LAUSR

Background Tender and swollen joint counts are part of rheumatoid arthritis (RA) disease activity assessments. While subclinical synovitis is now a well-known entity, the relationship between tender and swollen joints and ultrasound (US) detected inflammation has not been well explored. Objectives To compare US detected inflammation (synovitis and/or tenosynovitis) with joint swelling and/or tenderness of the wrist, an important joint in RA. Tendons are included as tenosynovitis on US can be mistaken for joint involvement clinically. Methods Wrist outcome groups (Groups) 1–4 were identified: 1=S0 T0 (not swollen, not tender); 2=S0 T1 (not swollen but tender; 3=S1 T0 (swollen but not tender); 4=S1 T1 (swollen and tender). Power Doppler (PD) and grey-scale (GS) US were used to grade (a) synovitis semi-quantitatively (0–3) at the following joint recesses: dorsal radiocarpal and intercarpal, dorsal ulnocarpal and volar radiocarpal, and (b) tenosynovitis dichotomously (0=no, 1=yes) at the following tendon sites: extensor digitorum, extensor carpi ulnaris, and flexor digitorum. Scores on each wrist consisted of a PD score, a GS score and a combined (PD +GS) US (CUS) score. Positivity (+ve) for PD, GS and CUS scores was analysed using a generalised linear repeated measures mixed model with binary distribution and logit link. Scores were analysed using a general linear repeated measures mixed model assuming Gaussian errors. In both analyses, patients were modelled as random effects, and wrist (R/L) and follow up visit (baseline, 3 months) as fixed effects. Pairwise comparisons on +ve and scores were carried out among the 4 groups in the context of the models. P-values were not adjusted for multiple comparisons. Results 122 wrist assessments resulted from 32 RA subjects (87.5% female; 78.1% Chinese; mean (SD) disease duration of 42.8 (52.9) months) who either started or escalated on systemic corticosteroids and DMARDs. All subjects were assessed at baseline and 29 at 3 months. Significant differences among Group scores were: 4 vs 1 (PD, p=0.0031; GS, p=0.0159; CUS, p=0.0045), 4 vs 2 (PD, p=0.0176; GS, p=0.0160; CUS, p=0.0074), and 4 vs 3 (CUS, p=0.0374). Significant differences among +ve were: 4 vs 1 (PD, p=0007), 4 vs 2 (PD, p=0.0234), and 3 vs 1 (PD, p=0.0202). No significant Group differences were found for 2 vs 1 (for +ve and scores) and when comparing the 4 groups for GS +ve and CUS +ve. Table 2 shows the frequency distribution of patients by wrist and follow up. There were no significant effects attributable to differences in wrists or follow up visit (p-values all >0.05).Abstract SAT0676 – Table 1 Analysis summary of ultrasound scores and positivity in the wrist.Abstract SAT0676 – Table 2 Frequency distribution of subjects by follow-up and wrist Disclosure of Interest None declared