- Score the patient on arrival, before endoscopy, using admission vital signs and the first chemistry and complete blood count.
- Enter blood urea either as urea (mmol/L) or as blood urea nitrogen, BUN (mg/dL); the calculator converts BUN → urea using the factor 0.357.
- Choose sex so the correct hemoglobin thresholds apply (men use the 12–12.9 g/dL band; women do not).
- Check off melena, syncope, known hepatic disease, and cardiac failure as applicable.
- Read the total score and its risk band; use the Next Steps panel to decide on admission, early endoscopy, or safe outpatient management.
All computation runs in your browser; no values are stored or transmitted.
When to Use
Use the Glasgow–Blatchford Score (GBS) in an adult presenting with a suspected acute upper gastrointestinal (GI) bleed — hematemesis, melena, coffee-ground emesis, or unexplained drop in hemoglobin with a suggestive history — to risk-stratify the patient before endoscopy. The score was derived specifically to identify patients at very low risk, who may safely skip admission and undergo outpatient endoscopy, and to flag high-risk patients who are likely to need transfusion, endoscopic therapy, or surgery.
Appropriate population
Adults in the emergency department or admitting unit with symptoms or signs of acute upper GI bleed. GBS is most useful at first contact, when the clinical and laboratory picture is complete but endoscopy has not yet been performed; it supports the triage decision (discharge, admit, resuscitate). International validation studies show a score of 0 has a very high negative predictive value for requiring intervention or dying from the bleed.
When NOT to rely on it
GBS is a risk score, not a diagnosis: it does not tell you the source of the bleed, and it can underestimate risk in patients whose vitals are being maintained by vasopressors, beta-blockers, or recent fluid resuscitation. It is not validated in lower GI bleeding, in pediatric patients, or in a patient who has already received a transfusion before scoring. It assumes a first-presentation bleed — in a rebleed after inpatient endoscopy, use the full post-endoscopy Rockall instead.
Pearls & Pitfalls
A score of 0 is the useful decision point
GBS is powerful precisely because it is good at the low end. A score of 0 means none of: urea ≥ 6.5 mmol/L, hemoglobin below the sex-adjusted cut-off, SBP < 110, pulse ≥ 100, melena, syncope, hepatic disease, or heart failure. In external validations, these patients almost never need intervention and can usually go home for outpatient endoscopy — the most actionable result the score provides.
Urea conversion
GBS uses blood urea in mmol/L. Philippine chemistry panels usually report BUN in mg/dL; the calculator converts automatically using urea (mmol/L) = BUN (mg/dL) × 0.357. A BUN of 20 mg/dL ≈ 7.1 mmol/L; 28 mg/dL ≈ 10 mmol/L; 70 mg/dL ≈ 25 mmol/L.
Pitfalls
(1) Masked shock: a patient on beta-blockers or vasopressors may have a falsely reassuring heart rate and SBP; interpret the score in context. (2) Delayed hemoglobin drop: a very acute bleed can present with a near-normal hemoglobin before equilibration — do not use a one-off low score to discharge a patient who continues to bleed. (3) Pre-existing anemia (chronic kidney disease, hematologic disease) raises the hemoglobin points without an acute bleed; treat the chronic value as the baseline. (4) A high score tells you the patient is sick, not why — proceed with resuscitation, PPI and endoscopy; the score does not substitute for clinical judgment.
Why Use It
Acute upper gastrointestinal (GI) bleed is common, costly to admit, and rarely trivial to triage at first contact. The Glasgow–Blatchford Score was derived from a large unselected emergency department cohort to predict the need for an intervention or death — transfusion, endoscopic therapy, surgery, or in-hospital mortality — using only data available at the door. A score of 0 reliably identifies a small but important group who can be discharged for outpatient endoscopy, sparing a hospital bed and an unnecessary admission. At the other end, a score of ≥ 7 is the trigger for early (within 24 h) endoscopy with a prepared therapeutic team. Multiple international validations have confirmed that GBS discriminates risk better than the pre-endoscopy (clinical) Rockall score, which is why most modern UGIB guidelines endorse it as the preferred first-contact tool.
