Glasgow-Blatchford Bleeding Score
A pre-endoscopy risk stratification tool for upper gastrointestinal bleeding that predicts the need for clinical intervention — including transfusion, endoscopic therapy, or surgery — and identifies very-low-risk patients who may be safely managed as outpatients.
Calculate Glasgow-Blatchford Score
Enter the patient’s presenting parameters. The score uses sex-specific haemoglobin thresholds, so patient sex is required. All values should reflect the initial presentation — before resuscitation or transfusion. A score of 0 identifies patients who may be safely considered for outpatient management.
The Glasgow-Blatchford Score is a pre-endoscopy risk stratification tool — it is applied before any endoscopic findings are available. Its primary clinical value is identifying GBS = 0 patients who may not require urgent inpatient endoscopy, and flagging high-score patients who are likely to need intervention. It does not predict the endoscopic diagnosis or rebleeding risk.
Understanding the Glasgow-Blatchford Score
The Glasgow-Blatchford Bleeding Score (GBS) was developed by Blatchford, Murray, and Blatchford in 2000 at the Glasgow Royal Infirmary, derived from a prospective study of 1,748 consecutive patients presenting with upper GI haemorrhage across multiple Scottish hospitals. The primary outcome was the need for clinical intervention — defined as any of: blood transfusion, endoscopic therapy, surgical intervention, or death.
The score ranges from 0 to 23 and uses only data available at the point of first clinical contact — laboratory values (urea, haemoglobin), vital signs (systolic blood pressure, pulse), clinical features (melaena, syncope), and two comorbidities (hepatic disease, cardiac failure). No endoscopic data is required, making it an ideal triage tool for the emergency department.
Scoring Components
| Parameter | Range | Points |
|---|---|---|
| Urea (mmol/L) | 6.5–8.0 | 2 |
| 8.0–10.0 | 3 | |
| 10.0–25.0 | 4 | |
| ≥ 25.0 | 6 | |
| Hb — Male (g/dL) | 12.0–13.0 | 1 |
| 10.0–12.0 | 3 | |
| < 10.0 | 6 | |
| Hb — Female (g/dL) | 10.0–12.0 | 1 |
| < 10.0 | 6 | |
| SBP (mmHg) | 100–109 | 1 |
| 90–99 | 2 | |
| < 90 | 3 | |
| Pulse ≥ 100 | Yes | 1 |
| Melaena | Yes | 1 |
| Syncope | Yes | 1 |
| Hepatic disease | Yes | 2 |
| Cardiac failure | Yes | 2 |
Worked Example
A 58-year-old male presents with haematemesis and melaena. SBP 108 mmHg, pulse 104 bpm, Hb 10.8 g/dL, urea 9.2 mmol/L. No syncope, no liver or cardiac disease.
Urea 9.2 (8.0–10.0) → +3
Hb 10.8 (male) (10.0–12.0) → +3
SBP 108 (100–109) → +1
Pulse 104 (≥100) → +1
Melaena → +1
Syncope → 0 | Liver → 0 | Cardiac → 0
GBS = 9 → High risk. This patient requires admission, urgent inpatient endoscopy, and probable transfusion and/or endoscopic intervention.
The GBS = 0 rule: The single most important clinical application of the Glasgow-Blatchford Score is that a score of exactly 0 identifies patients at very low risk of needing intervention (< 1%). Multiple validation studies have confirmed that GBS = 0 patients can be safely discharged for outpatient endoscopy within 24–72 hours, with appropriate safety-net advice. This can reduce unnecessary hospital admissions by up to 16–25% of all UGIB presentations.
Interpretation & Risk Stratification
| GBS | Risk Category | Need for Intervention | Recommended Management |
|---|---|---|---|
| 0 | Very low risk | < 1% | Consider discharge for outpatient endoscopy within 24–72 h with safety-net advice |
| 1–4 | Low risk | ~5–15% | Admit; inpatient endoscopy within 24 h; most will not require endoscopic therapy |
| 5–8 | Moderate risk | ~25–50% | Admit; urgent endoscopy; anticipate need for transfusion or endoscopic haemostasis |
| 9–12 | High risk | ~50–75% | Urgent admission; early resuscitation; expedited endoscopy; consider ICU-level monitoring |
| ≥ 13 | Very high risk | > 75% | Critical care setting; emergent resuscitation and endoscopy; prepare for potential surgery/IR |
Some studies and guidelines have explored using a GBS ≤ 1 threshold (rather than strictly GBS = 0) for identifying low-risk outpatient candidates. A 2012 international multicentre validation found that expanding to GBS ≤ 1 identified approximately 25% of patients as low-risk versus 16% with GBS = 0, while maintaining a very high negative predictive value. However, the GBS = 0 threshold remains the most widely endorsed and safest cut-off in current international guidelines (NICE 2012, ESGE 2015, ACG 2021). Clinicians should follow their local institutional protocol.
