Predictors of urgency in a pediatric psychiatric emergency service.

OBJECTIVES: To test the hypothesis that youth present to a psychiatric emergency service (PES) at least 25% of the time for nonurgent reasons, to examine the demographic characteristics that distinguish urgent from nonurgent visits, and to develop a model to predict urgency.

METHOD: Psychiatric emergency visits of all patients under 18 years from July 1, 1997, through June 30, 1998, were ed and coded as to level of urgency using Rosenn's classification system. Age, gender, ethnicity, arrival status, social service involvement, violence, substance abuse, and diagnosis were examined with respect to urgency in bivariate and multivariable analyses.

RESULTS: Forty percent of visits to the PES were not urgent. Demographic factors distinguished urgent from nonurgent visits. Independent predictors of urgency for the entire sample were age, diagnosis, arrival status, social service involvement, and violence. Only violence remained an independent predictor of urgency for the attention-deficit/hyperactivity disorder subgroup.

CONCLUSIONS: A large percentage of pediatric visits to the PES were for nonurgent reasons. This suggests that there are considerable unmet mental health needs of children and adolescents. These findings can be applied to improve and design appropriate services.