| |
Baseline only
|
Baseline and follow-up
|
|---|
|
Age, per year
|
0.99 (0.98, 0.99)
|
1.05 (1.02, 1.08)
|
|
Male vs. female
|
0.91 (0.75, 1.11)
|
1.03 (0.48, 2.20)
|
|
Disease duration, years
|
0.99 (0.99, 1.01)
|
1.06 (1.03, 1.08)
|
|
Serologic status*
|
|
Negative
|
Reference
|
Reference
|
|
Positive
|
0.66 (0.56–0.78
|
0.11 (0.04, 0.29)
|
|
Missing
|
0.42 (0.29–0.60)
|
0.88 (0.38, 2.06)
|
|
DMARD use**
|
|
csDMARD, baseline
|
0.71 (0.59–0.86)
|
0.18 (0.12, 0.27)
|
|
TNFi
|
–
|
0.48 (0.26, 0.90)
|
|
nonTNFi nonTCZ bDMARD
|
–
|
4.59 (2.91, 7.25)
|
|
Only csDMARD
|
–
|
5.83 (3.73, 9.13)
|
|
No DMARDs
|
–
|
0.40 (0.18, 0.88)
|
|
Glucocorticoid use
|
–
|
0.87 (0.61, 1.23)
|
|
CDAI, Severe
|
–
|
Reference
|
|
Moderate
|
–
|
0.65 (0.46, 0.91)
|
|
Low
|
–
|
0.48 (0.33, 0.70)
|
|
Remission
|
–
|
0.20 (0.11, 0.37)
|
|
Comorbidities, count#
|
–
|
1.16 (1.11, 1.22)
|
- Notes: Odds ratios (95% confidence intervals). Baseline models used logistic regression. Models with follow-up data used mixed generalized linear regression
- TCZ tocilizumab; DMARD disease-modifying anti-rheumatic drug; csDMARDs conventional synthetic DMARDs; nonTNFi nonTCZ bDMARD includes all JAK inhibitors, abatacept, and rituximab
- ---Not considered at baseline or not significant in univariable screen so not considered in multivariable model or withheld from the model because of problems with convergence
- *Serologic status defined as a positive if either rheumatoid factor or anti-CCP antibody were ever positive up to the relevant reference point
- **Reference category for DMARD use is the non-use of a give DMARD
- #Comorbidities are noted in Table 1. Clinical disease activity index (CDAI) categories defined as remission (CDAI < 2.8), low (CDAI 2.9–10.0), moderate (CDAI 10.1–22.0), and high (CDAI > 22.1)
