Independent predictors of ICU admission, mortality, and intervention requirements in acute upper gastrointestinal bleeding

Abstract Background: Upper gastrointestinal bleeding (UGIB) is a critical emergency with significant mortality. This study identifies independent predictors of ICU admission, mortality, and intervention requirements in acute UGIB in a tertiary care hospital. Methods: A prospective cohort study was conducted at King Saud Medical City, Riyadh, enrolling 276 adults with UGIB, between January and December 2022. Multivariable logistic regression analysis assessed predictors for primary outcomes (ICU admission and mortality) and secondary outcomes (endoscopic, radiological, surgical intervention, and rebleeding). Results: The cohort (72.1% male; mean age 53.4 years) had high-risk features (mean Glasgow–Blatchford Score GBS 7.6; Clinical Rockall Score 2.9). Endoscopy identified a bleeding source in 89.9%. Endotherapy was required in 42.0%, while surgical or radiological interventions were rare (1.1% each). Rebleeding occurred in 5.8%, ICU admission in 17.0%, 7-day mortality in 9.1%, and 30-day mortality in 14.9% of patients. Independent predictors were: for ICU admission—higher GBS (aOR 1.25 per point; 95% CI 1.14-1.37; P < 0.001) and rebleeding (aOR 3.82; 95% CI 1.68-8.69; P = 0.001); for endoscopic intervention—variceal bleeding (aOR 8.91; 95% CI 4.83-16.42; P < 0.001) and lower hemoglobin (aOR 0.88 per g/dL; 95% CI 0.81-0.96; P = 0.004); for mortality—rebleeding (early aOR 4.95; overall aOR 3.65; P < 0.001) and higher Total Rockall Score (aOR 1.38 per point; P < 0.001). Conclusion: In this Saudi cohort, rebleeding and ICU admission were the strongest mortality predictors. Validated scores reliably predicted outcomes, while varices determined intervention. These findings highlight the importance of protocolized care, dynamic risk stratification, and resource allocation in regions with high liver disease burden.