A computer language model of group clinical decision making at an agency for developmentally delayed children was developed. Five client-flow decisions were examined: accept or refer clients for services, assign clients to multidisciplinary teams or individual disciplines for services, select the subset of the II disciplines which evaluate clients, select subset of disciplines which are recommended to do treatment, and accept or reject clients referred for placement in special classes. The model processed information gathered from client records pertaining to background data and presenting complaints. The decision rules were developed from staff interviews. The model's output matched the actual process at a statistically significant level. The clinical decisions modeled in this study appeared to be primarily based on readily observable client attributes. Organizational attributes appeared to be important only when there was a shortage of professional staff. There was considerable overlap in the provision of services by several disciplines. When asked to describe their decision processes, clinicians listed more client attributes than they actually utilized.
展开▼
