Afrocentric-Enhanced Two-Step Flow Communication for Cervical Cancer Screening Among Pastoralist Women in Kenya: A Cluster-Randomized Mixed-Methods Study

Cervical cancer remains a leading cause of cancer-related morbidity and mortality among women in sub-Saharan Africa despite being largely preventable through vaccination, screening, and early treatment. In Kenya, screening uptake is especially low among pastoralist and other marginalized women because of limited awareness, sociocultural barriers, weak health communication, and poor access to preventive services, and conventional biomedical communication often overlooks indigenous knowledge systems and communal decision-making. This study examined the influence of an Afrocentric Two-Step Flow communication strategy delivered by Community Health Promoters on cervical cancer knowledge and screening uptake among nomadic pastoralist women in Isiolo County, Kenya. A mixed-methods design was nested within a cluster randomized trial involving 444 women aged 15-64 years from six Community Health Units; three units received the Afrocentric-enhanced Two-Step Flow intervention through trained Community Health Promoters and three received routine health education. Quantitative data were collected using interviewer-administered questionnaires adapted from the African Women Awareness of CANcer tool and analyzed with descriptive statistics, chi-square tests, and logistic regression in SPSS version 25, together with cluster-level and generalized estimating equations methods to account for randomization at the level of six Community Health Units, while qualitative data from focus group discussions were analyzed thematically using MAXQDA, and the two strands were integrated during interpretation. At endline, cervical cancer screening uptake was 45.9% (102/222) in the intervention arm compared with 18.0% (40/222) in the control arm. Because only six clusters were randomized (three per arm), the conventional individual-level chi-square test (χ² = 39.799; p < 0.001) and crude odds ratio (3.87; 95% CI 2.51-5.96) do not account for clustering and overstate precision; after adjusting for clustering (intracluster correlation = 0.18; design effect = 13.7) using an exact cluster-level randomization test and a multivariable generalized estimating equations model, the intervention remained associated with a substantially higher, though not statistically significant, odds of screening (cluster-adjusted odds ratio 2.8-3.9 depending on method; exact cluster-level p = 0.10), reflecting the limited statistical power of a six-cluster trial rather than an absence of effect. Qualitative findings showed that Community Health Promoters became trusted opinion leaders who embedded cervical cancer education in routine household visits and community dialogue; local language, culturally familiar examples, and repeated interpersonal engagement improved understanding, corrected misconceptions, and reduced fear and stigma. Afrocentric Two-Step Flow communication was associated with a substantial, culturally grounded improvement in cervical cancer knowledge and a large but not definitively confirmed increase in screening uptake; combining culturally grounded messaging with trusted community-based opinion leaders appears to be a potentially, culturally responsive strategy for reducing cervical cancer inequities in underserved populations, warranting confirmation in trials that randomize a larger number of clusters.