sJADAS10 Tracks Adult Still’s Disease Activity With Strong Agreement
Validation work in Still’s disease rarely comes down to a single score, but this one does. In adult patients followed at one reference centre, the systemic Juvenile Arthritis Disease Activity Score 10 was tested against adult disease-state definitions and performed well enough to support its use beyond paediatric practice. The study asked a simple but important question: can a tool created for children with Still’s disease accurately classify disease activity in adults who carry the same inflammatory syndrome into later life?
The answer from this cohort was yes, with meaningful caveats about the setting and design. Across 129 visits, disease activity was distributed across the full spectrum of the Rosina reference states: 38 visits met inactive disease, 19 low disease activity, 33 moderate disease activity, and 39 high disease activity. That spread matters because a validation study has to prove more than correlation in a narrow slice of severity; it has to separate quiet disease from active disease, and it has to do so when the clinical picture is mixed.
In that test, sJADAS10 posted the highest agreement with the reference classification. Weighted Cohen’s kappa reached 0.70, and accuracy reached 0.77. Those are not cosmetic numbers. In inflammatory disease, a score that can reliably map onto clinical state is what makes treatment escalation, tapering, and response assessment possible without leaning entirely on subjective impression. The score also showed strong discriminative performance on receiver operating characteristic analysis, with an area under the curve of 0.94 when identifying inactive disease versus all other states, and 0.97 when the comparison grouped inactive disease and low disease activity against moderate and high disease activity.
Why Cross-Age Validation Matters in Still’s Disease
Adult Still’s disease sits in a difficult clinical space. The disorder shares immunologic and phenotypic features with systemic juvenile idiopathic arthritis, yet it is managed in adult rheumatology clinics, where disease activity instruments have often been built around adult-onset inflammatory patterns. A score developed for paediatric patients can be attractive if it captures the core biology of the syndrome: quotidian inflammation, systemic features, joint burden, and the shifting balance between active cytokine-driven disease and relative quiescence.
That is the clinical appeal of the sJADAS10 here. A unified metric can help reduce fragmentation between paediatric and adult care, especially for patients whose disease biology does not change simply because they have aged out of one service and into another. When a score retains discrimination in adults, it becomes more than a paediatric artifact; it becomes a candidate language for disease control across the lifespan.
The study design supports that argument through direct comparison with four other adult-oriented measures, all tested against the Rosina disease-state framework. Rosina inactive disease, low disease activity, moderate disease activity, and high disease activity served as the anchor for comparison, giving the analysis a clinically grounded reference rather than an abstract statistical endpoint. The sJADAS10 outperformed the other tools on agreement, which suggests that the score’s components are aligned with the phenotype of Still’s disease rather than merely with generic inflammation.
What the Agreement Statistics Mean at the Bedside
A weighted kappa of 0.70 indicates substantial agreement, not perfection. In practical terms, the score is good at sorting adult visits into clinically meaningful categories, but it does not erase the uncertainty that comes with a systemic autoinflammatory disorder whose manifestations can shift rapidly. Fever, rash, arthritis, and laboratory activity do not always move in lockstep, and any disease activity index has to compress that complexity into a manageable clinical signal.
The ROC findings strengthen the case further. An AUC of 0.94 for inactive disease versus all other states indicates excellent separation at the quiet end of the spectrum, where treatment de-escalation decisions often hinge on whether residual symptoms represent low-grade activity or something more substantial. The 0.97 AUC for inactive and low disease activity versus moderate and high disease activity suggests the score can also identify the boundary where escalation becomes more likely. In a disease with substantial consequences from under-treatment, that boundary is not trivial.
The study supports a broader biologic intuition as well. Still’s disease is driven by inflammatory pathways that can remain clinically recognizable across age groups, and a score that performs well in both paediatric and adult patients likely reflects that shared pathophysiology. If the same cytokine-mediated disease architecture produces similar patterns of fever, arthritis, and systemic inflammation, then a harmonized activity measure becomes plausible rather than forced.
A Candidate for Harmonised Assessment Across the Lifespan
The authors’ conclusion follows directly from the data: the study supports the external validity of sJADAS10 in adults with Still’s disease, and the score appears to be a relevant candidate for harmonised disease activity assessment across paediatric and adult populations. That is an important statement for a condition that often straddles age-based classification systems without losing its clinical identity.
Still, the dataset comes from one unique reference centre, and the analysis used 129 visits rather than a multicentre cohort. Those features do not weaken the result, but they do define its range. Validation in a specialised centre can capture expert assessment unusually well, which is useful for benchmarking a tool and equally useful for identifying where it still needs testing. Broader confirmation in different care environments would be the next logical step if clinicians are to use the score as a standard measure in routine adult practice.
What this study delivers is a workable bridge between paediatric and adult rheumatology. The bridge is not speculative; it is grounded in concordance statistics, disease-state separation, and a reference framework built around clinically recognisable activity tiers. For adult patients with Still’s disease, that may help make treatment decisions more consistent, response tracking more reproducible, and future research easier to compare across age groups.