Risk factors and development of a predictive model for post-traumatic cognitive impairment in patients with frontal...

To identify factors associated with post-traumatic cognitive impairment in patients with frontal lobe cerebral contusion and to develop a predictive model to estimate the risk of cognitive dysfunction. This study aims to facilitate early identification of high-risk patients and support the development of individualized clinical interventions.

Cognitive function was assessed using the Mini-Mental State Examination (MMSE) and the Montreal Cognitive Assessment (MoCA) at 1 week, 1 month, and 6 months after injury in patients with frontal lobe cerebral contusion. Overall recovery and functional independence were evaluated using the Barthel Index (BI) and the Extended Glasgow Outcome Scale (GOSE) at 3 and 6 months post-injury. Correlation analyses were performed to explore the associations between overall cognitive performance and clinical variables, including injury severity, anterior skull base fracture, and anosmia. Potential predictors were initially screened using univariate logistic regression analysis, and variables with P < 0.05 were included as candidate predictors. These variables were subsequently incorporated into a Firth logistic regression model to construct a predictive model for cognitive impairment. Model discrimination was evaluated using the receiver operating characteristic (ROC) curve, while calibration was assessed using the Hosmer-Lemeshow goodness-of-fit test and a calibration curve. In addition, decision curve analysis (DCA) was performed to evaluate the clinical utility and net benefit of the model in clinical decision-making.

A total of 206 patients with mild-to-moderate frontal lobe cerebral contusion were included. Univariate logistic regression analysis demonstrated that age, unilateral versus bilateral injury, contusion and hemorrhage volume, presence of anterior skull base fracture, cerebrospinal fluid (CSF) rhinorrhea, anosmia during treatment, and mechanism of injury were significantly associated with the occurrence of cognitive impairment (P < 0.05). Multivariate logistic regression analysis revealed that age, anosmia during treatment, and mechanism of injury were independent predictors of cognitive impairment. A predictive model was subsequently developed based on these variables. ROC analysis showed that the area under the curve (AUC) for predicting cognitive impairment was 0.83 (95% CI: 0.773-0.887, P < 0.05). The optimal cutoff value was 0.369, yielding a sensitivity of 79.8% and a specificity of 73.5%. The calibration curve demonstrated good agreement between predicted probabilities and observed outcomes. Decision curve analysis further indicated that the model provided a favorable net clinical benefit across a range of threshold probabilities.

Age, anosmia during treatment, and mechanism of injury are independent risk factors for cognitive impairment in patients with mild-to-moderate frontal lobe cerebral contusion. The predictive model constructed based on these factors demonstrated good discrimination and acceptable calibration, indicating potential clinical value for risk stratification and early intervention.