急性脑梗死患者溶栓后再出血转化风险的预测模型及验证

Prediction model and validation of hemorrhagic transformation after thrombolysis in patients with acute cerebral infarction

目的探讨急性脑梗死(ACI)患者接受溶栓治疗后发生再出血转化的风险因素,明确血清中性粒细胞与淋巴细胞比值(NLR)在预测ACI方面的作用。方法收集2020-03-01—2023-05-01山东大学齐鲁医院德州医院神经内科连续入住的132例ACI患者作为研究对象,按溶栓后再出血转归情况分为出血组与未出血组,并以同期住院的79例ACI患者作对照。在此基础上,采用多因素Logistic回归模型确定出血相关因素,采用R软件构建线性关系图模型,对其进行验证,最后通过训练集合和确认集合的ROC曲线对其有效性进行验证。结果训练集中再出血组53例,未出血组79例,再出血转化率40.15%。单因素和多因素Logistic回归分析显示,患者CT早期梗死面积(OR=3.766,95%CI:1.009~14.066)、基线NIHSS评分(OR=1.298,95%CI:1.010~1.669)以及NLR(OR=3.730,95%CI:1.733~8.030)是ACI患者溶栓后再出血转化的危险因素,溶栓药物剂量(OR=0.755,95%CI:0.666~0.855)是ACI患者溶栓后再出血转化的保护因素。ROC曲线显示血清NLR水平预测ACI患者溶栓后再出血转化的AUC=0.809(95%CI:0.730~0.887)。通过构建列线图模型对风险预测模型进行验证,结果显示训练集与验证集C-index分别为0.947和0.826。ROC曲线分析显示,训练集AUC=0.950(95%CI:0.914~0.987),敏感度和特异度分别为86.1%和92.5%;验证集AUC=0.869(95%CI:0.794~0.944),敏感度和特异度分别为92.5%和64.1%。结论基线NIHSS评分、CT早期梗死面积、NLR以及溶栓药物剂量为ACI患者溶栓后再出血转化风险的影响因素,以上述指标建立的疾病风险预测模型具有良好的效能。

Objective To investigate the risk factors of rebleeding transformation after thrombolysis in patients with acute cerebral infarction(ACI),and to determine the efficacy of serum neutrophil-lymphocyte ratio(NLR)in predicting ACI.Methods In this study,132 patients with ACI who were continuously admitted to Dezhou Hospital,Qilu Hospital,Shandong University from March 1,2020 to May 1,2023 were collected as the research subjects.Based on their rebleeding transformation after thrombolysis,these patients were divided into rebleeding group and non-bleeding group,and 79 ACI patients treated in our hospital during the same period were included as the controls.On this basis,multi-factor Logistic regression model was used to determine the bleeding related factors.The R software was used to build a linear relationship graph model to verify it.Finally,its effectiveness was verified by the ROC curves of the training set and confirming set.Results There were 53 cases in the rebleeding group and 79 cases in the non-bleeding group,and the bleeding conversion rate was 40.15%.Univariate and multivariate Logistic regression analysis showed that baseline NIHSS score(OR=1.298,95%CI:1.010-1.669),CT early infarct size(OR=3.766,95%CI:1.009-14.066)and NLR(OR=3.730,95%CI:1.733-8.030)were risk factors for hemorrhage transformation after thrombolysis in patients with ACI.Thrombolytic drug dose(OR=0.755,95%CI:0.666-0.855)was a protective factor for rebleeding transformation after thrombolysis in ACI patients.ROC curve showed that the AUC of serum NLR level in predicting hemorrhage transformation after thrombolysis in patients with ACI was 0.809(95%CI:0.730-0.887).Subsequently,the risk prediction model was verified by constructing a nomogram model.The results showed that the C-index of the training dataset and the verification dataset were 0.947 and 0.826,respectively.ROC curve showed that the AUC of the training dataset was 0.950(95%CI:0.914-0.987),and the sensitivity and specificity were 86.1%and 92.5%,respectively.The AUC of the validation dataset was 0.869(95%CI:0.794-0.944),and the sensitivity and specificity were 92.5%and 64.1%,respectively.Conclusion Our research demonstrated that baseline NIHSS score,thrombolytic drug dosage,CT early infarct size,and NLR were all influencing factors for the hemorrhage transformation after thrombolysis in patients with ACI.The nomogram model established by the above indicators has a good performance in predicting the conversion of rebleeding.