Upper gastrointestinal non-variceal bleeding : validation of Glasgow Blatchford score and its association with Forrest classification in Hospital Universiti Sains Malaysia
Loading...
Date
2018
Authors
Yusof, Mohd Razaleigh
Journal Title
Journal ISSN
Volume Title
Publisher
Pusat Pengajian Sains Perubatan, Universiti Sains Malaysia
Abstract
Introduction
Upper Gastrointestinal Bleeding (UGIB) requiring endoscopic intervention is a common scenario in hospitals. However, not all patients require emergency or urgent endoscopy to be done. This study will validate and associate Glasgow-Blatchford Score (GBS) with Forrest classification. This is to predict the severity of Non-Variceal Upper Gastro Intestinal Bleed according to Glasgow Blatchford Score and validate this scoring system in Hospital Universiti Sains Malaysia. This study will also associate between two scoring system of Glasgow Blatchford Score and Forrest Classification between Non-variceal high risk bleeding with major bleed and non-variceal low risk bleeding with minor bleed.
Methodology
Data collected from June 2016 till February 2017, 113 patients with Non-variceal bleed underwent emergency OGDS were retrospectively reviewed and were stratified according to Glasgow Blatchford Score to high and low risk. This stratified risk is associated with Forrest classification to determine its endoscopic findings.
Results
Majority are in the high risk group 107 (94.7%) patients and low risk group 6 (5.3%) patients. Patients in the low risk group was followed up for 30 days and showed no complications or mortality. This study showed patients has a median score of 10.27±3.54. From the data analysis of Glasgow Blatchford Score showed 95.45% sensitivity and 5.49% specificity. Endoscopic findings showed 22 (19.5%) patients had major bleed and 91 (80.5%) patients had minor bleed. Using Fischer Exact Test, there is no significant association between risk and outcome (p-value > 0.950).
Conclusion
Based on this study alone, we cannot suggest Glasgow Blatchford Score as a predictor for severity of Upper Gastrointestinal Bleeding. This is because as shown in the results, there were no significant association between GBS and Forrest classification. However, this nonsignificant result maybe compounded by several factors as discussed, namely due to underlying Chronic Kidney Disease with anemia, the timing of endoscopy and initiation of medication. With this biases identified, it can be used as a guide in designing and conducting a better study in the future in order to come to a better conclusion about GBS in our population.
Description
Keywords
Gastrointestinal hemorrhage