0～6岁儿童冠状动脉内径预测模型的构建

Construction of coronary artery diameter prediction model for 0-6 years old children

目的探讨0～6岁健康儿童冠状动脉内径与生长指标(身高、体重、体表面积)的关系,构建校正的冠状动脉内径的预测模型。方法选择2013年3月至2017年4月在我院就诊的心脏结构正常儿童310例,年龄1 d至6岁,采用超声心动图仪测量冠状动脉的内径。计算自变量生长指标与因变量冠状动脉内径的Pearson相关系数,建立两者对数、指数回归模型,检验模型的异方差性以及Z值是否服从正态分布,确定最佳模型方程。结果冠状动脉内径与生长指标显著相关(Pearson相关系数r值范围0.665～0.756,P<0.05)。指数模型的Z值呈正态分布,随机误差项方差相等,确定为最佳模型。结论儿童冠状动脉内径随体表面积增长而增长,采用体表面积校正方法来计算标准化的冠状动脉内径Z值,经过正态性、等方差性检验,指数模型符合统计学要求,可以用来定量评价川崎病儿童冠状动脉病变。

Objective To investigate the regularity of coronary artery diameter and growth indicators （height, weight,body surface area） in 0-6 years old healthy children, and to establish a corrected prediction model for coronary arterydiameter. Methods A total of 310 children with normal heart structure in our hospital from March 2013 to April 2017 wereselected, aged from 1 d to 6 years old, and the internal diameter of coronary artery was measured by echocardiography. ThePearson correlation coefficients of the independent variable growth indicators and the coronary artery diameter of thedependent variable were calculated, and the logarithmic and exponential regression model were established to test theheteroscedasticity of the model and whether the Z value was subordinate to the normal distribution, and the optimal modelequation was determined. Results There was a significant correlation between the diameter of the coronary artery and thegrowth indicators （Pearson correlation coefficient r value range was 0.665-0.756, P〈0.05）. The Z value of the exponentialmodel was normal distribution, and the variance of random error was equal, and which was identified as the best model.Conclusion The coronary artery diameter increases with the increase of body surface area. The normalized Z value of the coronary artery is calculated using the body surface area correction method. The exponential model can meet the statisticalrequirements through normality test and isovariance test and can be used to quantitatively evaluate coronary artery lesionsin children with Kawasaki disease.