Premature departure from the pediatric emergency department: a cohort analysis of process- and patient-related factors.

ABSTRACT

OBJECTIVE: Previous literature suggests that process-related factors (eg, time of day, patient volume) and patient-related factors (eg, acuity, socioeconomic status) are associated with premature departure from emergency departments. We sought to evaluate the relationship of these and other factors with premature departure in a large, unselected cohort of pediatric emergency department patients. METHODS: This study was a retrospective cohort analysis of visits to a single tertiary site during a 1-year period. Patients' zip codes determined assignment of census-based socioeconomic metrics. Multivariate regression identified factors associated with premature departure. Sensitivity and subset analyses were performed. Return visits within 48 hours after premature departure were also reviewed. RESULTS: There were 46,417 visits, of which 2164 were premature departures. In multivariate analysis, independent predictors of premature departures were arrival time, arrival month, arrival day of week, patient acuity, concurrent premature departures, arrival rate, arrival period average length of stay, and poverty rate. Aside from patient acuity and poverty rate, no patient-related factors were significant in multivariate analysis. These results were robust in sensitivity analysis across different multivariate models. Among premature departures, there were 120 return visits (5.5%), of which 15 were admitted (0.7%). There were no deaths. Acuity was similar between initial and subsequent visits. CONCLUSIONS: Process-related factors and individual patient acuity have the strongest influence on premature departure from the pediatric emergency department. Health care organizations concerned with premature departure should focus efforts on improving pediatric emergency process flow.