PredictED: An Explainable ESI Level Classification and Length of Stay Prediction Using Machine Learning

DOI

10.2991/978-94-6463-684-0_16How to use a DOI?

Keywords

ED; emergency; ESI; LOS; overcrowding; triage; XAI

Abstract

Even before the COVID-19 pandemic, many Emergency Departments (ED) were already dealing with overcrowding issues. This study investigates how machine learning (ML) can help ED triage patients and predict the length of their stay (LOS). Using the MIMIC-IVED dataset, this study’s findings include that Histogram-based Gradient Boosting was the top-performing model for the ESI Level Classification, achieving an AUPRC of 69.48% and an F1-Score of 63.89%. Explainable AI tools, SHAP and LIME, were used to clarify how the models arrived at their predictions. Moreover, Histogram-based Gradient Boosting also excelled in predicting the LOS, with an MAE of 3.458, an MSE of 40.439, and an RMSE of 6.359. This use of ML streamlines the decision-making process and attempts to enhance the accuracy and efficiency of patient care in EDs.

Copyright

© 2025 The Author(s)

Open Access

Open Access This chapter is licensed under the terms of the Creative Commons Attribution-NonCommercial 4.0 International License (http://creativecommons.org/licenses/by-nc/4.0/), which permits any noncommercial 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 license and indicate if changes were made.

Download article (PDF)

TY  - CONF
AU  - Dhone Matthews Calibuyot
AU  - Perlita Gasmen
PY  - 2025
DA  - 2025/04/30
TI  - PredictED: An Explainable ESI Level Classification and Length of Stay Prediction Using Machine Learning
BT  - Proceedings of the  Workshop on Computation: Theory and Practice (WCTP 2024)
PB  - Atlantis Press
SP  - 253
EP  - 263
SN  - 2589-4900
UR  - https://doi.org/10.2991/978-94-6463-684-0_16
DO  - 10.2991/978-94-6463-684-0_16
ID  - Calibuyot2025
ER  -