What is the clinical significance and management of a patient with upper gastrointestinal bleeding based on their Rockall Score (RS score)?

Rockall Score: Clinical Significance and Management in Upper GI Bleeding

Primary Clinical Use

The Rockall Score is most valuable for predicting mortality risk after endoscopy, with a complete score >8 indicating high risk of death and a score <3 indicating excellent prognosis with very low rebleeding and mortality risk. 1 The complete Rockall score has the best discriminative ability for mortality among risk stratification tools, with an area under the ROC curve of 0.73-0.81, validated across multiple countries. 2, 1

Score Components and Calculation

The Rockall Score incorporates five variables that predict outcomes: 2, 1

• Age: Mortality reaches 30% in patients >90 years versus rare in those <40 years 3
• Shock parameters: Pulse >100 bpm AND systolic blood pressure <100 mmHg 2, 3
• Comorbidities: Heart failure, coronary disease, renal/hepatic failure, disseminated cancer (liver failure receives 3 points) 3, 4
• Endoscopic diagnosis: Type of lesion identified 2
• Stigmata of recent hemorrhage: Active arterial bleeding in peptic ulcer with shock carries 80% risk of death 3

Risk Stratification and Management Algorithm

Low-Risk Patients (Rockall Score <3)

Patients with a complete Rockall score <3 have excellent prognosis and can be considered for early discharge. 1 However, discharge requires meeting ALL of the following criteria: 1

• Hemoglobin >100 g/L (10 g/dL)
• Low-risk endoscopic findings (clean-based ulcer, Mallory-Weiss tear, or normal endoscopy)
• Age <60 years
• Minimal or no comorbidities
• Hemodynamic stability (pulse <100 bpm AND systolic BP >100 mmHg)
• Adequate social support and hospital accessibility
• 4-6 hour post-endoscopy observation period completed 1

High-Risk Patients (Rockall Score ≥8)

Patients with Rockall score >8 require intensive monitoring and aggressive management due to high mortality risk. 1, 3 Management includes:

• Immediate resuscitation with crystalloids 4
• Blood transfusion threshold of 70 g/L (7 g/dL) in most patients, targeting 70-90 g/L 2
• Higher transfusion threshold of 80 g/L (8 g/dL) for patients with cardiovascular disease, targeting post-transfusion hemoglobin ≥100 g/L 1
• Endoscopic intervention for high-risk stigmata (active bleeding, non-bleeding visible vessel, adherent clot) 2
• Close monitoring in intensive care setting for patients requiring intensive care 5

Critical Limitations and When NOT to Use Rockall Score

The Rockall Score should NOT be used for initial triage or discharge decisions before endoscopy. 1 The Glasgow Blatchford Score (GBS) is superior for identifying very low-risk patients who can be safely discharged without hospitalization or endoscopy, with 99% sensitivity for high-risk patients compared to Rockall's 93-96%. 1 A GBS of 0-1 identifies patients safe for outpatient management. 1, 6

The Rockall Score has better discriminative ability for mortality than for rebleeding, with only modest performance for predicting rebleeding (AUROC 0.642). 2, 7 For predicting need for hospital-based intervention, the GBS performs similarly to the complete Rockall score (AUROC 0.705 vs 0.727). 7

Specific Clinical Scenarios

Patients with Chronic Liver Disease

Patients with liver failure automatically receive 3 points in the Rockall comorbidity section, indicating higher risk. 4 These patients require: 4

• Immediate vasoactive drug therapy (terlipressin, somatostatin, or octreotide) if variceal bleeding suspected
• Antibiotic prophylaxis with ceftriaxone 1g IV daily for up to 7 days
• Restrictive transfusion strategy (hemoglobin threshold 7 g/dL, target 7-9 g/dL) to avoid increasing portal pressure

Patients with Cardiovascular Disease

Patients with coronary artery disease require higher hemoglobin targets during acute management (≥80 g/L), with post-transfusion target ≥100 g/L. 1 Comorbid cardiac disease is closely related to mortality in GI bleeding, as these patients have poor tolerance for anemia and hypotension leading to cardiac ischemia. 3

Common Pitfalls to Avoid

• Do not use pre-endoscopic Rockall score for discharge decisions—it lacks the endoscopic variables needed for accurate risk stratification 2, 1
• Do not rely on Rockall score alone to predict rebleeding—its AUROC for rebleeding is only 0.642, indicating poor discriminative ability 7
• Do not discharge patients based solely on low Rockall score—multiple criteria including hemoglobin >100 g/L, hemodynamic stability, and social factors must be met 1
• Deaths occur almost entirely in patients with significant comorbidities—age and comorbidity assessment are critical components that cannot be overlooked 3