Skip to main content
. 2016 Sep 15;16:981. doi: 10.1186/s12889-016-3645-1

Table 2.

Effect estimates for interacting variables in the cervical cancer models limited to complete data (n = 4247)

Associations between educational level or type of health insurance and cervical cancer mortality for each age group (years)
20–24 25–29 30–34 35–39 40–44 45–49
IRR (95 % CI) IRR (95 % CI) IRR (95 % CI) IRR (95 % CI) IRR (95 % CI) IRR (95 % CI)
Model 1
Educational level
 No education vs. higher education 32.5 (7.70–137) 14.3 (6.14–33.5) 14.7 (9.30–23.2) 9.27 (6.30–13.6) 11.4 (7.96–16.3) 7.42 (5.42–10.1)
 Primary vs. higher education 14.1 (5.27–38.0) 12.4 (7.34–21.0) 5.77 (4.08–8.14) 4.98 (3.69–6.73) 5.96 (4.37–8.13) 4.69 (3.58–6.13)
 Secondary vs. higher education 3.87 (1.47–10.2) 3.62 (2.13–6.14) 2.97 (2.10–4.21) 2.65 (1.94–3.60) 2.89 (2.10–3.97) 2.06 (1.54–2.75)
 No education vs. primary education 2.30 (0.68–7.77) 1.15 (0.56–2.40) 2.55 (1.75–3.71) 1.86 (1.37–2.53) 1.91 (1.49–2.45) 1.58 (1.26–1.99)
Model 2
Type of health insurance
 No insurance vs. contributory insurance 1.50 (0.60–3.72) 2.18 (1.43–3.35) 1.54 (1.12–2.13) 1.48 (1.12–1.94) 1.57 (1.21–2.03) 2.21 (1.73–2.81)
 Subsidised vs. contributory insurance 1.96 (0.99–3.85) 1.79 (1.27–2.51) 1.74 (1.36–2.21) 1.55 (1.25–1.91) 1.96 (1.61–2.39) 1.96 (1.63–2.37)
 Special vs. contributory insurance 2.91 (0.64–13.1) 0.29 (0.04–2.10) 0.64 (0.30–1.39) 0.59 (0.32–1.10) 0.93 (0.61–1.42) 0.79 (0.52–1.19)
 Subsidised vs. special insurance 0.67 (0.16–2.82) 6.17 (0.86–44.4) 2.71 (1.26–5.81) 2.61 (1.41–4.83) 2.10 (1.40–3.16) 2.49 (1.6–3.74)
 No insurance vs. special insurance 0.51 (0.11–2.44) 7.54 (1.03–55.2) 2.40 (1.09–5.31) 2.49 (1.31–4.73) 1.68 (1.08–2.61) 2.80 (1.82–4.32)
 No insurance vs. subsidised insurance 0.77 (0.35–1.68) 1.22 (0.83–1.80) 0.89 (0.66–1.20) 0.95 (0.74–1.23) 0.80 (0.63–1.02) 1.12 (0.90–1.40)

IRR Incidence rate ratios, 95 % CI 95 % confidence intervals

Model 1 assessed differences in cervical cancer mortality rates by educational level and Model 2 evaluated differences in mortality rates by type of health insurance. Both multivariable models included fixed effects for age group, urban or rural residence, and region of residence, as well as interactions with age. Only women with complete data for the risk factors of interest were included in these analyses