摘要
收集306例急性心肌梗死(AMI)患者的临床资料及随访1年期间死亡发生情况。原始数据按7∶3的比例随机分为建模和验证组。建模组用LASSO、Logistic回归筛选术后1年死亡的风险预测变量,构建列线图模型,验证组用受试者工作特征曲线(ROC曲线)和Hosmer-Lemeshow偏差性检验对模型进行验证。病变支数、胱抑素C及左室射血分数(LVEF)是术后1年死亡影响因素(P<0.05)。模型预测死亡风险ROC曲线下面积为0.824(0.621~1.000),Hosmer-Lemeshow偏差性检验结果χ^2=6.090(P=0.637),提示列线图模型具良好区分度及校准度,可预测急性心肌梗死患者术后1年死亡风险。
To collect the clinical data of 306 patients with acute myocardial infarction(AMI) and the occurrence of death during 1 year follow-up. The raw data were randomly divided into modeling group and validation group according to the scale of 7 ∶3. The modeling group used the least absolute shrinkage and selection operator(LASSO) regression and Logistic regression to screen the risk predictors of death at 1 year after operation to construct a nomogram model. The validation group used receiver operating characteristic curve(ROC) curve and Hosmer-Lemeshow deviation test to evaluate the prediction efficiency of the nomogram model. The number of lesions, cystatin C, and left ventricular ejection fraction(LVEF) was the influencing factors of death at 1 year after operation in patients with AMI. The area under the ROC curve of the nomogram model for predicting the risk of death was 0.824(0.621~1.000)(P<0.05), and the Hosmer-Lemeshow deviation test result was χ^2=6.090(P=0.637), suggesting that the nomogram model had good discrimination and calibration, and could be used to predict the risk of 1 year postoperative death in patients with AMI.
作者
叶青
马礼坤
张杰
Ye Qing;Ma Likun;Zhang Jie(Dept of Cardiology,Anhui Provincial Hospital Affiliated to Anhui Medical University,Hefei 230001)
出处
《安徽医科大学学报》
CAS
北大核心
2020年第12期1959-1963,共5页
Acta Universitatis Medicinalis Anhui
基金
2018年国家自然科学基金资助项目(编号:81870192)。
关键词
急性心肌梗死
冠状动脉介入术
死亡
影响因素
列线图
acute myocardial infarction
coronary intervention
death
influencing factors
nomogram
