# Upper Gastrointestinal Bleeding Risk Stratification

Use Glasgow-Blatchford scoring at presentation to identify patients suitable for outpatient care, while recognizing that hemodynamic instability, ongoing hemorrhage, comorbidity, and suspected varices override low-complexity disposition pathways and require resuscitation and timely endoscopy.

**Clinical question:** How should physicians stratify acute upper gastrointestinal bleeding to determine disposition, transfusion, endoscopy timing, and escalation?

Updated: 2026-09-15T22:10:03.534021+00:00

## What matters in practice
- Calculate a Glasgow-Blatchford score (GBS) at initial presentation; GBS 0-1 identifies very-low-risk patients who may be discharged with outpatient follow-up if clinical assessment does not identify another reason for admission. [17][24]
- Do not use a low-risk score to defer resuscitation or admission in hemodynamically unstable patients; stabilize circulation and optimize important comorbidities before endoscopy. [1][2]
- For hospitalized UGIB, use a restrictive red-cell transfusion threshold of hemoglobin 7 g/dL; consider a higher threshold in patients with cardiovascular disease. [11][17][24]
- Perform endoscopy within 24 hours after presentation in admitted patients, after resuscitation; evidence does not establish routine benefit for endoscopy within 6 hours solely because GBS is at least 12. [1][9][17]
- GBS is most useful for predicting need for hospital-based intervention; AIMS65 is more oriented to mortality prediction and should not replace GBS for low-risk discharge selection. [5][16][23]

## Determine whether the patient needs immediate resuscitation or can enter a low-risk pathway

Disposition begins with physiology, not with an endoscopic diagnosis.

Treat hemodynamic instability as a resuscitation problem before applying a discharge pathway. Start resuscitation in acute UGIB with instability, reassess hemodynamics during intravenous fluid administration, and optimize consequential comorbidities before endoscopy. [1][2]

After initial stabilization, calculate the GBS using admission blood urea nitrogen, hemoglobin, systolic blood pressure, heart rate, melena, syncope, hepatic disease, and cardiac failure. The score predicts the likelihood of hospital-based intervention, including transfusion, endoscopic treatment, and surgery; it is not a definitive diagnosis and does not replace assessment for persistent bleeding, shock, or alternative causes of hematemesis or melena. [5][23]

Use GBS 0-1 to identify a very-low-risk group potentially appropriate for emergency-department discharge with outpatient follow-up. This threshold is intended to identify patients at very low risk of rebleeding or death and is supported by major guideline recommendations; outpatient management still requires reliable follow-up and no competing clinical indication for hospitalization. [11][17][22][24]
- Admit rather than discharge when instability persists, bleeding is clinically ongoing, or comorbidity requires inpatient monitoring even if a calculated score appears low. [1][2]
- Do not interpret GBS as a mortality score alone: it is principally validated for need for intervention, whereas low risk of recurrent bleeding is not synonymous with low mortality risk. [5]
- Use a documented GBS rather than an informal impression when deciding whether ED discharge is reasonable. [5][17]

*Risk-stratification tools answer different questions in acute UGIB. [5][16][23]*

| Tool | Most actionable use | Decision threshold or interpretation | Important limitation |
| --- | --- | --- | --- |
| Glasgow-Blatchford score | Identify patients unlikely to need hospital-based intervention. [5][16] | GBS 0-1 identifies very-low-risk patients who may be discharged with outpatient follow-up when clinically appropriate. [17][24] | Does not independently determine mortality risk, source of bleeding, or urgency beyond clinical hemodynamic assessment. [5][23] |
| AIMS65 | Estimate mortality risk. [16][23] | Includes albumin below 3.0 g/dL, INR above 1.5, altered mental status, systolic blood pressure 90 mm Hg or lower, and age 65 years or older. [23] | Less appropriate than GBS for selecting patients for discharge based on anticipated intervention need. [16][23] |
| Rockall score | Risk assessment after incorporating clinical and, for the full score, endoscopic variables. [13][23] | Pre-endoscopic Rockall uses age, hemodynamic status, and comorbidity. [23] | Do not substitute it for GBS when the immediate question is whether hospital-based intervention is likely. [5][16] |

## Apply the Glasgow-Blatchford score to the disposition decision

GBS is a rule-out tool for very-low-risk patients, not a mandate for early intervention.

