U.S. ex rel. Azam Rahimi v. Rite Aid Corp. ( 2021 )


Menu:
  •                                RECOMMENDED FOR PUBLICATION
    Pursuant to Sixth Circuit I.O.P. 32.1(b)
    File Name: 21a0147p.06
    UNITED STATES COURT OF APPEALS
    FOR THE SIXTH CIRCUIT
    ┐
    UNITED STATES OF AMERICA ex rel. AZAM RAHIMI,              │
    Relator-Appellant,        │
    >        No. 20-1063
    │
    v.                                                  │
    │
    │
    RITE AID CORPORATION,
    │
    Defendant-Appellee.        │
    ┘
    Appeal from the United States District Court
    for the Eastern District of Michigan at Detroit.
    No. 2:11-cv-11940—Stephen J. Murphy, III, District Judge.
    Argued: January 26, 2021
    Decided and Filed: June 29, 2021
    Before: BATCHELDER, GRIFFIN, and STRANCH, Circuit Judges.
    _________________
    COUNSEL
    ARGUED: Peter A. Patterson, COOPER & KIRK, PLLC, Washington, D.C., for Appellant.
    William R. Peterson, MORGAN, LEWIS & BOCKIUS LLP, Houston, Texas, for Appellee.
    ON BRIEF: Peter A. Patterson, Charles J. Cooper, Vincent J. Colatriano, COOPER & KIRK,
    PLLC, Washington, D.C., Arun Subramanian, William D. O’Connell, SUSMAN GODFREY
    LLP, New York, New York, for Appellant. William R. Peterson, MORGAN, LEWIS &
    BOCKIUS LLP, Houston, Texas, Kevin J. Biron, Michael J. Ableson, MORGAN, LEWIS &
    BOCKIUS LLP, New York, New York, for Appellee. Frederick M. Morgan, Jr., MORGAN
    VERKAMP, LLC, CINCINNATI, Ohio, for Amicus Curiae.
    No. 20-1063                           U.S. ex rel. Rahimi                                 Page 2
    _________________
    OPINION
    _________________
    GRIFFIN, Circuit Judge.
    Azam Rahimi believes Rite Aid Corporation defrauded the federal government of
    hundreds of millions of dollars by overcharging it for generic prescription drugs, so he filed suit
    under the False Claims Act (FCA) and several state-law analogues. The district court dismissed
    Rahimi’s FCA claim under the Act’s public-disclosure bar and declined to exercise supplemental
    jurisdiction over his remaining claims. We agree and affirm.
    I.
    A.
    Defendant Rite Aid Corporation operates a nationwide chain of pharmacies. Like many
    of its competitors, Rite Aid offers a discount program—which it calls the “Rx Savings
    Program”—that provides its members hundreds of generic prescription drugs at reduced prices.
    The program is free and widely available but excludes customers whose prescriptions are paid in
    full or in part by publicly funded healthcare programs like Medicare Part D, state-administered
    Medicaid programs, or TRICARE.
    Federal regulations require pharmacies such as Rite Aid to dispense prescriptions for
    beneficiaries of these healthcare programs at their “usual and customary charge to the general
    public”—which is often referred to as the “U&C” rate. See 
    42 C.F.R. § 447.512
    (b)(2). In
    general terms, this means that pharmacies cannot charge the government more than the “cash
    price” offered to the public to fill such prescriptions. See United States ex rel. Garbe v. Kmart
    Corp., 
    824 F.3d 632
    , 636 (7th Cir. 2016).
    Relator Azam Rahimi alleges that Rite Aid routinely overbilled these government
    programs because the amounts it charged did not “take into account either the lower Rx Savings
    Program prices or any lower prices that Rite Aid makes available to other payers.” In other
    words, while everyone agrees that Rite Aid was not required to allow beneficiaries of publicly
    No. 20-1063                            U.S. ex rel. Rahimi                                 Page 3
    funded healthcare plans to participate in the Rx Savings Program, Rahimi’s theory is that Rite
    Aid was required to offer the government programs an equivalent-or-better discount because of
    its obligation to provide the U&C rate.
    Rahimi, who is a pharmacist, says he first suspected Rite Aid of overbilling the
    government when he saw Rite Aid’s advertisements for the Rx Savings Program that announced
    publicly funded healthcare programs were specifically excluded from participating. He then
    called a former classmate and current Rite Aid pharmacist in New York, John Doe, to discuss his
    suspicions. Doe told Rahimi that at his pharmacy, ninety to ninety-five percent of Rite Aid’s
    non-insured customers were enrolled in the Rx Savings Program and that Rite Aid’s billing
    software “will only generate the ‘Rx Savings price’ for a customer if the pharmacy has enrolled a
    customer in the program,” and would not generate the price for beneficiaries of government-
    funded healthcare plans. But when Rite Aid generated bills for those covered by publicly funded
    health insurance, it still represented the price to be the U&C rate, even though it “did not include
    the discounts offered to cash-paying customers through the Rx Savings Program.”
    Doe also obtained specific examples of the alleged fraud from his cousin, a Rite Aid
    customer whom he knew to be a New York Medicaid beneficiary. By reviewing his cousin’s
    receipts, Doe confirmed that Rite Aid had charged his cousin more for prescriptions than it
    would members of the Rx Savings Program. Doe relayed this information to Rahimi, who
    further investigated by calling Rite Aid pharmacies in eight other states, inquiring as to the prices
    Medicaid beneficiaries would pay for certain generic medications. “In every single instance,
    [he] learned that the Rite Aid pharmacy was charging Medicaid significantly higher prices for
    the generic medications than the prices made available to their Rx Savings Program members.”
    Rahimi also obtained examples involving Medicare Part D and TRICARE.
    B.
