Validation of systemic Juvenile Arthritis Disease Activity Score 10 in adults with Still's disease

The systemic Juvenile Arthritis Disease Activity Score 10 (sJADAS10) was specifically developed for paediatric Still's disease (SD) and has high construct validity and sensitivity to change, with validated operable thresholds. This study aims to validate the sJADAS10 and its thresholds in adults with SD.

The study involved adult patients with SD, followed in one unique reference centre. We extracted data necessary to assess disease activity according to the sJADAS10 and 4 other measures developed for adults with SD. As a reference, disease activity was classified as inactive disease (ID) or low disease activity (LDA), moderate disease activity (MDA), and high disease activity (HDA) according to the Rosina definitions. Weighted Cohen's kappa coefficients (wKappa) were used to evaluate the agreement between disease states according to the sJADAS10, the 4 other tools, and the reference. We compared the discriminative power (ID vs LDA/MDA/HDA or ID/LDA vs MDA/HDA) of the different tools by receiver operating characteristic (ROC) curves.

Data were available for 129 different visits: according to the reference, 38 corresponded to ID, 19 to LDA, 33 to MDA, and 39 to HDA. The highest agreement with the reference was observed with the sJADAS10: wKappa (0.70) and accuracy (0.77). The sJADAS10 discriminative power was also high, with an area under the ROC curve for ID or a combination of ID and LDA at 0.94 and 0.97.

The present study supports the external validity of the sJADAS10 in adults with SD. The sJADAS10 appears to be a relevant candidate for harmonised disease activity assessment across paediatric and adult SD populations.