Key population HIV estimates in sub- Saharan Africa
File(s)
Author(s)
Stevens, Oliver
Type
Thesis or dissertation
Abstract
Key populations experience disproportionate HIV burden in sub-Saharan Africa, but their inclusion in national HIV estimates has been constrained by fragmented surveillance and limited analytical capacity. At the outset of this thesis, no system collated and validated key population surveillance data; reviews had weak indicator and population coverage; UNAIDS estimates of new infections by population group were poorly substantiated; national programmes had limited ownership and use of data; and methods for translating surveillance data into programme-actionable estimates, including HIV incidence, were unavailable outside data-rich settings. This thesis aimed to lay the foundations for country-specific key-population-stratified estimates of comparable rigour to total population estimates within the UNAIDS estimates process.
In Chapter 2, I collated and validated key population size (KPSE), HIV prevalence, and antiretroviral therapy coverage data collected across sub-Saharan Africa during 2010– 2023, incorporating grey literature omitted from reviews and databases, and produced national estimates for all mainland countries in the region. Chapter 3, detailed the success and failures of the UNAIDS Key Population Workbook, developed to enable national teams to review observation-level data and create nationally-owned key population estimates. Chapter 4 estimated HIV prevalence was 50% higher among transgender women than cis-gender men who have sex with men in sub-Saharan Africa. Chapter 5 triangulated triangulated empirical incidence observations, mathematical models, and surveillance data and indicated that 12% of new HIV acquisitions occurred among key populations, much lower than UNAIDS’ estimate of 25%. Chapter 6 developed a statistical model estimated HIV incidence among female sex workers in Cotonou, Benin, from age- and duration-stratified cross-sectional prevalence data without requiring client surveillance or sexual-mixing data.
In Chapter 2, I collated and validated key population size (KPSE), HIV prevalence, and antiretroviral therapy coverage data collected across sub-Saharan Africa during 2010– 2023, incorporating grey literature omitted from reviews and databases, and produced national estimates for all mainland countries in the region. Chapter 3, detailed the success and failures of the UNAIDS Key Population Workbook, developed to enable national teams to review observation-level data and create nationally-owned key population estimates. Chapter 4 estimated HIV prevalence was 50% higher among transgender women than cis-gender men who have sex with men in sub-Saharan Africa. Chapter 5 triangulated triangulated empirical incidence observations, mathematical models, and surveillance data and indicated that 12% of new HIV acquisitions occurred among key populations, much lower than UNAIDS’ estimate of 25%. Chapter 6 developed a statistical model estimated HIV incidence among female sex workers in Cotonou, Benin, from age- and duration-stratified cross-sectional prevalence data without requiring client surveillance or sexual-mixing data.
Version
Open Access
Date Issued
2025-12-18
Date Awarded
2026-09-01
Copyright Statement
Attribution-NonCommercial 4.0 International Licence (CC BY-NC)
License URL
Advisor
Imai-Eaton, Jeffrey
Publisher Department
School of Public Health
Publisher Institution
Imperial College London
Qualification Level
Doctoral
Qualification Name
Doctor of Philosophy (PhD)