Noninvasive Doppler determination of cardiac output during submaximal and peak exercise

We compared pulsed Doppler (PD) measurements of stroke volume (SV) and cardiac output (CO) as a function of work load with previously reported values that were obtained by standard invasive methods. Suprasternal notch measurements of Doppler-shifted frequency (delta f) were obtained from the ascending aorta and SV calculated with the Doppler equation and an independent measurement of aortic diameter. Motion artifacts were minimized with the aid of a restraining table cycle ergometer. Signal aliasing was accommodated with manual summation of delta f waveforms. A total of 207 determinations were made in 10 sitting subjects exercising to exhaustion. Linear regression analysis of CO vs. work load was significant (P less than 0.001). The correlation coefficient (r = 0.95) and standard error of estimate value (1.21 1/min) were similar to values from the literature. Absolute values of CO and SV underestimated the literature values across all work loads. Technical reproducibility was assessed by comparing with paired t tests the differences between 65 duplicate serial measurements of CO and SV at rest and exercise. No significant differences (P less than 0.001) were found. We concluded that PD-determined SV and CO are reproducible and correlate linearly with work load in a manner consistent with reported invasive techniques. Thus the PD method appears suitable for use during submaximal and peak exercise.