JMIR Publications

ABSTRACT

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 to identify predictive factors for ICU delirium in patients with burns.

Methods:

This study focused on 82 patients with burns aged ≥18 years admitted to the ICU at Mie University Hospital for 24 h or more between January 2015 and June 2023. Seventy 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 h after ICU admission. Ten different machine-learning methods were employed to predict ICU delirium. Multiple receiver operating characteristic curves were plotted for various machine learning models, and the area under each curve (AUC) was compared. Additionally, Shapley Additive exPlanations (SHAP) analysis was used to identify the top 15 risk factors contributing to delirium onset in each model.

Results:

In this study, ICU delirium occurred in 32 out of 82 patients, representing 39% of the cohort. Key factors associated with the development of ICU delirium in burn patients include advanced age (77.0 years vs. 60.5 years; P<0.001), a higher intubation rate (1.0 vs. 0.0; P<0.001), larger burn area (16.0% vs. 8.0%; P=0.007), increased white blood cell count (13,380 cells/μL vs. 8,910 cells/μL; P=0.003), and significantly decreased daily urine output (412.5 mL vs. 1493.0 mL; P<0.001). These findings highlight the importance of these factors in predicting and managing ICU delirium among burn patients. Among the ten machine learning models tested, AdaBoost (AUC: 0.83), gradient boosting machine (AUC: 0.82), support vector machine (AUC: 0.79), logistic regression (AUC: 0.79), and random forest (AUC: 0.79) demonstrated high accuracy in predicting ICU delirium.

Conclusions:

The 24-hour urine output (from ICU admission to 24 hours), SaO2 (oxygen saturation), burn area, total bilirubin level, and intubation upon ICU admission were identified as the major risk factors for the onset of delirium. Additionally, 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. Clinical Trial: Not applicable