Improved estimation of left ventricular hypertrophy

Abstract

When considering the possibility of requiring a patient to take an additional exam, the doctor decides based on the amount of information that this exam will give to him or her about the patient's health and on the costs involved with this procedure. For that reason, it was decided to study the data obtained from a 24-hour ambulatory blood-pressure monitoring (ABPM-24h) in order to extract more information from it than is usually done. An important variable that indicates a well-functioning heart is the left ventricular mass index (LVMI). The test that measures this variable is the echocardiogram. But the costs involved with this exam are high when compared to the costs of performing an ABPM-24h. Chaves proposed two statistical models to approach the problem: a multiple regression model to quantify the LVMI and a logistic regression model to estimate the probability of a person having left ventricular hypertrophy. Chaves said that if other variables, especially the pulse wave velocity, were included in his models, their predictive power could be improved. This article presents the two models for estimating the left ventricular hypertrophy. An estimation of the arterial compliance based on a first-order approximation of the pulse cycle and on the systolic stroke volume is proposed. By including this variable in the models elaborated by Chaves, a substantial improvement in their power is obtained.