The decision in In re Bard IVC Filters Products Liability Litigation, 969 F.3d 1067 (9th Cir. 2020) (“Booker”), is yet another reminder that multidistrict litigation as it is currently conducted is a fundamentally flawed process, dedicated more to forcing settlements than to any of the goals envisioned by Congress when it passed 28 U.S.C. §1407.  The defendant had a strong preemption defense – particularly for a Class II, non-PMA medical device – but could not get it before an appellate court, due to the lack of a final appealable order, until years later, after having to go through (and lose) a “bellwether” trial.  If the plaintiffs had lost a preemption motion, of course, with a final judgment entered against at least some of them, they would have an immediate right to appeal.  In the Booker case, ongoing MDL settlement pressure continued and was effective to force the resolution of a plethora of claims that were questionable, at best.

The joy of MDLs:  what disfavors plaintiffs is appealable; what disfavors anyone else is negotiable.

The defendant’s preemption argument in Booker was primarily that the FDA’s §510(k) clearance process for the devices in question, which occurred after Congress had revamped and toughened “substantial equivalence” in the Safe Medical Devices Act of 1990, should be preemptive because:  (1) Congress then imposed the same “reasonable assurance of safety and effectiveness” standard on §510(k) devices that was held preemptive in Medtronic, Inc. v. Lohr, 518 U.S. 470 (1996), and Riegel v. Medtronic, Inc., 552 U.S. 312 (2008) (more or less the same argument we examined in depth here), and (2) the FDA in clearing these particular devices had in fact imposed numerous device-specific requirements, particularly with respect to their design.

Had that argument succeeded, it would have seriously reshaped the contours of that MDL litigation.  However, the district court rejected it.  See In re Bard IVC Filters Products Liability Litigation, 2017 WL 5625547 (D. Ariz. Nov. 22, 2017), and by the time an appeal was decided three years later, case peculiar circumstances prevented the most significant preemption arguments from being presented, let alone resolved.

Because there is no defense right to appeal critical MDL-wide issues such as preemption, those issues (when a defendant loses) are necessarily filtered through subsequent bellwether trials, assuming the litigation gets that far.  However, in Booker, the bellwether trial verdict in question, while hardly a win, was not a disaster for the defendants:

Four claims were presented to the jury at trial: strict liability design defect, negligent design, strict liability failure to warn, and negligent failure to warn.  The jury found Defendants liable on the negligent failure to warn claim and awarded Plaintiff $1.6 million in compensatory damages and $2 million in punitive damages.

In re Bard IVC Filters Products Liability Litigation, 2018 WL 3037161, at *1 (D. Ariz. June 19, 2018), aff’d, 969 F.3d 1067 (9th Cir. 2020).  The design claims had been the key.  They were the only claims in Booker against which the defendants did not move for summary judgment, 969 F.3d at 1072, but in the bellwether trial, “[t]he jury found for [defendant] on” them.  Id.

Bound by Lohr, Booker followed the Supreme Court’s holding that, interpretation of §360k(a) (the medical device preemption clause in the statute) was “substantially informed” by an FDA regulation “provid[ing] that state requirements are preempted only when the FDA has established ‘specific requirements applicable to a particular device.’”  969 F.3d at 1072 (quoting both Lohr, 518 U.S. at 495 and 21 C.F.R. §808.1(d)).  Riegel, of course, had later concluded “that §808.1(d)(1) can add nothing to our analysis but confusion,” 552 U.S. at 329, but since Riegel did not expressly overrule Lohr and Lohr was more directly applicable to §510(k) devices, Booker applied Lohr.

We further note that Booker did not mention any “presumption against preemption,” and indeed professed to follow Supreme Court precedent that, “[w]hen a federal law contains an express preemption clause, we ‘focus on the plain wording of the clause.’”  969 F.3d at 1073 (quoting two Supreme Court cases).  However, those were merely proud words on a dusty shelf, since the “plain wording” of §360k(a) contains nothing resembling “specific requirements applicable to a particular device.”  Rather, that was a gloss added by the FDA, to which Lohr deferred under the influence of the since-repealed presumption against preemption.  However, that presumption died in Puerto Rico v. Franklin California Tax-Free Trust, ___ U.S. ___, 136 S. Ct. 1938, 1946 (2016) (“because the statute contains an express pre-emption clause, we do not invoke any presumption against pre-emption”) (citation and quotation marks omitted), and the notion that the FDA could define the scope of preemption of its own organic statute met its demise in PLIVA, Inc. v. Mensing, 564 U.S. 604, 613 n.3 (2011) (“we do not defer to an agency’s ultimate conclusion about whether state law should be pre-empted.”) (citation omitted).  So the entire edifice is shaky.

