BACKGROUND

Emergency departments (EDs) in teaching hospitals have competing goals of timely patient care and supervised trainee education. Previous investigations have indicated that trainees add time to the length of ED patient encounters. However, no studies have quantified the effect of trainees on pediatric ED length of stay (LOS).

OBJECTIVES

The objectives were to measure the effect of trainees on pediatric ED LOS by comparing LOS for patients managed by a pediatric emergency physician (PEP) alone to LOS for patients seen by a trainee and a precepting PEP (Trainee+PEP). A secondary objective was to identify factors other than provider type associated with LOS differences observed in teaching hospital pediatric EDs.

METHODS

Data were extracted from a computerized ED tracking system in an urban tertiary care children's hospital with approximately 52,000 visits annually. All patients were seen by a PEP alone, an urgent care physician, or a trainee (a pediatric emergency medicine fellow; a pediatric, emergency medicine, or combined internal medicine/pediatrics resident; or a medical student) plus a precepting PEP. The primary comparison was the ratio of median LOS for the PEP group versus the Trainee+PEP group.

RESULTS

There were 92,193 visits eligible for inclusion over a 2-year period. Median patient age was 5.75 years (interquartile range [IQR] = 21 months to 12.9 years). The PEP group managed 9,141 patients (10%), while the Trainee+PEP group treated 72,135 patients (78%). Overall LOS for an ED visit was 221 minutes. The median LOS was 192 minutes for PEP patients and 225 minutes for Trainee+PEP patients (difference of means = 17%, p < 0.001). Laboratory and imaging studies were associated with LOS increases of 111 and 74 minutes, respectively; both were performed more frequently in Trainee+PEP patients (44% vs. 33% for laboratory studies and 41% vs. 39% for imaging studies, both comparisons p < 0.001). When LOS was analyzed after adjusting for confounding factors including patient acuity, laboratory or radiologic testing, and trainee year, LOS for Trainee+PEP was higher by 17 minutes, or 9% (95% confidence interval [CI] = 6% to 12%, p < 0.001). When LOS was examined for four specific diagnoses (asthma, gastroenteritis, appendicitis, foot/ankle sprain), there were no significant differences in LOS between the PEP and Trainee+PEP groups.

CONCLUSIONS

In the pediatric ED of a teaching hospital, ED LOS is on average 9% higher in patients seen by trainees. In an era of increasing efforts to accelerate throughput while training future providers, these findings provide an important metric for the delivery of pediatric emergency care.