Key result
Machine learning ensemble predicts 30-day hospital readmission in patients with diabetes with ~0.69 AUC.
Why the study?
Hospital readmission among patients with diabetes remains a major healthcare challenge, and early identification of high-risk patients could support targeted interventions.
Population
101,766 hospital encounters of patients with diabetes from the Diabetes 130-US Hospitals dataset
Comparison
Multiple machine learning models evaluated including Logistic Regression, Random Forest, XGBoost, LightGBM, and stacking ensemble
Design
Multi-institutional retrospective dataset analysis and machine learning model development study
Follow-up
30 days
Authors
Loading...
May facilitate targeted interventions in high-risk diabetes patients; leaves open outcome benefit pending prospective randomized evaluation.
Observational (n=101,766)
Yes
Effect estimate: Calibrated AUC 0.688
A machine learning stacking ensemble model can predict 30-day hospital readmission in patients with diabetes with moderate discrimination, potentially supporting clinical decision-making.
Salim et al. (2026) conducted an observational in Diabetes (n=101,766). Machine learning framework (stacking ensemble model) was evaluated on 30-day hospital readmission (Calibrated AUC 0.688). A stacking ensemble machine learning model predicted 30-day hospital readmission in patients with diabetes, achieving a calibrated AUC of 0.688 and a Brier score of 0.094.
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: