Development and internal validation of machine learning in predicting prognosis of acute kidney injury patients in resource-limited setting
Authors' conclusion
Machine learning models using routinely available clinical variables may offer useful prognostic information for AKI outcomes in resource-limited settings and outperform traditional scoring systems.
Study context
- Population
- Critically ill patients with acute kidney injury (AKI) in Southeast Asia (Southeast Asia AKI cohort)of 6993 ICU patients
- Participants
- 1,650 patients
- Comparison
- Logistic regression vs random forest vs XGBoost vs conventional clinical scores
- Outcome
- In-hospital mortality and major adverse kidney events within 28 days (MAKE-28)
Methodological detailsprospective multicenter cohort, machine learning, recursive feature elimination
- Prospective multicenter cohort
- Machine learning
- Recursive feature elimination
Performance
As reported in the abstract
Prediction of MAKE-28 and hospital mortality
AUC of the three models for MAKE-28
0.73-0.76
AUC of the three models for hospital mortality
0.71-0.75
Further findings
Observed outcomes in the cohort
- Died during hospitalization
- 778 (47.1%)
- Experienced MAKE-28
- 1204 (73.9%)
Original publication
Development and internal validation of machine learning in predicting prognosis of acute kidney injury patients in resource-limited setting. J Crit Care. Published 7 February 2026. doi:10.
Composed by AI from the published abstract and checked against it. How Briefs are made →