Abstract: SA-PO0087
Neutrophil Percentage-to-Albumin Ratio as a Predictor of AKI in Patients with Cirrhosis: A Novel Approach Using Artificial Intelligence
Session Information
- AKI: Clinical Diagnostics and Biomarkers
November 08, 2025 | Location: Exhibit Hall, Convention Center
Abstract Time: 10:00 AM - 12:00 PM
Category: Acute Kidney Injury
- 102 AKI: Clinical, Outcomes, and Trials
Author
- Younis, Sherif Elbaz, HCA Healthcare Graduate Medical Education, Kingwood, Texas, United States
Background
Liver cirrhosis affects approximately 2.2 million adults in the USA and is associated with significant morbidity, mortality, and economic burden. Acute kidney injury (AKI), occurring in up to 60% of hospitalized cirrhotic patients, is linked to increased mortality and treatment costs. Diagnosing AKI in cirrhosis is challenging due to altered serum creatinine and urine output parameters. The neutrophil percentage-to-albumin ratio (NAR), a novel inflammatory biomarker, has not yet been evaluated as a predictor of AKI in cirrhotic patients.
Methods
This prospective cross-sectional study included 322 patients with liver cirrhosis and ascites admitted from January to November 2024 at two hospitals in Egypt. AKI was defined per International Club of Ascites and Acute Disease Quality Initiative guidelines using Kidney Disease Improving Global Outcomes criteria. Blood samples collected upon admission measured NAR, serum albumin, creatinine, and inflammatory markers. Statistical analyses included logistic regression models adjusting for Model for End-Stage Liver Disease (MELD) score, age, diabetes, and chronic kidney disease (CKD). Machine learning methods (Random Forest model and Shapley Additive Explanations analysis) assessed predictor importance.
Results
AKI developed in 130 patients (40.4%). Patients with AKI had significantly higher NAR, C-reactive protein, neutrophil percentage, MELD scores, and lower albumin levels compared to patients without AKI (all p<0.001). NAR demonstrated superior diagnostic performance (area under the curve = 0.893), with 85.4% sensitivity and 72.3% specificity at a cutoff value >23.2. For severe AKI (stages 2–3), NAR’s predictive accuracy improved (area under the curve = 0.910). Logistic regression confirmed NAR (odds ratio=1.061, p=0.004), low serum albumin (odds ratio=0.095, p=0.004), and CKD as independent predictors. Random Forest analysis identified serum albumin and NAR as top predictors of AKI.
Conclusion
NAR, measured at hospital admission, reliably predicts AKI in cirrhotic patients with ascites. Its predictive power exceeds traditional biomarkers, particularly for advanced AKI stages, highlighting its clinical utility for early identification and proactive management of renal dysfunction.