Background/Purpose: JDM is a multisystem vasculopathic disease that primarily affects the skin and muscles. Most tools for assessment of disease activity in JDM are lenghty, complex, and centered on physician’s evaluation. We aim to develop a composite disease activity score for JDM and provide preliminary evidence of its validity.

Methods: A panel of experts devised the score, named Juvenile DermatoMyositis Activity Index (JDMAI), based on their clinical experience and a literature review. The JDMAI is composed of 4 clinical domains: 1) physician’s global assessment of overall disease activity on a 0-10 visual analog scale (VAS); 2) parent’s/child’s global assessment of child’s wellbeing on a 0-10 VAS; 3) muscle strength/endurance; 4) skin disease activity. Eight versions of the JDMAI were tested, which differed in the tools used to assess items 3 and 4. For item 3, two versions included the hybrid MMT/CMAS (hMC) with score in deciles (0-10), two the hMC with its original score (0-100), two the MMT-8 (0-80), and two the CMAS (0-52). For item 4, four versions included physician’s global rating of skin disease activity on a 0-10 VAS, and four included the cutaneous domain of the Disease Activity Score (DAS) (0-9). Validation was conducted on 275 patients included in a multinational dataset, evaluated at baseline and at 6, 12, and 24 months. Construct validity was assessed by calculating between-subject and within-subject correlations with JDM outcome measures not included in the JDMAI; internal consistency was assessed with Cronbach α and responsiveness to change with standardized response mean (SRM). Discriminant ability was determined in a different multinational dataset of 142 patients, by assessing the JDMAI score in patients rated in remission, low, moderate, or high disease activity by the attending physician.

Results: In between-subject exercise, all JDMAI versions showed strong (r>0.7) correlations with CHAQ (0.72-0.82), muscle VAS (0.77-0.87), muscle DAS (0.76-0.86) and total DAS (0.68-0.90), and moderate correlations (r=0.4-0.7) with pain VAS (0.50-0.57) and Myositis Damage Index (MDI) (0.51-0.60). Owing to the interrelatedness of longitudinal data from an individual patient, within-subject correlations were higher and were all strong (r=0.76-0.97). SRM was good (1.09-1.57) and was higher for JDMAI 1 and 2. Cronbach’s alpha was fair (0.70 and 0.69) for JDMAI 1 and 2, and poor for other JDMAI versions. All JDMAI versions discriminated strongly between patients in different disease activity states (Kruskal-Wallis test, p < 0.001).

Conclusion: Overall, the JDMAI1 and JDMAI2 revealed the best measurement properties in validation analyses. The JDMAI1 (score range: 0-40) may be preferred over the JDMAI2 as it weights equally its 4 components, whereas in the JDMAI2 (score range: 0-39) items 1 to 3 are scored on a 0-10 scale and skin disease on a 0-9 scale.