Socioeconomic, demographic and health-system determinants of the cervical cancer screening knowledge–practice gap in Rwanda: a cross-sectional analysis of the 2025 Rwanda demographic and health survey

Cervical cancer is a leading cause of cancer death among women in Rwanda and across sub-Saharan Africa, despite being largely preventable through vaccination, screening and early treatment. Rwanda has achieved near-universal cervical cancer awareness, yet screening uptake remains low. This study examined the demographic, socioeconomic and health-system determinants of screening uptake and the extent to which knowledge is statistically associated with socioeconomic inequality in uptake. We analyzed 2025 Rwanda Demographic and Health Survey data for 14,274 women aged 15–49 years with complete outcome data. The outcome was ever having been tested for cervical cancer by a health worker. Exposures spanned demographic, socioeconomic, health-system-contact and screening-related knowledge domains. We estimated weighted prevalence and bivariate associations, then fitted multivariable logistic regression in two nested blocks, before and after adding knowledge, to assess attenuation of socioeconomic gradients; standard errors were obtained from cluster-robust models and cross-checked against a full Taylor-linearized survey design incorporating weighting, clustering and stratification. A linear-probability-model decomposition quantified the share of each gradient statistically associated with knowledge and described patterns of statistical association. Data cleaning and analysis were performed using Python 3.14, with a sensitivity analysis conducted in R. 93.4% of women had heard of cervical cancer and 80.4% knew a screening test existed, yet only 19.2% had ever been screened. Screening was most strongly associated with parity (adjusted odds ratio [aOR] 2.74, 95% CI 2.08–3.61 for ≥ 5 versus 0 living children), age, HIV-positive status (aOR 2.89, 95% CI 2.11–3.97), health insurance (aOR 1.32, 95% CI 1.12–1.57) and recent health-facility contact (aOR 1.20, 95% CI 1.07–1.34); education showed no consistent gradient. Screening-related knowledge was the strongest predictor overall (aOR 9.95, 95% CI 6.47–15.29), corresponding to an adjusted probability of screening approximately 18% points higher among women aware of both cervical cancer and the screening test than among women with no awareness. Knowledge statistically accounted for 32–39% of the wealth gradient in screening but only about 9%, or not distinguishably more than none, of the geographic disparity between provinces, which ranged from 11.1% ever-screened in Southern Province to 25.1% in Eastern Province. Rwanda’s cervical cancer screening gap is only partly explained by knowledge. Screening appears to be delivered largely opportunistically through existing health-facility contact, and provincial disparities are statistically independent of knowledge, consistent with (though not proof of) a supply-side health-system capacity gap. Closing this gap and meeting national and WHO 90-70-90 elimination targets will likely require expanding screening access outside these contact pathways and directing targeted attention to underperforming provinces.