What is the clinical evidence from this study?

Study design: Cohort. Population: Cardiac surgery-associated acute kidney injury (CSA-AKI) (n=733). Intervention: LightGBM machine learning model vs. Logistic Regression. Primary outcome: Prediction of CSA-AKI (Area Under the Curve) (AUC 0.784, 95% CI 0.702-0.859).

What question did this study set out to answer?

This study aims to develop and validate a machine learning model to predict acute kidney injury after cardiac surgery in older adults.

June 12, 2026Clinical Interventions in Aging0 citationsOpen Access

Development and Validation of a Machine Learning Model for Predicting Serum Creatinine-Defined Acute Kidney Injury in Older Adults After Cardiac Surgery

YFYuhao FuSoochow University JQJiajie QianSoochow University YZYang ZhangSoochow University

Key Result

The LightGBM machine learning model effectively predicted cardiac surgery-associated acute kidney injury in older adults, achieving an AUC of 0.784.

Key Points

This study aims to develop and validate a machine learning model to predict acute kidney injury after cardiac surgery in older adults.
Retrospective study utilizing data from two centers with 623 patients for model training and validation, and 110 for external validation.
Key predictors were identified using LASSO, Recursive Feature Elimination, Boruta, and Random Forest methods.
Six machine learning models were developed, including LightGBM and assessed using area under the curve (AUC).
24.1% of patients developed acute kidney injury during the first week post-surgery.
LightGBM achieved the highest AUC of 0.784 (95% CI: 0.702–0.859) among the models evaluated.
Key predictors included lactate, surgical duration, APTT, transfusion volume, and PT, with analysis performed using SHAP techniques.

Study Design

Type

Cohort (n=733)

Multicenter

Yes

Structured PICO

Does a machine learning model (LightGBM) accurately predict the risk of cardiac surgery-associated acute kidney injury in older adults?

Population

733 older adults aged 65 and older undergoing their first cardiac surgery, evaluated retrospectively for acute kidney injury within 7 days postoperatively.

Exposure

LightGBM machine learning model for predicting cardiac surgery-associated acute kidney injury (CSA-AKI) using 16 key variables (including lactate, surgical duration, APTT, transfusion volume, and PT).

Comparator

Traditional Logistic Regression model and other machine learning models (KNN, SVM, RF, XGBoost).

Outcome

Development of cardiac surgery-associated acute kidney injury (CSA-AKI) defined by KDIGO Serum creatinine criteria within the first postoperative week.hard clinical

A LightGBM machine learning model incorporating intraoperative and baseline variables effectively predicts the risk of acute kidney injury in older adults following cardiac surgery, outperforming traditional logistic regression.

Main Result

Effect estimate: AUC 0.784 (95% CI 0.702-0.859)

Absolute Event Rate: 0.784% vs 0.775%

Limitations

Retrospective study design
Conducted in Chinese centers with a modest sample size, limiting generalizability to other ethnicities or healthcare systems
Perioperative urine output was excluded from renal function evaluation due to data unavailability

Abstract

Background: Cardiac surgery-associated acute kidney injury (CSA-AKI) is a frequent and devastating postoperative complication, particularly among older adults. Accurate risk stratification and early prediction of CSA-AKI are essential for guiding preventive strategies and optimizing clinical decision-making. Methods: In this retrospective study, data from two centers (n=623) were utilized for model training and internal validation, whereas data from a third, distinct center (n=110) were reserved for external validation. CSA-AKI was defined according to the Kidney Disease: Improving Global Outcomes (KDIGO) Serum creatinine criteria. Key predictors were identified using a consensus of four methods: Least Absolute Shrinkage and Selection Operator (LASSO), Recursive Feature Elimination (RFE), Boruta, and Random Forest-based filtering. Six machine learning (ML) models, including Logistic Regression (LR), K-Nearest Neighbors (KNN), Support Vector Machine (SVM), Random Forest (RF), Extreme Gradient Boosting (XGBoost), and Light Gradient Boosting Machine (LightGBM), were developed utilizing five-fold cross-validation. Predictive performance was assessed using the area under the receiver operating characteristic curve (AUC). The SHapley Additive exPlanations (SHAP) approach was applied to interpret the best-performing model. Results: Development of CSA-AKI was noted in 177 patients (24.1%) during the first postoperative week. In terms of comparative performance, LightGBM exhibited the greatest AUC (0.784, 95% confidence interval CI: 0.702– 0.859). The most influential features were lactate, surgical duration, activated partial thromboplastin time (APTT), transfusion volume, and Prothrombin Time (PT). SHAP-based summary and force plots interpreted the model at global and local levels. Furthermore, SHAP dependence plots elucidated non-linear effects of single features on CSA-AKI risk. Conclusion: Machine learning models demonstrate high efficacy in predicting CSA-AKI risk in older adults. The LightGBM model outperformed other algorithms; coupled with interpretability tools, it can assist clinicians to identify high-risk patients earlier and optimize perioperative management. Keywords: acute kidney injury, cardiac surgery, older adults, machine learning, SHAP

Ask AI

Helpful

Bookmark

View Full Paper