Clinical Context & Management Pathway
The GBS does not predict the endoscopic diagnosis, but understanding the common aetiologies of UGIB helps guide management. The most frequent causes include peptic ulcer disease (35–50% of cases, including gastric and duodenal ulcers), erosive gastritis/duodenitis (8–15%), oesophageal and gastric varices (5–15%, depending on the population), Mallory-Weiss tears (5–10%), oesophagitis (5–15%), and vascular lesions including Dieulafoy lesions and angiodysplasia (2–5%).
Malignancy (gastric or oesophageal carcinoma) accounts for 2–5% of acute UGIB presentations and should be considered particularly in older patients, those with weight loss, or those with dysphagia. Cameron lesions (ulcers within a hiatal hernia) and post-procedural bleeding (e.g., post-polypectomy, post-sphincterotomy) are important additional causes in specific clinical contexts.
Regardless of the GBS result, all patients with acute UGIB should undergo a structured initial assessment: Airway-Breathing-Circulation approach, two large-bore IV cannulae, blood sampling (FBC, U&E, LFTs, coagulation, group-and-save/crossmatch), and haemodynamic assessment. Patients with haemodynamic instability (SBP < 90 or tachycardia > 100 after initial fluid) should be considered for critical care input.
Transfusion strategy: current guidelines recommend a restrictive transfusion threshold (Hb < 7 g/dL) for most patients, as the TRIGGER and Villanueva trials demonstrated improved survival with restrictive versus liberal strategies. The exception is patients with acute coronary syndromes, where a threshold of 8 g/dL is typically used. Avoid over-transfusion in variceal bleeding, as it may worsen portal hypertension.
Endoscopy timing is stratified by clinical urgency. Very urgent (< 12 hours): suspected variceal haemorrhage or haemodynamically unstable despite resuscitation. Urgent (within 24 hours): all admitted UGIB patients who are haemodynamically stable — this is the standard recommendation from NICE, ESGE, and ACG guidelines. Outpatient (24–72 hours): GBS = 0 patients who have been discharged.
Evidence suggests that endoscopy within 24 hours for non-variceal UGIB improves outcomes primarily through earlier haemostatic intervention and reduced length of stay, rather than reduced mortality per se. For variceal bleeding, earlier endoscopy (within 12 hours) is associated with improved haemostasis and survival.
Proton pump inhibitors (PPIs): Pre-endoscopy IV PPI (e.g., pantoprazole or omeprazole) is commonly administered, though evidence for its benefit before endoscopic diagnosis is limited. Post-endoscopy, high-dose IV PPI (80 mg bolus followed by 8 mg/h infusion for 72 hours) is recommended for high-risk peptic ulcer stigmata (active bleeding, visible vessel, adherent clot).
Variceal bleeding: Terlipressin (or octreotide where terlipressin is unavailable) should be started as soon as variceal haemorrhage is suspected — before endoscopy. Prophylactic antibiotics (ceftriaxone or norfloxacin) are recommended in all cirrhotic patients with GI bleeding. Band ligation is the endoscopic treatment of choice for oesophageal varices.
Prokinetics: IV erythromycin (250 mg, 30–120 minutes pre-endoscopy) improves gastric visualisation by promoting gastric emptying and is recommended in selected patients with significant blood or clots in the stomach. Tranexamic acid is no longer recommended routinely for UGIB following the HALT-IT trial, which showed no mortality benefit and an increased risk of venous thromboembolism.
A GBS of 0 means all of the following are true: urea < 6.5 mmol/L, Hb ≥ 13 g/dL (male) or ≥ 12 g/dL (female), SBP ≥ 110 mmHg, pulse < 100, no melaena, no syncope, no hepatic disease, and no cardiac failure. These patients can be considered for early discharge with outpatient endoscopy typically within 24–72 hours.
Discharge criteria should also include: haemodynamic stability for ≥ 2 hours of observation, ability to tolerate oral fluids, reliable social support and transport, capacity to understand and follow safety-net advice, ability to return to hospital if symptoms recur, and arranged outpatient endoscopy appointment. Patients with significant comorbidities, anticoagulant use, or social concerns may require admission despite GBS = 0.
