Today’s guest post, about a bottom-ten RICO third-party payor action from (no surprise) the Ninth Circuit, is by long-time friend of the blog (and blogger in his own right), Jonah M. Knobler, of Patterson Belknap. We named one similar case, Kaiser Foundation Health Plan, Inc. v. Pfizer, Inc., 712 F.3d 21 (1st Cir. 2013), as our #1 worst of the worst for 2013, and a second, In re Avandia Marketing, Sales Practices & Product Liability Litigation, 804 F.3d 633 (3d Cir. 2015), was our #4 worst case of 2015. This case involves a different drug and a different defendant, but a similarly miserable result. As always our guest posters are 100% responsible for their views, so Jonah deserves all of the credit (and any of the blame) for what follows.
When Congress enacted the Racketeer Influenced and Corrupt Organizations Act (RICO) in 1970, its express intent was taking on the Mafia. Among the furthest things from its mind was policing the labeling and promotion of FDA-regulated prescription medications. Yet here we are, 50 years later, defending a succession of RICO suits alleging that pharmaceutical companies engaged in “racketeering” by misstating a prescription drug’s safety or efficacy.
Most courts have dismissed these suits at the very threshold for lack of proximate causation. Last week, however, the Ninth Circuit breathed new life into this misguided liability theory, holding that proximate cause does not bar patients’ or insurers’ RICO claims against pharma companies for “omitt[ing] [to] mention … a drug’s known safety risk.” Painters & Allied Trades Dist. Council 82 Health Care Fund v. Takeda Pharms. U.S.A., __ F.3d ____, 2019 U.S. App. LEXIS 35844, at *4 (Dec. 3, 2019). Painters comes dangerously close to transforming RICO—a law meant to take down La Cosa Nostra—into a federal product-liability regime, contrary to the guidance of Congress and the Supreme Court.
By way of background, the first part of this post discusses how the Supreme Court has construed RICO, focusing on its element of proximate cause. The second part explains how the Supreme Court’s proximate-cause principles should apply to pharmaceutical RICO claims. The third and last part discusses Painters and explains how the Ninth Circuit deviated from those governing principles.
Part I: The Supreme Court And RICO Proximate Cause
As noted, Congress’s clear purpose in enacting RICO was fighting “criminal enterprises such as the Mafia and its many petty imitators.” United States v. Masters, 924 F.2d 1362, 1367 (7th Cir. 1997). Senator Hruska of Nebraska, who proposed the bills that eventually became RICO, explained that “the evil to be curbed” was “organized crime”—that is, “mobsters” who employ “threats of violence, extortion, and similar techniques.” Sedima v. Imrex Co., 473 U.S. 479, 514 (1985) (Marshall, J., dissenting). RICO was ultimately enacted as part of the Organized Crime Control Act of 1970, whose preface stated that “it was the declared purpose of Congress ‘to seek the eradication of organized crime in the United States….’” United States v. Turkette, 452 U.S. 576, 588 (1981) (quoting Pub. L. 91-452, 84 Stat. 923). References to the Mafia and “organized crime” in RICO’s legislative history are legion.
Yet, in Sedima v. Imrex Co., 473 U.S. 479 (1985), a bare majority of the Supreme Court held that RICO suits do not require a substantive “organized crime nexus.” Id. at 494. While the majority voiced its “concern over the consequences of an unbridled reading of the statute,” it found “no room in the statutory language”—as opposed to the legislative history—to demand a Mafia link. Id. at 481, 495. As the dissenters feared, this opened the door to RICO’s deployment against “legitimate businesses in ordinary commercial settings,” an outcome Congress plainly did not intend. Id. at 506 (Marshall, J., dissenting).
Seven years later, however, the Court narrowed RICO’s sweep in a different way. In Holmes v. SIPC, 503 U.S. 258 (1992), the Securities Investor Protection Corporation (SIPC) alleged that the defendant, Holmes, had engaged in a fraudulent stock-manipulation scheme. That scheme caused broker-dealers who were SIPC members to become insolvent, such that they could not meet their financial obligations to their customers. Their insolvency, in turn, triggered the SIPC’s duty under the Securities Investor Protection Act to advance funds to reimburse the broker-dealers’ customers. The SIPC sued Holmes under RICO, asserting that his fraudulent stock manipulation had injured it in the amount of the funds it was required to reimburse. See id. at 262-63.
Because of Sedima, the suit’s lack of an organized-crime nexus was no barrier. But the Holmes Court found another reason to toss the SIPC’s claim: proximate cause. RICO’s private-right-of-action provision states that “any person injured … by reason of a violation … may sue.” 18 U.S.C. §1964(c) (emphasis added). Read literally, “by reason of” would seem to require only but-for causation. And Holmes’ scheme was indeed a but-for cause of the SIPC’s losses. This time, however, the Court took an approach unlike the one it took in Sedima: noting the “unlikelihood that Congress meant to allow all factually injured plaintiffs to recover,” it held that a RICO claim requires “not only … ‘but for’ cause,” but “proximate cause as well.” Id. at 266, 268 (emphasis added).
