The statistical illusion of a global HIV endgame: disparities in regional progress toward 95–95–95 targets

Abstract

This study analyzes prevalence growth rates across regions using UNAIDS data and statistical modeling to forecast trends through 2050. Globally, the 95–95–95 targets are projected to be achieved by 2035, 2034, and 2025, respectively, largely due to the success of robust interventions in Eastern and Southern Africa. While some progress has also been made in other regions, including improvements in treatment coverage and testing uptake in parts of Latin America and Asia, significant gaps remain, with a continued high burden and sustained accumulation of new infections in certain countries, where persistent transmission and limited treatment coverage reveal stark disparities. These regional crises challenge the framework’s global applicability, suggesting Africa’s progress cannot be extrapolated universally. Relying solely on the 95–95–95 framework risks creating an illusory sense of achieving an “endgame” without tailored strategies that address diverse epidemiological contexts. This study used a random forest model analyzing feature correlations with HIV growth. The new infection is most strongly correlated, while the first 95 target is also highly correlated, indicating that prevention and high testing significantly reduce the growth rate. The LGBT score and urban population are positively correlated, although unmeasured confounders (e.g., disparities in testing access) may contribute to this association or reflect underlying social network density. The third 95 target displays weaker correlations, showing that HIV growth is driven by those who have not been tested. Prevention efforts should prioritize encouraging untested individuals in urban LGBT populations to undergo testing, underscoring the need for targeted initiatives alongside broader structural interventions.