Prognostic value of stress hyperglycemia ratio in critically Ill patients with acute kidney injury: a machine learning-driven retrospective cohort analysis - PubMed
4 hours ago
- #mortality prediction
- #acute kidney injury
- #machine learning
- The study explores the prognostic value of stress hyperglycemia ratio (SHR) in critically ill patients with acute kidney injury (AKI).
- A retrospective analysis of 5,555 ICU patients from the MIMIC-IV database was conducted, with patients divided into SHR quartiles.
- Primary outcomes measured were 30-day and 90-day all-cause mortality.
- A U-shaped relationship was found between SHR and mortality, with higher SHR levels significantly increasing the risk of death.
- Machine learning models, particularly LightGBM, demonstrated superior predictive performance (AUC = 0.864) compared to traditional ICU scoring systems.
- The study concludes that SHR is an independent, nonlinear predictor of short-term mortality in ICU patients with AKI.