Table 4.

Context variables associated with screening coverage and cervical cancer mortality in multivariate analyses.

Independent variable	N countries	OR	95% CI	p-value
A. Screening coveragea
Income level				<0.001
High	12	Ref	Ref
Upper-middle	20	0.58	0.38–0.89
Lower-middle	4	0.52	0.25–1.07
Low	1	0.05	0.01–0.29
Predominant health system type				0.02
No-fragmented (NHS/NHI)	19	Ref	Ref
Fragmented (SS/OP)	18	0.59	0.39–0.90
Independent variable	N countries	RR	95%CI	p-value
B. Cervical cancer mortalityb
Screen and treat approach				0.03
No/no program	18	Ref	Ref
Yes	12	1.38	1.03–1.84
Missing	1	–	–
Inequality-adjusted HDI				0.01
Very high/High	5	Ref	Ref
Medium	7	1.92	1.24–3.10
Low	4	1.91	1.15–3.27
Missing	5	–	–

OR: Odds Ratio, RR: Rate Ratio, CI: Confidence Interval, NHS: National Health system, NHI: National Health Insurance, SS: Social Security, OP: Out of pocket, HDI: Human Development Index. Data in bold correspond to statistically significant associations.

^aBeta-regression model.

^bPoisson-regression model. Coverage for women aged 25–65 years, 3-years interval. For cervical cancer mortality screening coverage is included as an independent variable and missing categories are not included in the regression models.