Of course, the devil is in the details. As Holmes noted, “proximate cause” is a concept that takes on “many shapes,” depending on the context. Id. at 268; see also Lexmark Int’l, Inc. v. Static Control Components, Inc., 572 U.S. 118, 133 (2014) (noting that “[t]he proximate cause inquiry … take[s] various forms” and that the particular form “is controlled by the nature of the statutory cause of action”). When most lawyers think of proximate cause, it probably calls to mind Mrs. Palsgraf and foreseeability. But “[a] plaintiff must make a different showing of proximate cause—one that is often more difficult to make—when bringing suit under the RICO statute than when bringing a common-law cause of action.” Lerner v. Fleet Bank, N.A., 459 F.3d 273, 278 (2d Cir. 2006). Specifically, Holmes held, RICO proximate cause requires a “direct relation between the injury asserted and the injurious conduct alleged.” 503 U.S. at 269. The aim of this directness requirement is “to prevent … intricate, uncertain inquiries from overrunning RICO litigation.” Anza v. Ideal Steel Supply Corp., 547 U.S. 451, 460 (2006).
This notion of a “direct relation” encompasses several related ideas. Most important is the number of “step[s]” in the causal chain leading from the defendant’s violation to the plaintiff’s injury. Holmes, 503 U.S. at 271-72. “The general tendency … is not to go beyond the first step” in that chain. Id. In other words, proximate cause is ordinarily lacking if the immediate “cause of [the plaintiff’s] asserted harms … is a set of actions … distinct from the alleged RICO violation.” Anza, 547 U.S. at 458; see also BCS Services v. Heartwood 88, LLC, 637 F.3d 750, 756 (7th Cir. 2011) (generally no proximate cause where “the defendant’s conduct and the plaintiff’s injury are separated by intermediate pairs of cause and effect”).
Holmes’s “direct relation” requirement also considers whether “other, independent factors” not attributable to the RICO violation may have contributed to the plaintiff’s injury. 503 U.S. at 269; see also Anza, 547 U.S. at 459 (direct relation lacking where plaintiff’s damages “could have resulted from factors other than [defendant’s] fraud”). This question is closely connected with the length of the causal chain, because the more “steps” between act and injury, the more uncertainty enters into the picture, and “the more difficult it becomes to ascertain” what portion of the harm is “attributable to the violation.” Holmes, 503 U.S. at 269.
It follows that the “direct relation” concept is also related to the simplicity or intricacy of the plaintiff’s damages model. “[M]assive and complex damages litigation,” Holmes cautions, “not only burden[s] the courts, but … also undermine[s] the effectiveness of treble-damages suits.” Id. at 274. In this respect, RICO’s directness requirement resembles the concept of antitrust standing, which bars claims where assessing damages would require “massive evidence and complicated theories.” Assoc. Gen. Contractors v. Cal. State Council of Carpenters, 459 U.S. 519, 542-33 (1983) (cited with approval in Holmes and Anza).
A final “consideration … relevant to the RICO ‘direct relationship’ requirement is whether better situated plaintiffs would have an incentive to sue.” Hemi Grp., LLC v. City of New York, 559 U.S. 1, 11-12 (2010); see also Holmes, 503 U.S. at 269-70. Where there are other ways “to vindicate the law,” a RICO suit by a remotely injured plaintiff is simply not worth the effort, as “the need to grapple with” thorny problems of causation and damages is “unjustified by the general interest in deterring injurious conduct.” Id.
In Holmes, the SIPC’s claim did not satisfy this “direct relation” test. The causal chain went “beyond the first step”: Holmes’s stock manipulation left the broker-dealers insolvent, and “that intervening insolvency”—not Holmes’s conduct—was the immediate trigger of the SIPC’s duty to reimburse the broker-dealers’ customers. Id. at 271. That extra step introduced causal uncertainty into the mix, for the broker-dealers might have become insolvent in part for reasons other than Holmes’s scheme: “say, [their own] poor business practices.” Id. at 272. For that reason, permitting the SIPC to sue “would open the door to ‘massive and complex damages litigation.’” Id. at 274. And the courts “would be shouldering these difficulties” for naught, because “the broker-dealers … could be counted on to bring suit for the law’s vindication.” Id. at 273.
Since Holmes, the Court has twice found RICO proximate cause lacking for similar reasons. In Anza, it found the requisite “direct relation” absent where the plaintiff’s causal theory had multiple steps; some of its alleged injury “could have resulted from factors other than [the defendants’] … fraud”; calculating damages would have required “complex” and “speculative” proceedings; and “the State c[ould] be expected to pursue appropriate remedies.” 547 U.S. at 458-61. And most recently, in Hemi, the Court found no “direct relation” where “[m]ultiple steps … separate[d] the alleged fraud from the asserted injury”; “independent factors” might have “account[ed] for” some portion of the alleged harm; and “[t]he State” could vindicate the law instead. 559 U.S. at 9-12, 15.