Calculate GBS before endoscopy for every patient with suspected UGIB once the necessary clinical data and initial laboratory values are available. GBS incorporates objective circulatory and laboratory abnormalities plus melena, syncope, hepatic disease, and heart failure, allowing risk assessment before the bleeding lesion is known. [5][23]

A GBS above 1 does not specify a transfusion trigger, an ICU threshold, or a need for endoscopy in less than 24 hours. It indicates that the patient does not meet the guideline-supported very-low-risk discharge threshold and should proceed through inpatient assessment according to hemodynamics, bleeding trajectory, comorbidity, and endoscopic availability. [1][11][17]

Do not use high GBS values alone to justify routine urgent endoscopy in under 6 hours. In patients with GBS of at least 12, comparison of urgent endoscopy within 6 hours with early endoscopy at 6-24 hours did not establish benefit from the earlier timing; resuscitation and clinical optimization remain the priority before the procedure. [1]
- GBS 0-1: consider discharge with outpatient follow-up only after confirming clinical stability and absence of another admission requirement. [17][24]
- GBS greater than 1: hospital-based evaluation is generally indicated; use physiologic status and clinical course, rather than a single high-score cutoff, to determine level of care. [1][5]
- Suspected liver disease or heart failure increases GBS and should also independently heighten attention to resuscitation tolerance and inpatient monitoring needs. [23]

### Avoid score-driven errors

Do not delay stabilization to obtain a complete score. Hemodynamic instability warrants immediate resuscitation, and endoscopy timing should be determined after adequate resuscitation rather than by a score alone. [1][2]

Do not infer that GBS 0-1 excludes all important disease. The threshold identifies a group with low likelihood of adverse outcomes requiring hospital-based intervention; it does not replace outpatient diagnostic planning when hematemesis, melena, anemia, or medication-related bleeding risk remains clinically concerning. [11][17][22]

*GBS-based disposition framework for suspected acute UGIB. [5][11][17][24]*

| Clinical state | GBS role | Next action |
| --- | --- | --- |
| Hemodynamic instability or need for active resuscitation | Score does not supersede physiologic urgency. [1][2] | Resuscitate, reassess hemodynamics and comorbidities, and arrange inpatient endoscopy after stabilization. [1][2][17] |
| Stable patient with GBS 0-1 | Very-low-risk category. [11][17][24] | Consider ED discharge with outpatient follow-up if no other clinical reason for admission exists. [17][22][24] |
| Stable patient with GBS greater than 1 | Does not meet the very-low-risk discharge threshold. [17][24] | Hospitalize for further assessment and endoscopy within 24 hours when admitted. [1][17] |

## Use risk stratification with restrictive transfusion and pre-endoscopic preparation

Hemoglobin thresholds and procedural preparation are separate from the GBS disposition threshold.

For hospitalized patients with UGIB, use red-cell transfusion at a hemoglobin threshold of 7 g/dL under the ACG recommendation. International consensus guidance uses a threshold below 80 g/L in patients without cardiovascular disease and recommends a higher threshold for patients with cardiovascular disease; therefore, individualize the trigger upward when cardiovascular disease changes tolerance of anemia. [11][17][24]

Do not equate a GBS-related predicted transfusion requirement with an automatic blood-product order. GBS helps identify patients likely to need intervention, whereas transfusion should follow the hemoglobin threshold, cardiovascular context, hemodynamic course, and ongoing clinical assessment. [5][11][17]

Consider erythromycin infusion before endoscopy in hospitalized UGIB when preparing for upper endoscopy, as recommended by the ACG guideline. The provided guideline excerpt supports its use but does not specify a dose; use institutional protocols for administration details. [17][24]

