When to Admit Patients with Upper Gastrointestinal Bleeding
Admit all patients with acute upper GI bleeding EXCEPT those with a Glasgow-Blatchford Score (GBS) ≤1, who can be safely managed as outpatients with outpatient endoscopy. 1, 2
Risk Stratification Framework
The decision to admit hinges on formal risk assessment using the Glasgow-Blatchford Score, which is the only validated tool recommended by international guidelines for identifying very low-risk patients who can avoid hospitalization 1, 2, 3.
Patients Who Can Be Discharged (No Admission Required)
- GBS ≤1: These patients are at very low risk for rebleeding, 30-day mortality, or need for hospital-based intervention 1, 2
- Can be managed with outpatient endoscopy follow-up
- This represents a conditional recommendation with moderate quality evidence
Patients Who MUST Be Admitted
All other patients require hospitalization, particularly those with:
- Hemodynamic instability (hypotension, tachycardia, shock index elevation) - these patients need immediate resuscitation 1, 4
- GBS ≥2 - indicates increased risk requiring inpatient monitoring and endoscopy
- Active bleeding (hematemesis, melena with ongoing blood loss)
- Significant comorbidities, especially cardiovascular disease requiring different transfusion thresholds 1
- Anticoagulation use - though endoscopy should not be delayed, these patients need inpatient management 1
Important Caveats
Do NOT use the AIMS65 score for discharge decisions - guidelines explicitly recommend against it for identifying low-risk patients 1. The Glasgow-Blatchford Score is superior and the only validated tool for this purpose.
Shock index matters for mortality prediction: Higher shock index correlates with increased mortality risk, reinforcing the need for admission in hemodynamically compromised patients 4.
Post-Endoscopy Discharge Considerations
Even among admitted patients, selected patients with acute ulcer bleeding who are at low risk for rebleeding based on clinical AND endoscopic criteria may be discharged promptly after endoscopy 1. This includes patients with clean-based ulcers or flat pigmented spots (low-risk endoscopic stigmata) who have stable vital signs and hemoglobin.
The evidence strongly supports using objective risk stratification rather than clinical gestalt alone, as the GBS has been validated across multiple international cohorts to safely identify the minority of patients who can avoid hospitalization entirely.