The effect of variations of pulsed Doppler sampling site on calculation of cardiac output: an experimental study in open-chest dogs.

Abstract

We measured aortic flow by two-dimensional Doppler echocardiography in an open-chest dog model to examine how variations in Doppler sample volume length and position influence aortic hemodynamic flow calculations. Fourteen dogs underwent right-heart bypass, in which venous return from the venae cavae drained by gravity to a reservoir. A variable-speed roller pump returned the blood to the pulmonary artery, fixing left-sided cardiac input and output. Echo Doppler measurements were performed using a 3.5 MHz transducer placed directly on the aortic arch to determine internal aortic cross-sectional area. The transducer was then directed to image the aortic arch for Doppler velocity measurements and the various sampling sites were investigated. Doppler cardiac output could then be determined for each of the various sample volumes over a range of known roller pump settings. Doppler velocity was analyzed using fast Fourier transform spectral analysis. Mean velocity over the cardiac cycle was obtained by planimetry of the area under the Doppler velocity curve with a minicomputer. Doppler-derived determinations of cardiac output achieved a correlation of r = 0.98-0.99 to values obtained by the roller pump over a range of cardiac outputs from 0.75-5 l/min. The standard error of the estimate was 0.21/min. In this laminar flow model, there was no difference between the predictive accuracy of any of the sampling sites over the range of roller pump flows. Our study shows that Doppler velocity measurements can be used to quantify aortic flow over a clinically useful range and that variations of sample length and position did not produce significant differences in calculated flows.