The upshot is this: even if RICO doesn’t require a substantive nexus to organized crime, the statute’s “direct relation” requirement limits RICO’s reach to cases with causal theories akin to what one might see in a Mafia case. Mobsters act through unsubtle tactics like extortion, assault, kidnapping, arson, and murder. Those tactics inflict immediate and obvious harm on their victims—harm that can be proven through direct evidence, without resorting to fancy econometrics. Holmes and its progeny teach that, even if RICO claims may stray substantively outside the Mafia sphere, RICO causal theories must remain just that straightforward.
Part II: RICO Proximate Cause in Pharma Cases
So how should these principles apply to pharma cases? The template for pharma RICO suits is well-established: the plaintiff is either a patient or, more commonly, a “third-party payor” (TPP), such as a health insurance company or union-affiliated welfare benefit fund. The plaintiff alleges that the pharmaceutical defendant misrepresented the safety or efficacy of its medication by promoting it for an off-label use, concealing a health risk, or both. Because traditional personal-injury damages are unavailable under RICO, the plaintiff claims economic loss: but for the manufacturer’s violation, it would not have paid for the drug (or would have paid less for it). To make such claims economically viable, these suits are almost always brought as putative class actions. Often, the claimed damages rise into the hundreds of millions or billions of dollars, even before trebling.
The fraudulent schemes alleged in these suits are anything but straightforward. The plaintiff generally describes a multi-pronged effort, consisting of misleading “detailing” (direct, in-office presentations) to prescribers; financing of biased clinical research and slanted publications in medical journals; paying “key opinion leaders” in the medical community to talk up the drug; conducting publicity campaigns to counter unfavorable press; and other forms of supposedly misleading promotion or advertising. Eventually, the plaintiffs allege, the sum total of these efforts “corrupts the medical discourse,” trickling down to the plaintiffs’ own physicians or those of their beneficiaries. Those physicians then write prescriptions that they should not have written, and the plaintiffs ultimately pay for them.
The vast majority of courts have agreed that “allowing [plaintiffs] to move forward” on such a theory “would present precisely the types of problems the Holmes Court sought to avoid” through its directness requirement. Ironworkers Local Union No. 68 v. AstraZeneca Pharms. LP, 585 F. Supp.2d 1339, 1344 (M.D. Fla. 2008), aff’d, 634 F.3d 1352 (11th Cir. 2011); see also Sidney Hillman Health Ctr. of Rochester v. Abbott Labs., 192 F. Supp.3d 963, 967-70 (N.D. Ill. 2016), aff’d, 873 F.3d 574, 575-78 (7th Cir. 2017); Sergeants Benevolent Ass’n Health & Welfare Fund v. Sanofi-Aventis U.S., L.L.P., 20 F. Supp.3d 305, 323 (E.D.N.Y. 2014), aff’d, 806 F.3d 71 (2d Cir. 2015); Employer Teamsters-Local Nos. 17/505 Health Welfare Trust Fund v. Bristol Myers Squibb Co., 969 F. Supp.2d 463, 473-76 (S.D. W. Va. 2013); In re Bextra & Celebrex Mktg., Sales Practices & Prods. Liab. Litig., 2012 U.S. Dist. LEXIS 111446, at *218-24 (N.D. Cal. Aug. 2, 2012); UFCW Local 1776 v. Eli Lilly & Co., 620 F.3d 121, 134 (2d Cir. 2010); In re Yasmin & Yaz (Drospirenone) Mktg., Sales Practices & Prods. Liab. Litig., 2010 U.S. Dist. LEXIS 80758, at *22-27 (S.D. Ill. Aug. 5, 2010); In re Schering-Plough Corp. Intron/Temodar Consumer Class Action, 2009 U.S. Dist. LEXIS 58900, at *86-93 (D.N.J. July 10, 2009), aff’d, 678 F.3d 235 (3d Cir. 2012); Se. Laborers Health & Welfare Fund v. Bayer Corp., 655 F. Supp.2d 1270, 1280-84 (S.D. Fla. 2009), aff’d, 444 F. App’x 401, 410 (11th Cir. 2011); Dist. 1199P Health & Welfare Plan v. Janssen, 784 F. Supp.2d 508, 523-25 (D.N.J. 2011); UFCW Cent. Pa. & Reg’l Health & Welfare Fund v. Amgen, Inc., 400 F. App’x 255 (9th Cir. 2010); In re Actimmune Mktg. Litig., 614 F. Supp.2d 1037, 1051 (N.D. Cal. 2009), summarily aff’d, 464 F. App’x 651 (9th Cir. 2011).
For starters, there is well more than one “step” in the causal chain leading from the alleged wrongdoing to the alleged injury. Here is how that chain looks in a typical TPP claim:
- First, the manufacturer makes false or misleading statements about its drug to the medical community.
- Second, members of the medical community accept those statements as truth and propagate them until they “corrupt the medical discourse.”
- Third, based on that corrupted knowledge base, the TPP’s pharmacy benefit manager (PBM)—the large outside company that manages most TPPs’ prescription drug plans—recommends adding the drug to the TPP’s formulary.
- Fourth, the TPP accepts the recommendation of its PBM to place the drug on formulary, rendering use of the drug eligible for reimbursement.
