SEARCH

SEARCH BY CITATION

Objective

The aim of this pilot study was to provide groundwork that could be utilized to optimize the global ultrasound (US) assessment of the whole joint for synovial pathologic features in patients with rheumatoid arthritis (RA).

Methods

US images of the second metacarpophalangeal joint in 8 predefined imaging planes, comprising regions that comprehensively capture the synovial pathologic features of the whole joint, were obtained from 30 patients with RA. Twelve experienced sonographers evaluated these images at the level of both the individual image and the whole joint, using a visual analog scale (VAS) to assess pathologic severity. Interrater reproducibility of the VAS scores was evaluated with intraclass correlation coefficients (ICCs), and factors that independently influenced the global assessment of the whole joint were identified using multiple linear regression analysis.

Results

A total of 14,276 VAS scores were analyzed. Interrater reproducibility of any eligible VAS assessment of synovial pathologic features was good (ICC 0.65). US assessment of synovial pathologic features in joints with mild inflammation was less reproducible than that in joints with severe inflammation. Although the most severely affected region in a joint did not always represent the average pathologic severity among the 8 regions, global assessment of the whole joint strongly correlated with assessment of the most severely affected region (P < 0.001). Importantly, the standard, midline imaging plane was not the most influential plane on the global assessment of the whole joint. Assessment of synovial fluid accumulation was not reproducible (ICCs 0.20–0.42) and did not substantially influence the global assessment of synovial inflammation (β = 0.06).

Conclusion

The results of this study provide a unique data set that could be utilized to optimize the global US assessment of synovial pathologic features of the whole joint in patients with RA.