All patients were risk stratified using the Glasgow-Blatchford bleeding score (GBS). Continuous data was assessed using the Mann-Whitney test and categorical data using Fisher’s exact test. Results: Of 373 patients with a primary diagnosis of UGIB, 56 (15%) presented with CGV. The mean age was 63 years (range 15–93) and 40 (71.4%) were male. 30 (54%) presented with isolated CGV while 26 (46%) presented with CGV plus melaena (22) and/or haematemesis (8, 5-Fluoracil research buy defined as haematemesis documented prior to ED presentation). No statistically significant differences in age or co-morbidity

burden were detected between the two groups. At endoscopy, patients presenting with isolated CGV were more likely to have a Mallory Weiss tear (4 vs 1, p = NS) or inflammatory or erosive disease of the oesophagus, stomach or duodenum (18 vs 14, p = NS). Patients presenting with

CGV plus haematemesis or melaena (CGV+HM) were more likely to have ulceration or malignancy (10 vs 2, p = 0.007). The two ulcers in isolated CGV patients were Forrest 3 and check details did not require endoscopic therapy. CGV+HM patients were more likely to be anticoagulated (19 vs. 10, p = 0.0038), require blood transfusion (15 vs. 7, p = 0.013), have a lower haemoglobin on presentation (110 vs. 128, p = 0.016) and have a higher GBS (7.8 vs. 4.7, p = 0.009). No differences were recorded in the number of patients on a Proton Pump Inhibitor (PPI) or treated with intravenous PPI during admission. One patient died nine days after gastroscopy from an unrelated condition. Conclusion: Patients presenting with coffee ground vomiting as the sole presenting symptom of an upper gastrointestinal bleed have a low risk of serious pathology being found at endoscopy. This implies that these patients do not require urgent medchemexpress endoscopy. M ROBERTSON,1 A MAJUMDAR,1

R BOYAPATI,1 W CHUNG,1 R TURBAH,1 J WEI,1 R VAUGHAN,1 S LONTOS1 1Department of Gastroenterology and Liver Transplant Unit, Austin Hospital, Heidelberg, Australia Introduction: Multiple algorithms predicting outcomes in upper gastrointestinal bleeding (UGIB) have been developed, the most widely used of which are the Glasgow-Blatchford (GBS) and Rockall scores. AIMS65 is a novel risk stratification score designed to predict inpatient mortality. The AIMS65 score assigns 1 point for each of the following: albumin level <30 g/L, INR > 1.5, altered mental status, systolic blood pressure <90 mmHg and age older than 65 years. Compared with existing scores, AIMS65 has the advantages of not being weighted and can be easily calculated with pathology values routinely obtainable in the emergency department. Objective: To assess the AIMS65 score as predictor of inpatient mortality in patients presenting with acute UGIB.