Objective: We evaluated overutilization or underutilization of inpatient resources to measure the emergency department (ED) decision-making process and its association with the following care factors: annual pediatric volume, presence or absence of a pediatric emergency medicine specialist; and presence or absence of ED residents.
Study design: Block random selection, using the three care factors, of 16 hospitals with pediatric intensive care units. The Pediatric Risk of Admission (PRISA II) Score was used to measure illness severity. Decision-making was evaluated for admissions (Admission Index: observed minus predicted admissions) and returns (Return Index: observed minus predicted 72-hour returns). The Combined Index was a weighted average of the Admission and Return Indexes.
Results: There were 11,664 patients enrolled. Residents but not volume or pediatric emergency medicine specialists were associated with the decision-making performance indexes in multivariable analysis (no residents versus residents: Admission Index: 2.5 of 1000 patients versus 34.8 of 1000, P = .082; Return Index: -3.0 of 1000 versus 33.6 of 1000, P = .039; Combined Index: 1.9 of 1000 versus 35.5 of 1000, P = .024.
Conclusions: There is significant variability in ED decision-making for children. Residents but not volume or presence of a pediatric emergency medicine specialist are associated with increased differences in admission decisions. The process by which these differences occur was not investigated.