Outcomes after emergency general surgery at teaching versus nonteaching hospitals

Previous analyses demonstrate teaching hospitals to have worse outcomes raising concerns for quality of care. The purpose of this study was to compare outcomes between teaching and nonteaching hospitals for emergency surgical conditions in a national sample.

The Nationwide Inpatient Sample (2005-2011) was queried for patients with emergency general surgery (EGS) conditions as determined by the American Association for Surgery of Trauma. Outcomes of in-hospital mortality, major complications, length of stay (LOS) and hospital cost were compared between patients presenting to teaching versus nonteaching hospitals. Propensity scores were used to match both groups on demographics, clinical diagnosis, comorbidities, and disease severity. Multivariate regression analyses were performed further adjusting for hospital-level factors including EGS volume. Small effect estimates were further tested using standardized differences.

A total of 3,707,465 patients from 3,163 centers were included. A majority of patients (59%) (n = 2,187,107) were treated at nonteaching hospitals. After propensity score matching and adjustment, teaching hospitals had a slightly higher odds likelihood of mortality (odds ratio, 1.04; 95% confidence interval, 1.02-1.06), slightly lower rate of major complications (odds ratio, 0.99; 95% confidence interval, 0.98-0.99), slightly decreased LOS (5.03 days [4.98-5.09] vs. 5.22 days [5.16-5.29]), and slightly higher hospital costs [$12,846 [$12,827-$12,865] vs. $12,304 [12,290-12,318]). Although these differences were statistically significant at p < 0.05, the absolute difference was very small. Further testing of these effect estimates using standardized differences revealed an insignificant difference of 0.5% for mortality, 0.4% for major complications, 0.2% for LOS, and 3.1% for hospital cost.

National estimates of outcomes for EGS conditions demonstrate comparable results between teaching and nonteaching hospitals. Concerns regarding quality of care and higher costs at teaching hospitals may be unfounded. Further research to test for differences by specific EGS conditions, operative management, and hospital costs are warranted.