Bleeding gastroduodenal ulcers: selection of patients for surgery.

Abstract

AbstractIn a prospective study of bleeding peptic ulcer from 1980 to 1985, endoscopic findings were found to be inferior to shock on admission in the prediction of further bleeding in the hospital. According to management policy, 40 patients (age greater than 50 years on admission) from a total of 376 cases were treated by emergency surgery for shock. Endoscopic methods of treatment were not used during this period of study. In the remaining 336 cases, 83 (25%) bled again in the hospital and 62 patients in this group underwent urgent surgery. Endoscopic findings were classified into 3 groups: active bleeding (including visible vessel), evidence of recent bleeding, and no stigmata of hemorrhage in the upper gastro‐intestinal tract. In patients who bled again in the hospital, there was no significant difference between the number of cases with active bleeding (44%) and those with signs of recent bleeding (30%). Visible vessel was an unusual finding, present in 4% of cases, but was associated with a risk of further bleeding of 38%, an incidence comparable to active bleeding on endoscopy. The incidence of further bleeding was low (7%) without stigmata and comparable to that found in 33 cases with prepyloric ulcer (10%). Shock on admission was associated with a significantly greater incidence of rebleeding in 70% of cases. The combination of shock on admission and active bleeding on endoscopy was no better a predictor of further bleeding than was shock alone. Age was a poor discriminator as the incidence of further bleeding was the same comparing those under and over 50 years. Active bleeding on endoscopy was a less common finding with gastric ulcer than with duodenal ulcer, but was associated with a significantly greater incidence of rebleeding of 58% of cases compared to 42% for duodenal ulcer.