Reproducibility of Brachial Artery Ultrasound Measurements

Recent findings suggest that systemic artery endothelial function is associated with the preclinical phase of vascular disease and related to traditional atherosclerosis risk factors. Brachial artery diameter changes in response to hyperemia have been proposed recently as a noninvasive tool to assess endothelial function. To evaluate the reproducibility of brachial artery diameter measurements using ultrasound, we studied 12 healthy subjects (eight men and four women, mean age 37 +/- 9 years). An ATL HDI 3000 machine with a 5- to 10-MHz broadband transducer was used to image the right brachial artery at rest, approximately 4 cm above the elbow. Gray scale ultrasound and color Doppler long-axis images were recorded. Brachial arterial outer diameter (i.e., from anterior adventitia to posterior adventitia) and inner diameter (i.e., from anterior lumen-intima interface to posterior lumen-intima interface) were measured in each subject by two observers. An offline analysis system was used to make measurements at end-diastole from four cardiac cycles. Interobserver and intraobserver measurement variabilities (technical error rates) for brachial artery inner diameter were excellent, ranging from 2.5% to 3.8%. However, interobserver technical error rates for outer diameter measurements were significantly greater than those for inner diameter measurements, ranging from 16.3% to 22.1% (P < 0.001), presumably related to the difficulty in accurately defining the adventitial lines. There were no significant differences in interobserver and intraobserver variability for measurements made using gray scale and color Doppler-aided techniques. We conclude that interobserver and intraobserver reproducibility for brachial artery inner diameter measurements made from ultrasound images is excellent.