On May 27, the U.S. District Court for the Southern District of Georgia set aside Clover Health’s 2026 Star Rating and ordered CMS to recalculate it. The court held that ten measures relied on data sources the statute does not authorize, and that ten more never went through the notice-and-comment rulemaking the Medicare Act requires. RISE covered the decision on June 5 in a webinar, The Star Ratings Shakeup: Court Challenges CMS Measure Authority, with Stars strategist Melissa Newton Smith, compliance and policy advisor Ana Handshuh, and WilmerHale partner Kevin Lamb. It drew one of the largest single-day audiences in RISE history. The slide deck is available in the RISE on-demand library, however the webinar was not recorded so that all questions could be asked in a closed-door format.
Attendees sent more questions than the session could hold. We grouped them by theme and put them to our faculty, so the answers reach everyone working through this ruling, whether or not they were in the room that day. As always, this is industry education, not legal advice. Consult your own counsel on how the ruling applies to your organization.
How the ruling works
These ten measures are not new. What changed for Clover this year?
Nothing changed in the measure set. What changed is that a plan litigated it. Clover expected roughly four stars, received 3.5, and put the difference at about $120 million in quality bonus and rebate payments. With that much at stake, Clover challenged the legal foundation of the measures rather than the scoring. Recent appellate decisions have made courts more receptive to procedural challenges against CMS, and Clover walked through that door. Expect other plans to study the same playbook.
If CAHPS is an authorized data source, why were the Part D CAHPS measures excluded?
The statute the court applied requires that Star Ratings be "based on" data collected under quality improvement program authority under Part C. The court read "based on" as exclusive. Part D measures rely on data collected under separate Part D authorities (consumer satisfaction surveys, PDE data from payment reconciliation, Part D Plan Reporting, and IRE data) not the Part C quality improvement program. Even when a Part D measure uses CAHPS data, it's collected under Part D authority, not the quality program authority under Part C. That puts it outside the data pipeline the statute names for Star Ratings.
Isn’t Timely Appeals also built on IRE data? And wouldn’t COB and Polypharmacy fall into the same bucket?
Yes, and that is precisely the point to watch. The court ruled only on the twenty measures Clover challenged. It said nothing about Appeals Timeliness, COB, Poly-ACH, or the other measures that draw on data outside HEDIS, HOS, and CAHPS because those measures were not part of the case. But the logic of the ruling does not stop at Clover’s list. Our briefing flagged a set of additional measures vulnerable to the same argument, and any plan weighing its own challenge would be wise to consider that framing.
Does the notice-and-comment holding open the door to challenging almost every measure, even the ones Clover never named?
That is the broader exposure. The court ruled only on the 20 measures Clover challenged, and only on the grounds it needed to decide the case. But the notice-and-comment reasoning does not depend on which measure sits in front of the court. If the annual measure specifications set a substantive legal standard tied to payment, and CMS publishes them as sub-regulatory guidance without rulemaking, the same defect attaches to almost every measure in the set. The faculty named this a structural concern: taken to its conclusion, the second holding reaches well past Clover’s list, which is why a redesign-the-program scenario should be considered a possibility.
If CMS can only use HEDIS, HOS, and CAHPS, doesn’t that invalidate practically all of Part D?
It puts a hard question to most of it. The statute the court applied is a Part C provision, and much of the Part D measure set runs on PDE data, plan-reported data, and Part D CAHPS. The ruling itself reaches only Clover’s 2026 rating, so no other measure is invalidated today. But if the reasoning survives appeal and spreads, the Part D side of the program faces the larger rebuild. That is why our briefing treated “too few measures to produce an Overall Rating” as a scenario worth modeling rather than a hypothetical.
The ACA went through Congress and codified Star Ratings. Why didn’t that update CMS’s authority?
Congress did act, and that is exactly the problem for CMS. When the ACA codified the five-star system, it tied ratings to “data collected under section 1395w-22(e),” the provision the 2003 Medicare Modernization Act had already narrowed to HEDIS, HOS, and CAHPS. Congress also dropped the older language that let the Secretary use whatever measures he or she found appropriate. The court read that as a deliberate choice: Congress tethered Star Ratings to a specific data pipeline, and the agency has no authority to quietly widen it.
