Delirium, an acute brain dysfunction, is a common and serious complication in burn patients. The occurrence of delirium increases the difficulty of patient treatment, is associated with various adverse outcomes, and increases the burden on the patient’s family. Many scholars have studied the factors that cause delirium, but the causes, pathogenesis, and treatment of delirium in burn patients have not been fully revealed. There is no effective pharmacological treatment for delirium, but active preventive measures can effectively reduce the incidence of delirium in burn patients. Therefore, it is necessary to study the relevant factors affecting the occurrence of delirium in burn patients. This study was conducted on December 20, 2021 by searching the PubMed database for a narrative review of published studies. The search strategy included keywords related to “burns,” “delirium,” and “risk factors.” We reviewed the characteristics of delirium occurrence in burn patients and various delirium assessment tools, and summarized the risk factors for the development of delirium in burn patients in terms of personal, clinical, and environmental factors, and we found that although many risk factors act on the development of delirium in burn patients, some of them, such as clinical and environmental factors, are modifiable, suggesting that we can estimate the exposure of burn patients to risk factors by assessing their likelihood of delirium occurring and to make targeted interventions that provide a theoretical basis for the prevention and treatment of burn delirium.

AimsMachine learning‐based identification of key variables and prediction of postoperative delirium in patients with extensive burns.MethodsFive hundred and eighteen patients with extensive burns who underwent surgery were included and randomly divided into a training set, a validation set, and a testing set. Multifactorial logistic regression analysis was used to screen for significant variables. Nine prediction models were constructed in the training and validation sets (80% of dataset). The testing set (20% of dataset) was used to further evaluate the model. The area under the receiver operating curve (AUROC) was used to compare model performance. SHapley Additive exPlanations (SHAP) was used to interpret the best one and to externally validate it in another large tertiary hospital.ResultsSeven variables were used in the development of nine prediction models: physical restraint, diabetes, sex, preoperative hemoglobin, acute physiological and chronic health assessment, time in the Burn Intensive Care Unit and total body surface area. Random Forest (RF) outperformed the other eight models in terms of predictive performance (ROC:84.00%) When external validation was performed, RF performed well (accuracy: 77.12%, sensitivity: 67.74% and specificity: 80.46%).ConclusionThe first machine learning‐based delirium prediction model for patients with extensive burns was successfully developed and validated. High‐risk patients for delirium can be effectively identified and targeted interventions can be made to reduce the incidence of delirium.