Empiric PPI therapy may be administered after resuscitation and before endoscopy in acute UGIB. Definitive high-dose PPI treatment is tied to successful endoscopic hemostasis of ulcers with high-risk stigmata rather than to pre-endoscopic risk score alone. [3][17]
- Hemoglobin below 7 g/dL in hospitalized UGIB: transfuse red cells unless patient-specific factors require a different threshold. [17][24]
- Cardiovascular disease: use a higher transfusion threshold than in patients without cardiovascular disease. [11]
- Suspected bleeding ulcer with successful endoscopic hemostasis: give high-dose PPI continuously or intermittently for 3 days, then twice-daily oral PPI for the first 2 weeks after endoscopy. [17][24]

*Pre-endoscopic and post-endoscopic decisions that should not be inferred from GBS alone. [11][17][24]*

| Decision | Actionable rule | What changes management |
| --- | --- | --- |
| Red-cell transfusion | Use hemoglobin 7 g/dL as the ACG threshold for hospitalized UGIB. [17][24] | Use a higher threshold when cardiovascular disease is present. [11] |
| Endoscopic preparation | Consider erythromycin infusion before endoscopy. [17][24] | Use institutional dosing protocols because a dose is not specified in the cited guideline excerpt. [17] |
| PPI after ulcer hemostasis | High-dose PPI continuously or intermittently for 3 days, then oral PPI twice daily for 2 weeks. [17][24] | This regimen applies after successful endoscopic hemostasis of bleeding ulcers with high-risk stigmata. [17] |

## Choose endoscopy timing and escalation according to instability and endoscopic findings

For admitted patients, early endoscopy means within 24 hours after presentation.

Perform endoscopy within 24 hours of presentation for patients admitted with acute UGIB. Complete resuscitation and optimization of important comorbidity first; the evidence base does not support a default strategy of endoscopy in under 6 hours for all patients with high GBS values. [1][9][11][17]

At endoscopy, treat ulcer bleeding with active spurting, active oozing, or a nonbleeding visible vessel. Recommended modalities include bipolar electrocoagulation, heater probe, and absolute ethanol injection; clips, argon plasma coagulation, and soft monopolar electrocoagulation have lower-quality supporting evidence. [17][24]

For recurrent ulcer bleeding after initially successful endoscopic hemostasis, repeat endoscopy is suggested. If endoscopic therapy fails, proceed to transcatheter embolization rather than relying on repeated risk-score reassessment. [17][24]

For suspected variceal bleeding, do not manage the patient as routine nonvariceal UGIB. Variceal bleeding guidance advises endoscopy within 12-24 hours, and the clinical priority remains resuscitation and preparation for definitive endoscopic management rather than use of a nonvariceal discharge algorithm. [1]
- Admitted UGIB: schedule endoscopy within 24 hours after presentation. [1][17]
- Ulcer with active bleeding or nonbleeding visible vessel: perform endoscopic hemostasis. [17][24]
- Rebleeding after successful ulcer hemostasis: repeat endoscopy. [17][24]
- Failure of endoscopic treatment: arrange transcatheter embolization. [17][24]

### Post-endoscopic risk is lesion-based

Once endoscopy identifies a bleeding ulcer, endoscopic stigmata—not the pre-endoscopic GBS—determine whether hemostatic therapy and high-dose PPI are required. High-risk stigmata after successful hemostasis warrant three days of high-dose PPI therapy followed by twice-daily oral PPI for two weeks. [17][24]

*Escalation after endoscopic evaluation of ulcer bleeding. [17][24]*

| Finding or event | Recommended next step | Risk-stratification implication |
| --- | --- | --- |
| Active spurting or oozing ulcer, or nonbleeding visible vessel | Endoscopic hemostasis using an evidence-supported modality. [17][24] | Endoscopic lesion risk supersedes pre-endoscopic discharge assessment. [17] |
| Successful hemostasis of ulcer with high-risk stigmata | High-dose PPI for 3 days, then oral PPI twice daily for 2 weeks. [17][24] | Monitor for recurrent bleeding during the high-risk post-hemostasis period. [17] |
| Recurrent ulcer bleeding | Repeat endoscopy. [17][24] | Rebleeding requires procedural reassessment rather than score-only management. [17] |
| Endoscopic hemostasis failure | Transcatheter embolization is suggested. [17][24] | Escalate to definitive hemostatic intervention. [17] |

## Document the score, the clinical override, and the disposition plan

A usable risk assessment records why a patient was discharged, admitted, or escalated.

