Ulcers and nonvariceal bleeding.

Peptic ulcer remains the commonest and most significant cause of nonvariceal upper gastrointestinal bleeding. The incidence of peptic ulcer bleeding is rising in elderly patients, particularly for duodenal ulcer. Patients presenting with upper gastrointestinal bleeding who have low Rockall scores are at low risk of rebleeding and death. These patients currently utilize considerable health-care resources, but could safely be managed at home. The Rockall score can be used to predict the risk of rebleeding and death following variceal bleeding, but for patients with ulcer bleeding, its ability to predict death is questioned. Acid suppression is effective in preventing rebleeding from peptic ulcer. Standard doses of intravenous omeprazole may be as effective as high-dose regimens. Oral omeprazole also reduces rebleeding following endoscopic therapy for peptic ulcer. Mallory-Weiss tears result in significant bleeding in 23 % of cases. Endoscopic therapy may only be required in cases in which active bleeding is present. Endoscopic therapy is effective and safe in patients with major peptic ulcer bleeding who are over 80 years old. For peptic ulcer, injection of larger volumes of epinephrine (adrenaline; mean 16.5 ml) are more effective than small volumes (mean 8 ml). Injection of normal saline alone is less effective than bipolar electrocoagulation. The addition of fibrin glue to epinephrine injection does not confer an additional benefit over epinephrine alone. Argon plasma coagulation can be used to treat a range of lesions in the gastrointestinal tract. It is also effective for treatment of bleeding ulcer, but is no better than established methods. Haemoclips may be useful in bleeding Mallory-Weiss tears, but their use is difficult in patients bleeding from peptic ulcer. The presence of a large ulcer and active bleeding at the time of endoscopy are independent predictors of failure of endoscopic therapy.