- Fifth, the TPP sets the premiums that it collects from its members at a level that anticipates fewer prescriptions of the drug than actually end up being written. See Teamsters Local 734 Health & Welfare Trust Fund v. Philip Morris, 196 F.3d 818, 824 (7th Cir. 1999) (“[I]nsurers [may] collect only for [their] net outlay …; but there will be such a net outlay only if the insurers’ actuaries are not calculating rates correctly.”).
- Sixth, based on the same corrupted knowledge base, physicians in the TPP’s network choose to prescribe the drug for the TPP’s beneficiaries, instead of a less expensive drug or (less commonly) no drug at all.
- Seventh, those patients choose to fill their prescriptions. See Jane E. Brody, The Cost of Not Taking Your Medicine, N.Y. Times, Apr. 17, 2017 (noting that 20–30% of prescriptions go unfilled).
- Eighth, if the drug has one or more generic equivalents, the pharmacist fills the prescription with the drug that is the subject of the lawsuit, rather than the cheaper generic version.
- Ninth, the patients take the drug and fail to receive the advertised health benefit or experience bodily injury. See Sidney Hillman, 873 F.3d at 577 (noting that TPPs suffer no cognizable RICO injury by paying for a medication that “helped the patient”).
- Tenth, those patients submit their prescription claims for payment or reimbursement by the TPP.
- Eleventh, the TPP reimburses those prescriptions—and, because it did not set its premiums high enough back at Step Five, it suffers a net out-of-pocket loss after considering both sides of the ledger.
This causal chain, involving various independent actions by multiple parties, is even lengthier and more attenuated than the one the Supreme Court held fatally indirect in Holmes. See, e.g., Eli Lilly, 620 F.3d at 134 (“Plaintiffs’ theory of liability rests on the independent actions of third and even fourth parties, as physicians, PBMs, and PBM Pharmacy and Therapeutics Committees all play in a role in the chain between [the manufacturer] and TPPs.”); Bristol Myers Squibb, 969 F. Supp.2d at 475 (“Between Defendants’ alleged misleading marketing and Plaintiffs’ prescription reimbursements lies a vast array of intervening events….”). “Because [this] theory of causation requires us to move well beyond the first step, that theory cannot meet RICO’s direct relationship requirement.” Hemi, 559 U.S. at 10. The chain may be slightly shorter in patient cases, as opposed to TPP cases, but no matter: that chain, too, goes “well beyond the first step.” Id.
Just as important as the number of intervening links is the nature of those links—especially the physician’s prescribing decision. In rejecting claims like these, courts have stressed that prescribing decisions are “multifaceted and individualized,” Sanofi-Aventis, 806 F.3d at 92, and are “influenced by a number of things … [besides] representations by [the] manufacturer,” AstraZeneca, 585 F. Supp.2d at 1344. This includes years of study and training, word-of-mouth, first-hand clinical experience with the drug, independent literature, and patient-specific medical history. See AstraZeneca, 634 F.3d at 1362-63 (“The prescription choice the physician makes is … an individualized medical judgment bottomed on a knowledge of both patient and palliative…. The physician learns about a drug through multiple sources, only one of which might be the drug manufacturer’s promotions and literature.” (cleaned up)); Sidney Hillman, 873 F.3d at 577 (“[S]ome physicians … might have changed [their prescribing practices] in response to information that [the manufacturer] did not influence. The medical literature contains not only double-blind clinical studies but also case studies and even anecdotes that affect physicians’ prescribing practices. For some physicians, these may have dominated over anything [the manufacturer] did.”); Eli Lilly, 620 F.3d at 135 (“[The manufacturer] was not, however, the only source of information on which doctors based prescribing decisions. An individual patient’s diagnosis, past and current medications being taken by the patient, the physician’s own experience prescribing [the drug], and the physician’s knowledge regarding the side effects of [the drug] are all considerations that would have been taken into account….”).
The physician’s prescription decision—to say nothing of all the other intervening steps—makes it “difficult … to ascertain” what portion of the plaintiff’s out-of-pocket costs are “attributable to the [defendant’s] violation, as distinct from other, independent factors.” Anza, 547 U.S. at 458; see Sidney Hillman, 873 F.3d at 577 (noting the “difficulties” of “[d]isentangling the effects of the improper promotions from the many other influences on physicians’ prescribing practices”). And the need for such a “complex assessment” “implicates [the] fundamental concerns expressed in Holmes.” Anza, 547 U.S. at 459; see AstraZeneca, 585 F. Supp.2d at 1344 (finding that the “highly complex damages assessment” in a pharma RICO suit “strongly weighs against a finding” of directness).
