(P188) ASSESSMENT OF VARIABILITY IN THE ECHO-DOPPLER MEASUREMENT OF AORTIC PULSE WAVE VELOCITY

Background: Aortic pulse wave velocity (APWV) is an indirect measure of elastic arterial stiffness and the earliest marker of cardiovascular risk in adults. Although APWV has been measured in mostly cross-sectional studies by many different methods, its variability has not been studied. The purpose of this study was to identify the optimal number of measurements needed for the least variability in the Echo-Doppler measurement of APWV.

METHODS AND RESULTS: Seventeen patients with a dilated aorta and 23 age- and sex-matched healthy controls (10 to 18 years) were evaluated. Four sets of imaging were performed at two different study visits. At the first visit, one sonographer repeated the imaging on two occasions at least 10 minutes apart (short-term variability). A second sonographer repeated the same imaging protocol (inter-sonographer variability). At the second visit one month later, the first sonographer repeated the imaging protocol (medium-term variability). APWV was calculated as aortic length/Doppler transit time in cm/s and measured by two reviewers over 1, 3 and 10 cardiac cycles. Inter-rater (one set of measurements, two reviewers), intra-rater (two sets of measurements, one reviewer), intra-sonographer (short-term) and medium-term test-retest, and inter-sonographer variability were compared using interclass correlation coefficient (ICC) and Bland-Altman plots.
Height, weight, body surface area and systolic blood pressure were no different, but diastolic blood pressure was higher in patients with a dilated aorta compared with controls (p=0.029). ICCs consistently improve when measured from 1 to 3 to 10 cardiac cycles in all cases except for medium-term test-retest variability. The combined group ICCs for 10 cardiac cycles were: inter-sonographer (.823); intra-rater (.978); short-term test-retest (.743); inter-rater (.729); and medium-term test-retest (.723). Bland-Altman plots showed good agreement for each comparison.

Conclusion: Inter- and intra-rater, short- and medium-term test-retest, and inter-sonographer variability for the Echo-Doppler measurement of APWV are optimized when measuring over a greater number of cardiac cycles in healthy children and in patients with a dilated aortopathy. We recommend that at least 10 cardiac cycles are measured to minimize variability of PWV in cross-sectional and longitudinal studies.