Special Populations
Anticoagulant and antiplatelet use is not a variable in the GBS, yet these medications significantly increase the risk of persistent and recurrent bleeding. A patient with a GBS of 0 who is on warfarin (INR therapeutic or supratherapeutic) or a DOAC may still be at significant risk and should not be automatically discharged. Consider anticoagulation status as a separate clinical factor in the disposition decision. Reversal agents may be needed for unstable patients.
The GBS was derived from a general UGIB population that includes variceal bleeding. However, variceal haemorrhage has a distinct pathophysiology, management pathway (terlipressin, antibiotics, band ligation, TIPS), and mortality profile. The GBS predicts the need for intervention but does not distinguish between variceal and non-variceal causes. A high GBS in a patient with known cirrhosis should trigger immediate variceal bleeding protocols, including vasoactive drugs before endoscopy.
Age is notably absent from the GBS (unlike the Rockall score). While this makes the GBS purely physiology-based, older patients with UGIB have higher comorbidity burden, polypharmacy, and reduced physiological reserve, meaning even low GBS scores may underestimate risk in the very elderly. Clinicians should weigh age, frailty, and comorbidities alongside the GBS when making disposition decisions, particularly for patients near the low-risk threshold.
Patients with pre-existing chronic anaemia (iron deficiency, chronic disease, thalassaemia trait) may present with a low haemoglobin that generates a high GBS even with minor acute blood loss. The score cannot distinguish acute from chronic anaemia. In patients with a known low baseline Hb, the GBS will overestimate the severity of the acute bleed. Consider the delta from baseline haemoglobin and the clinical presentation (volume status, haemodynamic parameters) when interpreting the result.
GBS vs Rockall & Other UGIB Scores
| Feature | Glasgow-Blatchford (GBS) | Pre-Endoscopy Rockall | Full Rockall | AIMS65 |
|---|---|---|---|---|
| Timing | Pre-endoscopy | Pre-endoscopy | Post-endoscopy | Pre-endoscopy |
| Predicts | Need for intervention | Mortality | Rebleeding & mortality | Mortality |
| Score range | 0–23 | 0–7 | 0–11 | 0–5 |
| Includes age | No | Yes | Yes | Yes (≥65) |
| Includes endoscopy | No | No | Yes | No |
| Identifies safe discharge | Yes (GBS = 0) | Moderate | N/A (post-endoscopy) | Moderate |
| Key strength | Best for identifying low-risk patients; widely validated | Simple; includes age | Includes endoscopic stigmata | Simple; 5 items; predicts mortality |
| Key limitation | Does not predict mortality; no age or endoscopy data | Less validated for safe discharge | Requires endoscopy results | Less validated for intervention need |
Use the GBS at presentation to identify GBS = 0 patients for possible outpatient management and to risk-stratify the remainder for urgency of endoscopy. After endoscopy, use the full Rockall score to assess rebleeding risk and guide post-endoscopic management (discharge timing, PPI duration, follow-up planning). The two scores are complementary, not competing — GBS for admission decisions, Rockall for post-endoscopy decisions.
Common Pitfalls & Limitations
The GBS was designed to use the patient’s presenting parameters — before IV fluid resuscitation or blood transfusion. Using values obtained after resuscitation (e.g., a systolic BP that has improved from 88 to 115 mmHg after 2 litres of saline) will produce a falsely reassuring lower score. If the initial presenting vitals are not available, use the worst recorded values rather than post-resuscitation values. This is particularly important for the systolic BP and heart rate components.
While a GBS of 0 identifies a very-low-risk group, it does not guarantee zero risk. Approximately 0.5–1% of GBS = 0 patients will still require intervention. Discharging these patients requires a structured approach: confirm haemodynamic stability for ≥ 2 hours, arrange outpatient endoscopy within 24–72 hours, provide written safety-net instructions (return immediately for further haematemesis, significant melaena, dizziness, or collapse), ensure the patient has reliable transport and social support, and document the discharge plan clearly.
Patients should not be discharged to situations where they cannot access emergency care rapidly if symptoms recur.
Neither anticoagulant nor antiplatelet use is a variable in the GBS. A patient on triple antithrombotic therapy (aspirin + clopidogrel + apixaban) with a GBS of 0 has a fundamentally different bleeding risk profile from an otherwise identical patient on no medications. The GBS should be considered alongside anticoagulation status, INR, and the urgency of antithrombotic therapy for the underlying indication (e.g., mechanical heart valve, recent coronary stent, atrial fibrillation). The decision to reverse, hold, or continue anticoagulation requires separate clinical judgement.
