Table 3

Multivariable associations between various factors and receiving contraception counselling

Model by outcome*
Covariables	Comparison	Model 1: any contraception counselling	Model 2: at least two contraception counselling specialties	Model 3: rheumatology contraception counselling	Model 4: cumulated over multiple to no counselling specialties
Covariables	Comparison	aOR (95% CI)	aOR (95% CI)	aOR (95% CI)	aOR (95% CI)
Age	Age: 10 years older	0.65 (0.48 to 0.87)	0.65 (0.51 to 0.83)	0.72 (0.56 to 0.91)	0.66 (0.53 to 0.81)
Race	Caucasian (ref=African American)	0.58 (0.34 to 0.99)	0.75 (0.47 to 1.19)	0.80 (0.51 to 1.25)	0.73 (0.49 to 1.09)
	Other/Asian/AI/NA (ref=African American)	0.58 (0.31 to 1.08)	0.97 (0.58 to 1.60)	0.75 (0.45 to 1.25)	1.03 (0.66 to 1.62)
Immunosuppression	Hydroxychloroquine (ref=not)	1.24 (0.66 to 2.35)	1.31 (0.76 to 2.28)	1.12 (0.65 to 1.93)	1.29 (0.80 to 2.08)
	Monoclonal Ab (ref=not)	1.18 (0.62 to 2.24)	1.06 (0.62 to 1.79)	1.52 (0.89 to 2.59)	1.02 (0.65 to 1.61)
	Any teratogenic (ref=not)	1.97 (1.20 to 3.24)	2.11 (1.43 to 3.11)	1.81 (1.22 to 2.67)	2.09 (1.48 to 2.95)

Results are based on 25 multiple imputed datasets.
Monoclonal Ab includes rituximab and belimumab.
Any teratogenic immunosuppression includes cyclophosphamide, methotrexate, mycophenolate mofetil and mycophenolic acid.
aORs and 95% CIs are shown in bold when the null effect is excluded at alpha=0.05. aOR >1 indicates higher odds of contraception counselling defined; aOR <1 indicates lower odds of contraception counselling defined, compared with the reference group.
*Models 1–3 results are from three separate multivariable logistic models and model 4 is from a multivariable, cumulative proportional odds logistic model for a four-level, ordinal, polytomous outcome of no contraception counselling, one contraception counselling subspecialty, two contraception counselling specialties, and three or more counselling specialties.
Ab, antibody; AI/NA, American Indian/Native Alaskan; aOR, adjusted OR.