594 Weekend Admissions Do Not Affect Management of Lower GI Bleeding

• Published in: The American journal of gastroenterology Vol. 114; no. 1; p. S346
• Main Authors: Li, Brian, Schwartz, Jeffrey, Lipcsey, Megan, Stein, Daniel J., Feuerstein, Joseph
• Format: Journal Article
• Language: English
• Published: 01.10.2019
• ISSN: 0002-9270; 1572-0241
• DOI: 10.14309/01.ajg.0000591908.34645.15

INTRODUCTION: Lower gastrointestinal bleeding (LGIB) is a common condition requiring GI consultation and management. Current ACG guideline recommends colonoscopy within 24 hours of LGIB. However, while colonoscopy is the most common intervention, it may not be readily available at all hospitals on the weekend when many endoscopy units are closed. Previous work has found that in non-variceal upper GI hemorrhage, patients admitted on weekends had lower total endoscopy rates and high hospitalization charges. We explored these outcomes in LGIB. METHODS: This national inpatient sample (NIS) from 2016 was queried for all cases of confirmed lower GI bleeding as the primary diagnosis code based on ICD-10. This included bleeding AVMs, diverticular hemorrhage, rectal ulcers, radiation proctitis and other anal/rectal hemorrhage. Outcomes for patients admitted on a weekend (Saturday/Sunday) were compared to weekday admissions using survey-adjusted Chi-square and correlation for comparisons and logistic/linear regression to examine mortality, length of stay, and hospitalization costs. Adjustment was made for the Elixhauser mortality score (a validated measure of comorbidities), colonoscopy, and transfusion in the regression. RESULTS: In 2016, 92,065 patients were admitted for LGIB overall. Of these, 23,070 (25%) were admitted on a weekend. See table 1 for causes of bleeding. For weekend admissions, mortality (0.8% vs 0.9%, P = 0.626), transfusion rate (28.8% vs 30.1%, P = 0.113), colonoscopy rate (58.4% vs 58.1%, P = 0.749), angiography rate (3.3% vs 3.3%, P = 0.894), mean days to colonoscopy (1.8 vs 1.8, P = 0.157), length of stay (4.1 vs 4.2 days, P = 0.068), and charges ($40,552 vs $40,604, P = 0.956) were all similar to weekday admissions. In the multivariable regression, colonoscopy (OR=2.7, P < 0.001) and Elixhauser score (OR=1.09, P < 0.001) but not weekday admission (OR=1.13, P = 0.535) or freedom from transfusion (OR=0.78, P = 0.140) predicted mortality. Length of stay (0.1 day difference, P = 0.053) and charges ($14 difference, P = 0.987) were also not affected by weekend admission after adjustment for the same covariates. CONCLUSION: In this estimate of all U.S. lower GI bleeding admissions from 2016, weekend admission did not affect outcomes including mortality, length of stay or total charges. Time to colonoscopy was similar. Undergoing colonoscopy was associated with a 2.7 times lower odds of mortality.