But wait! Couldn’t an expert economist come up with a statistical model that purports to isolate and quantify the influence of the challenged practices on prescribing rates? Indeed, when pharma RICO cases have survived motions to dismiss, the plaintiffs have usually proffered such a model—almost always a methodologically questionable regression analysis. As Bexis has noted, such models essentially propose “an aggregate trial by formula that deprives the defendants of contesting the individual merits of any of the thousands—heck, millions—of individual claims at issue.” Such models, moreover, answer only part of the question: even if they could isolate and quantify the influence of improper promotion on prescribing, they shed no light on (for example) how many of those prescriptions actually helped or hurt the patients who received them; what drug (if any) would have been prescribed for each patient but for the alleged fraud; or how much that alternative drug would have cost. See Sidney Hillman, 873 F.3d at 577. And in all events, the very fact that Ph.D.-level statistical analyses must be brought to bear, in and of itself, is proof that the causal theory in these cases is not “direct.” See SEIU Health & Welfare Fund v. Philip Morris Inc., 249 F.3d 1068, 1074-75 (D.C. Cir. 2001) (“Reliance on aggregate statistical proof … compounds the difficulties and does not alter the speculative [and indirect] nature of the claimed damages.”).
If that weren’t enough, in pharma RICO cases, there are always “better situated plaintiffs [with an] incentive to sue,” Hemi, 559 U.S. at 11-12—or, at minimum, less problematic ways than RICO for the same plaintiffs to “vindicate the law,” Holmes, 503 U.S. at 269-70; see Ore. Laborers-Emprs. Health & Welfare Trust Fund v. Philip Morris, Inc., 185 F.3d 957, 964 (9th Cir. 1999) (noting that the question is not who is the best RICO plaintiff, but whether there are better ways to “promote ‘the general interest in deterring injurious conduct’”). The federal and state governments both have a strong interest in ensuring that pharmaceutical companies label and promote their products properly, and neither has hesitated to vindicate that interest under the FDCA, False Claims Act, and other theories. See, e.g., Sidney Hillman, 873 F.3d at 578 (“Public prosecution avoids these problems [of causation and damages], so [Defendant’s] criminal conviction [for misbranding] and $1.6 billion [penalty] were the proper remedies.”). Moreover, physically injured patients routinely bring state-law product-liability claims against pharmaceutical companies that allegedly mislabel or falsely promote their products. Those suits already serve to “deter [the] injurious conduct” at issue, rendering RICO claims superfluous. Holmes, 503 U.S. at 269; see Ore. Laborers, 185 F.3d at 964, 966 (“Although the smokers cannot recover under … RICO …., they can seek recovery under other state law theories for personal injury and the associated medical costs…. [This] weigh[s] in favor of barring plaintiffs’ [RICO] claims.”); Bayer, 655 F. Supp.2d at 1283-84 (“The existence of [a product-liability] MDL demonstrates that … patients who suffered physical injury … are able to ‘vindicate the law … without any of the problems attendant upon [RICO] suits.’”).
In sum, the theories asserted in pharma RICO cases have far too many steps in their causal chain, and those steps are far too individualized. For that reason, isolating the harm caused by the defendant and calculating the plaintiff’s damages are nightmarishly difficult, if not impossibly speculative. Again, “[t]he element of proximate cause recognized in Holmes is meant to prevent these types of intricate, uncertain inquiries from overrunning RICO litigation.” Anza, 547 U.S. at 460. That’s why courts have overwhelmingly rejected these suits.
Part III: Critiquing Painters
The complaint in Painters largely hewed to the template described above. Like many other pharma RICO cases, it arose out of a product-liability MDL containing numerous personal-injury suits. See In re Actos (Pioglitazone) Prods. Liab. Litig., 274 F. Supp.3d 485, 503 (W.D. La. 2017). The plaintiffs were a union-affiliated welfare benefit fund (Painters) and a handful of individual patients. They alleged that the pharmaceutical defendants had “engaged in a decade-long scheme” to “sell the diabetes medication Actos … while concealing the bladder cancer risks associated with Actos from consumers, prescribers, third-party payors, and the [FDA].” Second Am. Compl. ¶1. The plaintiff sought class certification and treble damages “for the consumers and [TPPs] who were tricked into purchasing and/or reimbursing Actos prescriptions.” Id.
True to form, the alleged chain of causation was anything but simple. First, to “deceive the FDA” into approving Actos, the defendants allegedly promoted a sham medical “hypothesis” to “explain away” certain troubling test results. Id. ¶¶29-33, 48-50. Later, after approval, the defendants allegedly “resisted … label changes” desired by the FDA and “convince[d] the FDA,” using “numerous ‘experts,’” that cancer warnings were unnecessary. Id. ¶61-63. Meanwhile, the defendants allegedly “instruct[ed] [their] sales force to pitch [a misleading] message” about Actos’s safety profile in face-to-face meetings with prescribers. Id. ¶62. They allegedly decided not to “conduct market research on possible label language around bladder cancer,” lest they “risk public awareness.” Id. ¶ 67. They allegedly “conspired … to misrepresent the data” on Actos’ safety in medical journal publications. Id. ¶85. They allegedly “generat[ed] scientific materials about Actos, which despite the appearance of independence, were designed to persuade doctors to prescribe Actos.” Id. ¶40. They allegedly refused to “issue a Dear Doctor Letter to warn the medical community of the risk of developing cancer while taking Actos.” Id. ¶92. And they allegedly “misrepresent[ed] and conceal[ed] the ties between Defendants and other enterprise participants.” Id. ¶239.
