Abstract.Objective: To determine the number of ED patients with non-variceal upper gastrointestinal hemorrhage (NVUGIH) who could have been managed as outpatients through application of previously developed clinical guidelines. Methods: Descriptive study based on retrospective chart review of patients who presented with acute upper gastrointestinal hemorrhage (UGIH) to the ED of an urban teaching hospital from July 1 to December 31, 1996. Applying the clinical guidelines published by a health maintenance organization (HMO) group (no high-risk endoscopic features/varices/portal hypertensive gastropathy, no debilitation, no orthostatic vital sign change, no severe liver disease, no serious concomitant disease, no anticoagulation or coagulopathy, no fresh, voluminous hematemesis or multiple episodes of melena on the day of presentation, no severe anemia, and adequate home support), patients who could have been managed as outpatients after esophagogastroduodenoscopy (EGD) were identified and analyzed. Results: 145 UGIH patients were seen in the ED, of whom 128 (88%) were admitted and 111 (77%) underwent EGD. 21 (19%) had varices, leaving 90 (81%) with NVUGIH. 18 of these 90 patients (20%, 95% CI = 12% to 28%) fulfilled guidelines for outpatient management and had the following characteristics with p < 0.05: younger age [mean 54.2 Ϯ 5.5 (SEM) vs 63.8 Ϯ 1.9 years], less transfusion (0.9 Ϯ 0.3 vs 3.7 Ϯ 0.4 units), and shorter length of stay (2.1 Ϯ 0.4 vs 5.3 Ϯ 0.7 days). None of the 18 outpatient management patients had any complications. Conclusion: In a non-HMO urban teaching hospital, 18 patients with NVU-GIH met criteria for outpatient management in a six-month period and none developed a complication during a mean in-hospital stay of 2.1 days. Key words: acute upper gastrointestinal hemorrhage; clinical guidelines; outpatient care; bleeding. ACA-DEMIC EMERGENCY MEDICINE 1999; 6:196 -201 I N THE LAST three decades, a number of studies 1 -5 have used clinical and endoscopic features to predict patients' outcomes after upper gastrointestinal hemorrhage (UGIH). It was, however, with the advent of endoscopic hemostasis in the last decade that the management of non-variceal upper gastrointestinal hemorrhage (NVUGIH) changed radically. From numerous studies 5 -11 it was clear that endoscopic features of peptic ulcers enabled clinicians to predict the risk of rebleeding, and appropriate endoscopic intervention helped to reduce this risk. Following this development, at least two studies 12,13 have shown that selected pa- From an emergency medicine point of view, we were particularly interested in the practice guidelines developed by Longstreth and Feitelberg 12 to select patients with UGIH for outpatient care. These guidelines were developed in a health maintenance organization (HMO) setting and we wanted to know whether they would be applicable in a non-HMO urban hospital. The study by Lai et al. 13 addressed only patients with duodenal ulcers and all of them were admitted to the gastroenterology unit before assessment...