    Rahimi contends this practice violates the False Claims Act, 
    31 U.S.C. § 3729
    (a), and
    other state-law analogues because Rite Aid knowingly caused claims to be submitted for
    reimbursement by the government “that are materially false because they misrepresented
    No. 20-1063                                  U.S. ex rel. Rahimi                                          Page 4
    Defendant’s ‘usual and customary charge’ to the general public.”1 This is because, Rahimi says,
    “Rite Aid consistently charged government health programs higher amounts for the generic
    medications on the Rx Savings Program list than the amounts Rite Aid charged its cash-paying
    customers for the same medications.” So, he commenced this litigation on May 3, 2011.
    A few more pre-litigation facts are in order. First, as required by the FCA, Rahimi
    disclosed the alleged fraud to the government on May 2, 2011. He explained that he learned in
    September 2010 that Rite Aid advertised a savings program, and that “it was possible that Rite
    Aid was not passing on these discounts to Medicaid, as required by many states’ ‘usual and
    customary charge’ billing rules.” But he was uncertain of his theory because he did not know “i)
    the percentage of Rite Aid’s non-insured customers enrolling in the program; or ii) whether Rite
    Aid might be charging Medicaid a similar or even lower charge compared to the Rx Savings
    Program charge.” Rahimi thus recounted the steps he took to investigate his theory, including
    his interactions with John Doe.
    But his disclosure about the program was far from the only one.
    First, from the beginning of Rite Aid’s Rx Savings Program (and before Rahimi’s
    disclosures), Rite Aid had clearly stated in its advertisements and announcements that
    “[p]rescriptions paid for in whole or in part by publicly funded health care programs [were]
    ineligible” for the discounted drug prices.
    Second, the State of Connecticut learned in 2010 (again before Rahimi’s disclosures) that
    several pharmacies were charging Connecticut Medicaid recipients more than their membership
    discount prices, which authorities believed to violate existing law. Nevertheless, Connecticut
    then amended its law to require that pharmacies account for any membership discount program
    when establishing its U&C price for government billing. But rather than reduce the rate it
    1
    The reimbursement aspect of Rahimi’s FCA claim stems from the government-sponsored program’s
    insurance coverage for prescription drug costs. For example, Medicare Part D is a federal, voluntary prescription
    drug benefit program available to persons eligible for Medicare. It is overseen by the Centers for Medicare and
    Medicaid Services (CMS), which contracts with private companies (called Part D Sponsors) to handle claim
    submissions and payment processes for Medicare Part D beneficiaries. Under this model, Part D Sponsors work
    directly with retail pharmacies to provide covered prescriptions at negotiated rates. The Part D Sponsor then tenders
    payment to the pharmacy and seeks reimbursement from CMS. State-administered Medicaid programs and
    TRICARE similarly operate on a reimbursement system.
    No. 20-1063                           U.S. ex rel. Rahimi                               Page 5
    charged the government to match the Rx Savings price, Rite Aid raised the prices charged under
    the Rx Savings Program, in Connecticut only. Other major pharmacies threatened to discontinue
    their membership discount plans in Connecticut entirely. The Connecticut Attorney General
    issued a press release concerning Rite Aid on August 25, 2010 to declare that it was subpoenaing
    Rite Aid for information about changes to its discount drug pricing, which it “falsely blamed on
    a new state law.” The press release summarized the events surrounding the newly enacted law as
    follows:
    The law requires pharmacies to provide Medicaid and other state programs the
    same prescription drug discounts they offer consumers. Apparently in response,
    Rite Aid increased prices and made other changes to its Rx Savings discount drug
    program in Connecticut. The drug store chain posted signs that falsely blamed the
    higher prices and program changes on the new law. . . . All . . . benefits remain
    unchanged for consumers outside Connecticut.
    This development was widely reported in the United States by national and local media.
    Third, the Inspector General of the U.S. Department of Health and Human Services
    announced in October 2009 and October 2010 (yet again before Rahimi’s disclosures) that it
    would be “review[ing] Medicaid claims for generic drugs to determine the extent to which large
    chain pharmacies are billing Medicaid the usual and customary charges for drugs provided under
    their retail discount generic programs.”
    Fourth, a qui tam action was unsealed by the United States District Court for the Central
    District of California the month prior to Rahimi’s complaint being filed, which alleged that
    Kmart Pharmacies were engaging in an identical scheme to overcharge the government for
    prescriptions dispensed to beneficiaries of Medicaid and Medicare Part D by failing to apply a
    discount equal to their membership discount program rate when calculating the U&C charge.
    With this in mind, we turn back to Rahimi’s lawsuit. After he filed his complaint, the
    district court administratively closed the action while the federal government investigated his
    allegations and determined whether to intervene in the suit. More than five years later, the
    government formally declined to intervene in the suit, and the district court lifted the seal and
    authorized Rahimi to serve Rite Aid with the complaint. Rite Aid then moved for judgment on
    the pleadings under Federal Rule of Civil Procedure 12(c), contending the pre-filing facts set
    No. 20-1063                           U.S. ex rel. Rahimi                                 Page 6
    forth above, “[t]aken together,” demonstrated Rahimi could not overcome the Act’s prohibition
    on complaints based on publicly disclosed information, known as the “public disclosure bar.”
    The district court agreed, granted Rite Aid’s motion for judgment on the pleadings on Rahimi’s
    federal claim and declined supplemental jurisdiction over Rahimi’s eighteen state-law claims.
    Rahimi then sought reconsideration, which the court denied. Rahimi timely appealed.
    II.
