Table 2.

Observed radiographic changes from baseline to week 104

					Between-group differences in net % patients with progression, Etanercept—Control (95% CI)
Endpoint	Cohort	Improved n/N (%)	Worsened n/N (%)	Net % patients with progression,*† unadjusted (95% CI)	Unadjusted analysis	Adjusted analysis‡
∆ in mNY criteria	Control Etanercept	3/193 (1.6) 4/162 (2.5)	6/193 (3.1) 1/162 (0.6)	1.6% (−1.3 to 4.4) −1.9% (−4.9 to 1.2)	−3.4% (−7.6 to 0.8) p=0.11	−4.7% (−9.9 to 0.5) p=0.07
∆ ≥1 grade in ≥1 SIJ	Control Etanercept	21/193 (10.9) 19/162 (11.7)	36/193 (18.7) 16/162 (9.9)	7.8% (0.6 to 15.0) −1.9% (−9.7 to 6.0)	−9.6% (−20.3 to 1.0) p=0.08	−18.2% (−30.9 to −5.6) p=0.005
∆ ≥1 grade in ≥1 SIJ; shift from 0 to 1 or 1 to 0 considered no Δ	Control Etanercept	16/193 (8.3) 15/162 (9.3)	29/193 (15.0) 14/162 (8.6)	6.7% (0.3 to 13.2) −0.6% (−7.6 to 6.4)	−7.4% (−16.9 to 2.2) p=0.13	−16.4% (−27.9 to −5.0) p=0.005

Based on two of three readers assigning same category; otherwise considered no change.

Some patients started with lowest possible score and could not improve.

*Net % patients with progression=number of patients with worsening minus the number of patients with improvement, divided by the study population.

†One-way analysis of variance.

^‡Adjusted for these covariates at baseline: sex, symptom duration, smoking status, human leucocyte antigen-B27 status, Ankylosing Spondylitis Disease Activity Score, Spondyloarthritis Research Consortium of Canada MRI SIJ score and SIJ mNY grade.

∆, change; mNY, modified New York; SIJ, sacroiliac joint.