102例癫痫患者预后的危险因素分析

目的:确定预测癫痫预后相关变量,建立模型判断癫痫的预后。方法:回顾性研究102例随访大于24个月的癫痫患者的各种危险因素,用Microsoft Access2003建立数据库,应用单变量分析和多项Logistic回归分析,判断癫痫患者的预后的危险因素,并应用Kaplan-Meier乘积极限估计法估算随访时间点的发作率。结果:在单因素分析中,三组在性别、起病年龄、随访时间、病程、发作类型、全面强直-阵挛发作、睡眠发作、热惊厥史和癫痫家族史无统计学意义,而在病因、发作频率、EEG背景、EEG放电、影像学表现、单、多药治疗、神经系统体征等方面有统计学差异。多项Logistic回归分析结果:根据单因素的分析结果,对病因分级、单药或联合治疗、脑电图慢波、脑电图放电、发作频率的分级、神经系统体征、影像学改变等7个因素进行多项Logistic回归分析,建立回归模型。结论:病因明确、发作频率多、EEG慢波、放电、影像学异常、多药治疗和神经系统有阳性体征的癫痫患者的预后不佳。

Predict the prognosis through analysis the risk factors in 102 patients with epilepsy

Objective:To identify relevant variables predicting the prognosis of epilepsy patients and develop models to assess the individual prognosis.Method:Retrospectively studying the risk factors of 102 patients followed-by over 24 months.A database was made with Microsoft Access 2003.Through univariate analysis and multivariate Logistic regression analysis,to develop reliable models for the calculation of the individuai prognosis,and Kaplan-Meier survival curve to estimate the rate of occurrence.Result:In univariate analysis,three groups have no statistic difference in sex,initial age,course of diseases,follow-up time,seizure type,onset in sleep,history of febrile convulsions,family history,generalized tonic-clonic seizures,whereas definite etiology,frequency of onset,EEG slow wave,EEG discharges,imaging abnormalities,single or multi-drug therapy,nervous system signs have statistic difference.And applying multivariate Logistic regression analysis to establish regression equation.Conclusion:The patients with definite etiology,excessive frequency of onset,EEG slow wave,EEG epileptiform discharges,imaging abnormalities,multi-drug therapy,nervous system signs had a worse prognosis.