How individual clinicians make admission decisions in psychiatric emergency rooms

The goal of this study was to understand how individual clinicians make admission decisions in the psychiatric emergency room. Clinical and demographic data were collected on 7485 consecutive visits to four psychiatric hospitals in Israel during 1991 and 1992. This was about one-third of visits to psychiatric emergency rooms in Israel during this time. Twenty-one decision makers who made at least 50 decisions and admitted more than 15 and less than 85% of patients were included. Decision to admit patients was modeled using step-wise discriminant analysis. Clinicians examined different numbers of patients and admitted at different rates. In one hospital one group of clinicians appears to rely primarily on diagnosis and another group on other variables, predominately previous history, and social or referral factors. The most influential background variable was self-referral which favored not admitting patient. The most influential diagnoses were schizophrenia and affective disorder which favored admission. In another hospital the most important variables in most models were legal status of admission, number of previous hospitalizations, violence and suicide as presenting problems and referral source. All variables except referral source favored admission. Some referral sources favored admission and others did not. In the last hospital studied the most salient variable was self-referral which favored not admitting patient. Decisions in one hospital were not modeled because no clinician met study inclusion criteria. There appears to be variability between clinicians in what information they use to make their decisions in the psychiatric emergency room and what percent of patients they admit.