How does CMS conduct call center monitoring?
Call center monitoring data is collected by CMS’s contractor from February through May of each year. The contractor monitors whether TTY services and foreign language interpretation were available when calling the health plan’s prospective enrollee customer service phone line. Because the data comes from CMS’s contractor rather than from data plans collect under their quality programs, the court found it outside the authorized pipeline.
What is the difference between notice-and-comment and the Advance Notice and Rate Announcement process CMS already runs?
They are not the same thing, and that gap sits at the heart of the second holding. Notice-and-comment rulemaking under the Administrative Procedure Act and 42 U.S.C. 1395hh means CMS publishes a proposed rule, takes public comment, responds to it, and finalizes a binding regulation. CMS did exactly that for the Star Ratings framework in 2018. The annual measure specifications, the exact numerators, denominators, codes, and survey items, travel a different road: technical notes and the Advance Notice and Rate Announcement. The court held that if the specifications set a substantive standard tied to payment, they need full notice-and-comment, and that step never happened.
What happens to Clover and to ratings mechanics
Could Clover end up with a “Not Rated” designation?
That is the unintuitive result the faculty walked through. Remove the first 10 measures and Clover loses its Part D measures, which means no Part D improvement measure, then no Part D summary rating, and with no Part D summary rating an MA-PD contract has no overall rating. A contract in that position shows a “not enough data” result rather than a star number. So the relief Clover sought to lift a 3.5, run to its logical end, leaves the contract with no overall rating at all. How CMS handles that is one of the unknowns.
Does CMS have leeway to calculate an Overall Rating without the 20 measures? Does Clover still clear the 50 percent threshold?
Mechanically, the rules make it hard. CMS needs a minimum share of scored measures to produce each summary rating, and a contract needs a Part D summary rating to earn an overall rating. Strip the disputed measures and the Part D side falls below what the calculation requires. The faculty were blunt: what is not in dispute is that if the first 10 or all 20 come out, there is no Part D improvement measure, no Part D summary rating, and no overall rating. CMS has discretion in how it implements the order, but the standing methodology offers no obvious way to produce a normal overall rating, absent other methodological adjustments, once those measures are gone.
If a plan ends up with no overall rating, would CMS pay the QBP at the 3.5-star new-plan level?
Unsettled, and the faculty named it as an open question. New plans without a rating receive QBP treatment at a set level while unrated. Whether a previously rated contract that loses its rating through this recalculation gets the same treatment is not clear. One reading says such a contract would not qualify for a bonus at all. Another reads it as a “not enough data” situation that defaults to new-plan QBP treatment. Real dollars turn on which reading CMS and the courts adopt, so model both.
Will CMS recalculate the Part C reward factor for Clover?
It should, because the reward factor depends on the measures that remain. The reward factor rewards consistently high performance with low variation across the measure set. Pull 10 or 20 measures out and the inputs change, which changes whether a contract earns the factor and how large it is. A faithful recalculation reruns the reward factor and the improvement measures on the reduced set rather than bolting the old factor onto new measures. Threshold changes here are material, and any scenario model that skips them understates the swing.
Clover’s claims focus on measures it scored poorly on. How might CMS treat the measures Clover scored well on?
This is the open risk in asking for a recalculation. The court set aside the entire 2026 rating and ordered CMS to redetermine it consistent with the order rather than directing a specific score. A clean recalculation removes the measures the court found unauthorized, which helps Clover. But once CMS reopens the rating, nothing in the order freezes the measures where Clover performed well. Any plan weighing its own challenge should model the full recalculation, not only the measures that hurt it, and know where the rating lands when the math runs in both directions.
Does this apply to all plans, or is it opt-in? If a recalculation dropped a 4.5-star plan to 4, would the plan have to accept it?