    The False Claims Act “prohibits submitting false or fraudulent claims for payment to the
    United States, and authorizes qui tam suits, in which private parties bring civil actions in the
    Government’s name.” Schindler Elevator Corp. v. United States ex rel. Kirk, 
    563 U.S. 401
    , 404
    (2011) (internal citations omitted). The Act encourages relators “to act as private attorneys-
    general in bringing suits for the common good,” United States ex rel. Poteet v. Medtronic, Inc.,
    
    552 F.3d 503
    , 507 (6th Cir. 2009) (internal quotation marks omitted), and provides lucrative
    incentives to those who do so, see 
    31 U.S.C. § 3730
    (d)(1)-(2). However, “[t]o guard against
    potential ‘parasitic lawsuits’ and ‘opportunistic plaintiffs,’ Congress included a public-disclosure
    bar in the FCA.” United States ex rel. Maur v. Hage-Korban, 
    981 F.3d 516
    , 521–22 (6th Cir.
    2020) (quoting Poteet, 
    552 F.3d at 507
    ). That provision “bars qui tam actions that merely feed
    off prior public disclosures of fraud.” United States ex rel. Holloway v. Heartland Hospice, Inc.,
    
    960 F.3d 836
    , 843 (6th Cir. 2020). As most recently amended in 2010, the FCA’s public-
    disclosure bar directs that:
    The court shall dismiss an action or claim under [the FCA], unless opposed by the
    Government, if substantially the same allegations or transactions as alleged in the
    action or claim were publicly disclosed—
    (i) in a Federal criminal, civil, or administrative hearing in which the
    Government or its agent is a party;
    (ii) in a congressional, Government Accountability Office, or other
    Federal report, hearing, audit, or investigation; or
    (iii) from the news media,
    unless the . . . person bringing the action is an original source of the
    information.
    
    31 U.S.C. § 3730
    (e)(4)(A) (2010).
    No. 20-1063                              U.S. ex rel. Rahimi                                     Page 7
    We generally apply a three-part test to determine whether the public-disclosure bar
    precludes an otherwise valid FCA claim. “First, we ask whether, before the filing of the qui
    tam complaint, there had been any public disclosures from which fraud might be
    inferred.” Maur, 981 F.3d at 522. Second, we assess how closely related the allegations in the
    complaint are to those in the public disclosures. Id. “And third, we ask whether the qui
    tam plaintiff is nevertheless an original source of the information.” Id. (internal quotation marks
    omitted). We review de novo a district court’s dismissal of a complaint under Rule 12 by way of
    the public-disclosure bar. United States ex rel. Harper v. Muskingum Watershed Conserv. Dist.,
    
    842 F.3d 430
    , 435 (6th Cir. 2016).
    We face one additional complication here. While Rahimi has only one claim under the
    False Claims Act, the alleged fraud occurred both before and after the FCA was amended in
    2010.       Because those amendments made substantive changes to the law, which are not
    retroactive, there are similar but distinct legal tests for pre- and post-amendment conduct. See
    Holloway, 960 F.3d at 843–44. Accordingly, while we retain our usual three-step framework,
    we will identify areas where the legal tests diverge. But “[u]nder either version of the public-
    disclosure bar, [a relator] must demonstrate ‘(1) that the factual premise of [his] claim was not
    publicly disclosed before [he] filed the lawsuit, or (2) even if it was, that [he] was the original
    source of the information.’” Holloway, 960 F.3d at 843 (quoting United States ex rel. Advocates
    for Basic Legal Equal., Inc. v. U.S. Bank, N.A., 
    816 F.3d 428
    , 430 (6th Cir. 2016) (ABLE)).
    A.
    Under either version of the public-disclosure bar, we must first determine whether “there
    had been any public disclosures from which fraud might be inferred” before the filing of the qui
    tam complaint. Maur, 981 F.3d at 522; Poteet, 
    552 F.3d at 511
     (pre-amendment framework).
    A disclosure is public “if it appears in ‘the news media’ or is made ‘in a criminal, civil, or
    administrative hearing, [or] in a congressional, administrative, or Government Accounting Office
    report, audit, or investigation.’” Poteet, 
    552 F.3d at 512
     (quoting § 3730(e)(4)(A) (1986)).2
    2
    The list of potential “public” disclosures shrank with the 2010 amendments to exclude filings and
    rulings associated with state-court proceedings. Compare 
    31 U.S.C. § 3730
    (e)(4)(A) (2010), with 
    31 U.S.C. § 3730
    (e)(4)(A) (1986). That distinction is not relevant to this case.
    No. 20-1063                           U.S. ex rel. Rahimi                                   Page 8
    And a statement or allegation satisfies the inference-of-fraud element if “the information is
    sufficient to put the government on notice of the likelihood of related fraudulent activity.” 
    Id.
    (citation and internal quotation marks omitted).        In other words, the “publicly disclosed
    documents need not use the word ‘fraud,’ but need merely to disclose information which creates
    ‘an inference of impropriety.’” United States ex rel. Burns v. A.D. Roe Co. Inc., 
    186 F.3d 717
    ,
    724 (6th Cir. 1999) (quoting United States ex rel. Jones v. Horizon Healthcare Corp., 
    160 F.3d 326
    , 332 (6th Cir. 1998)).
    Disclosures of fraud generally fall into two categories:
    First, if the information about both a false state of facts and the true state of facts
    has been disclosed, we [will] find that there has been an adequate public
    disclosure because fraud is implied. . . . Second, if there has been a direct
    allegation of fraud, we will find a public disclosure because such an allegation,
    regardless of its specificity, is sufficient to put the government on notice of the
    potential existence of fraud.
