Skip to main content
. 2016 Jan 18;18:12. doi: 10.1186/s13075-016-0920-6

Table 2.

Results from multivariate GEE model to identify predictors of response

Variable β (95 % CI) P value
Age at baseline, years −0.003 (−0.01, 0.004) 0.35
RA duration, years 0.001 (−0.01, 0.008) 0.84
Sex, malea/female −0.13 (−0.41, 0.16) 0.39
DAS28 at baseline 0.39 (0.30, 0.49) <0.0005
Duration of anti-TNF therapy, weeks 0.02 (0.01, 0.03) <0.0005
RAPA seropositivity, yes/noa −0.07 (−0.39, 0.25) 0.66
ACPA seropositivity, yes/noa 0.10 (−0.28, 0.48) 0.61
Smoking status, nevera/past/current −0.10 (−0.24, 0.05) 0.18
Concurrent methotrexate use, yes/noa 0.43 (0.19, 0.68) <0.0005
Concurrent prednisolone use, yes/noa −0.08 (−0.28, 0.11) 0.40
Type of anti-TNF therapy, ETNa, ADA or INF −0.08 (−0.18, 0.02) 0.12

β regression coefficient. aReference category for these variables in the generalized estimating equations (GEE) model, RA rheumatoid arthritis, DAS28 Disease Activity Score in 28 joints, RAPA rheumatoid arthritis particle agglutination, ACPA anti-citrullinated protein antibodies