Mo1125 A Prospective Analysis of the Glasgow-Blatchford Score as a Predictor of Length of Hospital Stay for Upper Gastrointestinal Bleeding

Abstract

Background Antithrombotic drugs including low-dose aspirin, antiplatelet drugs and anticoagulants increase upper gastrointestinal bleeding (UGIB). This study aimed to clarify the features of UGIB in 11,919 Japanese patients taking antithrombotic drugs, and examine the temporal changes in the pathogenic mechanism for UGIB. Methods The characteristics of the cohort with those of the 11,919 patients, who were prescribed antithrombotic drugs in Saga Medical School Hospital during 2002 and 2011, were analyzed for examination of relationships between the use of acid-secretion inhibitors and antithrombotic drugs in the UGIB patients. During the period (2002-2011), retrospective chart review revealed 430 patients who underwent emergency endoscopy for UGIB in Saga Medical School Hospital. Results Numbers of UGIB patients were 186 in 2002-2006 and 244 in 2007-2011. In the second period, the proportion of UGIB patients infected with Helicobacter pylori decreased (76.9% in the first period vs. 49.6% in the second, p<0.0001), and the proportion of UGIB patients who took antithrombotic drugs increased (21.5% vs. 32.8%, p=0.012). The most common cause of UGIB was peptic ulcer disease, and the incidence of other causes, including reflux esophagitis, and gastric antral vascular ectasia, has increased in a time dependent manner. Significant risk factors for antithrombotic-related UGIB were hypertension (OR 6.546, 95%CI 4.091-10.477, p<0.001) and diabetes mellitus (OR 2.947, 95%CI 1.7534.956, p<0.001). The survey indicated the risk of UGIB in patients taking antithrombotics was 1.01% in 11,919 patients, whichwas 5-10-fold higher than that in the general population. The increased risk in UGIB caused by antithormbotics was significantly suppressed by acidsecretion inhibitors (OR 0.146, 95%CI 0.089-0.242, p<0.001). Conclusions The present large number study of 11,919 Japanese patients indicated the prevalence of UGIB in patients, who were prescribed antithrombotics, was 1.01%, and the increased risk of UGIB was attenuated by acid-secretion inhibitors.