Background Recently, a composite score named Juvenile Arthritis Disease Activity Score (JADAS) was found to be a valid instrument for assessment of disease activity. JADAS was developed with erythrocyte sedimentation rate (ESR) because C-reactive protein (CRP) values were not available in all databases used to validate the tool. However, as the authors of the JADAS highlighted, CRP is a direct measure of the acute-phase response and is less confounded by other factors. Nordal et al compared in a Nordic population the JADAS based on CRP with JADAS based on ESR and concluded that these instruments correlated closely, indicating that both scores can be recommended for assessing disease activity in JIA.
Objectives Determine JADAS CRP and compare its performance to the JADAS ESR, in a Portuguese population with JIA.
Methods A National based cohort of patients with JIA according to ILAR criteria, registered in Rheumatic Diseases Portuguese Register, Reuma.pt, was selected. Patients were included in the study when all disease activity measures were available for JADAS ESR and CRP calculation and one visit per patient was randomly selected. JADAS-CRP was adapted by replacing ESR with CRP as the inflammatory marker. JADAS 27-CRP was calculated similarly to JADAS 27-ESR as the simple linear sum of its four components, yielding a global score of 0-57. Pearson correlations and K statistics were used in analyses.
Results From the 729 patients with JIA included in the database reumaPT, 358 children had full data to calculate JADAS. From the 358 patients, 65.4% were female, mean disease duration 1.75 ± 9.03 years. 134 (37,5%) were classified as persistent oligoarticular, 53 (14,8%) extended oligoarticular, 51 (14,2%) polyarticular rheumatoid factor (RF) negative, 30 (8,4%) polyarticular RF positive, 39 (10,9%) systemic, 35 (9,8%) enthesitis-related arthritis, 11 (3,1%) psoriatic arthritis and in 5 (1,4%) patients we could not have access to the subtype of JIA. The correlation coefficient at the visit with all JADAS items available between JADAS-CRP and JADAS-ESR was 0,967, p< 0,0001), though the correlation coefficient between CRP and ESR was only 0,335 (p<0,0001). When comparing the JADAS-ESR and JADAS-CRP within each subtype of JIA, the strong correlation was maintained (all values of correlation > 0.9 and all p-values < 0.0001). The agreement between JADAS ESR and CRP across the 4 activity states (inactive disease, minimal disease activity, acceptable symptoms and active) assessed by K statistics was very good, showing 91.1% of the observations in agreement, K=0.867 (95%CI 0.824-0.91).
Conclusions In conclusion, in our study the JADAS27 based on CRP and ESR correlated closely, indicating that both measures can be used in clinical practice.
Disclosure of Interest None Declared