Number of Accountable Care Organizations Declined During COVID-19

Accountable care organization (ACO) growth has hit a snag again, with the COVID-19 pandemic impacting the number of ACOs in public and private contracts, according to a new analysis.

The analysis published in the Health Affairs blog found that ACO growth has plateaued over the last couple of years, starting in 2019 when CMS introduced higher downside financial risk to its flagship ACO model, the Medicare Shared Savings Program (MSSP). ACOs experienced another decline by the first quarter of 2021 after the COVID-19 pandemic and the early closure of the application period for the MSSP’s 2021 performance year.

The trend is a marked difference after a time of rapid growth from 2010 through 2018. By the latter end of the growth period, there were over 1,000 ACOs in public and private contracts. Currently, there are slightly less than 1,000 ACOs.

What’s more troubling for ACO growth is the recent exist of organizations. The analysis showed that ACO exits in the past two full years have outweighed entrants, which had not happened prior to 2019.

In total, over a quarter of all ACOs (27 percent) have dropped out of participation in all known ACO contracts, the analysis stated.

This trend has particularly impacted Medicare ACO contracts, which have plateaued. Additionally, Medicare’s share of total ACO contracts has declined from 41 percent in early 2018 to just 34 percent by the first quarter of 2021. In contrast, there has been steady growth in both commercial and Medicaid ACO contracts during the period.

Researchers, who hail from Leavitt Partners and the Duke-Margolis Center for Health Policy, attribute the plateau in ACO growth to greater downside risk requirements in the MSSP, which has prompted many ACOs to exit the program altogether, as well as an increase in other value-based payment model options.

Despite stunted ACO growth, the move to value-based payment is strong, they explained. For example, CMS’ Innovation Center (CMMI) has supported dozens of value-based payment models, including the Primary Care First demonstration which has 827 distinct participants and a similar level of accountability as the ACO model, they reported.

Additionally, CMMI introduced Direct Contracting Global and Professional demonstrations this year. The demonstrations build on the ACO model—Next Generation ACO Model, more specifically—to engage more providers in the move to value-based payment. ACOs make up a large portion of the 53 participants of the new demonstrations.

But still, “CMS needs to take specific actions to demonstrate its continued support for value-based payment, which has enjoyed strong support from prior administrations, both Republican and Democrat,” the researchers advised.

“For example, the Innovation Center could announce a path forward for NextGen ACOs and additional support for providers who aim to address social drivers of poor health that have contributed to health inequities,” they added.

Additionally, they suggested that CMS identify more opportunities for multi-payer ACOs models, especially in light of declining Medicare ACO contracts.

“Some models have been developed to encourage multiple payers to come together, with mixed success. Building on lessons learned from its prior multipayer models, and with the support of collaborating organizations such as the Health Care Payment Learning and Action Network, CMS should consider more flexible models and processes that make it easier for multiple payers to adopt similar models,” they wrote in the analysis.

Other recommendations for continued value-based payment success included developing models that address social needs, implementing ACOs based on prospective payments, and increasing the number of mandatory alternative payment models.

“Voluntary participants drop out if they don’t expect economic gains, making it difficult for reforms to reduce costs,” the researchers explained. “Lessons learned from past models can be used to create better mandatory models that are applicable across all Medicare providers. This would also help with evidence generation as it helps to remove selection bias and creates cleaner ‘intervention’ versus control groups.”