization in a comprehensive center: an evaluation of three alternative methods.

Three alternative methods for obtaining anticipated resource utilization information for comprehensive mental health centers are proposed. The methods differ on two dimensions, sample selection and statistical technique. Using an admission cohort rather than a discharge cohort and eliminating patients who terminated treatment against medical advice permitted a refined prediction. Although no significant differences were found between the chi-square and multiple-regression techniques, the latter resulted in a substantially greater degree of flexibility.