    Poteet, 
    552 F.3d at
    512–13 (internal citations and quotation marks omitted; first alteration in
    original). However, a public disclosure can also be piecemeal so long as the multiple sources of
    information reveal the allegation of fraud and its essential elements. “Courts use the following
    formula to explain that concept: If X + Y = Z, Z represents the allegation of fraud and X and Y
    represent its essential elements.” Holloway, 960 F.3d at 844 (quoting Jones, 160 F.3d at 331)
    (internal   quotation   marks    and    brackets   omitted).       “In   order    to   disclose   the
    fraudulent transaction publicly, the combination of X and Y must be revealed, from which
    readers or listeners may infer Z, i.e., the conclusion that fraud has been committed.” Id. (quoting
    Jones, 160 F.3d at 331).
    Turning to the facts of this case, the district court found a public disclosure of fraud for
    two reasons. It first observed that,
    the Connecticut Attorney General’s Office issued a press release on August 25,
    2010, recounted in the news media, announcing an investigation of [Rite Aid].
    The press release stated that Defendant increased its Rx Savings discount program
    prices in Connecticut “[a]pparently in response” to a new Connecticut law
    “requir[ing] pharmacies to provide Medicaid and other state programs the same
    prescription drug discounts they offer consumers.” The press release further
    stated that Defendant “posted signs that falsely blamed the higher prices and
    No. 20-1063                            U.S. ex rel. Rahimi                               Page 9
    program changes on the new law” by claiming that the law required defendant to
    “impose these drug prices increases on Connecticut consumers.”
    Second, it found “impossible to ignore” similarities to United States ex rel. Winkelman v.
    CVS Caremark Corporation, 
    827 F.3d 201
     (1st Cir. 2016), which concluded that the public-
    disclosure bar foreclosed a nearly identical claim brought by a relator against CVS Pharmacy. In
    the words of that court, the press release and coverage “dwelt, with conspicuous clarity, upon
    CVS’s persistent practice of not giving Medicaid the [membership discount program] price.
    Indeed, once the Connecticut legislature amended its Medicaid statutes to mandate that CVS
    provide the [discount rate] to the state’s Medicaid program, CVS threatened to end [the program]
    entirely.” 
    Id. at 209
     (emphasis omitted). The district court thus took its cue from Winkelman
    and concluded that “[t]he revelation that, immediately after Connecticut passed its 2010 law,
    Defendant raised the prices for its discount program only in Connecticut and publicly blamed the
    2010 law for the price increases—just as CVS threatened to terminate its discount program in
    response to the same law” disclosed a fraud.
    Rahimi challenges this conclusion on appeal, reasoning that the whole of the Connecticut
    Attorney General’s press release and surrounding news coverage (the Connecticut Publicity) did
    not adequately disclose the essential elements of any fraud. He says that unlike the similar press
    release involving CVS, which specifically alleged that CVS’s billings to the government violated
    the law, “[t]he issue with Rite Aid was . . . that it had raised prices to its Rx Savings Program
    customers in the state, but was falsely claiming to the public that Connecticut’s new law required
    it do so.” In other words, Rahimi says that the Rite Aid press release was only about “the
    pharmacy’s pretextual use of the new Connecticut law to justify raising prices to customers.”
    Thus, he claims that “[t]here was no suggestion of billing fraud against Rite Aid, let alone any
    disclosure of the ‘essential elements’ underpinning the present lawsuit.”
    We are not persuaded by this view of the facts. The following information was disclosed
    to the public and can be considered together for determining whether there was a public
    disclosure of the essential elements of a fraud:
    No. 20-1063                              U.S. ex rel. Rahimi                              Page 10
    •      Rite Aid excluded Medicare and Medicaid beneficiaries from participating
    in its Rx Savings Program.
    •      Connecticut believed that its pre-existing rules required pharmacies to bill
    its Medicaid program at the lowest drug price they offered consumers,
    including their membership discount programs, and passed the new law
    when some pharmacies disagreed.
    •      In 2010, directly in response to Connecticut’s mandating that pharmacies’
    U&C price track membership discount prices, Rite Aid raised its Rx
    Savings Program prices in Connecticut only.
    As the First Circuit concluded in Winkelman, these facts were sufficient to publicly disclose a
    fraud:
    [I]t requires hardly an inferential step to connect the allegedly true and allegedly
    misrepresented facts. The publicly disclosed materials revealed, quite plainly,
    that CVS was not providing its [membership discount] price as its U&C price to
    Connecticut’s Medicaid program. That is precisely why the Connecticut
    legislature essayed a statutory fix. So, too, those materials revealed Connecticut’s
    belief that the [membership discount] prices should have been provided to the
    state’s Medicaid program even before the statutory change. The allegations and
    transactions that comprised the essential elements of the claimed fraud were in
    plain sight after these disclosures.
    
    Id. at 209
     (internal citation omitted).
    Notwithstanding the minor differences between the CVS and Rite Aid press releases, we
    agree with our sister circuit that the essential elements of the alleged fraud were in plain sight
    after the Connecticut Publicity. The press release and surrounding national news coverage
    disclosed that pharmacies doing business in Connecticut were more explicitly required by the
    newly amended law to charge to the government a price equal to or lower than the discounted
    price paid by participants of the Rx Savings Program. Accordingly, unless Rite Aid had already
    matched its Rx Savings Program price to the U&C price (as Rahimi alleges it had always been
    required to do), it was forced down one of two paths: It either had to lower the U&C rate charged
    to government healthcare plans to match the Rx Savings rate, or it could raise its discount rate,
    so that members of the Rx Savings Program paid an amount equal to the government rate. The
    Connecticut Publicity establishes that Rite Aid took the latter path and falsely blamed the change
    in law for forcing it to raise prices for the Rx Savings Program. But in either circumstance, a
    No. 20-1063                           U.S. ex rel. Rahimi                               Page 11
    change in how Rite Aid priced its generic prescription drugs would reveal both the
    misrepresented facts (that Rite Aid was billing the government at its real U&C rate) and the true
    state of facts (that Rite Aid was charging less for the same drugs when dispensed to members of
    the Rx Savings Program). So even if the Connecticut Publicity did not put a numerical value on
    the difference between the U&C rate charged to the government programs and the lower rate
    Rite Aid charged to members of the Rx Savings Program, it sufficiently disclosed the fraud by
    allowing readers to infer that Rite Aid requested reimbursement from the government for
    prescription drugs at higher prices than it offered through the Rx Savings Program.
