Defining Objective BASDAI Cut-Offs for Disease Activity States and Improvement Scores in Axial Spondyloarthritis: A Multicentric Collaboration.

Abstract

Objective: To estimate objective cut-off values for Bath ankylosing Spondylitis Disease Activity Index (BASDAI) corresponding to Ankylosing Spondylitis Disease Activity Score with CRP (ASDAS) cut-off values (1.3, 2.1, and 3.5), and for interval changes (ΔBASDAI) corresponding to ΔASDAS (1.1 and 2).

Methods: In this multicentric study, adult patients with active axial spondyloarthritis (axSpA) treated with either tofacitinib 5 mg twice daily or adalimumab 40 mg subcutaneously fortnightly were recruited. Available paired data on BASDAI and ASDAS and paired interval change of these parameters (taken at least 3-months apart) were analyzed. Cut-off values for BASDAI were determined from a multinomial logistic regression and that for ΔBASDAI were determined with ordinal logistic regression with predicted probabilities. Diagnostics were assessed with correct classification rate (CRR), polychoric correlations (PCC), Goodman and Kruskal Gamma (GKG), and Matthew's correlation coefficient (MCC).

Results: Total 962 observations of paired data and 670 interval change data were available from 266 patient (mean age 35 years, mean disease duration 7 years, mean BASDAI 2.7, and mean ASDAS 2.1). Following ASDAS classes were observed: inactive disease 131 (13.6%), moderate disease activity 322 (33.5%), high disease activity 420 (43.65%), and very high disease activity 89 (9.25%). The three cut-offs generated for BASDAI were 0.8, 2.5, and 6 for BASDAI and those for ΔBASDAI were 2 and 4. The CRR for BASDAI cut-offs was 65.59 and that for ΔBASDAI was 54.9. For the BASDAI cut-offs, both PCC and GKG showed high values (> 0.85) and the MCC was 0.471.

Conclusion: In patients with axSpA, without any clinical or serological confounders, BASDAI values 0.8, 2.5, and 6 corresponded to ASDAS-CRP values 1.3, 2.1, and 3.5, respectively, and ΔBASDAI values of 2 and 4 correspond to ΔASDAS cut-offs of 1.1 and 2.