Skip to main content
. 2018 Mar;138(3):534–541. doi: 10.1016/j.jid.2017.10.005

Table 3.

Crude Cox proportional hazards model and adjusted model using inverse probability treatment weighting by propensity score, showing hazard ratios from a multinomial model involving etanercept, adalimumab, and ustekinumab versus non-biologic therapy1

Etanercept Adalimumab Ustekinumab All Biologics
Comparison against all non-biologic systemic therapies, hazard ratios (95% confidence intervals)
 Crude 1.11 (0.79–1.57) 0.98 (0.74–1.29) 1.04 (0.70–1.54) 1.02 (0.80–1.31)
 Adjusted 1.10 (0.75–1.60) 0.93 (0.69–1.26) 0.92 (0.60–1.41) 0.96 (0.73–1.27)
 Concomitant immunosuppressants 1.05 (0.67–1.64) 1.09 (0.70–1.68)
Adjusted 0–6 months 2.18 (0.95–5.01)
Adjusted 6–12 months 1.20 (0.51–2.81)
Adjusted 1–2 years 0.73 (0.35–1.53)
Comparison against methotrexate
 Crude 1.37 (0.90–2.07) 1.19 (0.83–1.71) 1.26 (0.80–1.99) 1.31 (0.94–1.84)
 Adjusted 1.47 (0.95–2.28) 1.26 (0.86–1.84) 1.22 (0.75–1.99) 1.29 (0.90–1.85)
 Concomitant immunosuppressants 1.00 (0.64–1.57) 1.04 (0.67–1.62)
1

Exposure time with concomitant (methotrexate, cyclosporine, fumaric acid esters, hydroxycarbamide) immunosuppressive medication use is adjusted for, with exposure time to two immunosuppressive therapies classed as concomitant in the non-biologic cohort.