    B.
    The second step in the public-disclosure framework is determining whether the public
    disclosures were sufficiently related to the allegations of fraud contained in the qui tam
    complaint. Maur, 981 F.3d at 522.
    Under the pre-amendment FCA, this means the qui tam claim is “supported by the
    previously disclosed information” such that a “substantial identity exists between the publicly
    disclosed allegations or transaction and the qui tam complaint.” Poteet, 
    552 F.3d at 514
    . “In
    applying the substantial-identity test, we held that the relator’s claims are based on prior public
    disclosures where ‘essentially the same . . . scheme’ was ‘the primary focus’ of the prior
    disclosure and the complaint.” Holloway, 960 F.3d at 847 (quoting United States ex rel
    McKenzie v. BellSouth Telecomm., 
    123 F.3d 935
    , 940 (6th Cir. 1997)). Pre-amendment qui
    tam actions are barred if they are “based even partly upon public disclosures.” 
    Id.
     (quoting
    McKenzie, 
    123 F.3d at 940
    ). Under the updated FCA, we assess “whether the allegations in the
    complaint are ‘substantially the same’ as those contained in the public disclosures.” Maur,
    981 F.3d at 522. (quoting Holloway, 960 F.3d at 849). Thus, we have recognized that the post-
    amendment bar is “more lenient” to relators because it requires more similarity between the
    public disclosures and the qui tam allegations. Holloway, 960 F.3d at 849–51.
    1.
    Rahimi contends his qui tam claim was not sufficiently related to the publicly disclosed
    allegations to fall within the ambit of the public-disclosure bar under either version of the FCA.
    No. 20-1063                           U.S. ex rel. Rahimi                                Page 12
    In his view, the district court “wrongly extended its public-disclosure ruling about U&C fraud on
    Connecticut’s Medicaid Program to every single Medicaid program alleged in the complaint . . .
    and federally administered programs[.]” He posits that if “a single news article about a single
    investigation by a single state attorney general constitutes a ‘public disclosure’ of any fraud that
    could be committed by that company in any state in the country—under any state’s rules, even if
    significantly different—the purpose of the FCA is defeated, and relators will have no incentive to
    report fraud in other jurisdictions.” In short, Rahimi emphasizes the breadth of his allegations—
    he says Rite Aid was defrauding the federal government and 18 states, but the Connecticut
    Publicity involved only one of those states and only one of the three healthcare programs
    (Medicaid, and not Medicare Part D or TRICARE). Therefore, in his view, the Connecticut
    Publicity was too narrow to be sufficiently related to the fraud he alleged, so the public-
    disclosure bar does not apply.
    But circuit precedent and the First Circuit’s persuasive opinion in Winkleman foreclose
    this argument. In Holloway, we concluded, as here, that a relator’s claims could not survive the
    public-disclosure bar because his “allegations add[ed] some new details to describe essentially
    the same scheme by the same corporate actor” as the publicly disclosed fraud. 960 F.3d at 851–
    52. We are compelled to reach the same result here.
    And even if not bound by Holloway, Winkleman persuasively explains that once “the
    same fraudulent scheme [] was laid bare in the Connecticut disclosures, the identification of
    additional government programs does nothing more than add a level of detail to knowledge that
    was already in the public domain.” 827 F.3d at 210. It further explained:
    The relators labor to distinguish their complaint from the public disclosures by
    emphasizing its breadth: the Medicare Part D program was never mentioned in the
    Connecticut disclosures, nor did those disclosures aver that CVS was allegedly
    playing fast and loose with the Medicaid program in other states. This argument
    elevates form over substance. When it is already clear from the public disclosures
    that a given requirement common to multiple programs is being violated and that
    the same potentially fraudulent arrangement operates in other states where the
    defendant does business, memorializing those easily inferable deductions in a
    complaint does not suffice to distinguish the relators’ action from the public
    disclosures.
    Id. We see no reason to disagree.
    No. 20-1063                            U.S. ex rel. Rahimi                                Page 13
    2.
    Rahimi adds an additional reason to reject the district court’s conclusion on this issue,
    claiming that variations in the way administering agencies define their U&C rate put his claim
    outside the reach of the public-disclosure bar. But before we can consider the merits of that
    contention, Rahimi has a procedural hurdle—whether he forfeited our consideration of this new
    position by making it for the first time in his motion for reconsideration before the district court.
    See Evanston Ins. Co. v. Cogswell Prop., LLC, 
    683 F.3d 684
    , 692 (6th Cir. 2012) (“Arguments
    raised for the first time in a motion for reconsideration are untimely and forfeited on appeal.”).
    Rahimi first says there is no forfeiture because “Rite Aid . . . did not clearly advance the
    rationale on which the district court relied, [and] did not clearly argue that the Connecticut
    disclosure was itself sufficient to defeat all of those other claims.” We are not persuaded. Once
    Rite Aid raised the public-disclosure bar and pointed to the disclosures it thought were related to
    the fraud Rahimi alleged, he had an opportunity to explain why the court should conclude that
    his claim was not related—for instance, by explaining that each program applied unique U&C
    rules.
