BACKGROUND: Randomised controlled trials (RCTs) evaluating the clinical effect of proton pump inhibitors (PPIs) in peptic ulcer (PU) bleeding yield conflicting results. OBJECTIVES: To evaluate the efficacy of PPIs in acute bleeding from PU using evidence from RCTs. SEARCH STRATEGY: We searched CENTRAL, The Cochrane Library (Issue 4, 2004), MEDLINE (1966 to November 2004), EMBASE (1980 to November 2004), proceedings of major meetings to November 2004, and reference lists of articles. We contacted pharmaceutical companies and experts in the field. SELECTION CRITERIA: RCTs of PPI treatment (oral or intravenous) compared with placebo or H(2)-receptor antagonist (H(2)RA) in acute bleeding from PU. DATA COLLECTION AND ANALYSIS: Two reviewers extracted data independently, assessed study validity, summarised studies and undertook meta-analysis. The influence of study characteristics on the outcomes was examined by subgroup analyses and meta-regression. MAIN RESULTS: Twenty-four RCTs comprising 4373 participants in total were included. Statistical heterogeneity was found among trials for rebleeding (P = 0.04), but not for all-cause mortality (P = 0.24) or surgery (P = 0.45). There was no significant difference in all-cause mortality rates between PPI and control treatment; pooled rates were 3.9% on PPI versus 3.8% on control (odds ratio (OR) 1.01; 95% CI 0.74 to 1.40). PPIs significantly reduced rebleeding compared to control; pooled rates were 10.6% with PPI versus 17.3% with control treatment (OR 0.49; 95% CI 0.37 to 0.65). PPI treatment significantly reduced surgery compared with control; pooled rates were 6.1% on PPI versus 9.3% on control (OR 0.61; 95% CI 0.48 to 0.78). There was no evidence to suggest that results on mortality and rebleeding were dependent on study quality, route of PPI administration, type of control treatment or application of initial endoscopic haemostatic treatment. PPIs significantly reduced surgery compared with placebo but not when compared with H(2)RA. There was no evidence to suggest that study quality, route of PPI administration or application of initial endoscopic haemostatic treatment influenced results on surgery. PPI treatment appeared more efficacious in studies conducted in Asia compared to studies conducted elsewhere. All-cause mortality was reduced only in Asian studies; reductions in rebleeding and surgery were quantitatively greater in Asian studies. Among patients with active bleeding or non-bleeding visible vessel, PPI treatment reduced mortality (OR 0.53; 95% CI 0.31 to 0.91), rebleeding and surgery. AUTHORS' CONCLUSIONS: PPI treatment in PU bleeding reduces rebleeding and surgery compared with placebo or H(2)RA, but there is no evidence of an overall effect on all-cause mortality.
