Related Experiment Video
Updated: Nov 10, 2025
Implementation of a Real-Time Psychosis Risk Detection and Alerting System Based on Electronic Health Records using CogStack
Published on: May 15, 2020
Development and external validation of a predictive nomogram model of posttraumatic epilepsy: A retrospective
Xue-Ping Wang1, Jie Zhong2, Ting Lei3
1Department of Neurology, West China Hospital, Sichuan University, No. 37, Guo Xue Xiang, Chengdu, 610041, Sichuan Province, China.
Objective:
We aimed to develop and validate a predictive model of posttraumatic epilepsy (PTE).
Methods:
The training cohort was patients registered at West China Hospital and diagnosed as traumatic brain injury (TBI) between January 1, 2011, and December 31, 2017. On the basis of multivariable cox proportional hazards model using a forward stepwise method, the nomogram was generated. We externally validated this instrument in 834 participants from two independent cohorts to assess its performance.
Results:
The nomogram was built based on the results of multivariable cox proportional hazards regression analysis of 1301patients from West China Hospital. The prevalence of PTE was 12.8% (95% confidence interval [CI], 10.9-14.6%) in training cohort, 10.5% (95% CI, 7.5-13.4%) in the testing 1 cohort, and 6.1% (95% CI, 3.7-8.4%) in the testing 2 cohort. 7 independent predictors of PTE composed the nomogram (sex, time of loss of consciousness, subdural hemorrhage, contusion sites, early posttraumatic seizures, TBI severity, and treatment). The C-index was 0.846 (95% CI, 0.817-0.876), and the corresponding sensitivity and specificity were 0.867 and 0.738. External validations showed good discrimination in overall testing cohorts with a C-index of 0.895 (95% CI, 0.859-0.930), in the testing 1 cohort (C-index 0.897, 95% CI, 0.855-0.938) and testing 2 cohort (C-index, 0.883, 95% CI, 0.814-0.952). Calibration of this model was also good since the calibration plots were close to the ideal line.
Conclusions:
This nomogram was developed and validated in a large cohort for individualized prediction of PTE, which can identify individuals at high risk of epilepsy and help us find preventive drugs based on these targeted population.
