They say that “states are the laboratories of democracy.” In no place is this more evident than the state of Vermont’s efforts to radically change their healthcare delivery system.On October 26, CMS and Vermont’s Green Mountain Care Board (GMCB) jointly announced the Vermont All-Payer Accountable Care Organization (ACO) Model. This new, first of its kind initiative is aimed at accelerating healthcare delivery reform for the entire state and its population by establishing a statewide ACO that is responsible for the health outcomes of its entire population.
After failing to create a single-payer healthcare system due to the $2.6 billion the plan would cost in new taxes, state leaders have pursued more modest proposals that would change how healthcare services in the state are paid with an aim to improve care and reduce costs. Under the new All-Payer ACO model, the GMCB will supervise payment to providers based on patient outcomes instead of the traditional fee-for-service method. The GMCB will oversee the statewide ACO’s payment calculations, quality assurance efforts, and decide which providers are responsible for specific patient populations. The statewide ACO is also risk-bearing – meaning that the state will keep any savings the ACO generates, but it will bear the cost if the ACO is over-budget.
Statewide officials estimate that the all-payer model could hold cost increases to about 3.1 percent, down from the estimated 6 percent growth, saving the state approximately $10 billion over the next decade.
To support this statewide experiment, Vermont will receive more than $50 million in Medicare funding over six years to support its reform efforts with nearly $10 million coming in 2017 as startup funding. According to an October presentation made by the GMCB, the federal government is also willing to provide over $200 million in Medicaid funding capacity to support investments in the ACO and community-based providers.
Perhaps Vermont is on to something. Perhaps not.
While Vermont may have convinced federal regulators, success is not guaranteed. For starters, the ACO model, while it has produced better quality, has not consistently shown that it can reduce costs. A Harvard University study published in the New England Journal of Medicine (NEJM) in April found promising results but cautioned that savings may be slow to develop based on the different types and cohorts of ACOs. The study, which analyzed the spending and care quality of the first 220 ACOs that entered the Medicare Shared Savings Program (MSSP), found that the early MSSP participants reduced Medicare spending, but not among 2013 entrants. The study’s lead author, J. Michael McWilliams, noted that the “results suggest that ACOs with no downside risk can achieve savings, but that savings to Medicare and society may be slow to develop.”
The most recent data published by CMS on the performance of ACOs participating in the MSSP and the Pioneer ACO Model in 2015 support the findings of the NEJM study. While more ACOs are successfully receiving payments from CMS for hitting quality and financial benchmarks, as a whole the MSSP and Pioneer ACOs produced a net loss to CMS of $216 million with wide variation among individual ACOs in their spending and quality performance.
While some question the effectiveness of the model, others are concerned about the potential administrative costs associated with a “top down” ACO model led by state regulators. Some feel that this approach will stifle the local innovation required for success.
Others question neither the model nor the philosophy but point to the failure of Vermont’s Medicare ACOs to generate savings in the past and are concerned that savings will not be realized because Vermont already has a low-cost, efficient system.
Whether you are a believer or not, we can all agree that no silver bullet has yet been found that will lower costs and improve the quality of care.
Only time will tell if Vermont’s model is a success that can be emulated by the rest of the country.
Read more about Vermont’s All-Payer ACO Model here.
Read more on this topic: