缺血性脑卒中患者出血转化的危险因素分析及预测模型研究

发布时间：2018-04-11 03:08

  本文选题：缺血性脑卒中 + 出血转化　； 参考：《青岛大学》2017年硕士论文

【摘要】：目的通过对缺血性脑卒中患者发生出血转化(Hemorrhagic transformation,HT)的相关危险因素进行统计学分析,建立同样适用于非溶栓患者的预测模型,以期为临床医生在筛选HT高危患者时提供一定的参考依据。方法筛选2015年1月-2016年9月青岛大学附属医院神经内科收治的缺血性脑卒中患者,纳入符合条件的327例,对患者的卒中病因、临床表现、一般资料等进行回顾分析。采用Logistic回归分析等方法筛选出HT的危险因素,建立新的预测模型,通过绘制受试者工作特征曲线(receiver operating characteristic curve,ROC)获得评分界值,大于评分界值的为HT高危组,低于评分界值的为HT低危组。结果筛选符合条件的急性缺血性脑卒中患者,根据患者CT、MRI等影像学复查结果,分为出血转化组(HT组)和非出血转化组(NHT组)。与NHT组相比,HT组房颤病史、收缩压、血糖、尿蛋白、白细胞数(WBC)、凝血酶原时间(PT-S)、国际标准化比值(INR)、溶栓治疗、大面积梗死、美国国立卫生研究院卒中量表(NIHSS)评分≥10分水平显著偏高,血小板数、胆固醇、阿司匹林治疗、氯吡格雷治疗水平显著偏低,差异均有统计学意义(P0.05)。HT组与NHT组比较,入院TOAST分型差异有统计学意义(P0.05)。对收缩压组间进行逐层比较,140-159mm Hg、160-179 mm Hg组间差异有统计学意义,余逐层比较无统计学差异。对血糖组间进行逐层比较,7.0mmol/L、7.0-11.0mmol/L组间差异有统计学意义,余逐层比较无统计学差异。对收缩压(160mm Hg,≥160mm Hg)、血糖(7.0mmol/L、≥7.0mmol/L)进行再分组,带入回归分析。排除人为因素影响过大的阿司匹林、氯吡格雷治疗因素,以及存在交叉的TAOST分型(与房颤病史)、PT-S(与INR)影响因素。Logistic回归分析结果示:房颤病史(OR=4.94,95%CI:1.879~12.988,p=0.001)、收缩压≥160mm Hg(OR=2.762,95%CI:1.175~6.494,p=0.02)、大面积脑梗死(OR=3.796,95%CI:1.495~9.635,p=0.05)、NIHSS评分≥10分(OR=5.665,95%CI:2.289~14.021,p0.05)、溶栓(OR=3.086,95%CI:4.981~96.257,p0.05)为HT的独立危险因素。根据Logistic回归分析结果建立新的预测模型:Logit P=-3.588+1.579×X1+1.016×X2+1.334×X3+1.734×X4+3.086×X5。Logit P值越大,HT的风险越高,所建ROC曲线下的面积(AUC)为0.912(95%CI:0.873~0.950,P0.001),预测模型诊断价值较高。综合判断各点灵敏度加特异度之和,取youden指数(敏感度+特异度-1)最大时所对应的最佳分界点为界值(cut-off值),界值为-2,此时灵敏度88.33%,特异度为82.02%。结论1.房颤病史、收缩压≥160mm Hg、大面积脑梗死、NIHSS评分≥10分、溶栓治疗为HT的独立危险因素。2.当评分大于-2时,HT的发生的风险显著增高。
[Abstract]:Objective to establish a predictive model for non-thrombolytic patients by statistically analyzing the risk factors associated with hemorrhagic transformation in patients with ischemic stroke.In order to provide some reference for clinicians in screening HT high-risk patients.Methods from January 2015 to September 2016, 327 patients with ischemic stroke were selected from Department of Neurology, affiliated Hospital of Qingdao University. The etiology, clinical manifestation and general data of stroke were analyzed retrospectively.The risk factors of HT were screened by Logistic regression analysis, and a new prediction model was established. By drawing the operating characteristic curve of receiver operating characteristic curveveroc, the critical value of HT was obtained, and the high risk group of HT was found to be higher than the threshold of HT.The group with low risk of HT was lower than the score limit.Results the patients with acute ischemic stroke were selected and divided into two groups according to the imaging findings such as CT MRI. The patients were divided into two groups: Hemorrhage transformation group (HT group) and non-hemorrhagic conversion group (NHT group).Compared with NHT group, atrial fibrillation history, systolic blood pressure, blood glucose, urine protein, white blood cell count, prothrombin time (PT-SN), international standardized ratio (INRN), thrombolytic therapy, large area infarction were observed in HT group.NIH stroke scale score 鈮，

本文编号：1734137