Improved pediatric ICU mortality prediction for respiratory diseases: machine learning and data subdivision insights

Sep 17, 2026

Pediatric mortality in intensive care settings requires stronger clinical preparedness, particularly for respiratory-related admissions that form a large share of pediatric ICU cases. This study uses data from 1,188 patients to build and evaluate machine-learning models for predicting pediatric ICU mortality, including class-balancing strategies. A CatBoost model reaches an AUC of 72.22%, while stacking ensemble achieves 60.59%. A proposed data subdivision technique improves performance to an AUC of 85.2% and accuracy of 89.32%.

LanguageEnglish

CategoryResearch & Report

FormatPDF

Pages16

Reading Time40 min

AccessFree to download

Abstract

Introduction

Pediatric respiratory diseases and PICU mortality

Clinical and economic impact

Methods

Prithula et al. Respiratory Research (2024) 25:216 [https://doi.org/10.1186/s12931-024-02753-x](https://doi.org/10.1186/s12931-024-02753-x) Respiratory Research RESEARCH Open Access Improved pediatric ICU mortality prediction for respiratory diseases: machine learning and data subdivision insights Johayra Prithula1, Muhammad E. H. Chowdhury2*, Muhammad Salman Khan2, Khalid Al‑Ansari3, Susu M. Zughaier4, Khandaker Reajul Islam5 and Abdulrahman Alqahtani6,7 Abstract The growing concern of pediatric mortality demands heightened preparedness in clinical settings, especially within intensive care units (ICUs) . As respiratory‑related admissions account for a substantial portion of pediatric ill‑ nesses, there is a pressing need to predict ICU mortality in these cases. This study based on data from 1188 patients, addresses this imperative using machine learning techniques and investigating different class balancing methods for pediatric ICU mortality prediction. This study employs the publicly accessible“Paediatric Intensive Care database”to train, validate, and test a machine learning model for predicting pediatric patient mortality. Features were ranked using three machine learning feature selection techniques, namely Random Forest, Extra Trees, and XGBoost, result‑ ing in the selection of 16 critical features from a total of 105 features. Ten machine learning models and ensemble techniques are used to make accurate mortality predictions. To tackle the inherent class imbalance in the dataset, we applied a unique data partitioning technique to enhance the model’s alignment with the data distribution. The CatBoost machine learning model achieved an area under the curve (AUC) of 72 . 22%, while the stacking ensemble model yielded an AUC of 60 . 59% for mortality prediction. The proposed subdivision technique, on the other hand, provides a significant improvement in performance metrics, with an AUC of 85 . 2% and an accuracy of 89 . 32% . These findings emphasize the potential of machine learning in enhancing pediatric mortality prediction and inform strate‑ gies for improved ICU readiness. Keywords Pediatric mortality, Respiratory diseases, Pediatric ICU, Mortality prediction, Early recognition, Machine learning *Correspondence: Muhammad E. H. Chowdhury [mchowdhury@qu.edu.qa](mchowdhury@qu.edu.qa) 1 Department of Electrical and Electronics Engineering, University of Dhaka, Dhaka 1000, Bangladesh 2 Department of Electrical Engineering, Qatar University, 2713 Doha, Qatar 3 Emergency Medicine Department, Sidra Medicine, Doha, Qatar 4 Department of Basic Medical Sciences, College of Medicine, Qatar University, 2713 Doha, Qatar 5 Department of Physiology, Faculty of Medicine, University Kebangsaan Malaysia, 56000 Kuala Lumpur, Malaysia 6 Department of Biomedical Technology, College of Applied Medical Sciences in Al‑Kharj, Prince Sattam Bin Abdulaziz University, 11942 Al‑Kharj, Saudi Arabia 7 Department of Medical Equipment Technology, College of Applied, Medical Science, Majmaah University, 11952 Majmaah, Saudi Arabia © The Author(s) 2024. Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit [http://creativecommons.org/licenses/by/4.0/](http://creativecommons.org/lice

FAQs

What problem does the study address?

It addresses the need to predict pediatric ICU mortality for respiratory disease admissions in order to support clinical preparedness and targeted intervention.

What dataset and modeling approach are used?

The study trains, validates, and tests models using publicly accessible Paediatric Intensive Care database data from 1,188 patients, applying multiple machine-learning models and ensemble techniques.

How does the proposed data subdivision method affect results?

The subdivision technique improves performance metrics substantially, reporting an AUC of 85.2% and an accuracy of 89.32% for mortality prediction.

Sep 17, 2026

Uploaded by Ava Thompson