个体化预测颈动脉易损斑块的列线图模型的建立及验证

Development and validation of a nomogram model for predicting carotid Vulnerable plaque

目的分析急性脑梗死(acute cerebral infarction,ACI)患者颈动脉易损斑块的独立危险因素,建立并验证个体化预测颈动脉易损斑块的列线图模型。方法纳入2014年1月-2017年6月于河北北方学院附属第一医院神经内科诊治的ACI患者128例,收集临床资料。应用单因素及多因素Logistic回归模型,分析ACI患者颈动脉易损斑块的独立危险因素。应用R软件构建预测颈动脉易损斑块的列线图模型,并进行验证。结果多因素Logistic回归分析显示,高血压病(OR=2.966,95%CI:1.267-6.945),2型糖尿病(OR=3.701,95CI%:1.555~8.813)、LDL(OR=1.006,95CI%:1.001-1.011)、TNF-α(OR=1.281,95CI%:1.066-1.540)及MMP-9(OR=2.320,95CI%:1.322-4.070)均为颈动脉易损斑块的独立危险因素(P<0.05)。对列线图模型进行验证,ROC曲线显示该模型预测颈动脉易损斑块的曲线下面积为0.804(95%Cl:0.724~0.884);校准曲线图形为斜率接近于1的直线,Hosmer-Lemeshow拟合优度检验(χ^2=7.560/=0.478)均显示该模型预测颈动脉易损斑块具有良好准确度。结论本研究基于高血压病、2型糖尿病、LDL、TNF-α及MMP-9这5项ACI患者颈动脉易损斑块的独立危险因素,构建的预测颈动脉易损斑块的列线图模型,具有良好的区分度与准确度,可为临床个体化防治ACI提供科学指导。

Objective To determine risk factors of carotid vulnerable plaque in patients with acute cerebral infraction(ACI),and develop a nomogram model to predict the risk of carotid vulnerable plaque.Method A total of 128 ACI patients hospitalized at the department of neurology at the First Affiliated Hospital of Hebei North University from January 2014 to June 2017 were included.Uni-and multivariate logistic regression analyses were performed to identify independent risk factors of carotid vulnerable plaque in ACI patients.A nomogram was developed by R software and validated to predict the risk of carotid vulnerable plaque.Results Multivariate logistic regression analysis revealed that hypertension(OR=2.966,95%CI:1.267-6.945),type 2 diabetes(OR=3.701,95CI%:1.555-8.813),LDL(0R=1.006,95CI%:1.001-1.011),TNF-α(0R=1.281,95CI%:1.066-1.540)and MMP-9(0R=2.320,95CI%:1.322-4.070)were independent risk factors of carotid vulnerable plaque(P<0.05).Validation of the nomogram:ROC curve revealed that the model predicting carotid vulnerable plaque was with an area under the curve 0.804(95%CI:0.724—0.884);the slope of the calibration plot was almost one and the model passed Hosmer-Lemeshow goodness of fit test(χ^2=7.560,P=0.478)which demonstrated that the model was with good accuracy.Conclusions The nomogram predicting carotid vulnerable plaque constructed based on hypertension,type 2 diabetes,LDL,TNF-α,and MMP-9 was with good discrimination and accuracy,which could provide scientific guidance for individualized prevention and treatment of ACI in clinical practice.