Related Experiment Video
Updated: Jul 2, 2025
Author Spotlight: A Neonatal Heterotopic Rat Heart Transplantation Model for the Study of Endothelial-to-Mesenchymal Transition
Published on: July 21, 2023
Using machine learning to predict five-year transplant-free survival among infants with hypoplastic left heart
Andrew H Smith1, Geoffrey M Gray2,3, Awais Ashfaq4
1Division of Cardiac Critical Care Medicine, The Heart Institute, Johns Hopkins All Children's Hospital, 501 6th Avenue South, St. Petersburg, FL, 33701, USA. asmit356@jhmi.edu.
Insights
Machine learning models can predict transplant-free survival in hypoplastic left heart syndrome (HLHS) patients beyond infancy. These models offer valuable insights for families and healthcare providers during staged surgical palliation.
Area of Science:
- Pediatric Cardiology
- Congenital Heart Disease
- Machine Learning in Medicine
Background:
- Hypoplastic left heart syndrome (HLHS) presents significant morbidity and mortality, often managed with palliative surgery.
- Current risk stratification models for HLHS typically rely on survival analyses limited to infancy.
- Predicting long-term outcomes beyond infancy is crucial for informed decision-making in staged palliation.
Purpose of the Study:
- To develop and validate machine learning (ML) models for predicting transplant-free survival (TFS) in HLHS patients.
- To assess the performance of ML models incorporating data from different stages of surgical palliation.
- To provide individualized survival probability predictions for HLHS patients undergoing staged palliation.
Main Methods:
- Utilized data from the Pediatric Heart Network (PHN) Single Ventricle Reconstruction (SVR) trial and its extension study (SVR II) with 5-year follow-up.
- Developed ML-driven models to predict TFS, incrementally incorporating features from pre-Stage 1 palliation (S1P) through Stage 2 palliation (S2P) hospitalization.
- Evaluated model performance using time-dependent area under the curves (td-AUC).
Main Results:
- ML models incorporating features up to S1P hospitalization demonstrated robust predictive performance.
- Models showed time-dependent AUCs exceeding 0.70 through 5 years post-S1P.
- A model using features through S1P hospitalization achieved a td-AUC of 0.838 (95% CI 0.836-0.840).
Conclusions:
- Machine learning offers a powerful tool for individualized prediction of transplant-free survival in HLHS.
- ML models can provide valuable prognostic insights extending years beyond initial staged surgical palliation.
- These predictions can aid families and healthcare providers in managing HLHS during long-term care.
Abstract:
Hypoplastic left heart syndrome (HLHS) is a congenital malformation commonly treated with palliative surgery and is associated with significant morbidity and mortality. Risk stratification models have often relied upon traditional survival analyses or outcomes data failing to extend beyond infancy. Individualized prediction of transplant-free survival (TFS) employing machine learning (ML) based analyses of outcomes beyond infancy may provide further valuable insight for families and healthcare providers along the course of a staged palliation. Data from both the Pediatric Heart Network (PHN) Single Ventricle Reconstruction (SVR) trial and Extension study (SVR II), which extended cohort follow up for five years was used to develop ML-driven models predicting TFS. Models incrementally incorporated features corresponding to successive phases of care, from pre-Stage 1 palliation (S1P) through the stage 2 palliation (S2P) hospitalization. Models trained with features from Pre-S1P, S1P operation, and S1P hospitalization all demonstrated time-dependent area under the curves (td-AUC) beyond 0.70 through 5 years following S1P, with a model incorporating features through S1P hospitalization demonstrating particularly robust performance (td-AUC 0.838 (95% CI 0.836-0.840)). Machine learning may offer a clinically useful alternative means of providing individualized survival probability predictions, years following the staged surgical palliation of hypoplastic left heart syndrome.
