• No evidence of hemorrhage or vessel exposure on a second endoscopy
  
• No evidence of hemorrhage or vessel exposure on a third endoscopy
  

• Plasma ghrelin levels
  
• Serum gastrin levels
  

Gastrointestinal hemorrhage is a relatively common condition, with the source of the bleeding being most commonly from the upper gastrointestinal tract, especially from gastric and duodenal ulcers. It often requires emergency treatment. First, the site of bleeding is determined. If an exposed blood vessel is found in the hemorrhagic lesion, or in the case of oozing or projectile hemorrhage, endoscopic hemostasis is performed on the lesion. After hemostasis is achieved, prevention of re-bleeding is important; usually, an antacid or similar medication is administered and the course is monitored under fasting conditions.Suppression of gastric acid secretion is necessary to raise gastric pH levels and maintain normal blood coagulation, and to promote healing of hemorrhagic lesions. In Japan, intravenous preparations of H2 receptor antagonists and proton pump inhibitors have been commonly used.In foreign countries, drug therapy for patients with upper gastrointestinal hemorrhage emphasizes the maintenance of normal blood coagulation. High doses of these drugs have been established to constantly maintain a pH of 7 in the stomach (Daneshmend TK, et al., BMJ 1992, 304:143-147; Labentz J, et al., Gut 1997, 40:36-41; Hasselgren G, et al., Scand J Gastroenterol 1997, 32:328-333; Schaffalitzky de Muckadell OB, et al., Scand J Gastroenterol 1997, 32:320-327; Sung JJY, et al., Ann Intern Med 2003, 139:237-243). In a clinical study, proton pump inhibitors were superior to H2 receptor antagonists in terms of clinical efficacy (Labentz J, et al., Gut 1997, 40:36-41). In Japan, emphasis is placed on promoting healing of lesions since endoscopic hemostasis is a fairly common practice; doses have been established at levels similar to therapeutic doses for peptic ulcers. It cannot be said, however, that superiority of intravenous proton pump inhibitors over H2 receptor antagonists has been established at such doses. This can possibly be attributed to fact that in previous studies the study populations were not homogenous in terms of severity; for example, patients requiring endoscopic hemostasis and those that did not were both included.Against this background, this study will compare the hemostasis-maintaining effects of intravenous omeprazole and ranitidine in patients with upper gastrointestinal hemorrhage that have undergone endoscopic hemostasis, to establish which anti-secretory medication prior to the start of oral alimentation is effective in preventing re-hemorrhage after hemostasis.