TY  - JOUR
AU  - Esumi, Ryo
AU  - Funao, Hiroki
AU  - Kawamoto, Eiji
AU  - Sakamoto, Ryota
AU  - Ito-Masui, Asami
AU  - Okuno, Fumito
AU  - Shinkai, Toru
AU  - Hane, Atsuya
AU  - Ikejiri, Kaoru
AU  - Akama, Yuichi
AU  - Gaowa, Arong
AU  - Park, Eun Jeong
AU  - Momosaki, Ryo
AU  - Kaku, Ryuji
AU  - Shimaoka, Motomu
PY  - 2025
DA  - 2025/3/5
TI  - Machine Learning–Based Prediction of Delirium and Risk Factor Identification in Intensive Care Unit Patients With Burns: Retrospective Observational Study
JO  - JMIR Form Res
SP  - e65190
VL  - 9
KW  - burns
KW  - delirium
KW  - intensive care unit
KW  - machine learning
KW  - prediction model
KW  - artificial intelligence
KW  - AI
AB  - Background: The incidence of delirium in patients with burns receiving treatment in the intensive care unit (ICU) is high, reaching up to 77%, and has been associated with increased mortality rates. Therefore, early identification of patients at high risk of delirium onset is essential for improving treatment strategies. Objective: This study aimed to create a machine learning model for predicting delirium in patients with burns during their ICU stay using patient data from the first day of ICU admission and identify predictive factors for ICU delirium in patients with burns. Methods: This study focused on 82 patients with burns aged ≥18 years who were admitted to the ICU at Mie University Hospital for ≥24 hours between January 2015 and June 2023. In total, 70 variables were measured in patients upon ICU admission and used as explanatory variables in the ICU delirium prediction model. Delirium was assessed using the Intensive Care Delirium Screening Checklist every 8 hours after ICU admission. A total of 10 different machine learning methods were used to predict ICU delirium. Multiple receiver operating characteristic curves were plotted for various machine learning models, and the area under the curve (AUC) for each was compared. In addition, the top 15 risk factors contributing to delirium onset were identified using Shapley additive explanations analysis. Results: Among the 10 machine learning models tested, logistic regression (mean AUC 0.906, SD 0.073), support vector machine (mean AUC 0.897, SD 0.056), k-nearest neighbor (mean AUC 0.894, SD 0.060), neural network (mean AUC 0.857, SD 0.058), random forest (mean AUC 0.850, SD 0.074), adaptive boosting (mean AUC 0.832, SD 0.094), gradient boosting machine (mean AUC 0.821, SD 0.074), and naïve Bayes (mean AUC 0.827, SD 0.095) demonstrated the highest accuracy in predicting ICU delirium. Specifically, 24-hour urine output (from ICU admission to 24 hours), oxygen saturation, burn area, total bilirubin level, and intubation upon ICU admission were identified as the major risk factors for delirium onset. In addition, variables, such as the proportion of white blood cell fractions, including monocytes; methemoglobin concentration; and respiratory rate, were identified as important risk factors for ICU delirium. Conclusions: This study demonstrated the ability of machine learning models trained using vital signs and blood data upon ICU admission to predict delirium in patients with burns during their ICU stay. 
SN  - 2561-326X
UR  - https://formative.jmir.org/2025/1/e65190
UR  - https://doi.org/10.2196/65190
UR  - http://www.ncbi.nlm.nih.gov/pubmed/39895101
DO  - 10.2196/65190
ID  - info:doi/10.2196/65190
ER  -