Performance characteristics of the simplified version of ankylosing spondylitis disease activity score (SASDAS)

Abstract

Various types of disease activity measures are available for axial spondyloarthritis (axSpA), and there is no gold standard for all individual patients. The ankylosing spondylitis disease activity score (ASDAS) is highly discriminatory, sensitive to change, and associated with structural progression. A simplified version of the ASDAS (SASDAS) was proposed and found to be a simple and practical tool to assess disease activity. Our aim was to test the performance characteristics of the SASDAS and compare it with validated tools. In total, 97 consecutive ankylosing spondylitis (AS) patients were included in the study. Disease activity was assessed by the ASDAS-erythrocyte sedimentation rate (ESR), ASDAS-C-reactive protein (CRP), bath ankylosing spondylitis disease activity index (BASDAI), and SASDAS. The relationship among these activity indices and the level of agreement of various activity categories were tested. There was a strong correlation between the SASDAS and other activity indices, including the BASDAI (r = 0.916, p < 0.001), ASDAS-CRP (r = 0.847, p < 0.001), and ASDAS-ESR (r = 0.942, p < 0.001). Although the agreement between the ASDAS-ESR and SASDAS was good (weighted kappa of 0.744 and total agreement of 77 %), there was moderate agreement between the ASDAS-CRP and SASDAS (weighted kappa of 0.579 and total agreement of 66 %). The disagreement was particularly striking in moderate and high disease activity states. Approximately 40 % of patients classified as moderate activity according to the ASDAS-ESR and 45 % according to the ASDAS-CRP were differentially categorized by the SASDAS. The results of the present analysis suggest that the simplified version of the ASDAS-ESR should be further validated in various settings and populations due to a questionable level of agreement between the ASDAS-CRP and SASDAS.