    Next, Rahimi claims that he raised this argument in his response to Rite Aid’s motion for
    judgment on the pleadings. But neither of Rahimi’s cited examples explained that the existence
    of various formulations of a U&C price meant that the Connecticut Publicity did not reach the
    fraud he alleged. In his response to the motion for judgment on the pleadings, he argued that the
    Connecticut Publicity was insufficient to “lead[] to an inference that Rite Aid overbilled
    Connecticut Medicaid or other government programs.” In other words, he argued that the
    Connecticut Publicity did not meet the first step of the framework by disclosing any fraud—not
    that the allegations of his qui tam complaint were unrelated to the fraud in the public view. And
    in a court-authorized sur-reply, Rahimi referenced how some state’s Medicaid rules defined
    “usual and customary” to mean “a pharmacy’s lowest charge or require the charge to the
    government to reflect all discounts and special pricing.” But he went no further in explaining
    how any of these differences affected the public disclosure analysis. That is insufficient to
    preserve the argument for review. See FTC v. E.M.A. Nationwide, Inc., 
    767 F.3d 611
    , 630 (6th
    Cir. 2014).
    No. 20-1063                           U.S. ex rel. Rahimi                               Page 14
    Finally, Rahimi points to the district court’s reasoning in denying his motion for
    reconsideration. In his view, “when the district court denied reconsideration, it held that it was
    doing so because, in ruling on Rite Aid’s motion, it had already ‘explicitly considered in detail’
    the argument based on differences in program billing requirements.” But this mischaracterizes
    what the district court said:
    [T]he motion raises two issues, both of which were explicitly considered in detail
    and resolved in the Court’s initial order.
    First, Relator took issue with the Court’s analysis of his claims in light of
    [Winkelman]. . . .
    Second, Relator disagreed with the court’s discretionary decision not to exercise
    supplemental jurisdiction over state law claims under the laws of numerous states
    in a case in which there are no remaining federal claims.”
    The court did not say that it had “explicitly considered in detail” anything relating to the
    differing U&C definitions.
    We therefore conclude that Rahimi forfeited the argument that the public-disclosure bar
    does not apply because the unique application of varying U&C definitions meant that the fraud
    he alleged was not sufficiently related to the information in the public domain. Further, we
    decline Rahimi’s invitation to excuse the forfeiture because failing to consider his argument will
    not result in “a plain miscarriage of justice.” Scottsdale Ins. Co. v. Flowers, 
    513 F.3d 546
    , 552
    (6th Cir. 2008) (citation and internal quotation marks omitted).
    C.
    Having concluded Rahimi’s claim alleges fraud that is sufficiently close to transactions
    already in the public domain, the final step in the analysis is determining whether he may
    nonetheless pursue his claim as an original source. See 
    31 U.S.C. § 3730
    (e)(4)(A)–(B).
    1.
    We must first address an apparent conflict between the Supreme Court’s interpretation of
    the pre-amendment original-source exception and our caselaw. The text of the 1986 version of
    the FCA provided that an original source is an individual: (1) with “direct and independent
    knowledge of the information on which the allegations are based;” and (2) who has “voluntarily
    No. 20-1063                            U.S. ex rel. Rahimi                                 Page 15
    provided the information to the government before filing an action under [the FCA] which is
    based upon the information.” 
    31 U.S.C. § 3730
    (e)(4)(B) (1986). Our court has understood this
    provision to require an original source to disclose his information to the government before any
    public disclosure of the allegations, McKenzie, 
    123 F.3d at 942
    , an interpretation that we adopted
    after consideration of the D.C. Circuit’s opinion in United States ex rel. Findley v. FPC-Boron
    Employees’ Club, 
    105 F.3d 675
     (D.C. Cir. 1997). And it appears we are the only two circuits to
    read the 1986 version of the Act this way. See McKenzie, 
    123 F.3d at
    941–42 (noting circuit
    split).
    Subsequently, in Rockwell International Corporation v. United States, 
    549 U.S. 457
    (2007), the Supreme Court interpreted § 3730(e)(4) in a manner that was plainly inconsistent
    with McKenzie’s view of the text. Rockwell held that the original-source exception applies to
    “information upon which the relators’ allegations are based” rather than “information underlying
    the public disclosure.” Id. at 470–71. In so reasoning, the Court compared § 3730(e)(4)(A)’s bar
    on actions based on “the public disclosure of allegations or transactions” with subparagraph
    (B)’s original-source exception for persons having “direct and independent knowledge of the
    information on which the allegations are based.” § 3730(e)(4)(B). The Court first explained that
    subparagraph (B)’s “information on which the allegations are based” must mean the facts
    underlying the relator’s alleged fraud. Id. at 470–71. Second, it rejected the reading that the
    “allegations or transactions” referenced in subparagraph (A) must be the same as the
    “allegations” referenced in subparagraph (B), a reading which would make “original-source
    status [dependent] on knowledge of information underlying the publicly disclosed allegations.”
    Id. at 471. It reasoned that Congress would have used the identical phrase in subparagraph (B) if
    it had “wanted to link original-source status to information underlying the public disclosure.” Id.
    Thus, the Court drew a distinction between public allegations and the information underlying the
    public allegations.   Id. at 471–72.     In sum, the Court held a relator may make the same
    allegations as disclosed in a prior public disclosure and still qualify for original source status, so
    long as the relator had direct and independent knowledge of the information underlying his
    allegations.    See id. at 472 (“To bar a relator with direct and independent knowledge of
    information underlying his allegations just because no one can know what information underlies
    the similar allegations of some other person simply makes no sense.”).
    No. 20-1063                                  U.S. ex rel. Rahimi                                         Page 16
    Although the D.C. Circuit has subsequently recognized that requiring a relator to furnish
    his information to the government before any public disclosure is contrary to Rockwell (and
    therefore no longer follows its Findley decision), see United States ex rel. Davis v. District of
    Columbia, 
    679 F.3d 832
    , 838–39 (D.C. Cir. 2012), we have continued to apply the
    Findley/McKenzie rule. See United States v. Garman, 719 F. App’x 459, 463–64 (6th Cir.
