Doppler echocardiographic estimation of cardiac output: analysis of temporal variability

The temporal variability of combined cross-sectional and Doppler echocardiographic estimates of cardiac output was studied in 14 normal subjects. In each subject cross-sectional echocardiographs and Doppler velocities were recorded from the aortic, pulmonary and mitral valves. Recordings were repeated after 30-60 min and after 1-3 months to allow estimation of short-term and long-term temporal variability. A components of variance analysis showed that between-subject variability was significantly larger than within-subject variability for all measured and calculated variables. Long-term variability was larger than short-term variability for all variables except the mitral valve area. Calculation of flows from the three measurement sites were equally reproducible and the mean percentage error for CO measurements performed 1-3 months apart ranged from 6.4% to 7.4%. The 95% confidence intervals for Doppler measurements of CO at the three sites ranged from +/- 0.49 l min-1 to +/- 0.56 l min-1, suggesting that the temporal variability of flow measurements using the non-invasive Doppler method is sufficiently small for the method to be useful in serial haemodynamic studies.