For every suspected UGIB presentation, document the calculated GBS, hemodynamic status after initial resuscitation, hemoglobin-based transfusion decision, and planned timing of endoscopy. This links the validated low-risk threshold to the actual disposition decision and makes explicit when instability or comorbidity overrides a low-complexity pathway. [1][11][17]

For ED discharge at GBS 0-1, document outpatient follow-up and return precautions for recurrent bleeding. For GBS greater than 1 or any physiologic concern, document the admitting service, monitoring setting based on clinical severity, and endoscopy target within 24 hours. [17][22][24]

After endoscopy, update risk assessment using lesion findings and hemostatic response. Recurrent bleeding should trigger repeat endoscopy; failure of endoscopic therapy should prompt transcatheter embolization. [17][24]
- Record: GBS, blood pressure and heart rate trajectory, hemoglobin, cardiovascular disease status, transfusion decision, and endoscopy timing. [11][17][23]
- For discharge: record GBS 0-1, clinical stability, outpatient follow-up, and explicit instructions to return for recurrent bleeding. [17][22][24]
- For admission: record the reason GBS or clinical status places the patient outside the very-low-risk pathway. [1][17]

*Minimum decision documentation for acute UGIB risk stratification. [1][11][17][22]*

| Care transition | Document | Purpose |
| --- | --- | --- |
| Initial ED assessment | GBS and post-resuscitation hemodynamics. [1][5][23] | Separates potential low-risk discharge candidates from patients needing hospital-based evaluation. [17][24] |
| Transfusion decision | Hemoglobin value and cardiovascular disease status. [11][17] | Shows application of restrictive transfusion thresholds and justified exceptions. [11][17] |
| Endoscopy plan | Target within 24 hours for admitted patients and any reason for altered timing. [1][17] | Aligns procedure timing with resuscitation and clinical severity. [1] |
| Post-endoscopy plan | Stigmata, hemostatic therapy, PPI plan, and escalation pathway for rebleeding. [17][24] | Transitions from pre-endoscopic score-based risk assessment to lesion-based management. [17] |

