S1383 A Comparison of the ABC and AIMS65 Scores in Predicting Outcomes in Patients With Acute Upper Gastrointestinal Bleeding

Introduction: Upper gastrointestinal bleeding (UGIB) remains a significant medical emergency with mortality rates of 2%-15%. Early risk stratification is crucial for optimizing patient outcomes and resource allocation. While both the ABC score (Age, Blood tests, Comorbidities) and AIMS65 score are validated risk assessment tools, their comparative performance in predicting clinical outcomes beyond mortality requires further evaluation. Methods: We conducted a retrospective cohort study of adult patients (ages 18-89) admitted with UGIB to multiple Northwell Health hospitals between January 2019 and January 2024. Patients were identified using ICD-10 codes and confirmed by chart review for documented UGIB (hematemesis, coffee-ground emesis, bloody nasogastric aspirate, or endoscopic evidence). We calculated ABC and AIMS65 scores for each patient and compared their ability to predict primary outcomes (in-hospital mortality and 30-day readmission) and secondary outcomes (hospital length of stay [LOS] and ICU LOS). Statistical analysis included Spearman correlation coefficients and Wilcoxon rank-sum tests. Results: Among 2,272 patient encounters analyzed, in-hospital mortality was 4.89% (n = 111). The ABC score demonstrated superior correlation with hospital LOS compared to AIMS65 (r = 0.468 vs r = 0.335, both P < 0.0001). Similarly, ABC showed stronger correlation with ICU LOS (r = 0.340 vs r = 0.199, both P < 0.0001). Both scores effectively discriminated mortality, with significantly higher scores in non-survivors (P < 0.0001 for both). However, neither score predicted 30-day readmission (ABC: P = 0.896; AIMS65: P = 0.620). Conclusion: The ABC score demonstrates superior performance compared to AIMS65 in predicting resource utilization metrics, particularly hospital and ICU length of stay, while maintaining comparable mortality discrimination. These findings suggest the ABC score may be more valuable for clinical decision-making regarding disposition and resource allocation in UGIB patients. Further prospective validation and direct ROC curve comparison is warranted to confirm these findings and develop optimized risk-stratification protocols.