TITLE:   Modeling to inform global health policy: Better use of resources in global HIV programs

SPEAKER: Salal Humair

ABSTRACT:

The fight against HIV has been a hallmark of global health efforts in the last decade. This fight takes place in an environment where evidence, advocacy and policy continually push and pull each other—new evidence about the efficacy of HIV interventions emerges at a rapid pace; pressure for revising policies grows quickly from advocacy groups; and policy makers struggle to manage aspirations within available resources. In this environment, models can help illuminate debates about better use of resources when considering new policies.

For instance, widely-celebrated evidence emerged in 2011 from a randomized controlled trial showing that antiretroviral treatment (ART), used for reducing mortality and morbidity in patients at a relatively advanced stage of HIV, can reduce HIV transmission by as much as 96% if given to patients at a less advanced stage. This was quickly hailed as a game-changer in the fight against HIV, resulting in significant pressure for using ART for prevention (coined treatment-as-prevention or TasP). Policy makers worried about the additional resources that would be needed, but modeling studies argued that TasP is worthwhile, as it will be cost-effective compared to an absolute benchmark of an intervention’s cost-effectiveness (e.g. three times the GDP per capita). I will present a model that allowed us to assess the cost-effectiveness of TasP not against an absolute benchmark but against combinations of other existing HIV interventions, such as ART and medical male circumcision. This model offered a valuable counter-argument to the prevailing narrative and found traction with policy makers.

A broader, and continuing, debate among HIV policy makers concerns allocative efficiency – how to optimally allocate resources among different HIV prevention and treatment interventions over time. While models have addressed optimal allocation of resources between different HIV treatment interventions, or between different HIV prevention interventions, consideration of treatment and prevention together is rare. This is partly because treatment and prevention are viewed as having different objectives—reducing mortality and morbidity vs. averting new infections—so it is not clear how to combine both. Studies based on cost-effectiveness typically suggest putting all resources into prevention, a result difficult to justify in practice. I will present a model in which we formulated the optimal allocation problem in a manner which reveals intuitively and practically appealing results. I will present insights from analysis of simple versions of the model, as well as the results of a computational model built from insights obtained from simpler cases.

I will end the talk by mentioning the breadth of opportunities that exist for modeling in the area of HIV as well as global health in general.