非ST段抬高型心肌梗死辅助诊断模型的构建与评价  

Construction and evaluation of auxiliary diagnostic model for non-ST segment elevation myocardial infarction

作　　者：安尼瓦尔·阿不里孜[1] 秦练[2] 热娜·热合木丁[1] 陈小翠[1] 马翔[1] Anniwaer Abulizi;QIN Lian;Rena Rehemuding;CHEN Xiaocui;MA Xiang(Cardiovascular center of the First Affiliated Hospital of Xinjiang Medical University,Urumqi 830054,China;The First Affiliated Hospital of Shihezi University Medical College,Shihezi Xinjiang 832008,China)

机构地区：[1]新疆医科大学第一附属医院心血管病中心,乌鲁木齐830054 [2]石河子大学医学院第一附属医院,新疆石河子832008

出　　处：《新疆医科大学学报》2022年第12期1403-1409,共7页Journal of Xinjiang Medical University

基　　金：新疆维吾尔自治区自然科学基金(2017D01C346)。

摘　　要：目的 基于胸痛中心临床数据对非ST段抬高型心肌梗死(Non-ST-segment elevation myocardial infarction, NSTEMI)患者的初步诊断进行准确性评价,进一步筛选诊断特征变量构建Logistic回归诊断模型以提升NSTEMI诊断的准确性。方法 纳入胸痛中心非ST段抬高急性冠脉综合征(Acute coronary syndrome,ACS)患者临床数据建立数据库,筛选在24 h内完成冠状动脉造影(Coronary angiography, CAG)并确诊患者的临床数据建立实验数据集,以CAG诊断结果作为衡量标准对NSTEMI患者初步诊断的准确性进行评价,基于实验数据集数据构建Logistic回归诊断模型并对模型性能进行评估。结果 NSTEMI患者初步诊断的灵敏度为88.59%,特异度为90.57%,约登指数为0.79,Kappa值为0.78,ROC曲线的AUC值为0.864。本研究所构建Logistic回归诊断模型的诊断灵敏度为93.07%,特异度为94.21%,约登指数为0.873,Kappa值为0.870,ROC曲线的AUC值为0.924。结论 Logistic回归诊断模型在NSTEMI诊断的准确性、灵敏度以及一致性方面均较胸痛中心临床数据库中NSTEMI患者初步诊断的评价指标有所提升,具有一定的临床应用价值。Objective To evaluate the accuracy of preliminary diagnosis of NSTEMI(Non-ST-segment elevation myocardial infarction, NSTEMI) patients based on the clinical data from chest pain centers. To improve the accuracy of NSTEMI diagnosis by constructing Logistic regression diagnostic model through the selection of diagnostic-specific variables. Methods The clinical data of the patients with non-ST-segment elevation acute coronary syndrome from chest pain center were extracted for analyses. Patients who completed CAG(Coronary angiography, CAG) and confirmed diagnosis within 24 hours were selected, and the accuracy of the preliminary diagnosis of NSTEMI patients were evaluated by using CAG diagnosis results as a standard crateria for measurement. Logistic regression models based on the data of include the patients were constructed to evaluate model performance. Results The sensitivity of the initial diagnosis of NSTEMI was 88. 59 %;the specificity was 89. 44%;the Youden index was 0. 79;the Kappa value was 0. 78;the AUC value of the ROC curve was 0. 864. The diagnostic sensitivity of the constructed Logistic regression model was 93. 07%;the specificity was 94. 21%;the Youden index was 0. 873;the Kappa value was 0. 870, and the AUC value was 0. 924. Conclusions Compared with the evaluation indicators for the initial diagnosis of NSTEMI patients, the Logistic regression model has improved the accuracy, sensitivity and consistency of NSTEMI diagnosis, therefore ithas the certain value for clinical application.

关 键 词：非ST段抬高型心肌梗死 急性冠脉综合征 诊断模型 评价 

分 类 号：R542.2[医药卫生—心血管疾病]