Skip to main content

CMS to MA Plans: Show Us Your Algorithms

Analysis  |  By Laura Beerman  
   August 05, 2022

All of the tech asks from the agency’s new Medicare Advantage (MA) request for information (RFI).

To paraphrase the well-known children’s book, 2022 has been a terrible, horrible, no good, very bad year for MA.

Not in terms of enrollment growth and carrier profits, mind you. But the PR has been dismal—noting that MA plans are overpaid, that prior authorization (PA) requirements contribute to healthcare staff burnout, and, according to an OIG report, that medically necessary services that fall within PA coverage rules are often delayed or denied.

Now, CMS has released a RFI asking MA plans how they can make the program more equitable, accessible, innovative, affordable, and collaborative.

The use of data for MA plan and program improvement is peppered throughout the RFI. Specific and detailed asks concern what plans can do—and what CMS can do—to better leverage data for better outcomes.

Data sought for every objective

  • The best data to advance equity. CMS seeks better data related to race, ethnicity, and language; sexual and gender identity; people with disabilities and language/communication hurdles; cultural identity and religious preferences; socioeconomic need; and people in rural and underserved communities.
  • A focus on socioeconomic data. CMS adds specific questions, including MA plan challenges in “obtaining, leveraging, or sharing such data.”
  • Supplemental benefit use and outcomes. To improve both, CMS asks what “standardized data elements” it could collect and how they would also aid DOH, equity, and cost-sharing burdens.
  • Applications for utilization management (UM). With a Senate bill aimed to improve PA headed to the house, CMS wants to know which of its data, if any, help with UM/PA application and how MA plan data could align for better efficiency.
  • Value-based contracting. Data to assess VBC models within the MA program.
  • Competition dynamics. CMS seeks data on vertical integration and its MA market impact.

Multiple questions on data exchange and interoperability

For its “Drive Innovation to Promote Person-Centered Care” pillar, CMS specifically asks:

  • What are the key technical and other decisions MA plans and providers face with respect to data exchange arrangements to inform population health management and care coordination efforts?
  • How could CMS better support efforts of MA plans and providers to appropriately and effectively collect, transmit, and use appropriate data?
  • What approaches could CMS pursue to advance the interoperability of health information across MA plans and other stakeholders?
  • What opportunities are there for the recently released Trusted Exchange Framework and Common Agreement to support improved health information exchange for use cases relevant to MA plans and providers?

CMS seeks algorithmic intelligence

Healthcare algorithms—understanding them and regulating them—are a growing focus for CMS. In its MA RFI, the agency seeks detailed information on MA plan algorithms, including:

  • The algorithms used to identify in-need members
  • Algorithm prediction targets, such as cost and utilization
  • Algorithm testing and bias related to differential outcomes and how plans mitigate
  • Differential test function, validity, independent evaluation, and reporting

Tech was certainly not the only focus of CMS’ MA RFI. Other components, including health equity, MA plan design, VBC, and market dynamics, will be addressed in depth in the future.

Laura Beerman is a contributing writer for HealthLeaders.


Following a year of intense MA scrutiny, CMS seeks input to improve the program.

Better data is one of the agency's targets, with categories ranging from equity and competition dynamics to supplemental benefits, utilization management, and value-based care.

In addition to other RFI questions on data exchange and interoperability, CMS seeks intel on health plan algorithm use, prediction capabilities, testing, validity, and potential bias.

Get the latest on healthcare leadership in your inbox.