Sa1607 Localization of Angioectasias in the Small Bowel Using Capsule Endoscopy

Abstract

Predicting Acute Non-Variceal Upper Gastrointestinal Bleeding Which System Triages Best? Jack Kian Ch’NG, Yirong Sim*, Weng Hoong Chan, Yaw Chong Goh, Hock Soo Ong, Wai Keong Wong General Surgery, Singapore General Hospital, Singapore, Singapore Background: Acute non-variceal upper gastrointestinal bleeding (UGIB) remains a significant health problem and constitutes a significant proportion of emergency surgical admissions. The frequency and severity of this problem and its associated costs impose a significant burden on health care resources. Therefore, a triage system that can differentiate high and low risk patients for UGIB at initial presentation before offering emergency endoscopy is vital in guiding in hospital treatment. Several scoring systems have been devised to predict the severity of UGIB and the need for intervention, but have not been validated in our local (Singapore) setting. A highly predictive model for significant UGIB will guide the allocation of an appropriate level of care at admission, including emergent endoscopy. Aim: To validate and compare the Rockall and Blatchford score in predicting outcome after acute non-variceal upper gastrointestinal bleeding and the need for intervention. Methods: This is a single centre retrospective study conducted in Singapore General Hospital. All patients diagnosed with UGIB between January to December 2010 were included into the study. Data collected include demographics, mode of presentation, comorbidities, initial haemodynamic status, and initial haemoglobin level and blood biochemistry before transfusion. The 3 outcomes of interest were intervention (blood transfusion, endoscopy or surgery), re-bleeding and mortality. External validation of Clinical Rockall, Complete Rockall and Blatchford scoring schemes were evaluated by assessing discrimination and calibration. Area under the receiver operating characteristics curve (AUROC) was used as a measure of discrimination; calibration was evaluated using chi-squared Statistical software, SPSS version 17.0 was used for all analysis. Results: In the year 2010, 348 patients presented to our institute with UGIB and were analysed. There were almost twice as many males (65.2%) than females (34.8%) patients who presented with UGIB, with a mean age of 65 years. The most common mode of presentation is malaena (66.4%). Re-bleeding and mortality rate were 30.7% and 4.0% respectively. Clinical Rockall score, complete Rockall score and Blatchford score were significant associated with rebleed, mortality and intervention (p 0.05). Comparing AUROCs, the Blatchford scoring system was superior in identifying UGIB patients who require intervention, and are at risk of rebleeding. When a cut-off score value of 2 was used, the sensitivity and specificity of the Blatchford scoring system were determined to be 100% and 18.9% respectively. Conclusion: In our local population, the Blatchford score is superior in predicting low-risk patients who do not require urgent therapeutic endoscopy and who may be suitable for outpatient management.