Here we go again.  The winter solstice is upon us.  The days are short; the nights are long; and we have to rely on holiday lighting to keep the darkness – if not the cold – at bay.  Speaking of cold and darkness, it is now time for us to look back upon the results of drug and medical device litigation during 2023 and to select the ten worst decisions of the year.  But this year, as bad as the bottom ten opinions we discuss below are, our sense is that they are somewhat overshadowed by even worse non-prescription medical product judicial malarkey.

In particular, it’s hard to avoid the disastrous Mallory v. Norfolk Southern Railway Co., 600 U.S. 122 (2023) (here).  Although Mallory involves railroads, it threatens to undo much of what we’d thought we’d won with Bristol-Myers Squibb Co. v. Superior Court of California, 582 U.S. 255 (2017) (2017+1).  Who would have thought that defense side would prefer Justices Ginsburg and Breyer to the current, purportedly more “conservative” Supreme Court.  But here we are.

Nor do several adverse federal law decisions count, because they either, like Mallory, do not involve prescription medical products or do not involve product liability:  Alliance for Hippocratic Medicine v. FDA, 78 F.4th 210 (5th Cir. 2023), was the Fifth Circuit’s blatantly politicized attack on the FDA’s regulation of abortion-related drugs.  It utilized a dumbed down “arbitrary and capricious” standard while simultaneously taking by an anti-choice group’s pleaded “facts” as true (here, here, and here) for purposes of injunctive relief.  The questionable allegations of that single complaint thus effectively trumped many years of the FDA’s science-based decision-making.  We also consider the same court’s equally politicized attack on FDA authority in Apter v. Dep’t of HHS, 80 F.4th 579 (5th Cir. 2023) (here), to be worse than most of the decisions we’re discussing here.  Apter would deny the FDA the power to warn the public about the hazards of widespread off-label uses.  See also In re LTL Management, LLC Bankruptcy Proceeding, 64 F. 4th 84 (3d Cir. 2023), rejecting, as not in “good faith,” a so-called “Texas Two-Step” mass-tort based bankruptcy filing in a talc case.

But even putting those goliaths aside, it’s an unpleasant enough job to chronicle the ten worst prescription medical product liability decisions of the year.  We’ve been diligently preparing bottom ten annual lists since 2007, even though it’s distasteful, because if we don’t do it nobody else is likely to, and these abominable decisions deserve to be called out for what they are.  While we know that a late-breaking holiday horror, such as T.H. v. Novartis, 407 P.3d 18 (Cal. Dec. 21, 2017), in 2017), or Bausch v. Stryker Corp., 630 F.3d 546 (7th Cir. Dec. 23, 2010) (here), can supersede our list, so far they’ve been thankfully uncommon.

So we’ll stop blathering and get on with it.  Here are the ten decisions we’d most strongly prefer that judicial Santa Clauses not have left under our (or plaintiffs’) tree.  If you’ve been burned by any of these judicial debacles, we sympathize.  Bexis’ amicus efforts this year in Mallory obviously came to naught, and we’ve had our own cases on this list before (see, e.g., 2013-2 and 2021-10).  A lawyer who never loses doesn’t litigate hard cases.  Moreover, the pain here is temporary, since next week we’ll trot out our top ten best decisions of 2023.

