Comparison of three methods for calculating the Bath Ankylosing Spondylitis Metrology Index in a randomized placebo‐controlled study

Abstract

To compare 3 methods for calculating the Bath Ankylosing Spondylitis Metrology Index (BASMI) score using patients who participated in the GO-RAISE study. Patients with active ankylosing spondylitis (AS) were randomly assigned in a 1:1.8:1.8 ratio to receive subcutaneous injections of placebo, golimumab 50 mg, or golimumab 100 mg every 4 weeks. Lumbar flexion, tragus-to-wall distance, lumbar side flexion, intermalleolar distance, and cervical rotation angle measurements at baseline, week 14, and week 24 were used to calculate the BASMI 2-step (BASMI(2) ), 10-step (BASMI(10) ), and linear (BASMI(lin) ) scores. BASMI(2) scores were generally lower than BASMI(10) and BASMI(lin) scores, which were nearly identical. Median changes from baseline to week 14 in the combined golimumab group were similar to those in the placebo group when using the BASMI(2) calculation method (0.00 versus 0.00; P = 0.288). The combined golimumab group showed significantly greater improvement from baseline to week 14 than the placebo group when using the BASMI(10) (-0.20 versus 0.00; P = 0.018) and BASMI(lin) (-0.31 versus -0.07; P = 0.015) calculation methods, with the latter showing the greatest difference between golimumab and placebo. Guyatt's effect size was better for the BASMI(lin) and the BASMI(10) versus the BASMI(2) in the combined golimumab group at week 14 (0.58 and 0.53 versus 0.42, respectively) and week 24 (0.76 and 0.69 versus 0.61, respectively), despite the relatively short period to assess changes in spinal mobility. The BASMI(lin) method was the most sensitive to changes in range of motion exhibited by patients with AS who received golimumab.