Policymakers need guidance about how to best allocate resources for maximum improvement of population health, particularly in regions of the world where resources are scarce. Choosing between policy options requires taking into account the stream of health gains and costs compared against alternative policy options. This process requires bringing together data, societal values, and health care constraints.
The Optimal Resource Allocation research team is developing a simulation and optimization tool to help decision-makers set priorities. The team’s work will provide guidance to decision-makers investing in prevention, treatment, interventions, and infrastructure, and the uncertainty associated with these decisions.
Key Activities
- Develop a simulation engine to help inform health policy priorities
Analyzing the impact of many choices available to policymakers related to resource allocation, constraints, and desired outcomes depends upon a robust simulation environment that can follow a population and health system simultaneously over time and display the evolution of characteristics and outcomes. Beyond existing simulation approaches in global health, a critical innovation is tracking both the health system and the population simultaneously with their interactions. By regulating the inputs, the user can assess changes in the population characteristics and outcomes accordingly.
The research team is building and testing such a simulation engine by identifying select conditions for which there are data available about the relevant range of possible intervention-platform combinations to address those conditions. The simulation engine will utilize demographic, epidemiologic, platform enhancement, and platform capacity inputs to model the health of a population over time. This research is a component of the Disease Control Priorities Network.
- Apply the simulation engine to assess optimal resource allocation
One of the key questions policymakers ask is, What outcome can I expect from investing in interventions X, Y, and Z in varying combinations, given the current health system? In this instance, any “intervention” is really an intervention-platform combination. For example, antimalarial bed nets could be distributed through different platforms – they could be made available at pharmacies, or they could be handed out door-to-door by community health workers. The impact in terms of disability-adjusted life years averted by each approach could be different. More importantly, policymakers are rarely interested in only one condition. Malaria is not their only concern; they may be focusing on malaria, antenatal care, and HIV. All of these have different relevant intervention-platform combinations; all could be delivered through the health system; and the need for each could vary dramatically from place to place, even within a country.
The research team will create a model, eventually incorporated into a software application, that will allow policymakers to assess the best course of action based upon predicted changes in population health outcomes over time and with a set of political and budgetary constraints.
The analytical challenge arises from the complexity of the task: If there are individual policy options, there are an exponential number of possible combinations. As soon as a plausible number of policy options are included in the calculation, a very large number of combinations should be assessed. Even with a highly efficient health system and population simulation tool, more sophisticated search algorithms will need to be developed to produce useful results. The research team will borrow innovations from other sectors and test their feasibility for the optimal health resource allocation problem.
Impact
The Optimal Resource Allocation research team’s work will provide policymakers with tools and information to compare bundles of policy options with different sets of constraints and in different contexts in order to optimally allocate resources to improve health outcomes.
