Outcomes of Upper Gastrointestinal Bleeding at United States Teaching and Non-teaching Hospitals: A National Inpatient Sample Analysis

Background/aims: Studies have varied results regarding the impact of the teaching and non -teaching status of hospitals on the outcomes for hospitalized patients with upper gastrointestinal bleeding (UGIB). To evaluate these outcomes, we conducted a retrospective cohort study using the 2014 National Inpatient Sample (NIS) database. Methods: We included all adult patients who were admitted with the principal diagnosis of UGIB. Patients admitted to rural and urban non-teaching hospitals were classified as non -teaching, whereas those admitted to urban teaching hospitals were classified as teaching. The main outcomes of interest were in -hospital mortality, percentage of patients requiring inpatient endoscopy, and endoscopic therapy, packed red blood cell (PRBC) transfusion, length of stay (LOS), and total hospitalization charges. Results: The study included 132,085 (97%) with nonvariceal UGIB (NVUGIB) and 4,200 (3%) with variceal UGIB (VUGIB). Of them, 62% were managed at teaching hospitals. Compared with admitted patients at nonteaching hospitals, patients with nonvariceal UGIB admitted at teaching hospitals had similar adjusted inhospital mortality rates (adjusted odds ratio (OR): 0.97, 95% confidence interval (CI): 0.79-1.19), inpatient endoscopy rates (OR: 0.98, 95% CI: 0.91-1.1), and early endoscopy rates (within 24 hours) (OR: 0.98, 95% CI: 0.91-1.1) and lower PRBC transfusion rates (OR: 0.87, 95% CI: 0.79-0.97) but higher endoscopic therapy rates (OR: 1.3, 95% CI: 1.2-1.4), length of stay (mean increase of 0.43 days) (P<0.01), and total hospital charges (mean increase of $4,369) (P<0.01). Patients with variceal UGIB had similar adjusted inhospital mortality rates (OR: 1.2, 95% CI: 0.61-2.3), inpatient endoscopy rates (OR: 0.97, 95% CI: 0.67-1.4), early endoscopy rates (within 24 hours) (OR: 0.97, 95% CI: 0.67-1.4), endoscopic therapy rates (OR: 2.5, 95% CI: 0.54- 11.2), and total hospital charges (P=0.45), and lower PRBC transfusion rates (OR: 0.63, 95% CI: 0.45-0.88) but higher length of stay (mean increase of 0.69 days) (P=0.02). Conclusions: Patients with nonvariceal UGIB treated at US teaching hospitals and non -teaching hospitals have similar mortality, rates of in -hospital endoscopy, and early endoscopy, but teaching hospitals have higher rates of in -hospital therapeutic endoscopy, length of stay, and total hospital charges. There was no difference in any of the outcomes for variceal gastrointestinal (GI) bleeding treated at teaching hospitals compared with those treated at non -teaching hospitals, except for length of stay, which was higher among patients admitted to teaching hospitals compared to those admitted to non -teaching hospitals.