    2017); United States ex rel. Antoon v. Cleveland Clinic Found., 
    788 F.3d 605
    , 617 (6th Cir.
    2015); Poteet, 
    552 F.3d at 515
    .3 So the district court considered itself bound to follow that
    precedent and concluded that Rahimi could not qualify as an original source under the 1986
    version of the Act.
    We agree with the D.C. Circuit’s conclusion in Davis and now hold McKenzie’s
    interpretation of the original-source exception is incompatible with Rockwell. And although we
    have applied McKenzie post-Rockwell, we are not bound by our continued, uncritical application
    of the Findley/McKenzie rule because intervening Supreme Court decisions allow a panel of our
    court to revisit prior precedent, “even in the unusual situation where binding circuit precedent
    overlooked earlier Supreme Court authority.”                  Ne. OH. Coal. for the Homeless v. Husted,
    
    831 F.3d 686
    , 720 (6th Cir. 2016). This is one of those unusual cases. Accordingly, when
    applying the 1986 version of the Act, we will no longer require that a relator provide information
    to the government prior to any public disclosure to qualify as an original source.
    2.
    While the district court relied on McKenzie to conclude that Rahimi could not qualify as
    an original source under the pre-amendment FCA, we affirm on the alternative grounds that
    Rahimi did not have direct knowledge of the information on which his allegations were based.
    See § 3730(e)(4)(B) (1986).
    We have explained that “direct” knowledge does not require that the relator be a
    corporate insider because “first-hand knowledge is not a necessary component of direct
    knowledge.” Antoon, 788 F.3d at 618 (emphasis omitted). “For instance, a relator may have
    3
    There appears to be a reason for this. Upon review of the briefs in those cases, we note the parties failed
    to bring the conflict between Rockwell and McKenzie to our attention, and so gave us no reason to reconsider our
    prior authority.
    No. 20-1063                            U.S. ex rel. Rahimi                              Page 17
    direct (but not first-hand) knowledge of the billing practices of an institution, and uncover fraud
    only after consulting a public document that reveals those practices are fraudulent.” Id. at 618–
    19. We thus interpret § 3730(e)(4)(B) to mean “direct knowledge is knowledge gained by the
    relator’s own efforts and not acquired from the labor of other people,” id. at 619 (internal
    quotation marks omitted), and a relator’s knowledge is “independent” if it “does not depend or
    rely upon the public disclosure.” Id. at 617 (internal citation marks and citation omitted).
    Finally, we are mindful that “[e]ach case is different and must be analyzed to assess the degree of
    the relator’s original input into the facts disclosed to the government.” Id. at 619.
    We conclude that the majority of Rahimi’s information was not gained by his own efforts
    and instead came from John Doe (and Doe’s cousin) and therefore cannot itself be considered
    “direct.” On this record, this information—including that a particular Rite Aid pharmacy in New
    York was charging Medicaid beneficiaries more than the Rx Savings Program rate and that the
    overwhelmingly majority of cash purchasers at the same pharmacy were members of the Rx
    Savings Program—was not the fruit of Rahimi’s labor. Id.; see also United States ex rel. Reagan
    v. E. Tex. Med. Ctr. Reg’l Healthcare Sys., 
    384 F.3d 168
    , 177 (5th Cir. 2004) (“The plain
    meaning of the term ‘direct’ requires knowledge derived from the source without interruption or
    gained by the relator’s own efforts rather than learned second-hand through the efforts of
    others.”).
    Beyond this second-hand information, Rahimi’s own investigation boiled down to calling
    various Rite Aid pharmacies and inquiring about particular drug prices charged to customers, not
    the prices submitted to the government for reimbursement. We have explained that “[t]he
    purposes of the Act would not be served by allowing a relator to maintain a qui tam suit based on
    pure speculation or conjecture.” Antoon, 788 F.3d at 620 (quoting United States ex rel. Aflatooni
    v. Kitsap Physicians Servs., 
    163 F.3d 516
    , 526 (9th Cir. 1999)). In Antoon, this principle meant
    that a patient did not qualify as an original source where he suspected that his doctor had not
    actually performed a medical procedure based on his own pre- and post-operation observations
    and medical records but had no further information to establish whether the doctor was
    personally involved with the surgery. 
    Id.
     at 619–20. We said that, although the relator “suspects
    No. 20-1063                            U.S. ex rel. Rahimi                                Page 18
    that [his doctor] submitted a false claim, mere suspicion is insufficient to qualify” that relator as
    an original source. 
    Id. at 620
    .
    The same holds true here. Until Rahimi obtained evidence of individual claims that Rite
    Aid submitted to state Medicaid programs (included in his Third Amended Complaint), Rahimi
    could only suspect that Rite Aid was submitting false claims to the government based on the
    rates quoted to him by Rite Aid pharmacists for the prices beneficiaries of publicly funded
    insurance would pay. And because that further evidence was not part of Rahimi’s disclosures to
    the government, it cannot be considered for determining whether he was an original source at the
    time he filed suit. See 
    31 U.S.C. § 3730
    (e)(B)(4)(B).
    While mere reference to public documents is not an automatic bar to original-source
    status, the inquiry is fact-dependent and must be analyzed as to the degree of the relator’s
    original input into the facts disclosed to the government. Antoon, 788 F.3d at 618 (collecting
    cases). Here, Rahimi’s original input consists solely of putting more flesh on the fraud scheme,
    of which the bones were already public. It would be contrary to the purpose of the Act to extend
    the original source exception to such activities. Accordingly, we hold that Rahimi does not
    qualify as an original source under the 1986 version of the Act.
    3.
