The Centers for Medicare & Medicaid Services (CMS) recently issued a Medicare-related final rule invoking the agency’s statutory authority to promulgate retroactive rules after finding that failure to apply the final rule retroactively would be “contrary to the public interest.” The final rule is expected to face vigorous legal challenges in the coming years.
Of note, such challenges may ultimately provide the Supreme Court of the United States with an opportunity to reexamine a constitutional question whose importance goes beyond just the Medicare program: namely, whether a “public interest” statutory standard—whereby Congress directs an agency to regulate according to what the agency determines to be in the public interest—complies with the constitutional prohibition against Congress delegating its legislative power to agencies.
As we reported back in 2020, CMS issued a proposed rule invoking CMS’s rarely used retroactive-rulemaking authority to essentially ensure that, despite the Supreme Court’s decision in Azar v. Allina Health Services, 139 S. Ct. 1804 (2019), CMS would apply the same Medicare payment methodology found procedurally improper in that decision.
The retroactive-rulemaking power comes from statute. In relevant part, the Medicare Act authorizes CMS to engage in retroactive rulemaking when the agency determines that failure to apply a rule retroactively “would be contrary to the public interest.” 42 U.S.C. § 1395hh(e)(1)(A). We warned then that CMS’s invocation of its retroactive-rulemaking authority set a potentially dangerous precedent that should not go unnoticed by all Medicare stakeholders.
CMS has yet to complete its post-Allina rulemaking. A final rule is expected on or before August of this year. Importantly, however, CMS recently completed a different Medicare rulemaking in which the agency expressly invoked its retroactive-rulemaking authority. Once again, all Medicare stakeholders—as well as those interested in administrative law generally—should take notice.
The RADV Final Rule
On February 1, 2023, CMS published a final rule involving the agency’s controversial Risk Adjustment Data Validation (RADV) program for Medicare Advantage (MA) organizations. It is this final rule that invokes the retroactive-rulemaking power.
In a nutshell, in the RADV program, CMS audits particular contracts between CMS and MA organizations by which MA organizations agree to operate health plans providing Medicare benefits to enrollees otherwise eligible for traditional fee-for-service Medicare. The MA organizations, in turn, contract with numerous unaffiliated third parties to provide health care items and services to the plans’ enrollees. CMS pays MA organizations a per member, per month capitated payment that is risk-adjusted based on enrollees’ diagnosis codes, with higher payments being made to MA organizations for enrollees with diagnosis codes that signify a need for more costly care.
In 2018, CMS released a proposed rule in which the agency explained that it anticipated using sampling and extrapolation in RADV audits dating back to payments made to MA organizations in 2011. In doing so, the agency stated that it would evaluate whether the diagnosis codes for a sample of plan enrollees were supported by the enrollees’ underlying medical records. If errors were found within the sample indicating that the MA organization received what CMS believed to be an overpayment, the proposed rule explained that CMS intended to apply that error rate across the entire contract with the MA organization, resulting in an even larger amount owed by the MA organization. The proposed rule specifically sought comment on whether CMS should invoke its retroactive-rulemaking authority because failure to apply the rule retroactively “would be contrary to the public interest.”
CMS’s recent final rule pulls back somewhat from the agency’s original proposal by announcing that CMS will not engage in contract-wide extrapolation for payment years dating back to 2011. Instead, the agency announced that it would engage in contract-wide extrapolation beginning with payments made in 2018—the same year in which the RADV proposed rule was published. In doing so, CMS expressly invoked the agency’s retroactive-rulemaking authority, stating: “We believe that recovering extrapolated improper amounts . . . advances the public interest by protecting the overall integrity of the MA program.”
RADV Challenges as a Nondelegation-Doctrine Vehicle?
Litigation challenging the RADV final rule is expected in the coming years as CMS begins the process of issuing final audit results to MA organizations and recouping alleged overpayments estimated to exceed $4.5 billion over the next ten years. For example, because the Medicare Act does not expressly authorize CMS to use extrapolation in the MA context, many have questioned whether CMS has the statutory authority to do so. But beyond the typical legal theories that arise in a challenge of any final rule lurks the possibility of something with much more potential significance: a constitutional question on which several Supreme Court Justices have expressed interest recently.
Article I of the United States Constitution provides that “[a]ll legislative Powers herein granted shall be vested in a Congress of the United States.” The nondelegation doctrine, in turn, “bars Congress from transferring its legislative power to another branch of Government.” Gundy v. United States, 139 S. Ct. 2116, 2121 (2019) (plurality opinion for an eight-member Court).
Until recently, the conventional wisdom was that the Supreme Court would find that a statutory delegation is constitutional so long as Congress lays down an “intelligible principle” for the agency in exercising the delegated authority. Id. But recent statements joined by several Justices indicate that there is interest in reexamining whether that is the appropriate legal standard. See Gundy, 139 S. Ct. at 2139 (Gorsuch, J., dissenting) (stating, in a dissenting opinion joined by Chief Justice Roberts and Justice Thomas, that the intelligible-principle standard “has no basis in the original meaning of the Constitution”); see also Gundy, 139 S. Ct. at 2131 (Alito, J., concurring in judgment) (“If a majority of this Court were willing to reconsider the approach we have taken [under the nondelegation doctrine] for the past 84 years, I would support that effort.”). And although the Supreme Court decades ago found that a “public interest” statutory standard did not run afoul of the nondelegation doctrine, see, e.g., N.Y. Cent. Sec. Corp. v. United States, 287 U.S. 12, 24 (1932), recent statements suggest a willingness to reexamine that constitutional question, see Gundy, 139 S. Ct. at 2140 & n.67 (Gorsuch, J., dissenting) (quoting criticism concluding that a “public interest” statutory standard is “is too vague a standard to be left to free-wheeling administrators”) (citation omitted). Moreover, two members of the four-member plurality in Gundy (Justices Ginsburg and Breyer) are no longer on the Court.
Reasonable minds can debate whether the Medicare Act’s public-interest standard for permitting retroactive rulemaking effectively delegates legislative power to CMS. What is in the “public interest” appears to be in the eye of the beholder. In the RADV context, CMS found that the public interest supports recouping what CMS believes to be are overpayments. But CMS just as easily could have determined that certainty in the rules governing the MA program, which involves an annual bidding process by MA organizations dependent on actuarial assumptions, is more important and will promote innovations in the scope of additional benefits offered to enrollees and payment models used by MA organizations with the providers and suppliers with whom the organizations contract.
It is not uncommon for Supreme Court cases to grow out of somewhat-arcane regulatory disputes. For example, the Supreme Court’s opinion striking down the constitutionality of the Line Item Veto Act grew out of President Clinton’s rejection of legislative language resolving an obscure Medicaid-financing dispute. See Clinton v. City of New York, 524 U.S. 417, 423 (1998). Therefore, it would be a mistake to believe that litigation over the RADV final rule—or litigation over the expected post-Allina final rule—cannot eventually produce a vehicle for the Supreme Court to reexamine whether a “public interest” statutory standard passes muster under the nondelegation doctrine. We shall see.