Mo1535 – Risk Factors for In-Hospital Mortality in Variceal Hemorrhage: A Nationwide Analysis

Background: Active variceal hemorrhage accounts for approximately one-third of all-cause mortality in cirrhosis.Previous studies of variceal hemorrhage have had limited power to identify risk factors for mortality.Using a nationwide database, we aimed to provide nationallevel estimates of mortality from variceal hemorrhage in the United States.Methods: We analyzed the 2009 National Inpatient Sample (NIS) database.Patients aged 16-75 with a diagnosis of variceal hemorrhage were identified via ICD-9 diagnosis code.Patients with a discharge diagnosis of alternative causes of upper gastrointestinal bleeding were excluded.We included only patients who underwent upper endoscopy.Predictors of mortality were assessed using multiple logistic regression performed by SAS, Version 9.4 (SAS Institute, Cary, NC).Results: An estimated 4,555 patients with a discharge diagnosis of variceal hemorrhage were identified, out of which 371 (8.1%) died while hospitalized.For regression analysis we controlled for the following confounders: age, gender, race, insurance, transfusion of red blood cell, requirement for paracentesis, the presence of spontaneous bacterial peritonitis, acute kidney injuries, and portal vein thrombosis.Independent predictors of in-hospital mortality were coagulation defects (Adjusted OR, 2.25; 95% CI, 1.06 -4.80), sepsis (Adjusted OR, 4.66; 95% CI, 1.61 -13.43), hepatocellular carcinoma (Adjusted OR, 7.42; 95% CI, 2.52 -21.86), hepatic encephalopathy (Adjusted OR, 3.29; 95% CI, 1.26 -8.58), and endotracheal intubation (Adjusted OR, 15.28; 95% CI,, Conclusion: The all-cause in-hospital mortality rate in variceal hemorrhage was 8.1%.The presence of coagulation defects, sepsis, hepatocellular carcinoma, hepatic encephalopathy, and endotracheal intubation were independent predictors of mortality.This data can be used to develop a risk assessment model for patients admitted with esophageal variceal bleeding.