15289-s-bos

2 27 | Validation CHS and PiU Discussion This study shows that the portable CHS generated reliable and reproducible data that were in large agreement with the PiU measurements, irrespective of the observer acquiring the images. Interestingly, the ICC for the intra-observer variability was similar to that of the inter-observer variability suggesting a highly reliable measurement technique. High mean C-IMT values were not associated with high measurement variation suggesting that the measurements are reliable at both low and high C-IMT. These results are in line with an earlier study showing that the fully automated C-IMT- measurements of the CHS were comparable to manual C-IMT measurements(6). The use of manual C-IMT measurements has considerably decreased with the evolvement of automated C-IMT measurement software. However, differences between these automated systems might exist. In our study, the CHS measured the C-IMT significantly smaller than the PiU, although it is questionable whether the mean 0.015 mm difference is clinically relevant. Vanoli et al. also found smaller C-IMT values with the CHS (0.012mm) although in their study this was not statistically significant (6). These results do emphasise caution when comparing C-IMT results obtained with different systems. Carotid images on the PiU, a semiautomated system, must be stored before the C-IMT can bemeasured by the software. Storage enables retrospective evaluation, which is not possible on the CHS. The CHS on the other hand is a fully automated portable system that immediately generates the C-IMT value during image acquisition, these images are also stored but in less quality than those of the PiU. Notably, the acquisition time is shorter on the CHS than on the PiU. Moreover, Aldridge et al. showed that training of nonsonographers on 60 subjects with the CHS was sufficient to give results comparable to an expert sonographer(10). Taken together, the CHS has advantages compared to the PiU for performing carotid ultrasound imaging during outpatient visits and as a high- throughput system in large studies. Whether C-IMT measurements are useful for individual risk assessment remains a matter of debate (11, 12). Recent population studies showed that C-IMT is associated with CVD(8), although C-IMT seems to have no additional value in cardiovascular risk prediction(11, 12). However, the current guidelines of the European Society of Cardiology states that C-IMT-measurements should be considered in asymptomatic adults at moderate risk(13). The discussion about the clinical usefulness of C-IMT is outside the scope of the current study. The presence of carotid plaques and especially plaque volume are strong predictors of