汇集队列方程预测非房颤性缺血性脑卒中和短暂性脑缺血发作患者复发风险和预后的验证研究

发布时间：2018-05-11 14:51

本文选题：缺血性卒中 + 汇集队列方程　；参考：《首都医科大学》2017年硕士论文

【摘要】：目的应用中国国家卒中登记数据库(China National Stroke Registry,CNSR)对汇集队列方程(Pooled Cohort Risk Equation,PCRE)预测非房颤性缺血性脑卒中(Ischemic Stroke,IS)和短暂性脑缺血发作(Transient Ischemic Attack,TIA)患者的复发风险进行外部验证,以明确该模型是否适用于中国缺血性卒中二级预防人群。同时探讨PCRE、Essen评分(Essen Stroke Risk Score,ESRS)、卒中预后工具II(Stroke Prognostic Instrument II,SPI-II)对非房颤性IS和TIA患者不良预后结局评估的价值。材料与方法在CNSR数据库中筛选得到符合本研究纳入和排除标准的子数据库。CNSR研究是一项全国多中心、前瞻性、住院登记研究,从2007年9月至2008年8月入组了全国132家医院的住院患者(年龄18岁,发病14天内)。本研究的主要终点事件是随访3个月内、12个月内出现卒中复发。次要终点事件是随访3个月内、12个月内出现预后不良和死亡。计算模型各风险分层终点事件发生率,比较终点事件发生率变化趋势。通过建立接受者工作曲线,并计算曲线下面积(the Area Under the receiver operating Curve,AUC),比较PCRE与ESRE、SPI-II的区分度。AUC越接近1,模型的辨识力越高;AUC为0.5提示模型的辨识力差,AUC在0.5~0.7时有较低辨识力,AUC在0.7~0.9时有一定的辨识力,AUC在0.9以上时有较高的辨识力。以上统计均采用双侧检验,以P0.05为差异具有统计学意义。统计分析软件采用SAS软件(版本9.1)结果1.PCRE对非房颤性IS和TIA患者3个月内、12个月内累计卒中复发风险的预测PCRE-WH(White or other)对IS和TIA患者3个月、12个月累计卒中复发的AUC值分别为0.541和0.537。PCRE-AA(African American)对IS和TIA患者3个月、12个月累计卒中复发的AUC值分别为0.539和0.537。2.PCRE、ESRS、SPI-II对非房颤性IS和TIA患者3个月、12个月预后不良和累计死亡风险的预测。PCRE-WH对IS和TIA患者3个月、12个月预后不良的AUC值分别为0.560和0.579。PCRE-AA对IS和TIA患者3个月、12个月预后不良的AUC值分别为0.536和0.551。ESRS对IS和TIA患者3个月、12个月预后不良的AUC值分别为0.578和0.630。SPI-II对IS和TIA患者3个月、12个月预后不良的AUC值分别为0.626和0.649。PCRE-WH对IS和TIA患者3个月、12个月累计死亡的AUC值分别为0.546和0.559。PCRE-AA对IS和TIA患者3个月、12个月累计死亡的AUC值分别为0.530和0.538。ESRS对IS和TIA患者3个月、12个月累计死亡的AUC值分别为0.596和0.600。SPI-II对IS和TIA患者3个月、12个月累计死亡的AUC值分别为0.610和0.622。结论汇集队列方程对我国非房颤性IS和TIA患者3个月和12个月卒中复发风险有较低预测能力。汇集队列方程、Essen评分、卒中预后工具II对我国非房颤性IS和TIA患者3个月、12个月预后不良和累计死亡风险有较低预测能力。
[Abstract]:Objective to evaluate the risk of recurrence in patients with nonatrial fibrillation ischemic stroke (Ischemic stroke) and transient Ischemic ischemic stroke (TIA) by using China National Stroke Registry National Stroke. To determine whether the model is suitable for secondary prevention of ischemic stroke in China. At the same time, to evaluate the value of Essen Stroke Risk Scoreen Scoreen and II(Stroke Prognostic Instrument II SPI-III in evaluating the adverse outcome of non atrial fibrillation in patients with non atrial fibrillation is and TIA. Materials and methods the subdatabase that met the criteria for inclusion and exclusion of this study was screened from the CNSR database. The study was a multi-center, prospective, hospital registration study across the country. From September 2007 to August 2008, the hospitalized patients (age 18 years, within 14 days of onset) in 132 hospitals throughout the country. The main endpoint events in this study were stroke recurrence within 12 months after 3 months follow-up. Secondary endpoint events were 3 months of follow-up and 12 months of poor prognosis and death. The incidence of endpoint events in each risk stratification model was calculated and the trend of endpoint event incidence was compared. By creating the receiver's work curve, The area under the curve is calculated and the area under the curve is calculated. The closer the distinction between PCRE and PCRE SPI-II is, the higher the identification power of the model is, and the higher the AUC is, the lower the discernability of the model is at 0.50.AUC is lower than that of AUC at 0.70.0.AUC has a certain degree of identification when AUC is 0.70. 9. When AUC is more than 0.9, it has higher identification power. All the above statistics used bilateral test, with P0.05 as the difference was statistically significant. Statistical analysis software using SAS software (version 9.1) results 1.PCRE predicted the cumulative risk of stroke recurrence in patients with non-AF is and TIA within 3 months and 12 months. Values of 0.541 and 0.537.PCRE-AA(African American.PCRE-WH for patients with is and TIA for 3 months and 12 months for cumulative stroke recurrence were 0.539 and 0.537.2.PCREESRS SPI-II for 3 months and 12 months of poor prognosis and cumulative risk of death in patients with non-AF is and TIA respectively .PCRE-WH for is and TIA The AUC values of 3 months and 12 months of poor prognosis in patients with is and TIA were 0.560 and 3 months, respectively; the AUC values of 12 months of poor prognosis of patients with is and TIA were 0.536 and 0.551.ESRS, respectively. The AUC values of 12 months of poor prognosis were 0.578 and 0.578, respectively. The AUC value of 0.630.SPI-II for is and TIA patients was 3 months, the AUC value for 12 months was 0.626 and 0.649.PCRE-WH for is and TIA patients was 3 months, the AUC value of cumulative death for 12 months was 0.546 and 0.559.PCRE-AA for is and TIA patients was 3 months, and AUC value for 12 months cumulative death. The AUC value of cumulative death in 12 months was 0.596 and that of 0.600.SPI-II in is and TIA was 0.596, respectively, and the AUC value of cumulative death in 12 months was 0.610 and 0.622, respectively. Conclusion the pooled cohort equation has low predictive ability for stroke recurrence in patients with non atrial fibrillation is and TIA at 3 and 12 months in China. According to the cohort equation and Essen score, Stroke Prognostic tool II has a lower predictive power for poor prognosis and cumulative death risk in patients with non-AF is and TIA for 3 months and 12 months in China.
【学位授予单位】：首都医科大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：R743.3

【相似文献】