The order applies to one plan and one rating: Clover’s 2026 Star Rating. No other plan’s rating changes automatically, and there is no opt-in mechanism. A plan that wants the same relief has to seek it. If CMS instead responds programmatically, the agency will define the terms, and history offers a useful precedent: in the Tukey recalculation, CMS held harmless any plan whose rating would have dropped. Nothing requires CMS to repeat that approach, but upside-only relief is the pattern the industry has seen before.
What happens to the program next
Can CMS do anything about the 2027 Star Ratings coming this October? Could it issue them unchanged while it appeals?
The runway is short. The measurement year is largely closed and the 2027 specifications were set the same way the invalidated ones were. A full cure by October is unrealistic. The likeliest path is that CMS issues the 2027 ratings on schedule, appeals the ruling, and accelerates rulemaking to protect future years. That path makes this October’s release the first Star Ratings published under a legal cloud. Plans should model their 2027 ratings under the scenarios from our briefing rather than wait to see which version arrives.
Could CMS simply swap in compliant data sources to power the same scores?
For some measures, eventually. Concepts like medication adherence could in principle be rebuilt on data collected through plan quality programs (e.g., measure stewards could develop HEDIS-based versions). But measure development runs on multi-year timelines. Some measures have no plausible HEDIS, HOS, or CAHPS analog, since there is no quality-program version of a CMS test call. And under the court’s second holding, any rebuilt measure still has to go through notice-and-comment. This is a years-long fix, not an October fix.
What is the timeline for resolving all the unknowns?
Watch four clocks. First, the appeal: CMS has a limited window to take the case to the Eleventh Circuit, and its decision to appeal or not is the single biggest signal. Second, October 2026: the 2027 Star Ratings release shows us how CMS behaves under the cloud. Third, early 2027: the next rulemaking cycle is CMS’s first clean opportunity to cure the procedural defect prospectively. Fourth, the courts: an appellate decision likely lands 12 to 24 months out, and copycat litigation can reshape the picture at any point in between. Prolonged uncertainty is the planning assumption.
What plans should do
Should we reconsider investing in the affected Part D measures, like MTM, for 2027?
Do not walk away from them. The ruling may be reversed, the underlying activities serve your members and your compliance obligations regardless of whether they score, and MTM program requirements exist independent of Star Ratings. What should change is how you size the investment: model your rating with and without the affected measures, understand where your bonus payments come from under each scenario, and put your incremental dollars in the places that pay off in every version of the future. Today that means the HEDIS, HOS, and CAHPS core, which carries most of the rating weight and the least legal risk.
What advice would you give new plans?
New plans hold an unusual advantage: they receive new-plan QBP treatment while unrated, and they get to build their Stars operating model in plain view of where the program is heading. Build around the durable core of HEDIS, HOS, and CAHPS. Treat Star Ratings as an enterprise priority with a direct line to the C-suite rather than a workplan buried two levels down. Design your data infrastructure to be audit-ready from day one. The plans that struggle through this period will be the ones retrofitting governance onto programs built for a world that just changed.
Does this affect Medicaid?
Not directly. The ruling construes Medicare statutes, and state Medicaid quality programs rest on different authority. The indirect effects are real, though: integrated D-SNP products feel Star Ratings economics, states pay attention to how federal quality measurement holds up in court, and plans that run shared quality infrastructure across lines of business will feel any retooling on the Medicare side.
Wouldn’t a class action or trade-association suit make more sense than every plan suing alone?
Industry-wide relief through a trade association is conceivable, and the economics favor pooling. But plans are not similarly situated. A 4.5-star plan and a 3-star plan want opposite things from a recalculation, which makes a single industry position hard to hold. Watch instead for waves of individual suits built on the Clover template, each tailored to the measures where that plan underperformed and lost money.
Slides from the June 5 briefing are available in the RISE on-demand library. RISE will continue covering the ruling, the appeal (if any), and what both mean for your organization as developments warrant.
RISE provides industry education, not legal or compliance advice. Consult your organization’s legal or compliance advisors regarding your unique circumstances.