But it did not fall in Booker – largely because the massive FDA regulatory record that the defendants had assembled concerning their devices was directed against the MDL plaintiffs’ design claims (against which fewer other defenses applied).  Thus:

[Defendant] argues that when the FDA reclassified [the product] . . ., the agency imposed specific requirements in the form of the three special controls applicable to such devices.  The preemption issue in this case therefore turns on whether the special controls constitute “specific requirements applicable to a particular device.”

969 F.3d at 1074 (citation omitted).  Booker “assumed without deciding” that these “special controls” imposed by the FDA “were legally binding.”  Id.  That alone is more consideration than many §510(k) preemption arguments receive.

However, Booker nonetheless found no preemption.  First, it held that, despite the FDA incorporating two guidance documents as “special controls” in its clearance of the device in question, they still did not meet the Lohr/808.1(d)(1) gloss on the express preemption clause because the guidances themselves were not “specific” to the defendant’s device.  “The former applies to 13 different kinds of devices, the latter to 18 kinds.”  969 F.3d at 1074.  Booker held that Lohr required very specific specificity.  “[T]he guidance does not impose “specific requirements applicable to a particular device,” 21 C.F.R. § 808.1(d) (emphasis added). . . .  Instead, it applies generally to every member of the class of intravascular filters.”  Id.

Second, the plaintiff’s loss at trial on his design claims transmogrified on appeal into a win on preemption:

[T]he requirements that the intravascular-filter guidance imposes are not relevant to [plaintiff’s] failure-to-warn claim.  State requirements cannot meaningfully be described as “different from, or in addition to, the specific [FDA] requirements” if the two requirements are not relevant to each other.  21 C.F.R. §808.1(d).

Id. at 1074-75.  While Booker cites §808.1(d) in creating this “relevance” requirement, no cognate of the word “relevant” in fact appeals anywhere in this regulation.  Booker effectively made up a new “relevance” gloss to add to the existing Lohr gloss on the statute’s very different wording – preemption whenever a state requirement is “different from or in addition to” a state requirement.  This is the antithesis of textualism, but it was made possible by the peculiarity of the bellwether result – the jury verdict against the design defect claims, as to which no “relevance” quibbles could have been asserted.

[T]he preemption argument fails because [plaintiff’s] claim rests on an asserted state-law duty to warn of the risks posed by the particular design . . ., and the FDA has not imposed any requirements related to . . . how a device of that design should be labeled.

Id. at 1076.

Thus the defendant in Booker lost the heart of its preemption defense because it won.  The bellwether jury returned a verdict against plaintiff’s design claims, making the preemption argument not “relevant.”  Such is the nature of MDL practice, and why, at a fundamental level, the MDL system is broken.

Another way in which the MDL system is broken is that it encourages far flung courts to contort state law in ways that a state court (or even most federal courts interpreting state law) would never do – almost always to the plaintiff’s advantage to maintain settlement pressure.  That happened twice in Booker.  First, as to warnings, no court applying the law of Georgia, the state in question, had ever extended the duty to warn to include adverse comparisons to the risks of competing products.  Indeed, Georgia courts have rejected analogous duties to warn about risks of competing products.  Davis v. John Crane, Inc., 836 S.E.2d 577, 584 (Ga. App. 2019) (adopting bare metal defense in asbestos litigation); PLIVA, Inc. v. Dement, 780 S.E.2d 735, 743 (Ga. App. 2015) (rejecting innovator liability).  Nonetheless, Booker held that any warning duty not expressly prohibited by Georgia courts would be allowed.  “Georgia has not adopted a categorical prohibition on basing a failure-to-warn claim on the absence of a comparative warning,” so Erie conservatism be damned, “we conclude that the district court correctly allowed the jury to decide the adequacy of the warning here.”  969 F.3d at 1076-77 (citations omitted).