The GBS predicts the need for intervention (transfusion, endoscopic therapy, or surgery) — it was not designed or validated to predict mortality or rebleeding. For mortality prediction, the Rockall score, AIMS65, or the full clinical assessment are more appropriate. This distinction matters: a patient with a high GBS may require transfusion but have excellent long-term survival, while a patient with a moderate GBS and severe comorbidities may have a higher mortality risk despite a seemingly lower score.
The original GBS was developed using blood urea in mmol/L (standard in the UK). In North America and some other regions, BUN (blood urea nitrogen) in mg/dL is the standard laboratory measurement. The conversion factor is: Urea (mmol/L) = BUN (mg/dL) × 0.357, or equivalently, BUN (mg/dL) = Urea (mmol/L) × 2.8. The GBS thresholds are: urea 6.5, 8.0, 10.0, and 25.0 mmol/L, which correspond to BUN approximately 18.2, 22.4, 28.0, and 70.0 mg/dL. Using the wrong unit without conversion will produce an incorrect score. This calculator handles the conversion automatically.
Quick Reference Summary
| GBS | Risk | Action |
|---|---|---|
| 0 | Very low | Consider outpatient endoscopy with safety-net |
| 1–4 | Low | Admit; endoscopy within 24 h |
| 5–8 | Moderate | Admit; urgent endoscopy; anticipate intervention |
| 9–12 | High | Urgent admission; expedited endoscopy; ICU if unstable |
| ≥ 13 | Very high | Critical care; emergent resuscitation and endoscopy |
The Golden Rule: GBS = 0 is the safest, most validated threshold for identifying UGIB patients who can be managed as outpatients. Every point above zero progressively increases the probability that the patient will need transfusion, endoscopic intervention, or surgery. Use the score for admission and triage decisions, not for predicting mortality — that is the Rockall score’s role after endoscopy.
Disclaimer & References
For Educational Purposes Only. This calculator and the accompanying clinical information are intended as educational tools for healthcare professionals. They do not replace clinical judgement. Results should be interpreted in the full clinical context. Lab reference ranges vary by institution — verify with your own laboratory. Drug dosages should be confirmed against current prescribing information.
References
- Blatchford O, Murray WR, Blatchford M. A risk score to predict need for treatment for upper-gastrointestinal haemorrhage. Lancet. 2000;356(9238):1318–1321. DOI: 10.1016/S0140-6736(00)02816-6
- Stanley AJ, Ashley D, Dalton HR, et al. Outpatient management of patients with low-risk upper-gastrointestinal haemorrhage: multicentre validation and prospective evaluation. Lancet. 2009;373(9657):42–47. DOI: 10.1016/S0140-6736(08)61769-9
- Stanley AJ, Laine L, Dalton HR, et al. Comparison of risk scoring systems for patients presenting with upper gastrointestinal bleeding: international multicentre prospective study. BMJ. 2017;356:i6432. DOI: 10.1136/bmj.i6432
- Laine L, Barkun AN, Saltzman JR, Martel M, Leontiadis GI. ACG clinical guideline: upper gastrointestinal and ulcer bleeding. Am J Gastroenterol. 2021;116(5):899–917. DOI: 10.14309/ajg.0000000000001245
- Gralnek IM, Dumonceau JM, Kuipers EJ, et al. Diagnosis and management of nonvariceal upper gastrointestinal hemorrhage: European Society of Gastrointestinal Endoscopy (ESGE) Guideline. Endoscopy. 2015;47(10):a1–a46. DOI: 10.1055/s-0034-1393172
- National Institute for Health and Care Excellence (NICE). Acute upper gastrointestinal bleeding in over 16s: management. Clinical guideline [CG141]. 2012 (updated 2016). Available at: nice.org.uk/guidance/cg141
- Rockall TA, Logan RF, Devlin HB, Northfield TC. Risk assessment after acute upper gastrointestinal haemorrhage. Gut. 1996;38(3):316–321. DOI: 10.1136/gut.38.3.316
- Villanueva C, Colomo A, Bosch A, et al. Transfusion strategies for acute upper gastrointestinal bleeding. N Engl J Med. 2013;368(1):11–21. DOI: 10.1056/NEJMoa1211801
- Roberts I, Shakur-Still H, Afolabi A, et al; HALT-IT Trial Collaborators. Effects of a high-dose 24-h infusion of tranexamic acid on death and thromboembolic events in patients with acute gastrointestinal bleeding (HALT-IT): an international randomised, double-blind, placebo-controlled trial. Lancet. 2020;395(10241):1927–1936. DOI: 10.1016/S0140-6736(20)30848-5