Assessing synovitis in the hands in patients with rheumatoid arthritis by ultrasound: an agreement study exploring the most inflammatory active side from two Norwegian trials

OBJECTIVE: To assess if the right hand, the dominant hand, or the hand with more clinically swollen joints (SwJ) is per se the most inflamed and exhibits the greatest change during treatment and hence preferred for unilateral scoring of synovitis by ultrasound in rheumatoid arthritis (RA) patients. METHODS: Using data from two previously published Norwegian RA patient cohorts initiating treatment, bilateral metacarpophalangeal joint 1–5, proximal phalangeal joint 2+3, and wrists were evaluated by ultrasound. Using a 0–3 scoring system a grey-scale (GS), power Doppler (PD) and global synovitis score (GLOESS) was calculated for each hand (0–30). For precision, a difference of < ± 3 in sum score was pre-specified as indicating clinically insignificant difference in inflammatory activity for all three scores. RESULTS: Four hundred thirty-seven RA patients were included. Baseline ultrasound inflammation was statistically significantly higher in hands with more vs fewer SwJ ([mean difference, 95%CI] GS sum score 2.21[1.30 to 3.12], PD sum score 1.70 [0.94 to 2.47] and GLOESS 2.31[1.36 to 3.26]) and also exhibited significantly more change for all sum scores at 3 months follow-up (GS sum score 1.34 [0.60 to 2.08], PD sum score 1.17 [0.44 to 1.91], and GLOESS 1.43 [0.63 to 2.22]). No such differences were found between the dominant and the non-dominant or the right and the left hands at any time points. CONCLUSION: The hand with clinically more SwJ is statistically more inflammatory active according to GS, Doppler, and GLOESS sum scores, exhibits a change during treatment, and is potentially the best choice for unilateral scoring systems. ELECTRONIC SUPPLEMENTARY MATERIAL: The online version of this article (10.1186/s13075-019-1930-y) contains supplementary material, which is available to authorized users.