The Patient Protection and Affordable Care Act (ACA) of 2010 is introducing sweeping changes to the public and private health insurance systems that affect HIV care providers and patients. Key changes include expanding health insurance coverage through Medicaid expansion, subsidized health insurance offered though new insurance “marketplaces,” and eliminating pre-existing condition restrictions; changes in Medicaid provider reimbursement rates; and eliminating copayments for HIV testing. This legislation is complex and will not be fully implemented until 2019, and there will be significant interstate differences in whether and how states will implement key provisions, such as Medicaid expansion. These ACA implementation differences will affect how people with HIV are tested, diagnosed, linked to care, and reimbursed for medical services, which will ultimately affect state-specific long-term HIV/AIDS epidemiological trends, demand for HIV services, and costs to state and federal government programs.

Many ACA forecasting models do not capture the impact of the complex dynamics of HIV treatment and care, and how these dynamics affect funding needs. For example, micro-simulation models, such as RAND’s COMPARE model, may calculate differences in projected insurance rates for Medicaid expansion and non-expansion states, but may not calculate the impact of other concurrent state policy changes on HIV testing and  treatment. Other forecasting models may calculate the future need for HIV services using HIV prevalence data, but cannot account for cascading changes in the testing, diagnosis, linkage to HIV care, and budget impact to public payers caused by state policy changes, and how those changes subsequently affect HIV transmission.

System dynamics modeling differs from other forecasting models by taking a holistic view of all organizations and processes involved in the system, and incorporating feedback loops, dynamic processes, and nonlinearities in the relationships among variables. Modelers work closely with key stakeholders and experts to develop the system structure and include data from a wide variety of sources. Models aim to understand dynamic implications of policies, explain why and how outcomes may change under different circumstances, and identify potential unintended consequences.

We will describe our development of a system dynamics-based interactive policy evaluation tool to help state and federal planners make resource allocation decisions and understand the long-term, dynamic, and possibly counter-intuitive effects of ACA policies that are implemented at the state level. The model will be able to: (1) evaluate how different state ACA policy options can affect short- and long-term state-specific HIV transmission, testing, and treatment trends; and (2) evaluate short- and long-term changes in demand for specific public programs that provide HIV care, capacity requirements, and (3) show how they may differ across states. We will also discuss how systems modeling can produce state- and national-level estimates of HIV epidemic outcomes that are more realistic than incremental HIV growth estimates based on linear extrapolation from past epidemiological trend data. Additional resources will be provided for people interested in learning more about this new area of evaluation practice.