    The FCA now provides:
    For purposes of this paragraph, “original source” means an individual who either
    (i) prior to a public disclosure under subsection (e)(4)(A), has voluntarily
    disclosed to the Government the information on which allegations or transactions
    in a claim are based, or (2) who has knowledge that is independent of and
    materially adds to the publicly disclosed allegations or transactions, and who has
    voluntarily provided the information to the Government before filing an action
    under this section.
    31 U.S.C. 3730(e)(4)(B) (2010). Therefore, the original source exception is now explicitly tied
    to public disclosures, but Congress also provided a safety valve for relators who do not furnish
    the information to the government before a public disclosure but have “knowledge that is
    independent of and materially adds to the publicly disclosed allegations.” Id. “Materiality in
    this setting requires the claimant to show it had information ‘[o]f such a nature that knowledge of
    No. 20-1063                          U.S. ex rel. Rahimi                                Page 19
    the item would affect a person’s decision-making,’ is ‘significant,’ or is ‘essential.’” ABLE, 816
    F.3d at 431 (quoting Black’s Law Dictionary 1124 (10th ed. 2014)). “In other words, the relator
    must bring something to the table that would add value for the government.” Maur, 981 F.3d at
    527.
    The district court determined that Rahimi’s allegations failed the materiality requirement
    because the general fraudulent scheme “was easily inferred from the Connecticut publicity[,]
    [a]nd Relator’s specific examples of Defendant’s implementation of the scheme [did] not
    materially add to the public disclosures.” Rahimi now argues that his allegations were material
    because “[n]othing in the public disclosures spoke to whether the Rx Savings price was the usual
    and customary price that Rite Aid charged, or whether Rite Aid’s billing systems took those
    prices into account when computing charges to government payers.” Thus, the argument goes
    that “[w]ithout that information, neither the public, nor the government payers, had sufficient
    information to determine whether Rite Aid was liable for violating U&C billing rules.”
    We first observe that Rahimi is attempting to rewrite what he disclosed to the
    government. He contends on appeal that his pre-suit disclosure “reveal[ed] whether the Rx
    Savings Program price was so prevalent that it had become the usual and customary price under
    the rules of almost every government health program.” But his pre-suit disclosure did nothing of
    the sort; he reported to the government that he thought the Rx Savings Program price became the
    U&C price once it was “offered” to plan participants. The only information he relayed about
    public participation in the Rx Savings Program was that John Doe had reported that ninety
    percent of cash buyers at his pharmacy were enrolled—he did not disclose any information about
    broader consumer participation in the Rx Savings Program.
    But more to the merits, relator’s allegations do not materially add to the public
    disclosures because the government was already on notice that Rite Aid had not been applying
    its Rx Savings Programs discount when making its U&C rate calculation by operation of the
    Connecticut Publicity. Even though Rahimi was able to add additional state-specific information
    and examples of government beneficiaries paying more than its Rx Savings Program rate,
    “[o]ffering specific examples [of the alleged fraud] does not provide any significant new
    information where the underlying conduct already has been publicly disclosed.” Winkelman,
    No. 20-1063                            U.S. ex rel. Rahimi                                 Page 20
    827 F.3d at 212; see also ABLE, 816 F.3d at 432 (“A qui tam plaintiff ‘is not allowed to proceed
    independently if it merely adds details to what is already known in outline.” (quoting United
    States ex rel. Bogina v. Medline Indus., Inc., 
    809 F.3d 365
    , 370 (7th Cir. 2016) (internal
    quotations marks and brackets omitted))); Maur, 981 F.3d at 528 (“[B]y merely providing
    additional instances of the same type of fraud . . . , Maur has failed to offer information of ‘such
    a nature that knowledge’ of it ‘would affect’ the ‘government’s decision-making.’” (quoting
    ABLE, 816 F.3d at 431–32)). In sum, we agree with the district court that Rahimi proffered no
    information to change the government’s thinking or decision-making with respect to the alleged
    fraud. He therefore does not qualify as an original source under the 2010 version of the FCA.
    D.
    For these reasons, the district court properly applied the FCA’s public-disclosure bar to
    Rahimi’s FCA claim.
    III.
    Finally, we consider whether the district court abused its discretion by declining
    supplemental jurisdiction over four of Rahimi’s state-law claims. In declining supplemental
    jurisdiction, the district court determined that it should, in most cases, avoid deciding novel
    issues of state law and further observed that the general practice of federal district courts is to
    dismiss supplemental state-law claims if the federal claims are dismissed before trial.
    Rahimi argues that even if the district court properly dismissed his federal claim, it
    should have exercised supplemental jurisdiction over his equivalent claims brought under state
    law (and for which, unlike the federal government here, the states waived application of the
    public-disclosure bars). Generally, he argues that the district court did not adequately balance
    the interests because it failed to give due weight to judicial economy (the case had been pending
    in district court for almost nine years) and the avoidance of multiplicity of litigation.
    We are unpersuaded. Even though the seal had been lifted for more than three years, the
    case was still about two years away from its scheduled trial date when the district court granted
    Rite Aid’s motion for judgment on the pleadings. And while the parties had exchanged some
    No. 20-1063                           U.S. ex rel. Rahimi                               Page 21
    discovery while the motion was under advisement, it was largely an exchange of documents—no
    depositions or expert discovery took place.        And finally, the district court was correct to
    highlight that resolution of the remaining state-law claims would require it to opine on the
    intricacies of various states’ false-claim statutes. See Landefeld v. Marion Gen. Hosp., 
    994 F.2d 1178
    , 1182 (6th Cir. 1993) (holding that a district court did not abuse its discretion by declining
    supplemental jurisdiction when novel state-law questions outweighed judicial-economy
    interests). For these reasons, the district court permissibly declined to exercise supplemental
    jurisdiction.
    IV.
    We affirm the district court’s judgment.