1. In re Valsartan, Losartan, & Irbesartan Products Liability Litigation, 2023 WL 1818922 (D.N.J. Feb. 8, 2023).  Somewhat indicative of the “overshadowing” phenomenon mentioned above, for the first time our worst drug/device decision of the year is an unpublished (thankfully) district court opinion.  But in terms of both massiveness and maleficience, it was a whopper – an abuse of discretion so vast it remains difficult to do Valsartan (in)justice.  Around the turn of the century, the Supreme Court essentially cast product liability claims out as candidates for class action status.  But in one fell swoop Valsartan certified four different nationwide class actions, two for medical monitoring and one each for consumer protection and third-party payers.  These classes embrace over 400 products made by almost 30 defendants and include claims under the differing laws of 52 American jurisdictions.  No jury of average citizens could possibly sift through all that garbage intelligently.  And for all that this 40,000+ word opinion discusses, what it leaves out is even more important – such as any discussion of how any member of the class could possibly prove causation or entitlement to medical monitoring.  There’s no science to back any of the allegations up – as demonstrated by the masterful (and even longer) take down of nearly identical substantive allegations in the Zantac (2022+4) MDL.  The FDA, for one, advised patients to keep using these drugs.  Valsartan, by contrast, responds with platitudes and 50,000 foot generalizations.  Valsartan is the antithesis of the “rigorous” scrutiny that the Third Circuit requires of putative class actions.  On legal issues, too, Valsartan is fatally flabby.  It claims that Pennsylvania law is representative of medical monitoring claims nationwide but fails even to describe accurately Pennsylvania’s idiosyncratic version of medical monitoring (requiring negligence and thus not an “independent” claim).  Even in Pennsylvania, a mere “possibility of an increased risk” is not enough to entitle anyone to medical monitoring.  Erie doctrinal conservatism (mandated by over a dozen Third Circuit decisions) is nowhere in sight – as Valsartan would allow medical monitoring in (among other places) Illinois, New Hampshire, and Delaware, the most recent three states whose high courts have flatly rejected such causes of action.  Moreover (as we suspected at the time), the unprecedented “express warranty” claim based on these drugs simply appearing in the FDA’s “Orange Book” that a previous Valsartan decision (2021-4) claimed to be the law in all fifty states, now appears as the linchpin for why that cause of action can be certified.  We can only repeat how we closed our discussion of this opinion:  “Valsartan is one of the most nakedly result-oriented mass tort decisions that we have ever read.”  Unfortunately, the Third Circuit declined a Rule 23(f) interlocutory appeal for unknown reasons.  This is the third time that a Valsartan MDL decision has made our bottom ten list (see 2020-10).  We hope that the Valsartan defendants can stick it out, rather than be forced to settle, because we can’t see this decision ultimately surviving, assuming it continues to be contested.  We vilified Valsartan here.
2. State ex rel. Shikada v. Bristol-Myers Squibb Co., 526 P.3d 395 (Haw. 2023).  Shikada wasn’t 100% bad, but in a year where there weren’t that many truly awful (or truly great) appellate cases, it was bad enough for a state high court decision.  Shikada’s implications are, if anything, more significant than its holdings.  Shikada is the first appellate decision creating a state-law warning-based duty (here, via a consumer protection claim, brought by the state rather than anyone actually claiming injury) based solely on  pharmacogenomics − racially/ethnically-based genetic variations − allegedly affecting the effectiveness of FDA-approved prescription drugs.  That’s significant because the labels for over 500 drugs already have such information, under a voluntary FDA program.  In Shikada, the ultimate result was a nuclear penalty of over $800 million.  The penalty did not hold up, but the duty did.  The logic of the consumer fraud claim made no sense.  Allegedly, consumers deprived of efficacy (not risk) information had their informed consent rights infringed, which in the case of persons without the genetic difference, meant that they were spared the expense of undergoing negative genetic testing.  That was “damage” despite those consumers in fact saving money.  In some ways that was even worse than medical monitoring claims, which at least require that the testing be “necessary.”  Shikada also neutered the relevant statutory safe harbor for conduct complying with administrative orders.  The safe harbor did not include a defendant’s decision to add information that the FDA encouraged but did not mandate.  Given that ruling, the defendant’s preemption defense was also a goner.  So Shikada let the state have it both ways – prosecuting “conduct” for purposes of the state-law safe harbor, but “labeling” for purposes of preemption.  Supposedly anything sufficient to trigger a state-law duty to update automatically qualified as preemption preventing “newly acquired information.”  In the end Shikada interpreted the Hawai’i consumer protection statute to allow liability to the state for what amounts to classic “negligence in the air” − disfavored conduct that never actually hurt anybody “even if the drug proves to be safe,” where the state contends that additional investigation should have been conducted.  So, even though various errors required reversal of the monetary penalty, this dangerously broad statutory construction remains.  Avoiding barnyard expletives – we shoveled Shikada here.
