COVID-19’s Impact on ACO REACH Deserves CMS’s Attention

March 20, 2023

CMS must account for inclusion of COVID years in 2023 ACO REACH benchmarks to avoid unfairly penalizing REACH ACOs. 


  • CMS will factor years affected by the COVID-19 pandemic into ACO REACH’s financial benchmarks, creating unfair performance standards because of data anomalies.
  • The current ACO REACH rate book will create arbitrary winners and losers based on vast regional variation in Medicare spending that was exacerbated by the pandemic.
  • CMS should consider approaches to mitigate the potential negative impacts. 

By Samuel Johnmeyer and David Pittman 

Comprised of 132 ACOs and caring for 2.1 million beneficiaries, the ACO REACH Model tests new policies around health equity incentives and primary care capitation, among other things. But preliminary analysis from NAACOS member Physicians of Southwest Washington (PSW) and MultiCare Connected Care in Olympia, Wash., calls attention to a looming issue in the model—including years impacted by the COVID-19 pandemic will create regional variation that could impede ACOs’ success. Specifically, CMS will factor years affected by the COVID-19 pandemic into REACH ACOs’ financial benchmarks, creating unfair performance standards for ACOs because of data anomalies. The pandemic caused preventive, routine care to be missed in 2020, elective surgeries were delayed, and not all of that care was made up for in 2021. Furthermore, there were dramatic variations in spending across the country due to some regions more quickly lifting stay-at-home orders and deferring elective procedures. 

How are ACO REACH benchmarks determined?
The model’s “rate book”, which spells out Medicare fee-for-service spending in each county in the country, is modified from what’s used in Medicare Advantage. CMS uses the rate book to adjust REACH ACOs’ benchmarks for regional spending. The REACH rate book is based on three years’ worth of data (2019–2021) and will include two pandemic-stricken years for Performance Year 2023. Last year, the rate book was based off Medicare spending between 2017 and 2019 and didn’t have this problem. Conversely, the Medicare Advantage rate book is five years (2016–2020), so 2023 only includes one year of pandemic data and the impact of COVID will be mitigated in later years when averaging across five years instead of three. REACH’s benchmarks in 2023 are made up of 60 percent of the historical spending of the ACO’s assigned patients and 40 percent regional expenditures, defined by counties from which the ACO draws patients and determined by the rate book 

What is happening?
In short, there is large variation in regional trends between the 2022 rate book and the 2023 rate book due the inclusion of 2020 and 2021. For example, an average-risk beneficiary in Maine would receive a 1 percent reduction in its regional rate using the 2023 rate book compared with the 2022 rate book. Contrast this with an average-risk beneficiary in Washington, D.C., that would receive an 11 percent increase in its regional rate in 2023 compared with 2022. Puerto Rico (-2 percent) and Maine (-1 percent) would see the smallest year-over-year change, while D.C. (11 percent), New York (7 percent), and California (6 percent) would see the largest increases.  

In the case of Maine, a -1 percent regional trend for 2023 is most useful when you compare it to the national average for the same time period, which was +4 percent. This means an average beneficiary in Maine became 5 percent less expensive relative to an average beneficiary in the United States from 2022 to 2023 and the ACO faces a 5 percent headwind using the 2023 rate book. This is problematic when you consider that regional health care expenditures in 2020 and 2021 were likely influenced by state and local COVID restrictions and cultural norms instead of by structural changes in the way care is delivered in these markets. 

What are the implications?
The current REACH rate book will create arbitrary winners and losers based on vast regional variation in Medicare spending that was exacerbated by the pandemic. Some ACOs will see lower benchmarks in 2023 because of CMS’s including pandemic-stricken years. Let’s say the Maine ACO sees a 2 percent decrease in its benchmark. This could make participation untenable when considering CMS keeps the first 3 percent of savings for itself. The Maine ACO would have to generate 5 percent in savings to collect any reward for its participation. ACOs in the model collectively generated 3.3 percent in savings in 2021, the only year for which data are available.  

Additionally, decreased savings reduces the opportunity for ACOs to reinvest earned shared savings back into patient care through care coordinators, quality improvement efforts, enhanced health information technology, home visits, and other programs that help provider better care at lower costs. 

Policy Considerations
CMS should consider approaches to mitigate the potential negative impacts, such as:

  • Limit losses for entities with negative rate book impacts if COVID years are present.
  • Introduce a rate book corridor, so trends could only increase or decrease by a limited percentage.
  • Expand number of years, such as the five employed by Medicare Advantage, or only include one COVID year.

Release rate book in advance and allow for stakeholder feedback via a comment period prior to finalizing as is also done in Medicare Advantage.