Introduction While managing patients with acute upper GI bleed (AUGIB) a simple numerical score can be helpful to identify high risk groups & need for intervention as recommended by NICE.Recent NCEPOD report for AUGIB reported that only 32% patients had a Pre-endoscopy risk assessment performed. Two commonly used scoring systems in UK are GBS (Glasgow Blatchford score) and RCS (Rockall score). Recent comparisons have shown that the GBS was superior in identifying need for hospital-based intervention¹ while RCS was better in predicting mortality.²

Methods We conducted a retrospective study of 893 patients, admitted to the hospital and treated in endoscopy unit of Cardiff & Vale health board between September 2010 to September 2013 with AUGIB. We calculated the GBS and pre & post endoscopy RCS for each and compared several outcomes.

Results Overall, GI bleed related mortality was only 3.1% in our study, with chronic liver disease being one of the main risk factor. GBS was superior in identifying patients suitable for safe discharge with outpatient management. The GBS was also better at predicting the need for endoscopic intervention. Our study also found that very early endoscopy (ie < 6 hours) compared to rapid endoscopy (6–24 hours), did not improve survival in the highest risk patients (ie GBS > 10), and in fact had a significantly worse mortality rate of 28% versus 7.4% .

• Download figure
• Open in new tab
• Download powerpoint

Abstract PTH-043 Figure 1

Conclusion The GBS score is superior to the pre-endoscopy Rockall in rationalising need & timeliness of intervention. The post-Rockall score is shown to be the better predictor of mortality. This study also reinforced the importance of access to rapid endoscopic intervention within 24 hours, but did not demonstrate the need for very early gastroscopy.

References 1 Camellini L, et al. Dig Liver Dis. 2004;36(4):271–7.

2 Rockall TA, et.al. Gut. 1996;38(3):316–21.

Disclosure of Interest None Declared