Glasgow–Blatchford Score (GBS) — Upper GI Bleed Risk
Enter the admission vitals and labs. The score, risk band, and a bedside interpretation update automatically. Default urea unit is mmol/L; switch to BUN mg/dL if your lab reports that.
⚕ Blatchford O et al. Lancet. 2000;356:1318–1321. The Glasgow–Blatchford Score predicts the need for intervention or death at first contact in suspected upper gastrointestinal (GI) bleeding; it does not identify the source of bleeding. Interpret with full clinical judgment, repeat vitals and hemoglobin, and local resuscitation and endoscopy pathways. For licensed clinicians; not a substitute for individualized assessment.
Next Steps
Match the risk band to the admission pathway, and never let a reassuring score override a worsening clinical picture.
- GBS = 0 — very low risk. Negative predictive value for intervention or mortality is about 99% in large validations. If vitals and hemoglobin are stable, the patient is otherwise well, and a reliable outpatient endoscopy can be arranged, discharge with safety-netting is reasonable.
- GBS 1–3 — low risk. Most do not need endoscopic intervention, but admission for observation is still the usual safe default in the Philippine setting where rapid outpatient endoscopy is not always available. Start IV access, type and screen, and a proton pump inhibitor; recheck hemoglobin.
- GBS 4–6 — moderate risk. Admit, resuscitate, transfuse to maintain hemoglobin ≥ 7 g/dL (≥ 8 g/dL in cardiovascular disease), start intravenous PPI, hold anticoagulants when possible, and plan endoscopy within 24 h.
- GBS ≥ 7 — high risk. Likely to need endoscopic therapy or surgery. Resuscitate aggressively, escalate care, notify endoscopy early, and aim for therapeutic endoscopy within 24 h (earlier in ongoing bleeding or hemodynamic instability).
- Pair with the Rockall Score after endoscopy for rebleed and mortality prediction, and with the Shock Index and Transfusion threshold at the bedside.
Evidence & References
Scoring table
| Variable | Threshold | Points |
|---|---|---|
| Blood urea (mmol/L) | 6.5 to < 8.0 | +2 |
| 8.0 to < 10.0 | +3 | |
| 10.0 to < 25.0 | +4 | |
| ≥ 25.0 | +6 | |
| Hemoglobin — men (g/dL) | 12.0 to < 13.0 | +1 |
| 10.0 to < 12.0 | +3 | |
| < 10.0 | +6 | |
| Hemoglobin — women (g/dL) | 10.0 to < 12.0 | +1 |
| < 10.0 | +6 | |
| Systolic BP (mm Hg) | 100–109 | +1 |
| 90–99 | +2 | |
| < 90 | +3 | |
| Pulse ≥ 100/min | Yes | +1 |
| Melena at presentation | Yes | +1 |
| Syncope at presentation | Yes | +2 |
| Hepatic disease | Yes | +2 |
| Cardiac failure | Yes | +2 |
Risk bands
| GBS | Risk | Typical disposition |
|---|---|---|
| 0 | Very low | Outpatient endoscopy may be considered (NPV ≈ 99%) |
| 1–3 | Low | Admit for observation; early endoscopy is rarely required |
| 4–6 | Moderate | Admit, resuscitate, IV PPI, endoscopy within 24 h |
| ≥ 7 | High | Likely need for endoscopic therapy; early endoscopy & escalation |
Unit conversion: urea (mmol/L) = BUN (mg/dL) × 0.357. Derived and validated in unselected emergency-department cohorts; external validations (Stanley AJ, BMJ 2017) show GBS discriminates risk of intervention and death better than the pre-endoscopy Rockall score.
References
- Blatchford, O., Murray, W. R., & Blatchford, M. (2000). A risk score to predict need for treatment for upper-gastrointestinal haemorrhage. The Lancet, 356(9238), 1318–1321. https://doi.org/10.1016/S0140-6736(00)02816-6
- Stanley, A. J., Laine, L., Dalton, H. R., Ngu, J. H., Schultz, M., Abazi, R., Zakko, L., Thornton, S., Wilkinson, K., Khor, C. J. L., Murray, I. A., & Laursen, S. B. (2017). Comparison of risk scoring systems for patients presenting with upper gastrointestinal bleeding: International multicentre prospective study. BMJ, 356, i6432. https://doi.org/10.1136/bmj.i6432
