This study tests the assumption that psychiatric diagnosis facilitates clinical evaluations of need in emergency care before and after controlling for danger.Method
The data are from structured crisis assessments completed by emergency clinicians in four ethnically diverse locales (N = 653). Clinician-assigned diagnosis was categorized as adjustment, disruptive, mood, psychotic, and other, and a Danger scale score reflected danger to self or others.Results
Mood and psychotic disorders significantly increased hospital rates in multivariate analyses which controlled for demographic characteristics, site, and danger when relevant. The model with the best fit included both diagnosis and danger.Conclusions
Decisions should be linked to verifiable ratings of need and attention to danger, and its measurement should complement the current focus on diagnosis. J. Am. Acad. Child Adolesc. Psychiatry, 1998, 37(10):1030–1037.