Objectives: The objectives of this study were to assess 27-joint Juvenile Arthritis Disease Activity Score (JADAS-27) responsiveness, JADAS-27 changes corresponding to clinically important differences and cut-off scores for low and high disease activity in a large prospective JIA cohort.
Methods: JADAS-27 responsiveness, using effect size and standardized response mean (SRM), and changes in the JADAS-27 corresponding to clinically important differences were determined for clinical improvement (ACRpedi30) and worsening (flare). To assess whether various degrees of change in the JADAS-27 could be used to demonstrate improvement or worsening in individual patients, diagnostic parameters were computed for cut-off score changes. Finally, cut-off scores for low and high disease activity and their diagnostic parameters were determined.
Results: In 228 patients with 529 consecutive visits, ACRpedi30 was detected in 109 and flare in 111 visits. Regarding responsiveness, the effect size was 0.93 and SRM was 1.26 for clinical improvement, while for clinical worsening the effect size was 0.65 and SRM was 0.60. Changes in the JADAS-27 corresponding to clinically important difference were -5.5 for improvement and +1.7 for worsening. Cut-off score changes in the JADAS-27 had 65-90% sensitivity and 67-86% specificity for improvement, and 31-64% sensitivity and 89-97% specificity for worsening. The JADAS-27 cut-off score for low disease activity was ≤2.7 with 76% sensitivity and 62% specificity, and the cut-off score for high disease activity was ≥6 with 77% sensitivity and 77% specificity.
Conclusion: The JADAS-27 had moderate to good responsiveness and was changed by clinically important differences. The JADAS-27 cut-off scores differentiated between low and high disease activity. These JADAS-27 interpretations could be potentially applicable in clinical care and trials.
Keywords: clinically important difference; disease activity; juvenile idiopathic arthritis; responsiveness.