Home // eTELEMED 2025, The Seventeenth International Conference on eHealth, Telemedicine, and Social Medicine // View article

Using Explainable Machine Learning for Diabetes Management in Emergency Departments

Authors:
Silas Majyambere
Tony Lindgren
Celestin Twizere
Gerard Nyiringango

Keywords: Explainable Machine Learning; Intensive Care Unit; Diabetes; Length of Stay; SHAP.

Abstract:
Uncontrolled diabetes can lead to severe complications and Intensive Care Unit (ICU) admissions. This study presents an explainable machine learning model using electronic health records to predict ICU admissions and estimate hospital stay duration for diabetic patients. AdaBoost model outperformed other models on ICU admission prediction, while CatBoost exhibited superior performance in estimating ICU length of stays among diabetic patients admitted to the emergency departments. The results demonstrate the potential of explainable machine learning in ICU risk assessment and can aid healthcare providers in early intervention and resource utilization. The clinician and the proposed model agree on the top 25 features identified by Shapley Additive exPlanations (SHAP) methods for predicting ICU admission, but they differ in the ranking of the top five most significant predictors.

Pages: 22 to 29

Copyright: Copyright (c) IARIA, 2025

Publication date: May 18, 2025

Published in: conference

ISSN: 2308-4359

ISBN: 978-1-68558-270-8

Location: Nice, France

Dates: from May 18, 2025 to May 22, 2025