脑梗死rt-PA溶栓后出血转化危险因素研究及风险预测模型比较

发布时间：2018-07-21 14:54

【摘要】：目的：对SEDAN、SITS-MOST、iScore和GRASPS4个急性脑梗死经静脉重组组织型纤维蛋白溶解酶原激活物（rt-PA）溶栓后出血转化的风险预测模型进行比较并分析各模型所包含的危险因素。方法：收集自2009年1月至2014年3月之间于重庆医科大学附属第二医院诊断为急性脑梗死，并接受静脉rt-PA溶栓治疗的所有患者资料，利用4个预测模型进行评分，分别对各模型所包含的危险因素与是否发生出血转化的相关性进行单因素分析及Logistic回归分析，，用Fisher确切概率法及受试者工作曲线（ROC）检验并比较各模型对溶栓后出血转化的预测能力。结果：单因素分析显示溶栓前头部CT平扫显示早期梗死征象（P0.01）、动脉高密度征（P0.01）及溶栓前血糖（P=0.014）与溶栓后出血转化相关。Logistic回归分析显示溶栓前头部CT平扫显示早期梗死征象（P=0.036）、动脉高密度征（P=0.022）与溶栓后出血转化相关，既往高血压病史可能与溶栓后出血转化相关（P=0.054）。SEDAN模型ROC曲线下面积=0.911，SITS-MOST模型ROC曲线下面积=0.83。结论：SEDAN与SITS-MOST模型能有效预测溶栓后出血转化。
[Abstract]:Aim: to compare the risk prediction models of bleeding transformation after thrombolysis with recombinant tissue plasminogen activator (rt-PA) in 4 acute cerebral infarction patients with SEDAN SITS-MOSTIScore and GRASPS, and to analyze the risk factors involved in these models. Methods: from January 2009 to March 2014, all the patients who were diagnosed as acute cerebral infarction in the second affiliated Hospital of Chongqing Medical University and received intravenous rt-PA thrombolytic therapy were evaluated with four predictive models. Univariate analysis and logistic regression analysis were carried out to analyze the correlation between the risk factors included in each model and the occurrence of hemorrhage transformation. Fisher exact probability method and receiver operating curve (ROC) were used to test and compare the predictive ability of each model for bleeding transformation after thrombolysis. Results: before thrombolysis, CT plain scan showed early infarction sign (P0.01), arterial high density sign (P0.01) and blood glucose before thrombolysis (P0. 014). Logistic regression analysis showed that prethrombolytic head CT plain scan showed early stage. Infarction sign (P0. 036), arterial high density sign (P0. 022) were associated with bleeding and transformation after thrombolytic therapy. The history of hypertension may be related to hemorrhage transformation after thrombolysis (P0. 054). The area under ROC curve of SEDAN model is 0. 911 and the area under ROC curve of SITS-MOST model is 0. 83. Conclusion the model of SITS-MOST and SITS-MOST can effectively predict bleeding transformation after thrombolysis.
【学位授予单位】：重庆医科大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：R743.3

【参考文献】