Upper Gastrointestinal Hemorrhage in Jordan: An Analysis of Causes, Characteristics and Outcome

Abstract

Two hundred twenty-four patients presenting consecutively with massive upper gastrointestinal hemorrhage were endoscoped within 24 hours to determine the site of bleeding. Outcome was evaluated in relation with presenting variables. Duodenal ulcer was found to be the most common source of bleeding. Endoscopy demonstrated 33% of all patients had more than one potential bleeding site. Patients with duodenal ulcer were significantly younger than patients with gastric ulcer (41 +/- 17 years versus 58 +/- 19 years, respectively). There was a tendency for upper gastrointestinal hemorrhage to present more often in the winter months. Eighty-one percent of duodenal ulcer patients and 57% of gastric ulcer patients had dyspepsia prior to the presentation of bleeding. There was a strong correlation between presence of concomintant disease and the age of the patient. Forty-three percent of patients had ingested aspirin or other nonsteroidal anti-inflammatory agents in the week prior to the bleed compared with 18% in a control group. All nine patients with bleeding diathesis (including that induced by anticoagulant maintenance therapy) had an underlying source of bleeding. Sixty-nine percent of the patients had spontaneous resolution of their bleeding within 24 hours while 25% underwent emergency surgery. The overall mortality rate was 10%. Age, underlying chronic liver disease, and transfusion requirements were found to be independent predictors of mortality assessed by multivariate regression analysis.