Peer-reviewed article

Population impact of South Africa's human papillomavirus (HPV) vaccination programme on HPV prevalence in adolescent girls with and without HIV: a repeat cross-sectional study.

Dorothy A Machalek, Dorothy C Nyemba, Danielle Travill, Kathy Petoumenos, Zizipho Z A Mbulawa et al. · 1 Apr 2026

Abstract

BACKGROUND: A school-based human papillomavirus (HPV) vaccination programme providing protection against types 16 and 18 of HPV was introduced in South Africa in 2014 for Grade 4 girls (aged ≥9 years), achieving 87% coverage among learners with at least one of the two recommended doses. We evaluated the programme's impact on HPV prevalence among adolescent girls in a setting of high HIV prevalence. METHODS: In this repeat cross-sectional study, girls aged 17-18 years were invited from 15 primary health-care clinics in four provinces (Free State, Gauteng, Mpumalanga, and North West) of South Africa to provide a self-collected vaginal sample for HPV testing (AnyPlex II HPV28, Seegene, Seoul, South Korea). A survey done from June 19 to Dec 11, 2019, estimated baseline (pre-vaccine) HPV prevalence. A repeat survey done from Feb 16 to Dec 6, 2023, estimated HPV prevalence in a group eligible for the vaccination programme (post-vaccine). Vaccination status was assessed through district registers and self-reported data. The primary outcome was the impact of the HPV vaccination programme, measured as the relative reduction in HPV prevalence between the two birth cohorts using generalised linear regression (estimating adjusted prevalence ratios), overall and by HIV status. FINDINGS: Of 2470 participants enrolled, 819 girls were recruited for the pre-vaccine survey (248 living with HIV) and 1538 for the post-vaccine survey (295 living with HIV). Prevalence of HPV vaccine types HPV-16 and HPV-18 declined by 83%, from 21·6% (177 of 819 participants) in the pre-vaccine group to 3·2% (49 of 1538 participants) in the post-vaccine group (adjusted prevalence ratio 0·17, 95% CI 0·12-0·24; p<0·0001). A similar reduction was observed among those living with HIV, with prevalence decreasing from 29·4% (73 of 248 participants) in the pre-vaccine group to 4·4% (13 of 295 participants) in the post-vaccine group (adjusted prevalence ratio 0·18, 95% CI 0·10-0·32; p<0·0001). No significant reductions were noted for other HPV types, except HPV-31 and HPV-45, which is consistent with cross-protection. INTERPRETATION: In this large-scale evaluation of South Africa's two-dose HPV vaccination programme, we observed impacts similar to those seen with three-dose programmes in high-income settings, including equivalent impacts among adolescent girls living with HIV. These findings underscore the substantial population-level benefits of high-coverage routine HPV vaccination in a high-HIV-burden setting. FUNDING: Australian National Health and Medical Research Council and the Gates Foundation.