Evaluations of Increasing State Investments in Primary Care: Effects on Health Care Spending, Utilization, and Quality

Harvard Medical School, Boston MA

Investigators

Linked publications & trials

Abstract

Primary care accounts for only 2-7% of U.S. health care spending across payer populations, leading to an underfunded primary care delivery system. In response, states are increasingly adopting policies to increase spending on primary care, with the goal to improve prevention, chronic disease management, and slow the growth of total health care spending.1-5 To date, 4 leading states—Rhode Island, Oregon, Delaware, and Colorado—have enacted legislation or regulations to increase primary care spending, generally by requiring private insurers to explicitly raise the percentage of health care spending spent on primary care. Despite the importance of these state efforts to invest in primary care, their effects on total spending, health care utilization, and quality of care remain largely unknown. This dearth of rigorous evidence is increasingly relevant, as additional states (including California, Maine, Massachusetts, and Pennsylvania) are now debating or designing similar legislation to increase primary care spending. To address this gap in knowledge, we will conduct rigorous quasi-experimental evaluations, using newer matching techniques to improve causal inference, of the Rhode Island, Oregon, Delaware, and Colorado models to raise primary care spending. The initial evaluation of the Rhode Island model, conducted by our team, focused on the model’s short-run effects on commercial insurer spending. We found that Rhode Island’s commercial spending decreased by 9% relative to enrollees from other states, with this reduction driven by the caps on hospital prices in the initial 4-6 years. However, as Rhode Island continues to sustain its primary care investments and other states look to do the same, long-run evidence on the policy’s effect is absent. Moreover, knowledge of its effects on quality of care, use of high and low value care, and spillovers in Medicare and Medicaid patients remains lacking. This project builds on our earlier work in four key ways. First, using large public and private claims data, we will investigate changes in spending, utilization, and quality (both within and outside of primary care) associated with the Oregon, Delaware, and Colorado models—all of which have been enacted. Oregon’s legislation aims to raise primary care spending to 12% of total spending starting in 2023. Delaware’s model, effective 2021, increases primary care spending by 1-1.5 percentage points each year until 2025. Colorado’s model similarly increases primary care spending by 1 percentage point each year through 2023. Second, we will extend our Rhode Island evaluation through a decade following the policy, with the key addition of utilization and quality outcomes. Because it took a few years for the delivery system to respond, as our first study showed, potentially important long-run effects have yet to be evaluated. Third, we use newer statistical methods in matching and weighting within a quasi-experimental design to better isolate the policy effects. Lastly, in all 4 states, we will assess spillover effects of their primary care investment across payers—into the Medicare and Medicaid populations—which is important for both states and the federal government.