Painters, the TPP plaintiff, claimed that the defendants’ alleged campaign of deception “caused [it] to make payments for Actos that, absent the fraud, would never have occurred.” Id. ¶134. It did not point to any actual instances where misleading statements were made to any of Painters’ beneficiaries or their physicians—let alone to Painters itself. Nor did it allege that any of its beneficiaries actually developed bladder cancer, or that Actos failed to treat their diabetes satisfactorily. The individual plaintiffs alleged that they paid out-of-pocket for a portion of their Actos prescriptions and that they “never would have purchased and ingested the drug” but for the defendants’ violations. Id. ¶151. Only one of the individual plaintiffs alleged that she read and relied on the Actos label. ¶139. None of them alleged that they actually developed bladder cancer or that Actos failed to treat their diabetes.
In a brief order, the district court dismissed the plaintiffs’ RICO claims with prejudice, “find[ing] persuasive” the many decisions that have held proximate cause lacking in “highly similar case[s].” No. 2:17-cv-07223 (C.D. Cal.), ECF No. 140 (filed Feb. 1, 2018). In a published, unanimous opinion, the Ninth Circuit reversed. In so doing, it aligned itself not with the long string of decisions cited above—which includes two unpublished Ninth Circuit decisions—but with a pair of outlier decisions from the First and Third Circuits.
The Ninth Circuit acknowledged RICO’s “direct relation” requirement. But in applying it, the Ninth Circuit did not focus on the many steps in the plaintiff’s causal chain, as Holmes, Anza, Hemi, and its own prior precedent command. See Canyon Cnty. v. Syngenta Seeds, Inc., 519 F.3d 969, 981 (9th Cir. 2008) (“Under [Holmes and] Anza, courts must scrutinize the causal link between the RICO violation and the injury…. Where the violation is not itself the immediate cause of the plaintiff’s injury, proximate cause may be lacking.”). Instead, Painters found the directness requirement met because (1) “sell[ing] more Actos” to TPPs and patients was the goal of the alleged scheme; and (2) it was “foreseeable” that TPPs and patients would be harmed by the scheme. 2019 U.S. App. LEXIS 35844, *17, *30-31 (emphasis added).
The Supreme Court, however, has specifically rejected both intent and foreseeability as yardsticks of RICO proximate cause. Start with intent. In Anza, the Second Circuit had found proximate cause satisfied because the plaintiff was the “intended victim of the [defendant’s] scheme.” Ideal Steel Supply Corp. v. Anza, 373 F.3d 251, 257-64 (2d Cir. 2004) (emphasis added). The Supreme Court reversed, criticizing the lower court’s reasoning “that because the [defendants] allegedly sought to [harm the plaintiff], it [was] immaterial whether they took an indirect route to accomplish their goal.” Anza, 547 U.S. at 460 (emphasis added). Anza warned that “[a] RICO plaintiff cannot circumvent” the directness requirement by alleging that “the defendant’s aim” or “motive” was to injure him. Id. And in Hemi, the Supreme Court again stressed that “the intended consequences of the defendant’s unlawful behavior” are not relevant to “the directness of the relationship between the conduct and the harm.” 559 U.S. at 12; see also SEIU, 249 F.3d at 1074 (noting that courts have “rejected the contention that [the defendant’s] specific intent … is sufficient” to establish directness).
The Supreme Court has also expressly rejected the argument that “RICO’s proximate cause requirement turn[s] on foreseeability.” Hemi, 559 U.S. at 12. More recently, in a Fair Housing Act case, the Supreme Court invoked its RICO precedents in holding that “foreseeability alone is not sufficient to establish proximate cause.” Bank of Am. Corp. v. City of Miami, 137 S. Ct. 1296, 1305 (2017). A foreseeability test, the Court warned, “would risk [t]he ‘massive and complex damages litigation’” that the proximate cause requirement is supposed to prevent. Id. at 1306; see also Conrail v. Gottshall, 512 U.S. 552-53 (1994) (“If one takes a broad enough view, all consequences of [a violation], no matter how far removed in time or space, may be foreseen. Conditioning liability on foreseeability, therefore, is hardly a condition at all.”).
So much for an immediate causal nexus. What about the “difficult[y] [of] ascertain[ing] damages?” Painters, 2019 U.S. App. LEXIS 35844, at *18-19. Without any explanation, the Ninth Circuit dubbed itself “not persuaded that it is so difficult here that Plaintiffs should be denied the opportunity to prove their damages.” Id. But it is hard to square this with binding precedent. In Anza, for example, a New York City business sued its chief competitor in the local market, alleging that the competitor’s failure to charge sales tax allowed it to undercut the plaintiff’s prices and steal its local New York City customers. 547 U.S. at 454. The Supreme Court ordered this claim dismissed on the pleadings, observing that the required damages assessments would be too “complex,” “intricate,” and “speculative” for a RICO claim. 547 U.S. at 459-60. But calculating a single business’s lost sales in a duopolistic local market would be a walk in the park compared to modeling the damages suffered by the nationwide patient and TPP classes in Painters. Just for example, the Ninth Circuit acknowledged that the fact-finder would have to determine which “alternative drug” each patient who received Actos would have been prescribed in the but-for world and whether that drug would have “cost less than Actos” at the relevant time. 2019 U.S. App. LEXIS 35844, at n.7. If the Anza plaintiff was properly “denied the opportunity to prove [its] damages,” it’s perplexing why the Ninth Circuit thought the Painters plaintiffs deserved that “opportunity.” Cf. Canyon Cnty., 519 F.3d at 982 (“[C]ourts need not allow RICO plaintiffs leeway to continue on with their case in an attempt to prove an entirely remote causal link.”).
