Glasgow Blatchford Score Calculator

The Glasgow-Blatchford score answers a narrow question about acute upper gastrointestinal bleeding: will this patient need an intervention, meaning transfusion, endoscopic hemostasis or surgery. It deliberately uses only what is available at first contact, so no endoscopy is required. Blood urea carries the most weight, up to 6 points, because blood digested in the small bowel is absorbed as a protein load and raises urea out of proportion to creatinine; urea of 6.5 to 8.0 mmol/L scores 2 while 25 mmol/L or more scores 6. Hemoglobin is scored by sex, with under 10 g/dL scoring 6 in both men and women and 12.0 to 12.9 g/dL scoring 1 in men only. Systolic pressure under 90 mmHg adds 3, pulse of 100 or more adds 1, melena 1, syncope 2, and known hepatic disease or cardiac failure 2 each, to a maximum of 23. The clinically decisive value is zero. Patients scoring 0 have under a 1 percent chance of needing intervention, and many health systems discharge them for outpatient endoscopy, which removes roughly a fifth of presentations from inpatient beds. Some protocols extend the low-risk threshold to a score of 1 or less. Above zero the score discriminates far less sharply, and it was never designed to predict death, rebleeding after successful therapy, or the source of the bleed. It does not account for anticoagulation or antiplatelet therapy, and the urea thresholds are expressed in mmol/L, so units must be converted where laboratories report blood urea nitrogen in mg/dL.

The score was derived by Oliver Blatchford, William Murray and Mary Blatchford and published in the Lancet in October 2000. Working from Scottish audit data covering consecutive admissions for upper gastrointestinal haemorrhage, they built a risk score from clinical and laboratory variables available before endoscopy and tested it in a separate validation cohort. Their stated aim was to fill a gap left by the Rockall score, published by Tim Rockall and colleagues in 1996, which needed endoscopic findings for its full form and was optimized to predict mortality rather than the need for treatment. Validation followed quickly and repeatedly. Adam Stanley and colleagues reported in 2009 that patients scoring zero could be managed as outpatients without adverse outcomes, and a large international comparison led by Stanley and published in the BMJ in 2017, covering thousands of patients across six countries, found the Glasgow-Blatchford score superior to the Rockall, AIMS65 and PNED scores for predicting intervention or death, with its clearest advantage at the low-risk end. Guidelines absorbed the evidence: the American College of Gastroenterology 2021 guideline on upper gastrointestinal bleeding, along with European and United Kingdom guidance, now recommend the score for triage at presentation and generally endorse outpatient management at a threshold of 0 or 1.