3. Painters & Allied Trades District Council 82 Health Care Fund v. Takeda Pharmaceutical Co., ___ F. Supp.3d ___, 2023 WL 4191651 (C.D. Cal. May 24, 2023) (“PATDC82”).  If any individual pharmaceutical plaintiff:  (1) had let their agent destroy most of the relevant documentary evidence, (2) was suing over an increased risk of three-one hundredths-of one percent (0.0003), (3) was alleging fraud on the FDA, and (4) had their experts ignore an obvious alternative cause, that plaintiff would have been laughed out of court (and probably sanctioned).  But in the topsy turvy land of Ninth Circuit class action law, PATDC82 permitted such a plaintiff to represent a nationwide third-party payor class action asserting claims under the much-abused RICO statute.  Letting plaintiffs use RICO (or the False Claims Act) to bring fraud on the FDA claims (preemption only applies to state law) is a distinct minority position, but the Ninth Circuit has.  This nationwide class action thus expanded that circuit’s peculiar law to the entire country.  The would-be TPP class representative avoided spoliation by blaming an administrator that it had hired, and by asserting that its magical statisticians didn’t even need actual evidence.  The decision simply ignored the minuscule increased risk, and didn’t compare it to the risks of any alternative treatment – or to the risks of leaving the disease, diabetes, untreated.  The efficacy of the drug in question in treating diabetes appears to be unchallenged, so the entire litigation amounts to a judicial tax on a safe and effective drug.  Finally, plaintiffs’ magic statisticians opined that a decline in prescriptions for the defendant’s drug at about the same time as the discovery of the minuscule risk demonstrated some sort of causation.  But that decline had an obvious alternative cause – generic competition began during the same time period – that the statisticians made disappear by, again simply by ignoring it.  This combination of a nationwide class action with disfavored fraud on the FDA claims, destruction of evidence, minimal risk, and overreliance on questionable statistics, gives PATDC82 our “show” position as the third worst decision of 2023.  We pilloried PATDC82 here.
4. Sykes v. Cook Inc., 72 F.4th 195 (7th Cir. June 23, 2023). Sykes involved the weaponization of one of our least liked MDL divergences from the normal federal rules − direct filing of complaints into the MDL itself − and nonexistent (to Fed. R. Civ. P. 7) “pleadings” called “master” and “short-form” complaints.  In Sykes, the plaintiffs’ MDL solicitation machine dredged up several totally asymptomatic plaintiffs, who had “no pain or other symptoms,” only scans that allegedly showed that the defendant’s IVC filters “had perforated their veins.”  This reminds us of pleural thickening plaintiffs in asbestos actions, and evidently the MDL court thought so as well, because plaintiffs’ claims had been dismissed for lack of present injury under these plaintiffs’ home states’ laws.  Systematic early vetting of the garbage cases being filed no doubt would have found many more uninjured plaintiffs, but hey, at least the defendant got these, right?  Wrong.  On appeal, the Seventh Circuit, effectively made early vetting of uninjured claims an exercise in futility.  It vacated the MDL court’s with-prejudice dismissals because “to a legal certainty” those uninjured plaintiffs could not allege the jurisdictional amount ($75,000) for them to establish subject matter jurisdiction.  So, instead of early vetting that would permanently rid this MDL of uninjured (and never exposed) plaintiffs, they simply walk away none the worse for wear – and perfectly free to refile their actions – whenever they can allege the requisite injury.  That’s a neat trick, available only to MDL plaintiffs – they could adopt a master complaint’s false (in their cases) injury allegations to sue, but once they got caught, those plaintiffs avoided any penalty by virtue of the contradictory facts pleaded in their short-form complaints.  Sykes thus endorsed fail-safe pleading for MDL plaintiffs and made early vetting mostly a waste of time.  Still, there might be a silver lining to Sykes, early motions to dismiss MDL plaintiffs who do not affirmatively plead injury (or exposure) for lack of subject matter jurisdiction, with accompanying demands for immediate discovery into this non-waivable defense might now find some purchase, at least in the Seventh Circuit.  We slammed Sykes here.
