Reproducibility of linear cardiac output measurement by Doppler ultrasound alone

Inclusion of a pig aorta in an artificial circulation with pulsed blood flow allowed correlation of minute distance, measured in the aorta by Doppler ultrasound, and absolute blood flow, measured by timed blood-volume collection. The correlation coefficient was 0.99 with a standard error of prediction that was 5.4% of the minute distance predicted at a standard flow rate of 5 litres per minute. The horizontal distance between 95% confidence limits for a single prediction expressed as a percentage of 5 litres per minute was 33%, and this corresponded to the range of flow rates of 1.65 litres per minute that could give rise to the same measurement. In 142 patients duplicate measurements of minute distance were made with repositioning of the ultrasound transducer between recordings. The mean difference between paired readings, expressed as a percentage of the average (SD) of each pair was 5.4 (4.7)%. Thus, the non-invasive measurement of linear cardiac output by Doppler ultrasound is similarly reproducible in vitro and in vivo and compares favourably with the measurement of volumetric cardiac output by thermodilution.