Diagnostic Value and Validity of Early Spondyloarthritis Features: Results From a National Spanish Cohort

To evaluate the validity of different spondyloarthritis (SpA) features included in the Berlin diagnostic algorithm and the Assessment of SpondyloArthritis international Society (ASAS) classification criteria in an early SpA cohort. This was a longitudinal multicenter study including patients from the ESPeranza program cohort who were suspected to have SpA. Subjects were ≤45 years old, and SpA symptom duration was 3–24 months. Patients with axial SpA symptoms were selected and categorized according to diagnosis (yes/no) of axial SpA. Descriptive analysis was performed, and the sensitivity, specificity, predictive value, and likelihood ratio (LR) of each feature were calculated. Of 775 patients suspected to have SpA, 665 had predominantly axial symptoms and 516 of these patients were diagnosed with axial SpA. The most useful SpA features were sacroiliitis on magnetic resonance imaging (positive LR 6.6) or radiograph (positive LR 31.1) and peripheral arthritis (positive LR 8.9). The features with the lowest diagnostic utility were a family history of SpA (positive LR 1.5) and good response to nonsteroidal antiinflammatory drugs (positive LR 1.6). Inflammatory back pain (IBP; according to ASAS criteria) was described in only 27% of SpA patients, with a positive LR of 2.3. HLA–B27 positivity was present in 245 (48%), and the positive LR was 2.8. The diagnostic value of SpA features in patients with early axial SpA seems to be different than in patients with longstanding disease. Chronic back pain is better than IBP as an entry point to the diagnostic algorithm. Sacroiliitis on imaging is very important for early diagnosis, while the use of HLA–B27 status as a key factor is questionable.