5. Holley v. Gilead Sciences, Inc., 2023 WL 6390598 (N.D. Cal. Sept. 28, 2023).  Given our deep distaste for novel allegations of some “duty to innovate” purportedly safer products, even absent any claim that their FDA-approved predecessors were defective, decisions that permit such regressive claims are likely to find their way to our bottom ten (2021-5; 2019-7).  Indeed, this is the second time down that highway to litigation hell for “Holley.”  It is worse than ironic that the drug plaintiffs have attacked, tenofovir, was the drug that, as much as any, converted AIDS from a death sentence to merely a chronic condition.  No state suffered more from the AIDS epidemic than California – talk about biting the hand the feeds (saves) you.  Holley denied summary judgment against duty-to-innovate claims.  Despite their alleging that the defendant’s FDA-approved drug should never have been approved, or even submitted to, the FDA, Holley denied defendant’s preemption defense against what are substantively “stop-selling” claims held preempted in Bartlett (2013+1).  But there was no alternative; at the time the FDA approved the defendant’s first form of tenofovir, the plaintiff’s proposed “alternative” was still 10+ years in the future.  Imagine how many tens of thousands more AIDS deaths would have occurred had the defendant actually done what plaintiffs demanded.  Holley also allowed, in “risk/benefit” design analysis, plaintiffs to compare the risks and benefits of one drug against another, separately FDA-approved drug, which was unprecedented, since alternative designs are not supposed to be separate products.  Holley also allowed a “pre-approval” warning claim to escape preemption, largely on the same rationale.  That makes no sense either, because at the time of FDA approval, not to mention before FDA approval, there is, by definition, no “newly acquired information” beyond what the FDA considered in its approval decision.  An allegation, as of that time, that FDA lacked any particular data is a fraud on the FDA claim preempted under Buckman.  There was, of course, no California state appellate authority for any of these novel Erie predictions at the time Holley was decided, which is another black mark.  He heckled Holley here.
6. Wilson v. CooperSurgical, Inc., 2023 WL 6216933 (S.D. Ill. Sept. 25, 2023).  Remember Riegel (2008+1)?  In Riegel, the United States Supreme Court held that all the usual product liability claims:  “strict liability; breach of implied warranty; and negligence in the design, testing, inspection, distribution, labeling, marketing, and sale of the [product], and their claim of negligent manufacturing claims,” were expressly preempted as to any FDA pre-market approved medical device.  Since then, plaintiffs have chipped away at Riegel, mostly by turning “parallel claim” dictum about an issue that Riegel declined to decide into an exception to preemption, but the core of Riegel, preemption of design, warning, and implied warranty claims, generally remained.  But if Riegel was the zenith of PMA preemption, Wilson reached the nadir.  The defendant’s motion to dismiss was denied entirely − not a single claim in the plaintiff’s blunderbuss 10-count complaint was held preempted.  Mostly, Wilson went from there to here by refusing to consider preemption at all on a Fed. R. Civ. P. 12(b)(6) motion to dismiss.  It strictly applied the Seventh Circuit’s unfortunate language in Bausch that preemption as an “affirmative defense” was not generally subject to Rule 12(b)(6), to stick its head in the procedural sand and avoid preemption as a “procedural tripwire.”  Never mind that, since Bausch, the Supreme Court has entertained preemption issues under Rule 12(b)(6) on multiple occasions, including the Mensing (2011+1) prescription drug preemption decision.  If Bausch was ever a correct statement the law, it has since been impliedly overruled.  Wilson added the notion – contrary to the basic premises of TwIqbal – that plaintiffs should have “discovery” before their claims were dismissed on the pleadings.  Wilson finished by making a hash of state (Illinois) law – ignoring state high court precedent that the FDCA’s express rejection of any private right of action foreclosed negligence per se and more recent intermediate appellate authority just as firmly rejecting FDCA-based failure to report claims.  Unfortunately, politics also likely played a part, as the product was a contraceptive, and the judge had been a member of a group devoted to banning abortion.  Thus, Wilson was this year’s worst example of the other side’s enlistment of product liability claims as a means to do away with reproductive freedom.  We went to war with Wilson here.
