The optimal dosage of intravenous proton pump inhibitors (PPIs) for the prevention of peptic ulcer rebleeding remains unclear.


To compare the rebleeding rate of high-dose and standard-dose PPI use after endoscopic haemostasis.


A total of 201 patients with bleeding ulcers undergoing endoscopic treatment with epinephrine injection and heater probe thermocoagulation were randomised to receive a high-dose regimen (80 mg bolus, followed by pantoprazole 8 mg/h infusion, = 100) or a standard-dose regimen (pantoprazole 40 mg bolus daily, = 101). After 72 h, all patients were given 40 mg pantoprazole daily orally for 27 days.


There were no statistical differences in mean units of blood transfused, length of hospitalisation ≦5 days, surgical or radiological interventions and mortality within 30 days between two groups. Bleeding recurred within 30 days in six patients [6.2%, 95% confidence interval (CI) 1.3–11.1%] in the high-dose group, as compared to five patients (5.2%, 95% CI 0.6–9.7%) in the standard-dose group (= 0.77). The stepwise Cox regression analysis showed end-stage renal disease, haematemesis, chronic obstructive pulmonary disease (hazard ratio: 37.15, 10.07, 9.12, 95% CI: 6.76–204.14, 2.07–49.01, 1.66–50.00 respectively) were independent risk factors for rebleeding and Helicobacter pylori infection was associated with lower risk of rebleeding (hazard ratio: 0.20, 95% CI: 0.04–0.94).


Following combined endoscopic haemostasis of bleeding ulcers, co-morbidities, haematemesis and H. pylori Status, but not PPI dosage, are associated with rebleeding (http://www.Clinical NCT00709046).