Reproducibility of ultrasound assessment of carotid plaque occurrence, thickness, and morphology - The Tromso Study

Background and Purpose Ultrasonography is increasingly used in vascular research, but there is limited information about the reproducibility of the ultrasound method for screening purposes. In this study the reproducibility of ultrasound assessment of carotid plaque occurrence, thickness, and morphology was examined within the setting of a population health survey. Methods In 1994/1995, 6720 participants in the Tromso Study, Norway, underwent B-mode ultrasound scanning of the right carotid artery. The between-and within-sonographer reproducibility of ultrasound assessment of plaque occurrence and thickness was estimated by repeated scanning of a random sample of 107 participants. The between-and within-sonographer reproducibility of plaque morphology classification (echogenicity, four categories and heterogeneity, two categories) was determined by repeated reading of videotaped images of 119 randomly selected arteries with plaques. Results Between-and within-sonographer agreement on plaque occurrence was substantial with kappa values (95% CI) of 0.72 (0.60 to 0.84) and 0.76 (0.63 to 0.89), respectively. Reproducibility of plaque thickness measurements was moderate, with mean absolute differences ranging between 0.25 and 0.55 mm (coefficients of variation between 13.8% and 22.4%). Agreement on plaque morphology classification was high, with kappa values ranging between 0.54 and 0.73. Conclusions Population screening using B-mode ultrasound provides a valuable means for the detection and morphological evaluation of carotid plaques, whereas measurements of plaque thickness are subject to considerable measurement error.