Creation and prospective assessment of a machine learning model for predicting vomiting in children undergoing cancer treatment and hematopoietic cell transplantation

By
Adam Paul Yan
Lin Lawrence Guo
Priya Patel
Tal Schechter
Santiago Eduardo Arciniegas
Jiro Inoue
Emily Vettese
Karim Jessa
Bren Cardiff
George A. Tomlinson
L. Lee Dupuis
Lillian Sung
October 31, 2025

Bmc Cancer

Overview

A machine learning model was developed to predict vomiting in pediatric cancer patients undergoing treatment and hematopoietic cell transplantation. The model was evaluated in a prospective silent trial, demonstrating its potential utility in clinical settings.

Background

Vomiting is a prevalent symptom in pediatric cancer and hematopoietic cell transplant patients, significantly affecting their quality of life and leading to increased healthcare costs. Current predictive capabilities for vomiting in this population are limited, necessitating innovative approaches such as machine learning to enhance prediction accuracy and patient care.

Data Highlights

No numerical data provided in the source material.

Key Findings

A machine learning model was developed using electronic health record data to predict vomiting risk within 96 hours of admission.
The study included both retrospective model development and a prospective evaluation phase.
Admissions for various reasons, including chemotherapy and supportive care, were analyzed.
The model aims to integrate seamlessly into clinical workflows through a silent trial approach.
Previous studies have shown the effectiveness of machine learning in predicting nausea and vomiting in adult populations.

Clinical Implications

The development of this machine learning model could significantly improve the ability to predict vomiting in pediatric oncology patients, potentially leading to better management strategies and enhanced patient outcomes. Clinicians may consider integrating such predictive tools into their practice to optimize antiemetic therapy.

Conclusion

The creation and prospective assessment of this machine learning model represent a promising advancement in managing vomiting in pediatric cancer patients. Further validation and integration into clinical practice are essential for maximizing its benefits.