The reproducibility of continuous wave Doppler measurements in the assessment of mitral stenosis or mitral prosthetic function: the relative contributions of heart rate, respiration, observer variability and their clinical relevance.

The reproducibility of continuous wave Doppler echocardiographic measurements of transmitral diastolic flow velocity were studied in terms of bias and random error in 40 patients with either mitral stenosis or a Björk-Shiley mitral valve prosthesis. Twenty-seven patients were in sinus rhythm; 13 patients had atrial fibrillation. Intra- and interobserver differences in bias were small for the Doppler parameters studied i.e. early peak velocity (0.6% vs 3.6%), mean diastolic velocity (1.1% vs 8.6%), mean temporal velocity (2.3% vs 14.5%) and pressure half-time (2.7% vs 4.8%). The overall random error of the measurements (in terms of twice the standard deviation) was estimated separately in patients in sinus rhythm and atrial fibrillation: early peak velocity 5.6% and 9.2%, respectively, mean diastolic velocity 9.4% and 22%, mean temporal velocity 8.6% and 19% and pressure half-time 34% and 46%. The relative contributions to the overall random error of observer variation, heart rate dependency and respiratory variation were also studied. Heart rate dependency was demonstrated for both the mean diastolic velocity and the pressure half-time. Respiratory variation was found in the early peak velocity. From the results of this study the number of measurements to reduce the random error of the final average could be determined. Our results indicate that for the measurements in which a respiratory effect is present it is advisable to average the measurements taken over complete respiratory cycles.