Dynamic Allocation of Limited HIV Prevention Funds Using Control Theoretic Methods

Background:  A robust and theoretically sound science of HIV prevention necessitates methods that predict the effects of real-word decisions, taking into account explicit constraints such as institutional costs and political reality in order to provide meaningful decision support to policy makers. We present an example of such an analysis as a model from the perspective of an institution that is tasked with dynamically allocating limited HIV prevention resources between treatment-as-prevention (TasP) and pre-exposure prophylaxis (PrEP) prevention strategies.    

Methods:  Using a moderately complex HIV transmission model, we modeled 24 scenarios defined by the level of episodic behavioral risk, acute stage infection risk, population stage of epidemic, population endemic risk, baseline population level of viral suppression, and the burden of PrEP cost coverage.  For each scenario, we used open loop methods from control theory to find the optimal control policy (OCP), which is defined by the levels of TasP and PrEP investment at each time point that minimize incident infections while respecting budgetary constraints. 

Results:  We found that the OCP largely depends on PrEP cost coverage.  When PrEP costs are solely covered by the institution, total investment in TasP is almost always the optimal policy. However, when PrEP costs are supplemented by external providers (e.g. insurance companies), the OCP becomes complex, with fluctuating investment levels between TasP and PrEP. The complex solutions are sensitive to the behavioral and population characteristics of the study population.

Conclusions:  We have illustrated how control theory can be applied to a real-world problem that focuses on the perspective of an institution with limited resources. This analysis demonstrates how the cost burden of PrEP interacts with sexual risk dynamics to determine PrEP’s feasibility as a public health intervention