Ulcers and nonvariceal bleeding.

Abstract

Although peptic ulcer continues to be the commonest cause of acute upper gastrointestinal bleeding, obscure haemorrhage from the Dieulafoy malformation and haemobilia must be considered, and may be amenable to endoscopic therapy. The patients who are at the highest risk of rebleeding and death are elderly, in shock at presentation, and have major stigmata of recent haemorrhage (SRH). The endoscopist must identify SRH, and identification may be made easier by washing the area with hydrogen peroxide. The natural history of SRH has been defined. There is wide interobserver variation in the interpretation of SRH, and there is probably therefore little value in the endoscopist describing subtle appearances. Although the value of endoscopic haemostatic therapy is established, it has still not been taken up by all institutions. Endoscopic injection of fibrin glue into the bleeding ulcer is a logical and relatively easy approach, and a systematic histological study of resected ulcers has shown that this does not adversely affect the ulcer healing process. Thermal therapies such as argon plasma coagulation and the heater probe have comparable efficacy. Although a combination of injection and thermal treatments may be logical, there are only trends suggesting that this is better than monotherapy. Nevertheless, the gold probe continues to be used in clinical practice. Experiments in an animal model of gastric bleeding suggest that the gold probe is effective, and that the version with a wide-gauge needle is best. Haemoclips may stop acute upper gastrointestinal bleeding from a range of sources. Patients who rebleed after initial endoscopic haemostasis have a tenfold increase in the risk of death. An important study from Hong Kong suggests that repeat endoscopic treatment after rebleeding has comparable morbidity and mortality to a policy of urgent surgery without endoscopic repeat intervention.