Analysis of risk factors and construction of risk prediction model for drug-resistant epilepsy
Objective By analyzing the relevant factors of drug-resistant epilepsy(DRE),to screen out the high-risk factors,and to built the risk prediction model for guiding clinical treatment.Methods The medical records of 404 patients with epilepsy admitted to department of Neurosurgery of Xijing Hospital from January 2015 to December 2019 were collected.According to the definition of DRE,patients were divided into DRE group(n=85)and the drug treatment effective group(n=319).Univariate and multivariate Logistic regression analysis were performed on the relevant factors such as the initial onset and treatment conditions of patients,respectively.According to the results,the risk prediction model of DRE was established and verified.Results Univariate Logistic regression analysis showed that there were no significant differences in gender,neurological dysfunction,cluster attacks,history of cranial infection,EEG abnormalities,and the time from onset to standardized treatment between the two groups(all P＞0.05).Meanwhile,there were statistically significant differences in age,age of first onset,perinatal events,history of febrile convulsion,brain imaging changes,etiological classification,attack type,frequency of initial onset and curative effect after initial treatment between the two groups(all P＜0.05).Multivariate Logistic regression analysis showed that young age of initial onset,brain imaging changes,symptomatic epilepsy and high frequency of initial onset were independent risk factors for DRE(all P＜0.05).The risk prediction model of DRE was successfully constructed and the ROC curve was drawn,in which the area under the training set curve was 0.873 and the area under the verification set curve was 0.851.Both curves showed good clinical consistency,confirming the accuracy of the prediction model.Conclusion Attention and intervention should be paid to epilepsy patients with independent risk factors such as young age of initial onset,brain imaging changes,symptomatic epilepsy and high frequency of initial onset as early as possible to predict and diagnose DRE and improve the prognosis of patients.
drug-resistant epilepsyrisk factorsage of first onsetseizure typerisk prediction model
国家科技期刊平台
NSTL
万方数据
