Failure of endoscopic therapy in upper gastrointestinal hemorrhage due to duodenal ulcers

to evaluate the efficacy of endoscopic treatment in patients with upper gastrointestinal (UGIH) due to duodenal ulcer with high risk of persistent or recurrent bleeding and to determine the associated failure factors of this procedure.

three hundred and thirty-six patients with UGIH due to duodenal ulcer requiring endoscopic treatment were analyzed between January 1992 and December 2001. The patients were classified according to the endoscopic findings: a) patients with limited bleeding; and b) patients with persistent and/or recurrent bleeding due to therapeutic failure. The clinical guidelines followed in patients with endoscopic treatment failure were previously established in the internal protocol. The variables that obtained statistical significance in the univariate analysis were included in the logistic regression model to identify those with an independent predictive value for failure of the endoscopic treatment.

mean age of the patients was 60 +/- 17 years, 271 (81%) were male. Bleeding with severe hemodynamic affectation was detected in 82 patients (24%). The most common location of the duodenal ulcer was on the anterosuperior part of the duodenal bulb (227 patients, 68%). In 43 patients (13%) the ulcer was larger than 2 cm. The bleeding stigmata were classified as: Forrest I in 125 (38%) and Forrest II in 211 (62%). It was initially reached in 297 patients (88%). Twenty-two patients required emergency surgery (6,5%) and the global mortality rate was 3%. Severe hemodynamic affectation at admission (OR 11.8, p>0.001), ulcers exceeding 2 cm (OR 6.95, p = 0.019) and the presence of active bleeding during endoscopy (Forrest I) (OR 3.55, p = 0.08) were the variables included in the multivariate analysis independently associated to endoscopic therapy failure.

endoscopic therapy is an efficient treatment of upper gastrointestinal bleeding due to duodenal ulcer. By means of a clinical variable, the hemodynamic status and two endoscopies, bleeding stigmata and the size of the ulcer, a group of patients with high risk of endoscopic treatment failure can be selected.