A similar distortion of Georgia law occurred in connection with punitive damages.  Georgia law, as interpreted by Georgia courts, has strongly resisted punitive damages where, as in Booker, the defendant complied with the relevant governmental regulatory scheme.  In Stone Man, Inc. v. Green, 435 S.E.2d 205 (Ga. 1993), the court held that compliance “tend[s] to show that there is no clear and convincing evidence of willful misconduct, malice, fraud, wantonness, oppression, or that entire want of care which would raise the presumption of a conscious indifference to consequences.”  Id. at 206.  Thus compliance with governing standards ordinarily renders punitive damages unobtainable: “[P]unitive damages, the purpose of which is to ‘punish, penalize or deter,’ are, as a general rule, improper where a defendant has adhered to [the relevant] safety regulations.”  Id.  Where a product complies with all applicable regulations and standards, Georgia law supports summary judgment.  See Bryant v. BGHA, Inc., 9 F. Supp.3d 1374, 1396 (M.D. Ga. 2014) (summary judgment under Stone Man; design conformed to industry standards); Taylor v. Mooney Aircraft Corp., 464 F. Supp.2d 439, 448-49 (E.D. Pa. 2006) (same; applying Georgia law), aff’d, 265 F. Appx. 87 (3d Cir. 2008).  Compliance with applicable government standards can only be overcome with “evidence that the manufacturer engaged in a deliberate course of conduct which knowingly endangered those using the product.”  Uniroyal Goodrich Tire Co. v. Ford, 461 S.E.2d 877, 884 (Ga. App. 1995), rev’d in part on other grounds, 476 S.E.2d 565 (Ga. 1996).  Even adverse events involving similar injuries cannot overcome regulatory compliance and permit punitive damages.

There is no evidence that would justify an award of punitive damages in this case.  [Defendant’s product] complied with . . . the relevant regulations promulgated by [the federal government].  Plaintiffs hang their hat on the fact that, prior to [plaintiff’s] injury, [defendant] was aware of 741 [similar] injuries to [product users] from [similar] accidents . . ., but that peg is too wobbly to support punitive damages.

Hernandez v. Crown Equipment Corp., 92 F. Supp.3d 1325, 1356 (M.D. Ga. 2015) (footnotes omitted).  See Mims v. Wright Medical Technology, Inc., 2012 WL 1681810, at *5 (N.D. Ga. May 11, 2012) (summary judgment under Stone Man against warning-based punitive damages claim); In re Stand ‘N Seal, Products Liability Litigation, 2009 WL 2145911, at *9 (N.D. Ga. July 15, 2009) (same despite compliant recall plan’s later apparent inadequacy); Welch v. General Motors Corp., 949 F. Supp. 843, 845 (N.D. Ga. 1996) (same; despite similar problems with other vehicles).

In Booker, all this law went by the wayside as supposedly “derivative of [defendant’s] argument that it had no duty to warn of comparative risks.”  969 F.3d at 1077.  Inadequacy of the warning alone was enough to sustain punitive damages under Booker’s perfunctory reading of Georgia law:

[T]he jury found that the warning was not adequate.  As the district court explained, “[t]he evidence supported a finding that despite knowing that [its product] placed patients at a greater risk of harm” than other available filters, “[Defendant] chose not to warn physicians and instead downplayed the risk.”  Although it would have been possible for the jury “to draw a contrary conclusion,” we conclude that the evidence was adequate to support the jury’s award of punitive damages.

Id. (citation omitted).  No aggravating conduct is mentioned in Booker, only the alleged failure to comply with the disputed “duty” to warn about comparative risks of other products.  Booker thus reduced the strong Georgia compliance defense to punitive damages to a nullity.  No Georgia court, subject to review by Georgia appellate courts, would have made such a ruling.  Once again the pressure to affirm a “bellwether” verdict in an MDL has led to state law being bulldozed in order to give plaintiffs another tool with which to leverage settlement.

For all these reasons, Booker is another poster child for what is wrong with the current MDL system.