Reliability exercise of ultrasound salivary glands in Sjögren's Disease: an international web training initiative

Abstract

Introduction: Major salivary gland ultrasonography (SGUS) demonstrated its good metric properties as an outcome measure for diagnosing primary Sjögren's disease (SD). The objective was to assess SGUS reliability among sonographers with different levels of experience, using web training.

Methods: Sonographers from expert centers participated in the reliability exercise. Before exercises, training was done by videoconferencing. Reliability of the two most experienced sonographers (MES) was assessed and then compared to other sonographers. Intra-reader and inter-reader reliability of SGUS items were assessed by computing Cohen's κ coefficients.

Results: All sets were read twice by all 14 sonographers within a 4-month interval. Intra-reader reliability of MES was almost perfect for homogeneity, substantial for Outcome Measures in Rheumatology (OMERACT) scoring system (OMERACTss). Among LES (less experienced sonographers), reliability was moderate to almost perfect for homogeneity, fair to moderate for OMERACTss, and fair to almost perfect for binary OMERACTss. Inter-reader reliability between MES was almost perfect for homogeneity, substantial for diagnosis, moderate for OMERACTss, and substantial for binary OMERACTss. Compared to MES, reliabilities of LES were moderate to almost perfect for both homogeneity and diagnosis, only fair to moderate for OMERACTss, but increased in binary OMERACTss.

Conclusions: Videoconferencing training sessions in an international reliability exercise could be an excellent tool to train experienced and less-experienced sonographers. SGUS homogeneity items is useful to distinguish normal from abnormal salivary glands parenchyma independently of diagnosis. Structural damage evaluations by OMERACT scoring system is a new comprehensive score to diagnose patients with SD and could be easily used by sonographers in a binary method.