## References
1. Update on the management of upper gastrointestinal bleeding — bmjmedicine.bmj.com — https://bmjmedicine.bmj.com/content/1/1/e000202
2. Guideline on Managing Acute Upper Gastrointestinal Bleeding Updated | NEJM Clinician — clinician.nejm.org — https://clinician.nejm.org/guideline-managing-acute-upper-gastrointestinal-bleeding-updated-nejm-jw.FW115950
3. Management of acute upper gastrointestinal bleeding - The BMJ — www.bmj.com — https://www.bmj.com/content/bmj/364/bmj.l536.full.pdf
4. Acute upper gastrointestinal bleeding in the UK: 2022 audit update — gut.bmj.com — https://gut.bmj.com/content/75/4/760
5. Asia-Pacific working group consensus on non-variceal upper gastrointestinal bleeding: an update 2018 | Gut — gut.bmj.com — https://gut.bmj.com/content/67/10/1757
6. Asia-Pacific working group consensus on non-variceal upper ... - Gut — gut.bmj.com — https://gut.bmj.com/content/gutjnl/67/10/1757.full.pdf
7. Assessment of upper gastrointestinal bleeding - Differential diagnosis of symptoms | BMJ Best Practice — bestpractice.bmj.com — https://bestpractice.bmj.com/topics/en-gb/456
8. Proton-Pump Inhibitors to Prevent Gastrointestinal Bleeding — evidence.nejm.org — https://evidence.nejm.org/doi/full/10.1056/EVIDoa2400134
9. Timing of Endoscopy for Acute Upper Gastrointestinal Bleeding — www.nejm.org — https://www.nejm.org/doi/full/10.1056/NEJMoa1912484
10. Diagnosis and management of acute lower gastrointestinal bleeding — gut.bmj.com — https://gut.bmj.com/content/gutjnl/68/5/776.full.pdf
11. Management of Nonvariceal Upper Gastrointestinal Bleeding: Guideline Recommendations From the International Consensus Group — annals.org — https://annals.org/aim/fullarticle/2753604/management-nonvariceal-upper-gastrointestinal-bleeding-guideline-recommendations-from-international-consensus
12. Association of pre-endoscopic fresh frozen plasma transfusion with clinical outcomes in patients with acute upper gastrointestinal bleeding and mild coagulopathy: a two-center retrospective cohort study | Scientific Reports — www.nature.com — https://www.nature.com/articles/s41598-026-41863-y
13. Risk stratification in upper and upper and lower GI bleeding: Which scores should we use? — www.sciencedirect.com — https://www.sciencedirect.com/science/article/abs/pii/S1521691819300162
14. Risk stratification in acute upper GI bleeding: comparison of the AIMS65 score with the Glasgow-Blatchford and Rockall scoring systems - ScienceDirect — www.sciencedirect.com — https://www.sciencedirect.com/science/article/abs/pii/S0016510715030370
15. Conduct and reporting of studies developing pre-endoscopic ... — www.sciencedirect.com — https://www.sciencedirect.com/science/article/pii/S0895435626003100/pdf?md5=2d1532b43d3d033d5679f14f34653325&pid=1-s2.0-S0895435626003100-main.pdf
16. Scoring systems for risk stratification in upper and lower gastrointestinal bleeding — www.sciencedirect.com — https://www.sciencedirect.com/science/article/abs/pii/S1521691823000549
17. ACG Clinical Guideline: Upper Gastrointestinal... : American Journal of Gastroenterology — journals.lww.com — https://journals.lww.com/ajg/Abstract/9000/ACG_Clinical_Guideline__Upper_Gastrointestinal_and.98805.aspx
18. Assessment of Prognostic Scores for Emergency Department ... — www.annemergmed.com — https://www.annemergmed.com/article/S0196-0644(24)00359-7/pdf
19. Overview | Acute upper gastrointestinal bleeding in over 16s: management  | Guidance | NICE — www.nice.org.uk — https://www.nice.org.uk/guidance/cg141
20. Prospective Comparison of the AIMS65 Score, Glasgow-Blatchford Score, and Rockall Score for Predicting Clinical Outcomes in Patients with Variceal and Nonvariceal Upper Gastrointestinal Bleeding - PMC — pmc.ncbi.nlm.nih.gov — https://pmc.ncbi.nlm.nih.gov/articles/PMC8039743
21. AIMS65 scoring system is comparable to Glasgow-Blatchford score or Rockall score for prediction of clinical outcomes for non-variceal upper gastrointestinal bleeding - PMC — pmc.ncbi.nlm.nih.gov — https://pmc.ncbi.nlm.nih.gov/articles/PMC6660932
22. External validation and comparison of the Glasgow-Blatchford score, modified Glasgow-Blatchford score, Rockall score and AIMS65 score in patients with upper gastrointestinal bleeding: a cross-sectional observational study in Western Switzerland - PMC — pmc.ncbi.nlm.nih.gov — https://pmc.ncbi.nlm.nih.gov/articles/PMC10405788
23. Diagnostic Performance of AIMS65, Glasgow-Blatchford, and Pre-endoscopy Rockall Scores in Predicting Clinical Outcomes Among Upper Gastrointestinal Bleeding Patients in the Emergency Department — pmc.ncbi.nlm.nih.gov — https://pmc.ncbi.nlm.nih.gov/articles/PMC12703674
24. ACG Clinical Guideline: Upper Gastrointestinal and Ulcer Bleeding - PubMed — www.ncbi.nlm.nih.gov — http://www.ncbi.nlm.nih.gov/pubmed/33929377

## Editorial note

Prepared from cited clinical literature using Astra's research workflow. Verify recommendations against current guidance and patient-specific factors.