Article Text


PTH-140 Risk Stratification of upper Gastrointestinal Bleeding with the Glasglow-Blatchford Score: Experience of a District General Hospital in East London, United Kingdom
  1. F W D Tai1,
  2. E Selvaraj1,
  3. Y H Siaw1,
  4. E Seward1
  1. 1Gastroenterology, Whipps Cross Hospital, Barts Health NHS Trust, London, UK


Introduction The Glasglow-Blatchford Score (GBS) is a risk stratification tool to assess the need for clinical intervention to prevent death in patients with suspected upper gastrointestinal (UGI) bleeding1. In the UK it has been validated in four centres, with 99.6% sensitivity for not requiring endoscopic or transfusion therapy in patients with a GBS of zero1,3. Latest NICE guidelines recommend the GBS for risk stratification in UGI bleeds2. We aimed to establish whether the East London population behaved in a similar fashion to published data.

Methods The GBS was calculated retrospectively from audit data collected from patients with suspected UGI bleeding seen by the emergency department (ED) at Whipps Cross University Hospital, London between November and December 2011. During this period, clinical notes for patients with emergency department attendances coded as haematemesis, coffee-ground vomiting and/or melaena were analysed. Patients who were subsequently found to have a different diagnosis were excluded from further analysis. In addition to basic demographic and admission data, we assessed how frequently the GBS was applied. GBS scores were then correlated with endoscopy findings, and the need for endoscopic therapy at the time of endoscopy, and the need for transfusion.

Results 97 sets of notes were identified and 42 patients included in the audit, age range 26–96 (median 66 years). 61% of patients were male and 57% of admissions occurred between the hours of 0900 to 1700. In 77% of patients a GBS was not considered by emergency and/or acute medical physicians. All patients with a GBS of 0 were admitted and subsequently discharged with outpatient endoscopy.

Abstract PTH-140 Table 1

Of patients scoring 1 to 5, 11% had UGI pathology, 9% (2 patients both GBS of 5) required transfusion only. In patients scoring 6 to 13, 47% of them had UGI pathology, 33% and 87% required endoscopic therapy and blood transfusions respectively.

Conclusion UGI bleeds were most commonly found in males over the age of 65. Locally, the GBS is an underused risk stratification tool in determining the need for admission. Our preliminary data suggests patients with GBS of 0 can be discharged with outpatient endoscopy, and patients with a GBS more than 6 represent a high risk population requring emergent endoscopy. We propose that patients with a GBS of 1 or 2 can also be managed as an outpatient as our data suggest that patients in this group do not require admission. Local data suggests this can reduce patient admission rates by up to 17%.

Disclosure of Interest None Declared.


  1. Blatchford O et al. Lancet 2000; 256(9238):1318–21

  2. Dworzynski K et al. BMJ 2012; p.e3412

  3. Stanley AJ et al. Lancet 2009; 373(9657):42–7

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.