Study objective: We characterize repeat pediatric emergency department visits and determine the cause for such visits as an indicator of potential need for quality improvement. We hypothesized that most repeat ED visits resulting in hospitalization do not represent medical errors.
Methods: The study was performed at a large, tertiary care, academic children's hospital. Patients who returned to the ED within 72 hours of a previous visit were identified by computerized registration data. The charts of these patients were then reviewed by a member of the ED medical staff to identify factors from the initial visit that contributed to the return visit. A multidisciplinary committee then reviewed each case until consensus was achieved regarding the cause for the repeat visit. In this study, we analyzed the quality improvement decisions from a 12-month period.
Results: Over a 12-month period, during which there were a total of 51,195 visits, 285 (0.56%) patients were hospitalized after a repeat visit. The repeat visit was determined to be unrelated to the first visit in 12 (4.2%) patients. In 12 (4.2%) cases, the cause for repeat visit and hospitalization could not be determined. This resulted in a total of 261 patients for analysis. In 234 (90.0%) patients, the return visit was determined to be a result of the progression of illness (no medical error). Ten (3.8%) patients had a missed diagnosis, whereas 2 (0.8%) patients had errors in their treatment (likely medical error). An incomplete workup was cited in 7 (2.7%) patients (potential medical error). Parenting factors (refusing admission, not filling prescriptions, not giving prescribed medications) were noted in 5 (1.9%) cases. Three (1.1%) patients did not follow up with appropriate subspecialists.
Conclusion: The overall rate of repeat visits resulting in hospitalization is small. In the majority of these cases (90.0%), the ED evaluation was appropriate and the admission was for progression of the patient's illness. Given the small number of patients and the infrequency of missed diagnoses, this may not be an efficient method for assessing ED performance.