. 2020 Apr 1;3(4):e201903. doi: 10.1001/jamanetworkopen.2020.1903

Table 2. Effect of the Intervention on Participation in CCS in the Total Study Population of 10 360 Immigrant Women in Bergen, Norway .

Model	Underwent CCS as of January 2018, No. (%)		OR (95% CI)^a	RD (95% CI)^b
Model	Control (n = 5133)	Intervention (n = 5277)	OR (95% CI)^a	RD (95% CI)^b
Model 1^c	2632 (51.3)	2906 (55.6)	1.24 (1.11-1.38)	2.6 (1.1-4.0)
Model 2^d			1.24 (1.11-1.38)	2.5 (1.2-3.9)
Model 3^e			1.19 (1.06-1.34)	2.0 (0.5-3.5)

Abbreviations: CCS, cervical cancer screening; OR, odds ratio; RD, absolute risk difference.

^{^a}

Odds ratio for intervention vs control for CCS status in January 2018 (after intervention) estimated using random intercept logistic regression to account for clustering.

^{^b}

Risk difference for intervention vs control for CCS status estimated using generalized linear model with identity link function and binomial distribution with clustered robust standard errors.

^{^c}

Adjusted for baseline CCS status (January 2017).

^{^d}

Adjusted for baseline CCS status (January 2017), woman’s age, marital status, income level, and region of origin.

^{^e}

Adjusted for all covariates in model 2 plus additional adjustment for general practitioner’s sex, age, and region of origin.