An evaluation of the reproducibility of the measurement of the intima-media thickness of carotid arteries.

The intima-media thickness (IMT) of carotid arteries was demonstrated to be a reliable measure for early stages of atherosclerosis. B-mode ultrasound may be used to measure carotid IMT. The measurements of the IMT of the carotid artery (CA) conducted by different investigators can be comparable and enable the implementation of clinical trial successfully while maintaining a high reproducibility value. The objective of the study was to evaluate the reproducibility of the measurements made by the same investigator on two separate occasions (intraobserver variability) and the reproducibility of the off-line measurements between four sonographers in our laboratory (interobserver variability). The IMT of CA in 25 subjects (15 post stroke and 10 healthy persons) was investigated with the use of high-resolution ultrasonography. The CA subdivided into the common, bulbs and internal segments were scanned twice with a 3-week interval. Additionally three other readers with different levels of experience and skills in ultrasonography were asked to perform the same measurements in duplicate with at least a 3-week interval between. A high concurrence for intraobserver variability was detected with a correlation coefficient ranging from 0.92 to 0.95; p < 0.0001, and maximal bias 0.019 mm. Interobserver variability for all four readers also demonstrated a high correlation coefficient ranging from 0.72 to 0.83; p < 0.0001, and the maximal bias of measurements did not exceed 0.08 mm. The analogue measurements performed by the team demonstrate a reliable reproducibility in terms of the results of morphologic measurements. The differences obtained in the study were less than the error of the method (i.e. 0.1 mm) and should not influence clinical decision-making. Additionally, this study demonstrated that interobserver concurrence increases with the increasing experience of the investigators.