News & Insights


May 24, 2021

Health Headlines – May 24, 2021

Federal Court Invalidates Medicare DGME Rule that Penalizes Hospitals for Training Fellows—On Monday, May 17, 2021, Judge Timothy Kelly of the United States District Court for the District of Columbia issued a decision setting aside a regulatory formula CMS has adopted to compute Medicare DGME payments to hospitals.  Milton S. Hershey Medical Center, et al., v. Becerra, No. 19-02680 (D.D.C. 2021).  The plaintiffs, consisting of nearly fifty teaching hospitals, alleged that the regulation is contrary to the Medicare statute and otherwise arbitrary and capricious because it penalizes hospitals for training “fellows”—i.e., residents who are training beyond their initial residency period.  The District Court held that the regulation violates the statute and remanded the case back to CMS to recalculate the DGME payments owed to the plaintiff hospitals.  King & Spalding represents thirty-two of the hospitals in the suit.

Medicare makes direct graduate medical education (DGME) payments to hospitals to compensate them for the direct costs they incur hosting physician residency programs, such as resident stipends and teacher salaries.  DGME payments are calculated annually.  One factor that directly affects the amount of DGME payments is the number of residents the hospital trained during the year, as measured in units of full-time equivalents (FTEs).  The number of FTEs a hospital trains is further “weighted” depending on whether residents are in their initial residency period—the minimum number of years necessary to satisfy the requirements for initial board eligibility—or are participating in a training program beyond the residency period to pursue a certification from a subspecialty board.  Residents who train beyond their initial residency period are commonly referred to as “fellows,” and the Medicare statute requires that they be given one-half the weight (0.5 FTE) of residents training in their initial residency periods. 

The Medicare statute also limits the number of resident FTEs that can be used in computing DGME payments.  This limitation is known as the FTE cap.  For most hospitals, the FTE cap is equal to the unweighted count of resident FTEs claimed by the hospital on its 1996 cost report. 

CMS has adopted a regulation to implement these two requirements—giving fellows half the weight of initial residents and applying the FTE cap.  Under that regulation, if a hospital trains resident FTEs in excess of its FTE cap, then a formula is used to adjust the count of FTEs.  This formula sets the hospital’s FTE count as the product of the hospital’s current weighted FTE count (i.e., the FTE count after applying the 0.5 weighting factor for fellows), multiplied by the ratio of the hospital’s historic FTE cap to its current unweighted FTE count.  The cap-adjusted weighted FTE count is used to determine the hospital’s DGME payment for the year.

The plaintiffs in Milton contend that CMS’s formula for calculating the cap-adjusted weighted FTE count is contrary to the Medicare statute and arbitrary and capricious because it has the effect of penalizing hospitals for training fellows—those residents who are in programs beyond their initial residency period.  It is a mathematical certainty that, if a hospital is training residents in excess of its FTE cap, and any of its residents are fellows, then the hospital’s cap-adjusted weighted FTE count will be lower than it would have been had the hospital trained fewer (and in many cases zero) fellows. 

The following hypothetical illustrates the penalty imposed by CMS’s regulation.  Hospitals A and B both have an FTE cap of 100, and both train 150 residents who are in their initial residency period.  Hospital A does not train any additional residents.  Hospital B, on the other hand, also trains 50 fellows who are not in their initial residency period.  After applying CMS’s FTE cap formula (illustrated below), Hospital A’s cap-adjusted weighted FTE count is 100.  Hospital B’s, on the other hand, is only 87.5.  CMS’s FTE cap regulation penalized Hospital B by 12.5 FTEs because it trained 50 fellows.  Had Hospital B trained zero fellows, its cap-adjusted weighted FTE count would have been the same as Hospital A’s.  


Each of the plaintiffs in Milton are teaching hospitals that train residents in excess of their FTE caps and also train fellows.  As a result, they are penalized by CMS’s formula every year.  This prompted the Milton plaintiffs to bring their suit challenging the regulation.

