A one-year PEPFAR PrEP pause models out to 6,671 extra HIV infections

In January 2025 the US government issued a directive pausing all foreign aid programmes, including a 90-day pause on President's Emergency Plan for AIDS Relief funding for oral HIV pre-exposure prophylaxis, with an exemption for pregnant and breastfeeding women [s1]. A modelling study published in The Lancet HIV on 4 October estimates what a year without that funding would cost in infections.

The answer, under the model's assumptions: 6,671 additional new primary HIV infections (95% uncertainty interval 5,032–8,192), of which 5,663 (4,146–7,074) would occur in key populations [s1]. Counting onward transmission, the authors estimate a one-year pause could produce 10,313 (7,796–12,921) additional infections over the following five years [s1].

The paper describes a return to PrEP funding as "looking increasingly unlikely" at the time of writing [s1].

Who was on PEPFAR-funded PrEP

Towards the end of 2024, 719,384 people who were neither pregnant nor breastfeeding were receiving PEPFAR-funded oral PrEP across 28 sub-Saharan African countries, including 205,868 people from key populations [s1]. The study defines key populations as men who have sex with men, female sex workers, transgender women, and people who inject drugs [s1].

The coverage figures are the part of this study most worth reading closely. Among HIV-negative individuals in key populations, estimated PEPFAR-funded PrEP coverage ranged from 2.6 percent (2.0–3.4) in people who inject drugs to 5.0 percent (4.5–5.9) in female sex workers [s1]. Among non-key-population men, coverage was under 0.1 percent; among non-key-population women, 0.1 percent [s1].

So the programme being paused was reaching, at most, one in twenty people in the groups at highest risk. That is the context for the projected infection numbers being in the thousands rather than the hundreds of thousands — the coverage base was small to begin with.

Where the projected harm concentrates

Relative increases in new HIV infections within key populations were estimated at 0.8 percent (0.3–1.5) for people who inject drugs, 1.4 percent (0.8–2.3) for transgender women, 2.2 percent (1.3–3.0) for men who have sex with men, and 2.9 percent (1.8–4.4) for female sex workers [s1].

Those national-average figures conceal the country-level range. In Zambia, the country with the highest PEPFAR PrEP coverage across all key populations, the estimated increase ranges from 7.8 percent (2.5–17.1) in people who inject drugs to 18.1 percent (9.7–33.2) in men who have sex with men [s1].

The pattern is intuitive once stated: the countries that had built the most PrEP coverage have the most to lose from its withdrawal. Success is what creates exposure to a funding cut.

What kind of study this is

This is a static HIV transmission model, not an observation of what happened [s1]. It was parameterised with estimates of population size, HIV prevalence and incidence, and PrEP effectiveness for different subpopulations in each PEPFAR-funded sub-Saharan African country, and it used PEPFAR's own reporting on people returning for oral PrEP from July to September 2024 as the estimate of how many people the programme was serving [s1]. It was funded by the Wellcome Trust [s1].

A static model does not capture behavioural adaptation, substitution to other prevention methods, or a partial restoration of funding. It assumes the modelled year of removed PrEP is otherwise like the year that preceded it. The authors' conclusion is correspondingly conditional: mitigation measures, whether through alternative international donors or domestic government budgets, are needed to maintain and increase existing coverage [s1].

The number that isn't in this paper

The projection covers oral PrEP only. It does not model treatment interruption, testing, or the wider programme architecture PEPFAR funds, and it should not be read as an estimate of the total health effect of the January 2025 directive.

It also arrives in a year of compounding withdrawals. The United States initiated a year-long withdrawal from the World Health Organization in January 2025, citing financial and governance concerns [s2]. An analysis published on 3 October notes that the US was WHO's largest financial contributor, providing USD 1.284 billion in 2022–23, and that the exit creates uncertainty for pandemic preparedness and health equity programmes on which low- and middle-income countries depend [s2]. That paper is a commentary rather than a quantitative estimate, and it says so — its prescription is innovative collaborative funding models, not a costed alternative [s2].

How to read a projection like this

The 6,671 figure is not a prediction that this many people will be infected. It is the model's answer to a specific counterfactual question — how many additional primary infections follow from removing one year of a programme that covered a defined number of people at a defined level of effectiveness. Change the coverage assumption, the effectiveness assumption, or the duration and the number moves.

What survives those assumptions is the structural point. PrEP coverage in sub-Saharan Africa's highest-risk populations was low, heavily donor-financed, and concentrated in a handful of countries that had built it up. Removing the financing does not distribute the loss evenly; it lands hardest where the programmes worked best.

What to watch

The empirical test of this model is national HIV surveillance data from high-coverage countries over the next two to three years, and whether domestic budgets or other donors filled the PrEP gap in 2025. Neither is available yet.

Sources

• The impact of cuts in the US President's Emergency Plan for AIDS Relief funding for HIV pre-exposure prophylaxis in sub-Saharan Africa: a modelling study — The Lancet HIV, 4 October 2025
• A global health crisis in the making: the US withdrawal from the World Health Organization and its impact on global health equity — Journal of Global Health, 3 October 2025