Development and Testing of Reduced Joint Counts in Juvenile Idiopathic Arthritis

Abstract

ObjectiveTo develop and test reduced joint counts in children with juvenile idiopathic arthritis (JIA).MethodsFour reduced joint counts including 45, 35, 27, and 10 joints were devised by a panel of experienced pediatric rheumatologists, who selected the joints to be included based on the ease of technical assessment, functional relevance, and frequency of involvement. Three large samples of patients with JIA (total n = 4353) who had a detailed joint assessment available were used to develop and test reduced joint counts. Performance of reduced counts was examined by comparing their Spearman correlation with the standard (i.e., complete) joint count. Construct validity was evaluated by calculating Spearman correlation with other JIA outcome measures. Responsiveness to clinical change was determined through the standardized response mean (SRM).ResultsSpearman correlations of reduced joint counts with the whole joint count and with the other JIA outcome measures were comparable, revealing that they had similar ability to serve as surrogate for the whole joint count and construct validity. Responsiveness to clinical change was also comparable across reduced counts (SRM 0.83–1.09 for active joint counts and 0.63–0.81 for restricted joint counts). Based on these results and considering the relative feasibility of the different counts, the 27-joint reduced count is proposed for use in JIA. This joint count includes the cervical spine and the elbow, wrist, metacarpophalangeal (from first to third), proximal interphalangeal, hip, knee, and ankle joints.ConclusionReduced joint counts appear to be as reliable as standard joint counts in assessment of the severity of joint disease and its change over time in children with JIA.