| Objective: To analyze the clinical risk factors of post-ischemic stroke epilepsy(PISE).At the same time,the value of SeLECT scale and PSEiCARe scale in predicting the occurrence of PISE was compared and evaluated to determine the predictive scale which is more suitable for these patients,and to further improve the clinical workers’ ability to identify PISE in early stage,which manage stroke patients reasonably and efficiently.Methods: In the stroke registration database,patients with ischemic stroke(IS)were hospitalized in the first affiliated Hospital of Hainan Medical University from January2015 to December 2018.The follow-up data of one year after hospital discharge were acquired.98 patients with PISE were included in the PISE group.Meanwhile 98 patients with first IS who were gender matched in the same period and had no epilepsy during the follow-up period were randomly selected as the control group,namely nonPost-ischemic stroke Epilepsy(NPISE)group.A matched case-control study was conducted.It was a retrospective survey that collected the basic information as well as clinical data of the two groups.Using binary Logistic regression analysis statistically summarize and analyze the incidence of PISE correlated with risk factors.Simultaneously the two groups of patients were scored by SeLECT score and PSEiCARe score,and authenticity as well as the best cut-off value of both scales were analyzed separately;Receiver Operating Characterstic(ROC)of the two scales were drawn by using R software,and Area Under the Receiver Operating Characteristic curve(AUC)was calculated to predict the risk of PISE within one year after IS,which was compared and analyzed.Eventually it is aim to assess the predictive value of the both scales for PISE.Results: 1.The first seizures of PISE were more common in general tonic-clonic seizures(53,54.1%)than partial seizures,in which the proportion of focal seizures(45,45.9%)was higher in 19 cases(19,19.4%)of secondary bilateral generalized tonic-clonus focal seizures.2.Taking 65 years old as the boundary,the PISE group was divided into the young and middle-aged group(56.1%)and the old group(43.9%).The general tonic-clonic seizures were mainly in the young and middle-aged group and the old group,respectively accounting for 54.5%(30/55)and 53.5%(23/43).And the proportion of status epilepticus in the young and middle-aged group was higher than in the elderly group,which respectively accounting for 10.9%(6/55)and 2.3%(1/43).3.Univariate analysis showed that early onset seizures(ES),cognitive impairment,pulmonary infection,NIHSS score,cortical involvement,subcortical involvement,middle cerebral artery territory and infarct size were related to PISE.4.The binary logistic regression model with multiple independent variables showed that cortical involvement,early-onset seizures,pulmonary infection and cognitive impairment were independent risk factors for PISE.5.The sensitivity and specificity of PSEiCARe which could predict the risk of PISE in patients with IS were 0.480 and 0.980(cut-off value = 5),respectively.While the sensitivity and specificity of SeLECT scores which could predict the risk of PISE in patients with IS were 0.786 and 0.837(cut-off value = 3.5),respectively.The ROC curve suggested that the AUC values of PSEiCARe and SeLECT were 0.753 and 0.859 respectively,and the Z value was-2.960 by Z test comparing area,the difference was statistically significant.Conclusions: 1.Cortical lesions,with early-onset seizures,pulmonary infection and cognitive impairment were significantly correlated with the occurrence of PISE.2.Both PSEiCARe and SeLECT scores have good predictive value for PISE within one year after cerebral infarction,and SeLECT score is superior to PSEiCARe score which could predict the risk of seizures in patients with cerebral infarction within one year. |
PDF Full Text Request