What about the concern that physicians’ prescription decisions are too multifaceted and individualized? The Ninth Circuit was not concerned about that: “since Actos was a prescription drug,” the court reasoned, it was “perfectly foreseeable that physicians who prescribed Actos would play a causative role in Defendants’ alleged fraudulent scheme.” 2019 U.S. App. LEXIS 35844, at *30-31. Again, that is not how the “direct relationship” requirement works. Again, consider Anza: in that case, it was “perfectly foreseeable” to the defendant that, by shirking sales tax and undercutting its competitor’s prices, some of its competitors’ customers would jump ship. But that was not good enough: “Businesses,” the Anza Court noted, “lose and gain customers for many reasons, and it would require a complex assessment to establish what portion of [plaintiff’s] lost sales were the product of [defendant’s] decreased prices.” 547 U.S. at 459. As discussed above, that same observation applies in spades to physicians’ decisions whether to prescribe a particular drug. Notably, the Ninth Circuit has faithfully followed Anza’s approach in non-pharma RICO cases. See Canyon Cnty., 519 F.3d at 983 (finding no proximate cause where intermediate steps in the causal chain might have “numerous alternative causes,” such that “the proceedings required to evaluate the [plaintiff’s] injury would be speculative” and “intricate”).
The Ninth Circuit was also willing to overlook the physician’s role in the causal chain because “physicians [are] not … the ones paying for the drugs they prescribe.” 2019 U.S. App. LEXIS 35844, at *31. In other words, physicians aren’t injured by the RICO violation, so their intervening role doesn’t count. For starters, it’s far from clear that physicians suffer no injury from a scheme like the one alleged. See Sidney Hillman, 873 F.3d at 576 (“Physicians affected by [improper] promotions thus may lose business and revenue.”). More importantly, under Supreme Court and Ninth Circuit precedent, the question is not who is the first injured party; it is whether the plaintiff’s injury requires more than one “step” to bring about. A causally remote injury does not magically become “direct” just because no one else in the causal chain has been injured yet. See Canyon Cnty., 519 F.3d at 983 (holding that RICO proximate cause bars claims where “[t]he asserted [causal] link … is … attenuated,” even if “the [plaintiff’s] injury is not ‘derivative of an injury suffered by any other party’”).
In disregarding the physician’s role, Painters also relied on Bridge v. Phoenix Bond & Indemnity Co., 553 U.S. 639 (2008). But Bridge does not hold that intervening links in a RICO plaintiff’s causal chain do not matter. Rather, it holds that an intervening link may “be discounted” if—and only if—it operates in a purely mechanistic fashion and introduces no uncertainty into the causal theory. Sidney Hillman, 192 F. Supp.3d at 970; see Lexmark, 572 U.S. at 139-40 (observing that a causal chain that “includes [an] intervening link” may be deemed direct in “unique circumstances” where that step operates “automatically”). The Bridge plaintiffs’ causal theory went roughly like this: (1) the defendants lied to Cook County to gain access to a series of county-run contests; (2) the county selected the contest winners; and (3) the defendants won some prizes that otherwise would have gone to the plaintiffs. True, there was one intervening link in this chain: the county’s selection of prizewinners. Crucially, however, the prizes were “allocate[d] … ‘on a rotational basis’ in order to ensure that [they were] apportioned fairly among” all participants. Bridge, 553 U.S. at 639 (emphasis added). The only intervening step, in other words, was neither volitional nor multifaceted; it was 100% deterministic. Thus, it presented none of the concerns underlying RICO’s “direct relation” requirement. As discussed above, the multiple intermediate steps in a pharma plaintiff’s causal chain are very different; doctors, for example, do not prescribe medications on a “rotational” basis. See Sergeants Benevolent Ass’n Health & Welfare Fund v. Sanofi-Aventis U.S. LLP, 20 F. Supp.3d 305, 322-23 (E.D.N.Y. 2014) (“In this case, [unlike Bridge], the intervening acts which interrupt the causal chain … cannot be readily predicted…. [T]he prescribing decisions of physicians are based on so many factors as to defy any efforts to categorically attribute them to a particular cause.”).
