Related Experiment Video
Updated: Jul 5, 2025
Author Spotlight: Simulating Pediatric Cardiac Surgery Using a Neonatal Piglet Model
Published on: May 26, 2023
Machine learning prediction model of major adverse outcomes after pediatric congenital heart surgery: a retrospective
Chaoyang Tong1, Xinwei Du2, Yancheng Chen3
1Department of Anesthesiology.
Insights
Machine learning models accurately predict major adverse postoperative outcomes (APOs) in pediatric heart surgery. Mechanical ventilation time is a key predictor, aiding clinical decisions for better patient care.
Area of Science:
- Cardiovascular Surgery
- Pediatric Cardiology
- Medical Informatics
- Machine Learning in Healthcare
Background:
- Adverse postoperative outcomes (APOs) significantly impact mortality, hospital stay, and quality of life in pediatric congenital heart surgery patients.
- Accurate prediction of APOs is crucial for optimizing patient management and improving surgical outcomes.
Purpose of the Study:
- To evaluate the performance of five machine learning (ML) algorithms in predicting four major APOs after pediatric congenital heart surgery.
- To identify clinically meaningful interpretations of ML models for risk factor identification.
Main Methods:
- Utilized a dataset of 23,000 pediatric patients undergoing congenital heart surgery, with 13,927 in the training cohort and 9,073 in the testing cohort.
- Investigated four major APOs: low cardiac output syndrome (LCOS), pneumonia, renal failure, and deep venous thrombosis (DVT).
- Applied five ML models (LightGBM, LR, SVM, Random Forest, CatBoost) using 39 clinical and laboratory features, evaluating performance with AUC and SHAP values.
Main Results:
- Machine learning models demonstrated high predictive accuracy for all four APOs, with AUCs ranging from 0.893 to 0.970 in the testing cohort.
- LightGBM and LR models showed superior performance for specific outcomes, achieving the highest AUCs for LCOS, pneumonia, renal failure, and DVT.
- Shapley Additive Explanations (SHAP) identified mechanical ventilation time as the most significant predictor across all four major APOs.
Conclusions:
- Established ML models accurately predict the risk of major APOs in pediatric congenital heart surgery.
- These models provide reliable interpretations, enabling identification of high-risk contributors and supporting informed clinical decision-making.
- The findings highlight the potential of ML to enhance patient care and outcomes in complex pediatric cardiac surgery.
Background:
Major adverse postoperative outcomes (APOs) can greatly affect mortality, hospital stay, care management and planning, and quality of life. This study aimed to evaluate the performance of five machine learning (ML) algorithms for predicting four major APOs after pediatric congenital heart surgery and their clinically meaningful model interpretations.
Methods:
Between August 2014 and December 2021, 23 000 consecutive pediatric patients receiving congenital heart surgery were enrolled. Based on the split date of 1 January 2019, the authors selected 13 927 participants for the training cohort, and 9073 participants for the testing cohort. Four predefined major APOs including low cardiac output syndrome (LCOS), pneumonia, renal failure, and deep venous thrombosis (DVT) were investigated. Thirty-nine clinical and laboratory features were inputted in five ML models: light gradient boosting machine (LightGBM), logistic regression (LR), support vector machine, random forest, and CatBoost. The performance and interpretations of ML models were evaluated using the area under the receiver operating characteristic curve (AUC) and Shapley Additive Explanations (SHAP).
Results:
In the training cohort, CatBoost algorithms outperformed others with the mean AUCs of 0.908 for LCOS and 0.957 for renal failure, while LightGBM and LR achieved the best mean AUCs of 0.886 for pneumonia and 0.942 for DVT, respectively. In the testing cohort, the best-performing ML model for each major APOs with the following mean AUCs: LCOS (LightGBM), 0.893 (95% CI: 0.884-0.895); pneumonia (LR), 0.929 (95% CI: 0.926-0.931); renal failure (LightGBM), 0.963 (95% CI: 0.947-0.979), and DVT (LightGBM), 0.970 (95% CI: 0.953-0.982). The performance of ML models using only clinical variables was slightly lower than those using combined data, with the mean AUCs of 0.873 for LCOS, 0.894 for pneumonia, 0.953 for renal failure, and 0.933 for DVT. The SHAP showed that mechanical ventilation time was the most important contributor of four major APOs.
Conclusions:
In pediatric congenital heart surgery, the established ML model can accurately predict the risk of four major APOs, providing reliable interpretations for high-risk contributor identification and informed clinical decisions-making.
