The Centers for Medicare and Medicaid Services (CMS) annually assigns star ratings of 1 to 5 to Medicare Advantage (MA) plans. These ratings can affect plans’ enrollment, continued participation in MA, and plan compensation. Plans with ratings below 3 stars for three consecutive years can be terminated from the program. Under the Affordable Care Act, plans with higher ratings get additional payments, some of which they use to reduce enrollee cost or to provide additional benefits not covered by Medicare, such as dental, hearing, or vision services. In 2021, 80 percent of MA beneficiaries enrolled in a plan with a score of 4 or higher, and these plans received more than $11 billion total in additional payments.

During the COVID-19 pandemic, CMS changed the rules regarding how star ratings were calculated. In June, a federal judge ruled against three Medicare Advantage plans that sought to overturn the CMS rule. In addition to allowing the changes to stand, the ruling could have broader implications on how CMS uses data, potentially altering the way star ratings are calculated. There have been concerns that high ratings are not a good indication of quality.

Changes to 2021 Star-Rating Calculations

In April 2020, CMS issued a rule that modified data submission requirements for 2021 star ratings to address disruption posed by the COVID-19 pandemic. At issue were two datasets that comprise nearly half the star ratings’ data and could not be gathered remotely: the Healthcare Effectiveness Data and Information Set (HEDIS) and the Consumer Assessment of Healthcare Providers and Systems (CAHPS). CMS explained that in-person collection could endanger staff, strain limited resources, and divert attention from patient care.

CMS decided against removing these data altogether since they make up a large part of the total star ratings. Instead, CMS decided to calculate 2021 ratings using HEDIS and CAHPS data collected in 2019. The other half of the 2021 ratings, from data that could be collected safely, would be updated.

From 2020 to 2021, AvMed, a Florida-based health plan with approximately 230,000 members, had an overall rating decline from 4 to 3.5 stars; Prominence HealthFirst, a health plan in Nevada and Texas, saw its rating decline from 3.5 to 3 stars. This resulted in reduced plan payments and bonuses from CMS. The change had even greater consequences for Prominence HealthFirst of Texas, which received an overall rating of 2.5 for the third year in a row, meaning CMS could terminate it from the MA program. In November 2020, all three plans filed a lawsuit against the U.S. Department of Health and Human Services (HHS) claiming they would have received higher scores if the more recent HEDIS and CAHPS data were considered.

Lawsuit Challenging Star Ratings Calculations

The plans asserted that the rule exceeded CMS’s authority under the Medicare statute and was issued in a way that was procedurally faulty; in legal terms, they said the rule change was “arbitrary and capricious.”

The first substantive legal question was whether the Medicare statute allows CMS to use older data to determine star ratings. The statute itself is self-contradictory. It limits permissible data to types collected as of November 1, 2003. In the next provision, however, it requires the Secretary of HHS to consult with MA plans and submit a report to Congress explaining any changes to data collected, implying such changes are possible. The Secretary did not submit a report and argued that the statute applies only to collection of new types of data, not changes to how data already collected are used.

In his decision, Senior U.S. District Judge John Bates agreed that the statutory language is confusing and ambiguous. However, he also found that the Secretary’s interpretation — that no report to Congress was needed because no new data were collected — was reasonable.

The second legal claim — that the new star rating calculation methodology was arbitrary and capricious — is a high legal bar for the plaintiffs to meet. Unsurprisingly, Judge Bates determined that the decision to suspend data collection in person during a pandemic was reasonable, as was the Secretary’s decision to use the “best information available for each measure,” namely new data where safely available, as well as the previous year’s data.

Implications

The plaintiffs could appeal this decision, but it would be a long shot. The agencies were attempting to operate as close to normal as possible during a public health emergency.

The decision may simply be a product of what is reasonable rulemaking during a pandemic, but it implies that CMS could continue to pause the collection of new data in person during the pandemic. The data collected in person include information about claims accuracy and patients’ experiences. Not collecting them could decrease plans’ motivation to invest in these dimensions. Plans that did relatively well based on 2019 data could get a boost — perhaps undeserved —compared to those that did not. At a certain point, CMS may find it hard to justify continued use of 2019 data and will have to devise a new approach, which could include adding new data that are safe to collect during a pandemic, after consulting with stakeholders and explaining changes to Congress.

The decision could have broader import. Judge Bates’s discussion of the Medicare statute’s ambiguity on data collection implies that CMS may, even in normal times, have flexibility in how it uses data to calculate star ratings. As long as no new types of data are collected, the agency does not need to consult with stakeholders and report to Congress. This flexibility could give CMS leeway to use existing data differently to address ratings inflation and research and concerns that high ratings do not proxy quality.

The increased legal scrutiny of the MA star-rating system also could open the door to broader scrutiny of Medicare star ratings, such as recent investigative reporting and litigation about nursing home star ratings. Using measures to signal quality may incorrectly convey information and have an outsize impact on plans’ — or nursing homes’ — enrollment and compensation.