Peripheral arthritis instruments in spondyloarthritis still lack robust validation

A systematic literature review from the ASAS-SPARADISE project found that measurement properties for instruments used to assess peripheral arthritis disease activity in spondyloarthritis remain thin, uneven, and in several settings simply inadequate for confident use. The review examined adults with axial spondyloarthritis and peripheral spondyloarthritis, then mapped available tools to the specific domain of peripheral joint inflammation rather than to overall disease burden. That distinction matters because peripheral arthritis behaves differently from axial pain, stiffness, and global disease activity, and the wrong instrument can blur those differences rather than measure them.

The search was broad, spanning Medline, EMBASE, and Cochrane through February 2025, yet only eight studies met eligibility criteria out of 4,364 screened records. Such a small evidence base is not a minor technical issue; it means clinicians and researchers are still trying to quantify a core inflammatory domain with tools that have often been borrowed from adjacent conditions or from broader composite scores. In practice, that leaves room for uncertainty when interpreting response, judging flare, or selecting endpoints for trials that specifically target peripheral arthritis.

What the review evaluated and why the framework matters

The review followed the Population-Instrument-Measurement framework and focused on instruments retained after a prior domain-match and feasibility screen. Those included Patient Global Assessment, swollen joint count using 66/44 joints, tender joint count using 68/44 joints, C-reactive protein, Bath Ankylosing Spondylitis Disease Activity Index, Axial Spondyloarthritis Disease Activity Score, Disease Activity Index for Psoriatic Arthritis, and Disease Activity Score using 44 joints. The question was not whether these instruments capture something clinically relevant in a broad sense, but whether they actually perform well when asked to measure peripheral arthritis activity in spondyloarthritis.

The evaluation centered on construct validity, including correlation hypotheses and known-groups discrimination, as well as discrimination measures such as test-retest reliability, responsiveness, clinical trial discrimination, and thresholds of meaning. Risk of bias was judged with the OMERACT Good Methods Checklist, and evidence was synthesized using OMERACT rules: GREEN for good performance, AMBER for adequate performance, and RED for poor performance. In a field where endpoints often drive treatment decisions and trial interpretation, this kind of structured appraisal is more than academic housekeeping. It determines whether a score can be trusted to reflect change in inflammatory burden rather than noise, symptom overlap, or measurement drift.

Axial disease evidence was sparse and mostly only adequate

Evidence in axSpA was limited to three instruments: PGA, CRP, and BASDAI. Across these measures, performance landed mostly in the AMBER range for construct validity, longitudinal construct validity, and discrimination in clinical trials. AMBER is not a failure, but it is not a vote of confidence either. It signals partial support without enough consistency or strength to call the measure fully established for this specific domain.

A single sentence captures the practical problem: the peripheral arthritis domain in axSpA remains under-measured. When a tool is designed around global disease experience or axial symptoms, it can miss the nuances of swollen or tender peripheral joints, especially if inflammatory activity is modest, intermittent, or clinically overshadowed by spinal symptoms. That leaves clinicians dependent on composite impressions that may not separate peripheral synovitis from the rest of the disease burden.

Composite scores performed better in peripheral spondyloarthritis

The picture was somewhat stronger in pSpA, where composite instruments performed better overall. BASDAI, ASDAS, and DAPSA were rated mostly GREEN, suggesting more convincing evidence for construct validity and discrimination in that population. Even so, the apparent advantage of composite measures should not be mistaken for complete validation of a peripheral arthritis-specific construct. Composite scores often benefit from integrating multiple dimensions of disease activity, which can improve stability and responsiveness, but they may still capture a broader inflammatory syndrome rather than the peripheral joint domain alone.

For bedside care, that creates a familiar tension. A composite instrument may be clinically useful because it is workable and responsive, yet still fail to isolate the exact question a rheumatologist is trying to answer: how active is the peripheral arthritis component today? This matters in treatment escalation, in tapering decisions, and in assessing whether a biologic or targeted therapy is improving the joint phenotype rather than only the patient’s overall symptom burden.

The missing reliability and meaning thresholds leave an important gap

No study in the review provided test-retest reliability or thresholds of meaning. Those omissions are clinically important. Without test-retest reliability, it is difficult to know whether a stable patient will score the same on repeat measurement. Without thresholds of meaning, it is harder to interpret whether a change is large enough to matter to patients and clinicians rather than just to statistical models.

These absences are not peripheral details. They affect how outcome measures are used in trials, registries, and routine practice. A score that appears to move can still be a poor score if its change is not reproducible or interpretable. Inflammatory arthritis research depends on measurement precision, and precision depends on evidence that goes beyond face validity and correlation alone.

What the current evidence means for ASAS-SPARADISE and future tool development

The review supports a clear conclusion: evidence for instruments assessing peripheral arthritis in spondyloarthritis remains scarce, especially in axSpA, where the available data are limited and mostly only adequate. Composite instruments in pSpA currently perform better, but even there the evidence is incomplete. The field still lacks a validated, domain-specific tool that can reliably capture peripheral arthritis activity across the spectrum of spondyloarthritis.

That gap has direct implications for instrument selection in future studies. If the goal is to detect peripheral joint inflammation with confidence, broad disease activity tools may not be enough, and the review suggests that currently available measures do not yet fully solve the problem. The ASAS-SPARADISE effort is therefore not just about choosing the best existing score; it is about defining what a fit-for-purpose peripheral arthritis instrument should measure, how it should behave over time, and what degree of change should count as clinically meaningful.

The review was registered in PROSPERO under CRD420251177596 and published in RMD Open in 2026. Its most consequential message is not that one score beats another by a wide margin, but that the evidence base is still too thin to support strong claims for several commonly used instruments. Until reliability and thresholds of meaning are established, clinicians and trialists will continue to work with measures that are better than guesswork, yet still short of fully validated endpoints.


Reporting by Dr. A. Nikhil Krishnan, MBBS for Medicine World Foundation, with additional data provided by Annals of the Rheumatic Diseases (BMJ).
PRIMARY CLINICAL EVIDENCE & REGISTRY DATA