In terms of etiology, 21 patients (27.3%) took NSAIDs or antiplatelet agents.

H. pylori infection was detected in 13 patients (16.9%). 45 patients (58.4%) had an idiopathic etiology. Ulcers were predominantly located at duodenal bulb (59.7%) or D1/D2 junction (28.6%), with either Forrest class Ib (33.8%) or IIa (42.9%) morphology. Although all patients were treated endoscopically, 9 patients required salvage therapy; angio-embolisation (6) or surgery (3). Surveillance was performed at a mean duration of 54.6 days (range 28–125). At surveillance, 68 (88.3%) had complete healing of duodenal ulcers. Diabetes mellitus (DM) was associated with persistence of ulcer at surveillance [Odds Ratio (OR) 5.6, 95% CI 1.2–24.6; p = 0.02]. DM patients had a mean HbA1C of 7.2%. When compared with Chinese race, Malay race had higher risk of persistent ulcer [OR 9.9, 95% CI 1.9–52.3; p = 0.007]. BMN 673 Following multivariate logistic regression, Malay race was the only statistically significant predictor of persistent ulcer [OR 6.9,

95%CI 1.2–39.5; XL184 in vivo p = 0.03]. Post-surveillance, 9 patients with persistent ulcer were given a longer course of PPI therapy (5) or changed to a more potent PPI (4). Conclusion: Following therapy, bleeding duodenal ulcers may have delayed healing, especially in the Malay patient with DM. Further larger prospective studies may establish the role of surveillance endoscopy in this group of patients. Key Word(s): 1. duodenal ulcer; 2. bleeding ulcer; 3. therapeutic endoscopy; 4. surveillance see more Presenting Author: HYEWON LEE Additional Authors: EUN JUNG JEON, WOO CHUL CHUNG, CHANG NYOL PAIK, KANG MOON LEE Corresponding Author: HYEWON LEE Affiliations: St. Paul’s Hospital, St. Vincent’s Hospital, St. Vincent’s Hospital, St. Vincent’s Hospital Objective: To evaluate the clinical outcomes and severity of peptic ulcer bleeding (PUB) according

to the etiology – Helicobacter pylori (H. pylori) and drug (aspirin and nonsteroidal anti-inflammatory drug). Methods: A consecutive series of patients who had PUB and admitted to the hospital between 2006 and 2012 were retrospectively analyzed. A total of 232 patients were enrolled in this study, and we compared the clinical characteristics and outcomes according to the different etiologies (H. pylori, drug, H. pylori with drug and idiopathic). We also evaluated the severity using Blatchford score and Rockall score between four groups. Results: When H. pylori associated PUB compared with drug induced PUB, it was male dominant. In drug induced PUB, the longer duration of admission and larger ulcer were observed. Also, Blatchford score and Rockall score were the higher than H. pylori associated PUB. When idiopathic PUB compared with H. pylori associated PUB, the larger ulcer and more frequent rate of re-bleeding were observed. When idiopathic PUB compared with drug induced PUB, it was distinct of male predominance. Re-admission rate and re-bleeding rate after initial hemostasis were higher in idiopathic PUB.