The objective of this study was to investigate the performance of classification criteria sets (Assessment of SpondyloArthritis international Society (ASAS), European Spondylarthropathy Study Group (ESSG), and Amor) for spondyloarthritis (SpA) in a clinical practice cohort in Colombia and provide insight into how rheumatologists follow the diagnostic path in patients suspected of SpA. Patients with a rheumatologist's diagnosis of SpA were retrospectively classified according to three criteria sets. Classification rate was defined as the proportion of patients fulfilling a particular criterion. Characteristics of patients fulfilling and not fulfilling each criterion were compared. The ASAS criteria classified 81 % of all patients (n = 581) as having either axial SpA (44 %) or peripheral SpA (37 %), whereas a lower proportion met ESSG criteria (74 %) and Amor criteria (53 %). There was a high degree of overlap among the different criteria, and 42 % of the patients met all three criteria. Patients fulfilling all three criteria sets were older (36 vs. 30 years), had more SpA features (3 vs. 1 features), and more frequently had a current or past history of back pain (77 vs. 43 %), inflammatory back pain (47 vs. 13 %), enthesitis (67 vs. 26 %), and buttock pain (37 vs. 13 %) vs. those not fulfilling any criteria. HLA-B27, radiographs, and MRI-SI were performed in 77, 59, and 24 % of the patients, respectively. The ASAS criteria classified more patients as having SpA in this Colombian cohort when the rheumatologist's diagnosis is used as an external standard. Although physicians do not perform HLA-B27 or imaging in all patients, they do require these tests if the clinical symptoms fall short of confirming SpA and suspicion remains.
Keywords: Classification criteria; Clinical practice; Cohort; Performance; Spondyloarthritis.