Poor correlation between the erythrocyte sedimentation rate and clinical activity in juvenile rheumatoid arthritis

Despite questions regarding its validity as an estimator of inflammatory disease activity, monitoring of the erythrocyte sedimentation rate (ESR) continues to be routine practice among pediatric rheumatologists caring for children with juvenile rheumatoid arthritis (JRA). We studied a large group of patients with JRA in order to determine the degree of correlation between clinically apparent inflammation and the ESR. regression and correlation analyses and descriptive statistical techniques were used to establish the relationship between 1) the ESR and the amount of clinically apparent inflammation at a point in time, and 2) changes in the ESR and the corresponding changes in apparent inflammation. One hundred fifty-nine children with JRA who were participants in a double-blind, controlled trial of two antirheumatic drugs were assessed for clinical and laboratory evidence of inflammatory disease activity at an initial visit, and then periodically for the duration of the one year study. Results showed that, at the initial assessment, neither the total number of joints with active arthritis nor the severity score correlated well with the ESR (r = .196 and .245 respectively). These findings were independent of the course type of JRA and age of the child. Changes from baseline in inflammation showed little correlation (r less than .25) with changes in the ESR. These findings suggest that the ESR is a relatively poor indicator of the amount of articular inflammation present, and that changes of disease activity are not reflected closely by changes in the ESR among children with JRA.