Related Experiment Videos
Predicting length of stay in the pediatric intensive care unit at a tertiary center in Saudi Arabia using machine
Mohammed Alowa1, Aqilah Alqassab2, Dina Al-Rumaih2
1Pediatric Critical Care Department, Qatif Health Network, Qatif, Eastern Province, Saudi Arabia.
Insights
Machine learning accurately predicts pediatric intensive care unit (PICU) length of stay (LOS) using admission data. This approach aids in resource allocation and early risk stratification for critically ill children.
Area of Science:
- Pediatric critical care medicine
- Machine learning in healthcare
- Health informatics
Background:
- Prolonged pediatric intensive care unit (PICU) stays increase mortality and healthcare costs.
- Accurate prediction of length of stay (LOS) is crucial for resource allocation and patient management.
- Traditional models struggle with the complexity of pediatric critical illnesses.
Purpose of the Study:
- Develop and validate machine learning (ML) models to predict PICU LOS.
- Utilize admission-time clinical data for prediction.
- Identify key predictors of LOS using explainable AI.
Main Methods:
- Retrospective cohort study of 6,090 PICU admissions (2013-2022).
- LOS categorized into short, intermediate, and prolonged using percentile binning.
- Supervised ML algorithms trained and validated; performance assessed with AUC and SHAP values.
Main Results:
- Light Gradient Boosting Machine and Categorical Boosting models achieved AUCs of 0.826 and 0.832.
- Accurate prediction for short and prolonged stays; moderate for intermediate.
- Key predictors of prolonged stay: early mechanical ventilation, admission source, post-operative status, physiological instability, comorbidity burden.
Conclusions:
- ML models reliably classify PICU LOS using admission data, especially for extreme durations.
- Explainable AI identifies critical predictors for prolonged stays.
- This data-driven approach supports early risk stratification and operational decisions in pediatric critical care.
Background:
Prolonged stay in pediatric intensive care units (PICUs) is associated with increased mortality risk, elevated healthcare costs, and diminished critical care capacity. Accurate early prediction of length of stay (LOS) may facilitate resource allocation, discharge planning, and family counseling. Traditional regression-based models have demonstrated limited performance because of the complex, non-linear nature of pediatric critical illnesses.
Objectives:
To develop and internally validate machine-learning models that predict PICU LOS using admission-time clinical data and to identify key predictors using explainable artificial intelligence techniques.
Methods:
This retrospective cohort study included all eligible PICU admissions at a tertiary center in Saudi Arabia (2013-2022). LOS was categorized into short, intermediate, and prolonged stay using percentile-based binning. Multiple supervised machine learning algorithms were trained on a stratified set with cross-validation hyperparameter tuning and evaluated on an independent held-out test set. Performance was assessed using accuracy and micro-averaged multiclass area under the curve and interpretability via SHapley Additive exPlanations.
Results:
Data from 6,090 admissions were analyzed. The Light Gradient Boosting Machine and Categorical Boosting models demonstrated the best performance, achieving micro-averaged multiclass areas under the curve of 0.826 and 0.832, respectively. Discrimination was strongest for short- and prolonged-stay categories, with lower performance for intermediate stays. Key predictors of prolonged stay included early mechanical ventilation, admission source, post-operative status, physiological instability, and comorbidity burden.
Conclusions:
Machine-learning models using admission-time data can reliably classify PICU LOS, particularly at the extremes of stay duration. This explainable, data-driven approach may support early risk stratification and inform operational decision-making in pediatric critical care.