In their federal court briefings, the Milton plaintiffs alleged that CMS’s regulation violates the statute in at least two ways.  First, it is contrary to the statutory weighting factors because it has the effect of weighting fellows at less than 0.5.  Second, it contradicts Congress’s intent to pay—not penalize—hospitals for training fellows, albeit at a lower rate than residents training in their initial residency period.  The Milton plaintiffs also contended that the regulation is arbitrary and capricious because there is no rational basis for penalizing hospitals that train fellows in excess of their FTE caps. 

In its briefing, the government countered that the plaintiffs had waived their claims by not raising them in comments when the regulation was first adopted in 1997.  The government also defended the regulation by arguing that the statute gives CMS wide latitude to adopt rules for calculating DGME payments and does not expressly prohibit CMS from modifying the weighting factors for hospitals that train fellows in excess of its FTE caps. 

The District Court rejected the government’s waiver argument, noting that failure to raise an issue during rulemaking does not foreclose future challenges to those affected by the application of the rule.  Turning to the merits, the court ruled that the regulation violates the statute because it applies a weighting factor to fellows that is lower than the factor prescribed by the statute.  The court acknowledged that the statute gives CMS discretion to adopt DGME payment rules but found that the statute expressly requires those rules to be “consistent with” the DGME provisions of the statute, including the weighting factors.   The District Court further acknowledged that the statute does not expressly prohibit CMS from changing the weighting factors of hospitals training fellows in excess of their FTE caps, but rejected the notion that this rendered the statute silent.  “A statute is not silent simply because it fails to address every conceivable set of circumstances to which it might apply.”

A copy of the District Court’s decision is available here

Reporter, Alek Pivec, Washington D.C., +1 202 626 2914,

OIG Issues Favorable Advisory Opinions Regarding Medigap Plan and Preferred Hospital Organization Arrangements– Last week, OIG posted Advisory Opinions Nos. 21-03, 21-04, and 21-05 relating to three near-identical arrangements between Medigap plans and Preferred Hospital Organizations (PHOs).  The arrangements involve (i) a discount on a policyholder’s Medicare Part A deductible, (ii) a policyholder premium credit, and (iii) a monthly administrative fee to the PHOs. The arrangements are designed to incentivize the Medigap plans’ respective policyholders to seek inpatient care from a hospital within the applicable PHO network.  Although OIG determined that the arrangements would generate prohibited remuneration under the federal Anti-Kickback Statute and the Beneficiary Inducements Civil Monetary Penalties (CMP) law, OIG concluded that the arrangements pose a sufficiently low risk of fraud and abuse and that OIG would not impose administrative sanctions under those laws in connection with the proposed arrangements.

With respect to the deductible discounts, for each arrangement, the PHO’s network hospitals would provide a discount on the Medicare Part A inpatient deductible that the Medigap plan otherwise would cover for any policyholder.  The discount would be established in advance, pursuant to a written agreement between the applicable PHO and each of its network hospitals, and also documented in a separate written agreement between the PHO and the Medigap plan(s).  The discount on the Medicare Part A inpatient deductible offered by individual network hospitals would vary and could be as high as 100 percent.  However, the discount would not vary based on the volume of policyholder claims.  Additionally, each network hospital’s discount would be applied uniformly to all policyholders for a term of at least one year.  Any accredited, Medicare-certified hospital is eligible to participate in the PHO networks if the hospital: (i) meets the licensing and other requirements of applicable state law and (ii) agrees to discount the Medicare Part A inpatient deductible costs on behalf of all licensed offerors of Medigap policies that contract with the PHO, including the Medigap plans.

For the premium credit, the Medigap plans would offer a $100 premium credit to each policyholder who selects a network hospital for a Medicare Part A-covered inpatient stay.  The premium credit would be applied to the next premium payment and would reduce the amount the policyholder would owe.  In nearly all circumstances, the premium credit would not be in the form of a check, deposit, or other affirmative payment.  Policyholders would be eligible to receive only one $100 premium credit per Medicare Part A benefit period.  The Medigap plans would also not advertise the proposed arrangement, in whole or in part, to potential enrollees but would provide information about network hospitals and the premium credit to policyholders on enrollment and through periodic mailings.

