Background and study aims: A second-look endoscopy is often performed to evaluate the efficacy of a prior injection therapy in patients with bleeding peptic gastric or duodenal ulcers. Although this strategy is widely established, it does not rely on unequivocal data from controlled studies. In a prospective, randomized, controlled multicenter trial we assessed the effect of programmed endoscopic follow-up examinations with eventual retreatment on the outcome of bleeding ulcers in these patients.
Patients and methods: One hundred and five patients with gastric or duodenal peptic ulcers presenting with active (Forrest type I) or recent (Forrest type IIa and IIb) bleeding upon endoscopy within four hours after admission were included in the study. Emergency treatment consisted of the sequential injection of both epinephrine (1:10,000 v/v) and up to 2 ml of fibrin/thrombin around the ulcer base. Fifty-two patients were randomized to receive programmed endoscopic monitoring with eventual retreatment in cases of Forrest type I, IIa, or IIb ulcers beginning within 16-24 hours after the index bleed. Follow-up endoscopies were continued until the macroscopic appearance revealed a Forrest type IIc or III ulcer. Fifty-three patients in the control group were closely monitored, and only received a second endoscopy when there was clinical or biochemical evidence of recurrent bleeding. The groups did not differ with respect to age, sex, site and severity of bleeding.
Results: The numbers of patients with recurrent bleeding were similar whether they were endoscopically monitored or not (21% versus 17%, P=0.80 chi-squared test). In addition, there was no statistically significant difference between the two groups with respect to the number of blood units transfused, need for surgical intervention, hospital stay or number of deaths (Mann-Whitney U-test). Improving local ulcer stigmata was not related to a better outcome.
Conclusions: Programmed endoscopic follow-up examinations with eventual retreatment in patients locally injected for an acute or recent hemorrhage from a gastric or duodenal ulcer did not influence their outcome when compared to patients receiving only a second endoscopic intervention upon evidence for recurrent hemorrhage. Scheduled control endoscopies cannot be recommended after an initial successful endoscopic treatment of peptic ulcer bleeding when selection of the patients for second-look endoscopy is directed by the Forrest criteria.