1. The reproducibility of measurements of the arterial wall thickness in both the carotid and femoral artery was investigated by means of high-resolution B-mode ultrasonography. For this purpose, subjects with normal and increased intima-media thickness were selected. Images were stored on an optical disk and were analysed with a semi-automatic software program by two readers. Individuals were scanned twice by two independent observers.
2. Measurements were performed of the far and near wall of the common carotid artery and bulbous in 30 healthy subjects and 19 patients known to have an increased intima-media thickness. Far-wall measurements were made of the internal carotid artery on both sides and common femoral artery on the right side only.
3. In healthy subjects the mean within-observer coefficient of variation was 1.8% and 3.0% for the far wall in the common carotid artery on the right side and left side, respectively. For the near wall the mean coefficient of variation of the common carotid artery was 2.8% on the right and 3.4% on the left side. The mean coefficient of variation was less than 4% for both far and near wall in the bulbous and far wall in the internal carotid artery. Even in patients with increased intima-media thickness the mean coefficient of variation of each segement was less than 4.5%. In the control subjects the between-observer coefficient of variation of the common carotid artery was 2.8% and 5.1% for the far wall on the right and left side, respectively, and 3.4% and 4.2% for the near wall on the right and left side. In healthy subjects a mean difference of 0.002 mm within observers was found in the right far-wall common carotid artery, with limits of agreement of −0.048 to 0.052 mm. The coefficient of repeatability was 0.050 mm. For patients with increased intima-media thickness the mean difference in this segment was −0.006 mm (−0.094 to 0.082) with a coefficient of repeatability of 0.088 mm. For the near wall in the common carotid artery and far and near wall in the bulbous and internal carotid artery the mean differences were larger, but were all below 0.1 mm. The differences and limits of agreements increased between observers. In patients the between-observer mean difference of the far wall of the common carotid artery was −0.055 mm (−0.255 to 0.145). For the common femoral artery of normal control subjects the within- and between-observer mean differences were 0.005 mm (−0.119 to 0.129) and 0.015 mm (−0.081 to 0.111), respectively.
4. In conclusion, the reproducibility of intima-media thickness measurements in the common carotid artery is reliable, even in patients with increased artery wall thickness. Also in other segments prone to atherosclerosis, such as the bulbous, internal carotid artery and common femoral artery, a good reproducibility was found. To obtain good reproducibility it is highly recommended to use the same ultrasonographer to scan patients in follow-up studies.