A controlled, double blind, multi-center trial of endoscopic hemostasis for bleeding benign peptic ulcers will be performed with two portable devices. A total of 150 hospitalized patients with severe UGI hemorrhage and endoscopically documented peptic ulcer active bleeding or non-bleeding visible vessels will be randomized to receive a) endoscopic hemostasis with bipolar electrocoagulation (BP), b) endoscopic hemostasis with heater probe (HP) or c) routine endoscopy management without endoscopic hemostasis (control-C). All patients who meet entry criteria and give informed consent will be randomized. Stratification will be made into good or bad prognosis categories. Randomization will be performed by a therapeutic endoscopist at the time of endoscopy by opening a sealed envelope that designates treatment group or control. A different physician, who is unaware of the endoscopic treatment given by the therapeutic endoscopist will manage the patient before and after the randomization. Intensive care and routine medical-surgical management will be similar for all patients. Data will be collected on standard forms during the hospitalization to determine if patients benefited from the endoscopic therapy. To evaluate the treatments, comparisons of the three groups (BP, HP, C) will be made prospectively at four times during this study: 5 mins, 24 hrs, 72 hrs, discharge or seven days (whichever is earlier). Specific outcomes compared during hospitalization will be: hemostasis rates; rebleeding rates, emergency surgery rates for control of bleeding; number of transfusions before and after randomization; hospitalization duration (including intensive care unit time and regular floor time); elective surgery rates; complications; and mortality. For these comparisons other than mortality, a graded outcome score will be used. Comparisons will be determined for all treatment groups (BP, HP, C); for active bleeders and non-bleeding visible vessels as subgroups; and for good and bad prognosis subgroups. The study will be carried out by four experienced investigators in different study centers and should be completed within 5 years. The advantages of this trial are the protocol and study design; the study's efficient use of financial resources; the likelihood of completing the study; and the investigators who are all experienced therapeutic endoscopists with training in controlled clinical trials and the use of these thermal coagulation devices.
