Impact of anticoagulation on rebleeding following endoscopic therapy for nonvariceal upper gastrointestinal hemorrhage.

OBJECTIVE: Endoscopic therapy for nonvariceal upper gastrointestinal hemorrhage achieves hemostasis in greater than 90% of patients, but up to 20% rebleed. The aim of this study was to determine the impact of anticoagulation on rebleeding in patients undergoing endoscopic therapy for nonvariceal upper gastrointestinal hemorrhage.

METHODS: Patients who underwent successful endoscopic therapy for nonvariceal upper gastrointestinal hemorrhage between July 1, 1999, and June 30, 2004, at a large, tertiary care teaching hospital were identified. The primary outcome was rebleeding within 30 days. Secondary outcomes were transfusion requirement, length of stay, surgery, and mortality. Baseline data were analyzed using t-tests and chi(2) tests. Multivariable logistic and linear regression analyses were carried out to calculate the adjusted odds ratios for the international normalized ratio (INR) predicting the primary and secondary outcomes. The multivariable analyses controlled for: age, Charlson comorbidity index, antiplatelet agent use, postprocedure heparin use, postprocedure proton pump inhibitor use, hypotension, ulcer as the bleeding source, and active bleeding at endoscopy.

RESULTS: The study included 233 patients. Forty-four percent of the patients had an INR >or=1.3. Ninety-five percent of the anticoagulated patients had an INR between 1.3 and 2.7. The rebleeding rate was 23% in the anticoagulated patients and 21% in the patients with INRs <1.3. On multivariable analyses, INR was not a predictor of rebleeding, transfusion requirement, surgery, length of stay, or mortality.

CONCLUSIONS: Mild to moderate anticoagulation does not increase the risk of rebleeding following endoscopic therapy for nonvariceal upper gastrointestinal hemorrhage, suggesting that endoscopic therapy is appropriate in these patients.