For the PHO administrative fee, the applicable PHO would enter into a written agreement with each Medigap plan pursuant to which the Medigap plan would pay the PHO a monthly administrative fee as compensation for establishing the hospital network and arranging for the network hospitals to discount the Medicare Part A inpatient deductible.  The administrative fee would be a percentage-based fee—the PHO would receive a percentage of the aggregate savings that each Medigap plan would realize from the network hospital’s discounts on each Medigap plan’s respective policyholders’ Medicare Part A inpatient deductibles in a given month.  Accordingly, the monthly fee would vary by: (i) the number of policyholder claims for which the network hospitals provided a discount on the Medicare Part A inpatient deductible, and (ii) the amount of the discount on the Medicare Part A inpatient deductible provided by the network hospitals, as established in their respective written agreements with the PHO.  

OIG concluded that all three streams of renumeration (the deductible discount, the premium credit, and the administrative fee) would implicate the federal Anti-Kickback Statute and the premium credit would also implicate the Beneficiary Inducements CMP but, ultimately, the proposed arrangements would provide a sufficiently low risk of fraud and abuse. 

In reaching this conclusion, OIG relied on several factors, including:

  • For both the deductible discounts and the premium credits, OIG believed it was unlikely that those two streams of remuneration would result in overutilization or pose a risk of increased costs to federal healthcare programs.  OIG reasoned that, because it is generally in a Medigap plan’s financial interest to ensure appropriate utilization and costs (because the Medigap plan has financial responsibility for all costs their policies may cover), it is unlikely that the Medigap plans would promote inappropriate utilization by their policyholders.  OIG also considered it unlikely that the premium credit would be an improper inducement to policyholders to use inpatient care because, among other things, the credit would reduce the amount the policyholder would owe, rather than being an affirmative payment such as cash or check. 

  • Additionally, OIG concluded the deductible discounts and premium credits posed minimal potential for patient harm.  The discounts would apply universally to all policyholders.  Additionally, patient choice would not be impacted because policyholders could elect to receive care at a hospital that is not a network hospital.

  • The deductible discounts and premium credits would be unlikely to significantly impact competition because, among other things, the Medigap plans would not advertise any aspect of the proposed arrangement to potential enrollees, and any interested hospital would be eligible to join the PHO network provided the hospital meets the criteria.

  • Although no safe harbor would protect the PHO administrative fee, OIG concluded the PHO administrative fee would be sufficiently low risk under the federal Anti-Kickback Statute.  The Medigap plans and the PHOs certified that the administrative fee would be consistent with fair market value.  Additionally, although the administrative fee would be determined in a manner that takes into account the volume or value of federal healthcare program business, OIG concluded there was a low risk that the methodology for calculating the administrative fee would drive overutilization or result in increased costs to any federal healthcare program because:

    • The PHO administrative fee, while tied to the volume or value of referrals between each Medigap Plan and the network hospitals, ultimately reflects a percentage of the savings realized by the applicable Medigap plan, not revenue generated by the network hospitals;

    • It would be contrary to the Medigap plans’ financial interest, as offerors of Medigap policies with financial responsibility for the cost of certain items and services furnished to their policyholders, to drive overutilization of inpatient hospital services paid for by Medicare Part A; and

    • The Medigap plans certified that they would not pass on or otherwise shift the cost of the PHO’s administrative fee to any federal health care program.

OIG Advisory Opinion 21-03 is available here, Advisory Opinion 21-04 is available here, and Advisory Opinion 21-05 is available here.

Reporters, Isabella E. Wood, Atlanta, + 1 404 572 3527,, and Christopher C. Jew, Los Angeles, + 1 213 443 4336,