Performance of the Assessment of Spondyloarthritis International Society criteria for the classification of spondyloarthritis in early spondyloarthritis clinics participating in the ESPERANZA programme.

OBJECTIVE

The objective of this study was to analyse the performance of the Assessment of SpondyloArthritis International Society (ASAS) criteria for the classification of SpA in early SpA clinics.

METHODS

We used a cross-sectional study of patients referred to early SpA units within the ESPERANZA programme (a Spanish nationwide health management programme designed to provide excellence in diagnosis and care for early SpA). Patients were eligible if they were <45 years of age and had any of the following: (i) a 2-year history of inflammatory back pain; (ii) back or joint pain with psoriasis, anterior uveitis, radiographic sacroiliitis, family history of SpA or positive HLA-B27; or (iii) asymmetric arthritis. We excluded patients for whom imaging (X-rays/MRI) or HLA-B27 results were not available. We analysed the performance (sensitivity and specificity) of different classification criteria sets, taking the rheumatologist's opinion as the gold standard.

RESULTS

The analysis included 775 patients [mean age 33 (s.d. 7) years; 55% men; mean duration of symptoms 11 (s.d. 6) months]; SpA was diagnosed in 538 patients (69.5%). A total of 274 (67.9%) patients with chronic back pain met the ASAS axial criteria, 76 (56.3%) patients with arthritis but not chronic back pain fulfilled the ASAS criteria for peripheral SpA and 350 (65.1%) fulfilled all the ASAS criteria. The sensitivity and specificity of the ASAS criteria set were 65% and 93%, respectively (axial criteria: sensitivity 68%, specificity 95%). The sensitivity and specificity for the ESSG and Amor criteria were 58% and 90% and 59% and 86%, respectively.

CONCLUSION

Despite performing better than the Amor or ESSG criteria, the ASAS criteria may be limited to detection of early forms, particularly in populations in which MRI is not extensively available or in populations with a low prevalence of HLA-B27.