Development and Validation of a Prognostic Model to Predict Late Seizures After Traumatic Brain Injury: A Retrospective Analysis

Abstract

Posttraumatic epilepsy (PTE) is considered to be one of the most severe and enduring outcomes that can arise from traumatic brain injury (TBI). The authors’ study aims to create and authenticate a prognostic model for forecasting the PTE occurrence after TBI. The clinical prognostic model was developed in 475 people who had a TBI history in Nanning using a multivariate logistic regression model. The score in the authors’ prognostic model participants was subjected to external validation from other cities in Guangxi and assessed its performance with the area under the receiver operating characteristic curve (area under the curve), calibration plots, and decision curve analysis. Six variables were selected to establish the nomogram for PTE, including time, Glasgow Coma Scale, location, cranial imaging (midline shift), intracranial infection, and titanium mesh cranioplasty. The area under the curve was found to be 0.860 in the training cohort and 0.735 in the validation cohort, revealing that the nomogram exhibited a satisfactory level of discriminative ability. The calibration plots exhibited a substantial degree of concordance between the prognostic predictions generated by the nomogram and the observed outcomes in both the training and validation groups. In addition, the decision curve analysis demonstrated the clinical utility of the nomogram. The cutoff value for the training cohort was determined to be 0.381, whereas for the validation cohort, it was 0.380. This suggests that patients with a probability >0.381 should be given special consideration. A prognostic nomogram was formulated and verified to aid health care clinicians in assessing the prognosis of patients with PTE.