Factors associated with cervical cancer screening among women attending maternal child health clinics in Kenya.

Background : Cervical cancer is the second leading malignancy among women in sub-Saharan Africa. Early detection through screening and treatment of pre-cancerous cervical lesions is paramount in preventing cervical cancer. We sought to examine factors associated with ever being screened for cervical cancer among women attending infant immunization clinics in Kenya. Methods: We conducted secondary data analysis from a cross-sectional survey that enrolled mother-infant pairs attending 55 infant immunization clinics in Kenya between 2018 and 2019. Clinics were selected from high and medium/low HIV-burden counties using probability-proportionate-to-size sampling. The primary outcome was self-reported cervical cancer screening. To align with the target age group for cervical cancer screening in Kenya, we excluded women less than 25 years old. Factors associated with cervical cancer screening were assessed using generalized Poisson regression models accounting for survey design. Results: Overall, 970 of 1488 women were aged ≥25 years and had cervical cancer screening information available. Of these, 24% reported history of cervical cancer screening. Of the women screened, 76.5% had one prior screening and 38.7% underwent screening within the previous year. Among women previously screened, 3% reported an abnormal screening result, of which 18% received treatment. Cervical cancer screening was reported in 32.3% of women living with HIV (WLWH) and 16.3% of HIV-uninfected women (P<0.001). In multivariable analysis, screening prevalence was higher in WLWH than HIV-uninfected women (adjusted Prevalence Ratio [aPR]=1.65, 95% confidence interval [CI]: 1.22-2.24, P=0.001), employed than unemployed women (aPR=1.33, 95% CI: 0.99-1.79, P=0.058), and those from high HIV-burden counties compared to medium/low HIV-burden counties (aPR=1.49, 95% CI: 1.11-1.99, P=0.007). Among WLWH, longer antiretroviral therapy duration (>2 years) and being from a high HIV-burden county were associated with screening prevalence, although these associations were attenuated after adjustment (P=0.07 and P=0.08, respectively). Among HIV-uninfected women, older age (≥30 years) and being employed were associated with higher screening prevalence. Conclusion: Cervical cancer screening and linkage to treatment for women with abnormal results was low among women aged ≥25 years attending immunization clinics in Kenya. Higher screening uptake in WLWH and in high HIV-burden counties was presumably due to the integration of cervical cancer screening within HIV care.