A Comparison of the ABC and AIMS65 Scores in Predicting Outcomes in Patients with Acute Upper Gastrointestinal Bleeding: A Retrospective Multicenter Study

Why It Matters

Accurate early risk stratification can streamline triage, allocate intensive resources efficiently, and potentially improve survival for UGIB patients.

Key Takeaways

• ABC score AUC 0.793 vs AIMS65 0.661 for mortality.
• One ABC point raises mortality odds by ~51%.
• ABC predicts length of stay, ICU time, vasopressor use better.
• Neither score forecasts 30‑day readmission.
• Prospective validation of ABC thresholds still needed.

Pulse Analysis

Upper gastrointestinal bleeding remains a high‑stakes emergency, accounting for thousands of hospital admissions annually and a mortality rate approaching 5 percent. Clinicians rely on pre‑endoscopy scoring systems to gauge severity before definitive treatment. The ABC score, which incorporates age, laboratory values, and comorbidities, and the older AIMS65 model, which focuses on albumin, INR, mental status, systolic pressure, and age, are the most widely used tools in U.S. hospitals. Understanding their comparative performance is essential for evidence‑based triage decisions.

The Northwell Health study examined over two thousand adult UGIB cases from 2019 to 2024, applying both scores to each patient. Statistical analysis revealed the ABC score’s area under the ROC curve at 0.793, markedly higher than AIMS65’s 0.661, indicating superior discrimination for in‑hospital death. Moreover, a single point increase in the ABC metric translated into a 50.7 percent rise in mortality odds, while correlations with secondary outcomes—hospital length of stay, ICU duration, and vasopressor requirement—were consistently stronger. Neither score, however, demonstrated predictive power for 30‑day readmission, highlighting a gap in post‑discharge risk assessment.

These findings suggest that hospitals could improve patient outcomes and resource utilization by integrating the ABC score into early assessment protocols. Faster identification of high‑risk patients may prompt earlier endoscopic intervention, intensive monitoring, and targeted therapies, potentially reducing mortality. Nonetheless, the study’s retrospective design and single‑system data set underscore the need for prospective, multi‑regional validation before universal adoption. Future research should also explore combining the ABC score with post‑discharge predictors to address readmission risk, thereby offering a more comprehensive risk stratification framework for gastroenterology teams.