This article is only available in the PDF format. Download the PDF to view the article, as well as its associated figures and tables.

In Reply—.Ìý To challenge our "assumption that different hospitals and clinicians may have different criteria for admissions," McRae described similarities between his results and parts of our results. He referred to the find ing that hospitalized patients had lower GAS scores than patients who were not hospitalized, were more likely to be diagnosed as psychotic, and were more often labeled as having chronic disease. However, the usual finding that hospitalized patients tend to be significantly sicker than nonhospitalized patients is not inconsist ent with our finding that different clinicians use different criteria in making the decision to hospitalize. For example, while all four clinicians in our study hospitalized at least some patients who had acute psychosis, one clinician only hospitalized those patients not in treatment, and another clinician hospitalized only those patients who were not either revisiting the emergency room or former state hospital patients undergoing treatment.