Simplified semi-quantitative ultrasound score to evaluate the synovitis of bilateral wrists and hands in patients with rheumatoid arthritis

Objective To develop the optimal simplified combination of joints for ultrasonographic assessment of joint inflammation of wrists and hands in patients with rheumatoid arthritis (RA). Methods US examination was performed using grey-scale (GSUS) and power Doppler (PDUS) semi-quantitative scoring systems with scores 0-3 in 22 joints of 705 RA patients, including all proximal interphalangeal (PIP), metacarpophalangeal (MCP), and bilateral wrist joints. Continuous variables were presented as mean and standard deviation if normally distributed, and dichotomous variables were presented as frequencies. T test and Wilcoxon signed test were applied for statistical analysis. All correlations among US variables were assessed using Spearman's rank correlation test. Candidate joint set was selected by multiple stepwise linear regression analysis. Results Through multiple linear stepwise regression analysis, the standard coefficient of wrist, MCP5, MCP2 and MCP3 joints under GSUS was higher than other joints. And the adjusted R2 of the model composed of wrist, MCP5, MCP2 and MCP3 joints was greater than 0.9. Among the sum GS and PD scores of various selected joint combinations, total score-8 (GS vs PD), including bilateral wrist, MCP2, MCP3, MCP5 joints, not only showed highest sensitivity and negative predictive value (97.35% vs 97.20%; 92.67% vs 97.21%, respectively), but also had the highest correlation with the total score-22 (GS, PD) (r=0.989, 0.972, P<0.01). Conclusion Total score-8, including bilateral wrist, MCP2, MCP3, MCP5 joints, is simple and efficient enough for monitoring active synovitis of wrists and hands in patients with RA in daily practice. Key words: Arthritis, rheumatoid; Wrists and hands; Ultrasonography; Synovitis; Evaluation