Biotika (Feb 2019)
Scoring system comparison to determine severity in non-variceal upper gastrointestinal bleeding
Abstract
Determining scoring values in cases of UGIB is important to do so it can be used as a basis to determine the mortality rate of UGIB patients. Despite the dependency of most risk scoring systems for this disorder on endoscopic findings, the Glasgow-Blatchford Bleeding Score (GBS) and Incomplete Rockall Score (IRS) are based on simple variables. Our purpose to differentiate about sensitivity and spesificity between IRS and GBS to predict the severity of patient with non variceal upper GI bleeding without the endoscopic score. Our study was undertaken in emergency department Saiful Anwar hospital, Malang. The sample consist of 102 patients with non variceal upper GI bleeding from November 2012 to April 2013. With receiving-operator characteristics (ROC), we compare the significance between IRS and GBS in predicting the severity of patient with non variceal upper GI bleeding. In next step, we compare both of them with the diagnostic test. ROC curve shows that GBS is slightly better than IRS (0,955 vs 0,744, p<0,05). IRS has better specificity (79,55% vs 70,45%) than GBS. But, GBS has better sensitivity (96,55% vs 65,52%). The final results is GBS as an alternative diagnostic score has better results than IRS in understanding the severity of patient with non variceal upper GI bleeding in Saiful Anwar hospital. Therefore, it can be used as a simple predictor replacing the endoscopic score.
