ABSTRACT
OBJECTIVES: Major salivary gland ultrasonography (SGUS) is a suitable diagnostic tool in Sjögren's syndrome (SS). We aimed to determine the more representative gland, projection and format most applicable for reproducible image analysis. METHODS: One investigator performed SGUS in patients with SS. Parotid and submandibular glands were examined in longitudinal and transverse planes and evaluated bedside using a simplified scoring system (0-3). Longitudinal and transverse images and videos of all glands were stored and later evaluated/graded by three investigators, at two time-points. Agreement was calculated using intraclass correlation coefficient (ICC). RESULTS: The ICC for static image and video scoring compared to bedside evaluation ranged from 0.131 to 0.882. Average ICC for longitudinal/transverse image was 0.667/0.662, and 0.683/0.510 for longitudinal/transverse video. Interobserver reliability was good to excellent (0.81-0.94). Intraobserver reliability scores ranged from fair to excellent (0.46-0.96). The correlation between image and video evaluations of all modalities and examiners was good to excellent (0.614-0.904). The best mean ICC was found for the longitudinal projection of the left parotid gland (0.861) and the lowest mean ICC was for the transverse projection of the left submandibular gland (0.66). CONCLUSIONS: Our study indicates a trend favouring longitudinal video of the parotid gland as preferred projection, gland and storage format.