Painters’ treatment of the physician’s intervening role leaves much to be desired. But at least the court gave reasons for disregarding it. By contrast, Painters did not even discuss the fact that the plaintiffs’ causal chain—unlike in most pharma RICO cases—involved deception of the FDA itself. See 2019 U.S. App. LEXIS 35844, at *5 (“[P]laintiffs allege that Defendants convinced the FDA that studies revealing that Actos increased the risk of bladder cancer were wrong.”); Second Am. Compl. ¶¶ 48-50, 61-63 (defendants allegedly “deceive[d] the FDA” into approving Actos and “convince[d] the FDA” that warnings were unnecessary). For one thing, the FDA’s decisions are at least as complex and multifaceted as those of prescribing physicians. But more fundamentally, the Supreme Court has held that causal theories involving “fraud on the FDA” are off-limits, because allowing private plaintiffs to pursue such claims would “skew” the FDA’s “delicate” balancing of regulatory objectives. Buckman Co. v. Plaintiffs’ Legal Comm., 531 U.S. 341, 348 (2001). Of course, RICO claims arise under federal law, so they cannot technically be “preempted” like the state-law claims in Buckman, but that should not change the outcome. See Se. Laborers, 444 F. App’x at 407, 410 n.4 (“As with its [state-law] claim, [Plaintiff] may not rely on a … fraud-on-the-FDA theory of causation for its RICO claim.”).
But perhaps the most telling sentence in Painters is this one: “If we were to hold … that prescribing physicians’ and [PBMs’] decisions … sever the chain of proximate cause …, drug manufacturers would be insulated from liability for their fraudulent marketing schemes….” 2019 U.S. App. LEXIS 35844, at *31-32 (emphasis added). In other words, we must find proximate cause, despite Supreme Court and Ninth Circuit precedent, because the unpopular pharmaceutical industry must be punished for its presumed misdeeds. Of course, this rationale assumes its conclusion—i.e., that there should be RICO liability for inaccurately labeling or promoting a prescription drug. It also fails to acknowledge that there are many ways of deterring wrongdoing by the pharmaceutical industry other than a treble-damages claim under a federal statute designed to combat violent mobsters. Notably, during the tobacco-litigation frenzy of the 1990s, the Ninth Circuit held that physically injured consumers’ ability to bring state-law product-liability claims against manufacturers “weigh[ed] heavily in favor of barring” RICO claims. Ore. Laborers, 185 F.3d at 964. Apparently, the rules have changed.
Finally, absent from Painters is any discussion of the drawbacks of opening up RICO—“the litigation equivalent of a thermonuclear device,” Miranda v. Ponce Fed. Bank, 948 F.2d 41, 44 (1st Cir. 1991)—to uninjured patients and insurers. Federal courts are already overwhelmed with product-liability litigation; removing the requirement of physical injury and adding the incentives of treble damages and attorney’s fees will make this problem far worse. Cf. Laborers Local 17 Health & Benefit Fund v. Philip Morris, Inc., 191 F.3d 229, 240 (2d Cir. 1999) (finding no proximate cause in insurer’s RICO case against tobacco company, and noting that “rul[ing] otherwise could lead to a potential explosion in the scope of tort liability”). Additionally, requiring the pharmaceutical industry to pay huge treble-damage awards to sprawling classes of uninjured patients and insurers may have “grave health policy consequences,” including “higher priced name brand drugs” and “fewer innovative drugs.” In re Darvocet, Darvon & Propoxyphene Prods. Liab. Litig., 756 F.3d 917, 946 (6th Cir. 2014). Finally, as noted above, introducing the blunt instrument of RICO to pharmaceutical litigation threatens to “skew” the FDA’s regulatory balancing act in ways Congress could never have envisioned when it enacted the FDCA. Buckman, 531 U.S. at 348. Suffice it to say, extending RICO liability to its conceptual limit is not the unalloyed good that Painters appears to assume.
It’s been said a million times before: “the civil provisions of RICO are the most misused statutes in the federal corpus of law.” Goldfine v. Sichenzia, 118 F. Supp.2d 392, 394 (S.D.N.Y. 2000) (cleaned up). As relevant here, RICO was not intended as a “surrogate for garden-variety [tort] actions properly brought under state law,” Midwest Grinding Co. v. Spitz, 976 F.2d 1016, 1022 (7th Cir. 1992), or as a “federal products liability scheme complete with treble damages and attorney fees,” Summit Props. v. Hoechst Celanese Corp., 214 F.3d 556, 557 (5th Cir. 2000).
By adopting the “direct relation” requirement, the Supreme Court tossed out a lifeline to lower courts adrift in a sea of civil RICO claims that Congress did not anticipate or intend. That requirement, again, was meant to cabin RICO’s scope to disputes with simple, black-and-white causal theories—similar to those in Mafia cases, the prototypical application of RICO. By stretching that “direct relation” test beyond recognition to punish the pharmaceutical industry, Painters not only spurned the Supreme Court’s lifeline; it affirmatively urged on the floodwaters.
So what now? Perhaps the en banc Ninth Circuit will see things differently. Perhaps the Supreme Court will step in and reinvigorate RICO’s proximate-cause requirement, as it did in Anza and Hemi. Perhaps Congress could be convinced to amend RICO to bar suits involving FDA-regulated drugs and medical devices, the way it prohibited RICO suits involving regulated securities transactions when they got out of control in the 1990s. In the meantime, however, pharmaceutical and medical-device companies should brace for the onslaught of bogus “racketeering” claims that Painters has invited.