Optimizing HIV Spending Among Key Populations in Five Latin American and Caribbean Countries: A Mathematical Modelling Study.

introductionFrom 2010 to 2024, the Latin America region experienced a 13% increase in new HIV acquisitions, while progress in reducing HIV acquisitions has slowed in the Caribbean. We conducted allocative efficiency studies for five countries in the Latin America and Caribbean region, aiming to estimate how HIV spending could be optimally allocated for maximal impact.

methodsBetween 2021 and 2025, we collaborated with national partners in Colombia, Costa Rica, Dominican Republic, El Salvador and Honduras to develop national Optima HIV models incorporating country-specific key populations. Prevention, testing and antiretroviral therapy (ART) interventions were parameterized using country-specific unit costs, coverage estimates and effectiveness values. For each country, we projected HIV incidence, mortality and treatment resource needs from 2025 to 2030 for multiple scenarios: fixed intervention coverage (status quo), fixed spending allocations (counterfactual), and most recent spending optimized across interventions to minimize cumulative HIV acquisitions and HIV-related deaths.

resultsWith the status quo continued, HIV incidence was projected to decline by 2030 in two countries, stabilize in two and increase in one country. Transgender women and men who have sex with men have the highest incidence rate. Due to ongoing increases in the number of diagnosed people living with HIV, by 2030, annual ART resources will need to increase by 1%-15% to maintain the existing treatment coverage. Compared with the counterfactual (fixed spending) scenario, optimization of existing resources could avert 18%-33% of new HIV acquisitions and 5%-31% of HIV-related deaths over 2025-2030 by prioritizing primarily ART. If modelled ART unit costs reduce by 25%, then 23%-69% of projected HIV acquisitions and 8%-61% of HIV-related deaths could be averted by reinvesting savings to scale-up interventions for country-specific key populations instead of lower-yield interventions such as untargeted HIV testing.

conclusionsThis multi-country modelling study demonstrates the potential to reduce new HIV acquisitions and HIV-related deaths across Latin America and the Caribbean without increasing total spending. Recent reductions in treatment costs provide greater flexibility to invest in key population-focused services which could amplify epidemic impact. The findings underscore the importance of aligning HIV resources with epidemic drivers and expanding targeted prevention for key populations.