[Risk factors associated to mortality by upper GI bleeding in patients from a public hospital. A case control study].

Identify and establish risk factors associated with mortality secondary to upper gastrointestinal bleeding up to 30 days after the episode, at the Hospital Nacional Cayetano Heredia. A retrospective analytic observational case-control study was made with a case: control proportion of 1:3, analyzing 180 patient from which 135 were the controls and 45 the cases. It was determined by biological plausibility as potential risk factors to 14 variables, with which were performed bivariate and multivariate logistic analyses. It was found in the bivariate logistic analysis as variables statistically related to mortality: age (OR=1.02), hematemesis (OR=2.57), in-hospital upper gastrointestinal bleeding (OR=4), cirrhosis (OR=2.67), malignancy (OR=5,37), admittance to intensive care unit/Shock-Trauma (OR=9.29), Rockall score greater than 4 (OR=19.75), rebleeding (OR=5.65), and number of packed red blood cell transfusions(OR=1.22). While in the multivariate logistic analysis, the only variables statistically related to mortality were: malignancy (OR=5.35), admittance to intensive care unit/Shock-Trauma (OR=8.29), and Rockall score greater than 4 (OR=8.43). The factors that increase the risk of mortality in patient with upper gastrointestinal bleeding are: Rockall score greater than 4, admittance to intensive care unit/Shock-Trauma, rebleeding, in-hospital upper gastrointestinal bleeding, cirrhosis, hematemesis, malignancy, the number of red blood cell transfusion, and age.