The tools are now available to effectively treat HIV; to implement interventions that could prevent it; and even, according to modeling studies, to potentially eliminate it. However, in the Sub-Saharan African (SSA) countries with generalized HIV epidemics where treatment and intervention is most needed, a lack of resources could limit the impact of such programs. The three sequential studies proposed in this application (Specific Aims 1-3, respectively) will combine modeling and statistical analyses to determine how limited resources can be deployed with maximum efficiency, and effectiveness, to achieve clearly defined goals for treatment and prevention goals. We propose, and will test, the Geographic Optimization (GeO) Hypothesis: the efficiency (and thus effectiveness) of HIV treatment programs and interventions in SSA countries with generalized epidemics can be increased by disproportionately allocating resources to areas where incidence is higher than the national average, i.e., by using geographic targeting. We will focus on two different types of generalized HIV epidemics in SSA: predominantly urban (the type found in Botswana) and predominantly rural (the type found in Lesotho). Both epidemics are among the most severe worldwide. We will maximize the significance of our work by collaborating with two NGOs (ACHAP and PIH) that are responsible for designing and implementing HIV treatment and prevention programs in Botswana and Lesotho, respectively. The project links two approaches new to HIV modeling: (1) Geographic analysis By estimating the number, and geographic location, of infected and at-risk individuals (which we will accomplish by analyzing detailed georeferenced HIV data available from both countries), we can target treatment programs and interventions to the communities in geographic regions where they are most needed; (2) Optimization We will use the georeferenced data from both countries to develop mathematical models, which we will analyze using optimization techniques to determine how limited resources can be utilized most effectively. The optimization approaches that we employ will be based on the GeO Hypothesis and tailored to each aim:
in Aim 1 to minimize the probability of interruptions in the treatment supply chain;
in Aim 2 to maximize (given resource constraints) the impact of two types of interventions (Voluntary Male Medical Circumcision and Treatment as Prevention (TasP)) on preventing HIV infections;
in Aim 3 to minimize the resources needed to use TasP-based interventions to eliminate HIV. An interdisciplinary collaborative team will conduct the research, including experts from the countries that are its focus. As we will use data from Botswana and Lesotho to parameterize our models, our results will be particularly informative for their policymakers, enabling evidence-based decision-making. Notably, the methods and models we will develop will be sufficiently general to be useful for the 23 other SSA countries that have generalized epidemics and georeferenced HIV data from Demographic Health Surveys. The questions we address (e.g., how to minimize stock-outs) are relevant for all 25 countries.
This grant will use mathematical models and statistical analyses to design optimal strategies that maximize the efficiency of HIV treatment programs, the effectiveness of HIV interventions for preventing transmission and the efficiency of HIV elimination for countries in Sub-Saharan Africa with limited resources. The project focuses on two different 'types' of generalized HIV epidemics in SSA: predominantly urban (the 'type' found in Botswana) and predominantly rural (the 'type' found in Lesotho). Both epidemics are among the most severe worldwide. We will maximize the significance of our work by collaborating with two NGOs that are responsible for designing and implementing HIV treatment and prevention programs in Botswana and Lesotho, respectively.
