Abstract

Background: Continued or recurrent bleeding after endoscopic treatment for bleeding ulcer is a major adverse prognostic factor. Identification of such ulcers may allow for alternate treatments.

Aim: To determine factors predicting treatment failure with combined adrenaline injection and heater probe thermocoagulation.

Methods: Consecutive patients with bleeding peptic ulcers who received endoscopic therapy between January 1995 and March 1998 were studied. Data on clinical presentation, endoscopic findings, and treatment outcomes were collected prospectively. Multiple logistic regression analysis was used to identify independent risk factors for treatment failure.

Results: During the study period, 3386 patients were admitted with bleeding peptic ulcers: 1144 (796 men, 348 women) with a mean age of 62.5 (SD 17.6) years required endoscopic treatment. There were 666 duodenal ulcers (58.2%), 425 gastric ulcers (37.2%), and 53 anastomotic ulcers (4.6%). Initial haemostasis was successful in 1128 patients (98.6%). Among them, 94 (8.2%) rebled in a median time of 48 hours (range 3–480). Overall failure rate was 9.6%. Mortality rate was 5% (57/1144). Multiple logistic regression analysis revealed that hypotension (odds ratio (OR) 2.21, 95% confidence interval (CI) 1.40–3.48), haemoglobin level less that 10 g/dl (OR 1.87, 95% CI 1.18–2.96), fresh blood in the stomach (OR 2.15, 95% CI 1.40–3.31), ulcer with active bleeding (OR 1.65, 95% CI 1.07–2.56), and large ulcers (OR 1.80, 95% CI 1.15–2.83) were independent factors predicting rebleeding.

Conclusions: Larger ulcers with severe bleeding at presentation predict failure of endoscopic therapy.