938 Experience Matters: Higher Hospital Volume Is Associated with Lower Mortality in Acute Non-Variceal Upper Gastrointestinal Hemorrhage

Abstract

Introduction: Acute non-variceal upper gastrointestinal hemorrhage (ANVUGIH) is a common medical emergency associated with significant morbidity, mortality and healthcare costs. It is among the leading gastrointestinal causes of hospitalization and death. Previous studies have shown a relationship between higher hospital or procedural volume and superior outcomes for complex surgical procedures. To our knowledge, ours is the first study to examine this relationship for ANVUGIH. Methods: We used data from the Nationwide Inpatient Sample (NIS) for the year 2004. Hospitalizations related to ANVUGIH were identified by the presence of a primary discharge diagnosis of bleeding peptic ulcer, gastritis with hemorrhage, Mallory-Weiss lesions, Dieulafoy's lesions, angiodysplasias of the stomach, hematemesis and melena through appropriate ICD-9-CM codes. Hospitals were then divided into low, medium, and high volume hospitals if they had 1 125, 126 250, or more than 250 annual hospitalizations for ANVUGIH. Results: There were a total of 135 366, 132 746, and 123 007 discharges with ANVUGIH from low, medium, and high volume hospitals respectively. Over half the patients in each volume category were 66 years or older. Bleeding peptic ulcer was the most common cause of the ANVUGIH, with a slightly higher proportion in high volume hospitals (35% vs. 32.4%, p=0.001). Patients at high volume hospitals were more severely ill with a higher proportion having acute renal failure (8.9% vs. 4.9%, p<0.001) or acute respiratory failure. Patients at high volume hospitals (OR 0.85, 95% CI 0.74 0.98) had significantly lower in-hospital mortality than patients at low volume hospitals. Patients at high volume hospitals were also more likely to undergo upper endoscopy than those at low volume hospitals on multivariate analysis (OR 1.52, 95% CI 1.36 1.69). A similar difference was seen in undergoing endoscopic intervention for control of hemorrhage (OR's 1.59 and 1.50 for medium and high volume hospitals respectively). In addition, patients at high volume hospitals were more likely to undergo endoscopy within 1 day of hospitalization compared to low volume hospitals (60.5% vs. 53.8%, Adjusted OR 1.28, 95% CI 1.02 1.61). Conclusion: Higher hospital volume is associated with lower mortality, and higher rates of endoscopy and endoscopic intervention in patients with ANVUGIH.