Early prediction of in-hospital complications following acute traumatic brain injury: a retrospective study using multiple machine learning models

Traumatic brain injury (TBI) is frequently complicated by in-hospital events that contribute to prolonged hospitalization and poor neurological outcomes. This study aimed to develop and validate a clinically practical model to predict in-hospital complications in patients with acute TBI. We retrospectively enrolled 186 adult patients with acute TBI admitted to the Neurocritical Care Unit (Neuro-ICU) and developed multiple regression and machine-learning–based models to predict in-hospital complications. Five modeling approaches, logistic regression, Lasso regression, random forest, support vector machine, and decision tree were used to screen predictors and construct prognostic models. Model performance was assessed using discrimination, calibration, and decision-curve analysis. Of the 186 patients enrolled, 136 (73.12%) experienced in-hospital complications. Four admission-level variables, HbA1c, Glasgow Coma Scale (GCS) score, stress-induced hyperglycemia (SIH), and severe extracranial injuries were identified as independent predictors. The final model incorporating these variables showed strong discrimination, reliable calibration, and meaningful clinical utility, with logistic regression performing best and achieving an AUC of 0.952 (95% CI, 0.905–0.992). A simple and practical model incorporating HbA1c, GCS score, SIH, and severe extracranial injuries effectively predicts the risk of in-hospital complications in acute TBI patients. This model may assist clinicians in early risk stratification and clinical decision-making. Further multicenter prospective studies are needed to validate and refine the model.