OBJECTIVE: To perform metaanalyses of studies on outcome of bleeding ulcers of different proton-pump inhibitors (PPIs) regimens, after stratification of patients by endoscopic stigmata, and analysis of studies with and without endotherapy. METHODS: A total of 35 randomized trials comparing PPIs to placebo and/or H2-receptor antagonists (H2RAs) in 4,843 patients with high-risk endoscopic stigmata were retrieved. Outcomes were rebleeding, surgery, and mortality. RESULTS: Monotherapy with oral or bolus PPIs was superior to placebo and H2RAs in reducing rebleeding in both bleeders and nonbleeders at index endoscopy; the need for surgery was reduced only when compared to H2RAs. In nonbleeders, PPI monotherapy was as effective as a combination of endotherapy with H2RAs. A combination of endotherapy with PPIs was superior to monotherapy in reducing bleeding and surgery, and superior to endotherapy alone in minimizing rebleeding, but not surgery; the benefit was lost when confronted to endotherapy plus H2RAs, whether PPIs were given as infusion or bolus. By pooling data from studies comparing high doses of PPIs as continuous infusion versus regular doses as intermittent bolus, rebleeding, surgery, and mortality were not significantly different. CONCLUSIONS: Combination of endotherapy with either PPIs or H2RAs is indicated for nonbleeding ulcers at endoscopy with the intent to reduce rebleeding and surgery. Its value may extend to bleeding lesions, but current data are scanty. The benefit appears to be independent from route and doses of PPIs, as oral, bolus, or infusional methods are all effective.
ASJC Scopus subject areas