LAUSR

OBJECTIVE  To assess the inter- and intraobserver reliability of ultrasound of major salivary glands in patients clinically suspected of having primary Sjögren's syndrome (pSS) as well as to assess sources of variation in outcomes of ultrasonographic evaluation. METHODS  80 consecutive outpatients with clinically suspected pSS underwent ultrasound evaluation. The following ultrasound variables of the parotid and submandibular salivary glands were assessed: echogenicity, parenchymal homogeneity, presence of hypoechogenic areas, hyperechogenic reflections and clearness of posterior glandular border, according to the scoring system of Hocevar et al. (total score range: 0 - 48). Images were scored independently by three blinded observers in two sessions. RESULTS  The intraobserver reliability of the total ultrasound score was excellent, with an intraclass correlation (ICC) ranging from 0.89 to 0.96. The interobserver reliability was good to excellent, with ICCs of 0.84 and 0.76 for the total ultrasound score in the two sessions. The kappa value ranged from 0.60 to 0.83 depending on the applied cut-offs (cut-off score ≥ 15 and ≥ 17). Hypoechogenic areas and homogeneity of parotid glands showed the highest interobserver reliability. The median kappa for echogenicity was low. The total ultrasound scores varied more between observers in patients with higher ultrasonographic scores (approximately scores ≥ 20). CONCLUSION  Ultrasound of major salivary glands is a reliable imaging technique for patients with clinically suspected pSS. Discrepancies between observers in assessing the severity of ultrasound findings may interfere with detecting 'true' changes over time. When monitoring the progression of pSS or treatment efficacy, it is advised that a particular patient be scored by the same ultrasonographer at every time point.