Rapid development of an online tracker to communicate the human impact of abruptly halting PEPFAR support

Abstract

On 24 January 2025, the global HIV community was confronted with the abrupt announcement that the U.S. President's Emergency Plan for AIDS Relief (PEPFAR), a programme with longstanding bipartisan support, would be paused for 90 days following an executive order to halt all foreign aid [1, 2]. For the over 20 million individuals worldwide receiving HIV treatment through PEPFAR-supported programmes, the potential consequences were severe and immediate. As the public health community sought mechanisms to respond, an accurate, quantitative and science-based understanding of the potential magnitude of this pause was urgently needed. Systematic quantification of the potential impacts would allow the public health community to plan harm reduction strategies, advocate for policy responses and effectively communicate the severity of the situation to stakeholders, including government officials, civil society leaders, media outlets and advocacy groups.

To provide a rapid assessment, we undertook a near real-time modelling effort. We leveraged existing modelling work, publicly available data and expert consensus to generate projections. Complex mathematical modelling typically requires weeks to generate robust projections, limiting its utility for real-time decision-making. Our focus was on answering the most pressing question—the potential health consequences of this funding pause—using the most parsimonious model possible.

To facilitate accessibility and dissemination, we collaborated with a product development expert to translate these calculations into an interactive website, now available at https://pepfar.impactcounter.com/. The site launched on 28 January 2025, 4 days after the funding freeze was announced [10]. The platform provides an intuitive tool for users to quickly understand the potential consequences of the PEPFAR funding freeze.

As the tool gained visibility and traction, we continued to refine the modelled estimates through additional expert review. Members of the HIV modelling community, including those affiliated with the HIV Modelling Consortium, independently assessed the calculations, refined assumptions and incorporated the latest available data. This collaborative process ensured that the estimates remained as accurate and reliable as possible while maintaining clarity in communication.

The current estimates on adult deaths presented on the website are informed by detailed modelling projections from five well-established HIV mathematical models [11]. We used estimates from Jewell et al. on the number of excess deaths expected to occur over a 1-year time horizon associated with complete ART service disruptions over a 90-day period—as a reduction in ART is what specifically drives short-term mortality. This number was multiplied by the percent of HIV programmes funded by PEPFAR (47%) [12]. We assumed a linear distribution of deaths over the year, resulting in one death every 3.3 minutes.

There may be a 1-month lag between the funding freeze and when deaths begin, given the time required for viral rebound. The tracker, however, provides an understanding of the net current impact of the funding freeze. Additionally, these projections are short-term, but the total long-term impact is likely significantly greater due to onward HIV transmission resulting from service interruptions—and given the time required to re-establish services and re-engage those who have disengaged from the health system. Once service availability is clearer, the increment of deaths will be adjusted accordingly. Modelling the number of new infections from specific prevention and treatment disruptions caused by the funding freeze will become increasingly important.

As of 19 February 2025, the tool had undergone 127 updates, incorporating new insights, improved assumptions and user feedback. This iterative refinement process was made possible by the concerted and collaborative efforts of HIV modellers, clinicians, public health professionals and web developers. It has also been updated in response to direct requests from U.S. congressional staffers to add additional information and impact numbers. Despite the lack of mainstream media coverage, the website has been viewed more than 18,000 times across 153 countries.

At present, the tracker provides estimates based on well-characterized impacts of ART interruptions. As more data become available on the “known unknowns,” further refinements will be incorporated. Key areas of (un)certainty include:

As the situation evolves, and there are updates on the delivery of HIV services, the trackers will continue to be refined with the support of the HIV Modelling Consortium. The online tool serves as a dynamic resource to inform advocacy, policy decisions and global HIV response efforts. Although given rapid policy shifts, the future is not knowable, modelling provides a science-based projection of the consequences of current policy on systems and social conditions. This can give the public health community a shared sense of the magnitude, timing and nature of the crises, enabling collective action.

The authors declare no competing interests.