Endoscopic management of the bleeding ulcer. A survey.

The study was performed to evaluate the prognostic and therapeutic value of endoscopy in patients with bleeding ulcer. Before endoscopic control of ulcer bleeding was introduced 373 patients with bleeding ulcer and a median age of 67 years were studied retrospectively. Emergency surgery was required in 155 patients. The surgical mortality was 11% in 37 low risk patients, but 36% in 118 poor risk patients. There was a trend to improved outcome after introduction of diagnostic endoscopy but only in patients with hemorrhage managed conservatively. Surgical mortality remained unchanged. Ranked in order of prognostic importance, a complicating disorder, postoperative complications, overtransfusion, and absent past history of ulcer dyspepsia were identified as the most important determinants of fatal outcome. Age, onset of hemorrhage at home or in hospital, previous ulcer surgery, previous bleeds, ulcer site, and sex had comparatively less bearing on outcome. It was concluded that although surgery was efficient in preventing exsanguination, it was poorly tolerated in poor surgical risks who constituted 76% of the patients in need of emergency control of ulcer bleeding. The results indicate that the search for non-surgical methods is justified. In a pilot study, endoscopic electrocoagulation was applied to control ulcer bleeding in 60 patients. The experience from the study served as basis for a prospective study in an attempt to assess the prognostic and therapeutic potential of endoscopy in the management of bleeding ulcer and to define indications for emergency endoscopy. The histomorphologic effect of electrocoagulation employed to stop bleeding from acute gastric ulcers in rabbits suggested that intravascular occlusive fibrin thrombosis is the probable mechanism of hemostasis. In the prospective study, 539 consecutive patients admitted with hematemesis and melena underwent emergency endoscopy. Peptic ulcer, identified as the bleeding source in 51% of these patients, was the predominant lesion most liable to hemorrhage assessed as requiring emergency surgery. Black hematemesis with melena, occurring in 13% of the patients, was the best predictor of ulcer bleeding. Red hematemesis with melena (26%) was the most important predictor of major bleeding. Black hematemesis with melena was found to be as important as red hematemesis without melena (22%) in predicting major ulcer bleeding. In contrast, melena (18%) and, in particular, black hematemesis without melena (20%) were poor predictors of ulcer bleeding as well as of major bleeding.(ABSTRACT TRUNCATED AT 400 WORDS)