Predicting mortality and readmission in UGIB: a comparative study of rockall, glasgow-blatchford, and AIMS65 scores

Abstract

Background Upper gastrointestinal bleeding (UGIB) is a frequent emergency requiring prompt risk stratification for effective management. This study evaluates the predictive accuracy of three pre-endoscopic scoring systems-Glasgow-Blatchford Score (GBS), AIMS65, and pre-endoscopic Rockall Score (pRS)-in forecasting 30-day mortality and hospital readmission among UGIB patients in Iranian tertiary hospitals. Methods A prospective observational study was conducted at Al-Zahra and Khorshid Hospitals (Isfahan, Iran) between April 2024 and April 2025. Adult patients presenting with UGIB symptoms were assessed using pRS, GBS, and AIMS65. Data were collected prospectively at the time of admission and during hospitalization using standardized clinical forms and patient records. Statistical analyses-including receiver operating characteristic (ROC) curves, area under the curve (AUC) comparisons, sensitivity, specificity, and predictive values-were performed to determine each system's prognostic utility. Results Among 290 enrolled patients, 30-day mortality occurred in 23.4, with shock, tachycardia, syncope, altered mental status, and severe comorbidities significantly associated with adverse outcomes (p < 0.001). pRS exhibited the highest predictive accuracy for mortality (AUROC = 0.815) and readmission (AUROC = 0.605), outperforming GBS and AIMS65, though the difference between pRS and AIMS65 was not statistically significant (p = 0.924). At low-risk thresholds (pRS < 1, GBS < 2, AIMS65 < 1), all three systems demonstrated high sensitivity and negative predictive value (NPV), with pRS showing the best balance of prognostic performance. Conclusion The pre-endoscopic Rockall Score proved most effective for identifying high-risk UGIB patients and guiding early management. Its strong sensitivity and negative predictive value support its role in safe discharge planning and risk stratification in emergency settings. Further validation in diverse clinical contexts is recommended to enhance UGIB management strategies.