Modified Glasgow-Blatchford Bleeding Score as an Alternative Predictors of Severity for Non-Variceal Upper Gastrointestinal Bleeding

BACKGROUND: Upper gastrointestinal bleeding (UGIB) is a frequent cause of emergency hospital admissions. Despite the dependency of most risk scoring systems for this disorder, the Glasgow-Blatchford bleeding score (GBS) is based on simple variables. This research intended for investigate the accuracy of a modified GBS (mGBS) to predict the severity of non-variceal UGIB. METHODS: Study conducted in Emergency Department of Dr. Saiful Anwar Hospital, Malang, from November 2012 to April 2013. Endoscopy performed between 12-24 hours after the patient stabilized. Sixty patients diagnosed were included. The accuracy of the mGBS in predicting the severity of non-variceal UGIB was compared with the full GBS using receiver operating characteristic (ROC) curve. The severity based on high risk in mGBS score compared by Forrest classification. RESULTS: For prediction of the severity of non-variceal UGIB, the GBS (AUC 0.947, 95% CI 0.87-1.03) had a slightly than the mGBS (AUC 0.943, 95% CI 0.86-1.02, p<0.01). Compared to the GBS, the mGBS was more specific (63% and 97%, respectively) but less sensitive (96% and 84%, respectively). CONCLUSION: The mGBS is an alternative diagnostic tool in predicting the severity of non-variceal UGIB.