PREDICTING LEVEL OF CARE IN MENTAL-HEALTH-SERVICES UNDER A CONTINUUM OF CARE

In an effort to control the spiraling costs of mental health care, pub lic and private sectors have moved rapidly toward managed mental healt h care. An important feature of many managed mental health care system s is the use of the alternative or ''intermediate'' services in place of costly residential services. The introduction of these intermediate services affords clinicians a wider range of treatment options. Littl e is known, however, about the decision making process in assigning ch ildren to these services. Using data on an innovative system that made those services available, this article identifies predictors of inter mediate service use. Multivariate analyses are based on different vari ants of logistic regression; the analysis illustrates the use of multi nominal and nested logistic regression. Copyright (C) 1996 Elsevier Sc ience Ltd