基于平板运动试验指标建立冠心病诊断模型

Construction of the diagnostic model of coronary heart disease on the basis of treadmill exercise testing

目的利用平板运动试验(TET)指标建立冠心病(CHD)预测模型,提高CHD无创诊断的可靠性。方法以3个研究中心经冠脉造影(CAG)证实的CHD患者156例和非CHD者76例为研究对象,分析TET中相关指标的变化。结果 (1)CHD组最大心率(HRmax)、变时指数(CRI)低于非CHD组,而典型胸痛构成比、QRS时限、ST段压低和上抬高于非CHD组,差异有统计学意义(P<0.05或P<0.01)。(2)多因素Logistic回归分析显示:ST段压低(OR=6.885),QRS时限(OR=1.090),典型胸痛(OR=6.356)是预测CHD的独立指标,(P=0.000)。(3)CHD预测模型:Logit(P)=1.929ST段压低(mm)+0.086QRS时限(ms)+1.849典型胸痛-7.891。其预测CHD的受试者工作特征曲线(ROC)下面积是0.880(95%CI:0.833～0.928),大于ST段压低〔0.751(95%CI:0.686～0.816)〕、QRS时限〔0.744(95%CI:0.677～0.811)〕和典型胸痛〔0.709(95%CI:0.639～0.779)〕3个独立指标的任何一个。预测模型、ST段压低、QRS时限和典型胸痛诊断CHD的敏感度分别是0.788、0.673、0.654和0.615,特异度分别是0.855、0.816、0.750和0.803。结论 TET中ST段压低、QRS时限延长和典型胸痛均可作为预测CHD的重要指标,以此3个指标构建的联合预测模型诊断CHD的可靠性优于单一指标。

Objective To establish the diagnostic model of coronary heart disease basis on the multiple index during tread-mill exercise testing, so as to improve the reliability for the non - invasive diagnosis of coronary heart disease. Methods 156 cases with coronary heart disease and 76 cases without coronary heart disease from three centers were included in the study,and coronary heart disease were confirmed by coronary artery angiography examinations. The indexes were analyzed during treadmill exercise testing. Results （ 1 ） The maximal heart rate and chronotropic index of the coronary heart disease groups were lower, and chest pain rates, QRS duration, ST - segment depression, ST - segment elevation were higher than that of the without coronary heart disease groups （ P 〈 0. 05 or P 〈 0. 01 ）. （2） Multivariable logistic regression analysis indi- cated that QRS duration （ OR = 1. 090）, ST - segment depression （ OR = 6. 885） and chest pain were the independent index for predicting coronary heart disease. （3） The model for predicting coronary heart disease ： Logit （P） = 1. 929ST - segment depression（mm） ＋ 0. 086QRS duration（ms） ＋ 1. 849chest pain -7. 891. For diagnosis of coronary heart disease, the area under the ROC curve for models [ 0. 880 （ 95% CI： O. 833 - 0. 928 ） ~ was larger than for ST - segment depression [ 0. 751 （95% CI：O. 686 -0. 816） ~ ,QRS duration[0. 744（95% CI：O. 677 -0. 811 ） ~ and chest pain[0. 709（95% CI：O. 639 - 0. 779 ） 1. The sensitivity of models, ST - segment depression, QRS duration and chest pain for coronary heart disease were 0. 788,0. 673,0. 654 and 0. 615 ,and the specificity were 0. 855,0. 816,0. 750 and 0. 803, respectively. Conclusion The ST - segment depression, QRS duration and chest pain during treadmill exercise testing are the useful index for diagnosis of coronary heart disease, and predictive model including above three indexes appears to be better for diagnosis of coronary heart disease than a single index.