Abstract:
Introduction: Upper gastrointestinal bleeding (UGIB) is a common emergency with
significant morbidity and mortality. Risk stratification scores such as Glasgow Blatchford
Score (GBS), pre-endoscopic Rockall Score, and AIMS65 are used to predict adverse outcomes
and guide clinical management. However, their comparative effectiveness in predicting both
in-hospital mortality and six-month readmission remains underexplored in the Indian setting.
Objective: This study aimed to compare the effectiveness of GBS, pre-endoscopic Rockall,
and AIMS65 scores in predicting in-hospital mortality and six-month hospital readmission in
patients presenting with UGIB.
Materials and Methods: This cross-sectional observational study included 78 patients
presenting to the emergency department with UGIB between May 2023 and January 2024.
Scores were calculated based on clinical presentation and laboratory parameters. Patients were
followed for mortality during admission and readmission up to six months post-discharge. Data
were analyzed using chi-square, Mann-Whitney U tests, and ROC curves to assess predictive
performance.
Results: The Rockall score showed the strongest correlation with mortality, rising from 0% at
score 0 to 100% at score 6. AIMS65 and GBS also predicted mortality but less consistently.
Conversely, lower scores across all tools were associated with higher six-month readmission
rates. For instance, AIMS65 scores of 0–2 showed >90% readmission, while scores of 5 had
0% readmission. This inverse trend likely reflects survival bias, where high-risk patients do not
survive to be readmitted.
Conclusion: The Rockall score was the most reliable for predicting both mortality and
readmission trends. It remains a robust tool for UGIB risk stratification in emergency settings.
