ABSTRACT
BACKGROUND: Peptic ulcer disease is considered the cause of upper-GI bleeding in 50% of cases. A recent decline in the proportion of cases of upper-GI bleeding because of a peptic ulcer was noted by us. The objectives of this study were to evaluate the frequency of peptic ulcer in patients with upper-GI bleeding and the proportion of bleeding peptic ulcers with a non-bleeding visible vessel. METHODS: Patients with upper-GI bleeding seen from December 1999 until April 2001 at a tertiary, university-affiliated medical center were studied prospectively. The Clinical Outcome Research Initiative (CORI) database was used to correlate the single institution data with nationwide data. Endoscopic data in the CORI database for patients who had endoscopy for upper-GI bleeding between December 1999 and July 2001 were retrieved and analyzed. RESULTS: A total of 126 patients were included in the prospective study. The endoscopic findings were: peptic ulcer in 31.8%: 95% confidence interval (CI) [23.7%, 40.6%] of patients; a non-bleeding visible vessel was present in 10%: 95% CI[2.8%, 23.7%] of these peptic ulcers. From the nationwide CORI database, data for 7822 patients with upper-GI bleeding were obtained. The endoscopic findings were: peptic ulcer in 20.6%:95% CI[19.7%, 21.5%] of patients with upper-GI bleeding; a non-bleeding visible vessel was present in 7.3%: 95% CI[6.1%, 8.6%] of the ulcers. CONCLUSIONS: The frequency of peptic ulcer in patients with upper-GI bleeding and the proportion of bleeding ulcers with a non-bleeding visible vessel are less than previously reported.
