Background Measurement of intima-media complex (IMC) using B-mode ultrasonography is nowadays more frequently used in clinical studies. So far all studies dealing with recurrence of the method were based on the comparison of results using the correlation method and on estimation of absolute difference values. In available medical publications we found only one study where the Bland and Altman method, generally considered most optimal for the assessment of the method error was used. The aim of our study was to estimate the reproducibility of manual carotid intima-media thickness (IMT) readings by two different readers, using the method of absolute difference values, Spearman correlation and Bland and Altman method.
Material and methods The study was based on 45 patients of both sexes, aged from 21 to 84 years old. They underwent B-mode ultrasonographic examination of both carotid arteries. IMC was examined on proximal and distal wall of common carotid artery (CCA), at the bifurcation and internal carotid artery (ICA).
Results The smallest mean of the absolute difference values was obtained for measurements from CCA (0.028 mm). Measurements from bifurcation (0.050 mm) and ICA (0.038 mm) had much worse reproducibility. For all measurements at 12 points the mean absolute difference was 0.037 mm. Highly essential Spearman correlation coefficients were obtained for all segments, while using the Bland and Altman method, good reproducibility was obtained only for CCA. The analysis of results among patients with IMT < and ≥ 1.3 mm showed better reproducibility for IMT < 1.3 mm (in Bland and Altman method all results closed in the allowed 5% range), however small number of patients in the group with atherosclerotic plaque (IMT ≥ 1.3 mm) made it impossible to draw reliable conclusions.
Conclusions The best measurements reproducibility is obtained for IMC thickness within CCA.