Objective: The early detection of acute kidney injury (AKI) in patients with traumatic brain injury (TBI) may enable the prompt initiation of preventive strategies and the development of personalized treatment plans, thereby enhancing prognostic outcomes. This study seeks to develop a predictive model for AKI in TBI patients. Methods: An analysis was conducted on data from 409 adult TBI patients treated at our hospital between January 2021 and December 2021, with the outcome being the incidence of AKI. A logistic regression model was constructed, using the least absolute shrinkage and selection operator (LASSO) regression to select optimal variables. The model’s performance was assessed using several metrics, including the Brier score, the area under the receiver operating characteristic curve (AUC), calibration plots, and decision curve analysis (DCA). Results: AKI was identified in 9.0% of the patients. The final model included 5 variables: estimated glomerular filtration rate, a Glasgow Coma Scale score of <6, mechanical ventilation upon admission, hyperuricemia, and serum cystatin C levels exceeding 1.16 mg/L. Upon validation using bootstrapping, the model exhibited robust performance, evidenced by a Brier score of 0.069 (95% CI: 0.066-0.074) and the AUC of 0.809 (95% CI: 0.760-0.823). The calibration plot demonstrated a high degree of concordance between observed outcomes and model predictions. In addition, DCA affirmed the model’s clinical utility. Conclusion: The predictive model proficiently identifies patients at elevated risk for AKI, thereby facilitating timely intervention and management.
Ding et al. (Thu,) studied this question.