Predictive validity of the model for calculating of echocardiographic parameters in healthy patients

Abstract

Aim.Development of a model for calculating of predicted values of key echocardiography (EchoCG) parameters in patients of different ages and staturelweight values.Material and methods.The study included 10604 apparently healthy patients aged from 1 day to 65 years; 5726 (54%) of them are female. In addition to the general clinical study, all patients underwent EchoCG with the measurement of standard indicators as recommended by the American Society of Echocardiography. We measured body surface area (BSA) and selected a regression model, which most adequately links the values of the EchoCG parameters and staturel-weight values.Results.All EchoCG parameters showed a significant correlation with BSA. The patients were divided into four groups to receive more homogeneous cohorts. We have identified newborns and adults. A group of children was additionally divided according to BSA, by less than 0,3 m2 and more than 0,3 m2. The calculated regression equations were reliable in both cases — before and after separation. Comparison of dispersion excesses showed a better dependence among the separated groups. The separation also significantly increased the prediction accuracy.Conclusion.The proposed mathematical models relevantly predict the normal values of variables. The method is well suited for calculating the Z-index of main EchoCG parameters.