This implementation research study evaluated the integration of HPV education into existing school-based health education initiatives in Cambodia and Kenya using a mixed-methods, cluster-level difference-in-differences design.
Study setting, school selection, and arm-assignment procedures varied by country. In Cambodia, Kampong Cham and Kratie provinces were purposively selected based on operational feasibility, relatively low HPV vaccination coverage, differing urban-rural and service-access contexts, and support from the School Health Department. From a frame of 730 schools, 46 schools with larger numbers of age-eligible girls were selected and allocated 1:1 using operational district, urbanicity, and random-number ordering. In Kenya, Nairobi, Kilifi, and Migori counties were purposively selected to represent diverse implementation settings. An initial 160 schools were randomly sampled from an operational school-health facility mapping list. In Nairobi and Kilifi, schools were grouped geographically by ward and the resulting groups were randomly assigned to intervention or control. Before baseline, stakeholder validation led to replacement of 25 Nairobi and 6 Kilifi schools, with replacements retaining the original study arm. In Migori, 18 of the original 52 schools were replaced following county validation; the corrected 52-school roster was then randomized 1:1 at school level without ward grouping. The final study included 46 schools in Cambodia (23 intervention, 23 control) and 160 in Kenya (80 intervention, 80 control).
The intervention was educational. In Cambodia, the package targeted Grades 3-5 and integrated HPV content into existing sexual and reproductive health education. In Kenya, it targeted Grades 4-8 during implementation and integrated HPV content into existing school health education; these cohorts progressed to Grades 5-9 by the January 2026 endline assessment. Trained teachers delivered the education during routine school activities. HPV vaccination itself was not assigned or modified by the study; national eligibility, products, schedules, supply, and delivery arrangements remained unchanged.
Quantitative outcomes were assessed at baseline and endline. HPV vaccination coverage among age-eligible girls was assessed primarily from school immunization records, supplemented by linked health-facility records in Kenya where needed; EPI data in Cambodia and KHIS data in Kenya were used for validation and sensitivity analyses. Student knowledge, awareness, attitudes, and sexual and reproductive health knowledge were assessed using independent repeated cross-sectional KAP questionnaires in six schools per country. Country-specific intervention effects were estimated using difference-in-differences analyses.
Qualitative interviews with caregivers, teachers, and healthcare workers assessed HPV-related knowledge and vaccination perceptions, accessibility, and implementation feasibility and acceptability. Cost data were collected to estimate the cost per immunization administered under the intervention model. Actual enrollment comprised participants contributing directly collected KAP or qualitative data; vaccination coverage analyses used de-identified routine records and did not require individual enrollment.
Four Migori schools had intervention exposure that differed from their assigned study arm: two control-assigned schools received the intervention and two intervention-assigned schools did not. All four were retained in their assigned arms for analysis under an intention-to-treat approach.
The study received country-specific ethics approval from the National Ethics Committee for Health Research in Cambodia and the Amref Ethics and Scientific Review Committee in Kenya. Subsequent amendments or continuations were approved before extended study activities, including January 2026 endline data collection.