7. In re Philips Recalled CPAP, Bi-Level Pap, & Mechanical Ventilator Products Litigation, 2023 WL 7019287 (Sp. Mstr. W.D. Pa. Sept. 28, 2023).  CPAP isn’t technically even a judicial opinion (had it been it would have been further up this list).  Instead, it was a long and for us lugubrious recommendation written by a “special master,” an office with less authority even than a magistrate judge.  Essentially, a special master serves at the pleasure of the district judge overseeing a case.  In CPAP, however, the special master would expand the scope of state tort law in numerous ways that even an Article III federal judge cannot under the Erie doctrine, as applied by the relevant appellate (Supreme and Third Circuit) courts.  Repeatedly, CPAP interpreted state law in ways having no state appellate support.  Repeatedly, CPAP permitted purported state-law claims unless they were “clearly” precluded by existing state law.  Both these approaches to state law were simply ultra vires under Erie, which does not allow federal “judicial pioneers” in state law matters.  The Supreme Court has held (and reiterated) for over eighty years that “the proper function” of a federal court “is to ascertain what the state law is, not what it ought to be.”  Goodbye to all that in CPAP.  Indeed, this type of disrespect for state authority over state law has been so prevalent in MDLs such as CPAP that we call it the “MDL treatment.”  Here are some of the major errors in the 50+ page CPAP recommendation:  (1) treating mere discovery (fact sheets) as if they were Rule 7 “pleadings” for purposes of a Rule 12 motion to dismiss; (2) Applying the same “presumption” against preemption that the Supreme Court expressly rejected in Buckman, and relying on inapplicable “different from, or in addition to” express preemption language to hold that plaintiffs’ multiple allegations of “failure to apprise the FDA” as something other than impliedly preempted fraud on the FDA claims; (3) citing no law from any state, only an uninformed (and incorrect) gut reaction, to “recommend” that the learned intermediary rule not apply in cases of “no” warning, as opposed to “inadequate” warning; (4) permitting state-law negligent misrepresentation and consumer protection claims unless “clearly” prohibited by state law; (4) restricting the comment k “unavoidably unsafe” defense to design defect claims in Pennsylvania and California, despite directly-on-point contrary appellate authority in both states; and (5) permitting totally unprecedented “battery” claims against product manufacturers despite that theory’s rejection by every prior decision in the prescription medical product context.  These recommendations are sufficiently contrary to controlling Erie authority that we believe that, should they actually be adopted by a real court, mandamus would be a proper remedy (and we would undoubtedly include such a decision on next year’s bottom ten list).  We castigated CPAP here.
8. Crockett v. Luitpold Pharmaceuticals, Inc., 2023 WL 2187638 (E.D. Pa. Feb. 23, 2023).  Crockett checks in as the worst purely Rule 702 expert admissibility opinion of the year.  Rule 702 was amended at the beginning of this month to specify that:  (1) the court, not juries, decide whether the rule’s four admissibility prerequisites are met; (2) proponents of expert testimony must establish each of these four elements by a preponderance of the evidence; and (3) to be admissible, expert opinions must reliably apply the relevant principles and methods to the case-specific facts.  Not a moment too soon, as Crockett demonstrates what can happen when Rule 702’s provisions are ignored.  The comments to the 2023 amendments make clear that there never was a “strong preference” for admissibility in Rule 702, as Crockett claimed.  Crockett allowed a physician, with no warning related qualifications other than “clinical and research expertise,” to opine that FDA-approved drug warnings were inadequate.  Rather than examine the actual bases of those opinions, as Rule 702(b) required, Crockett waved the testimony through with the single observation that the expert claimed to have “conducted a “detailed review of the literature.”  Nowhere did Crockett discuss what that “literature” might be.  Worst of all was Crockett’s treatment of the expert’s “hypothesis” that the relevant risk was “systematically underreported” in clinical trials, although the expert never said by how much.  The only basis for that opinion was supposed “extensive research and experience” that the risk occurred “way more” than what the trials stated.  Crockett admitted this mush with the observation that “arguments about reliability” merely “go to the weight” of the opinion.  The Rules Committee’s comments to the amended rule refute that result.  Another abdication of judicial responsibility under Rule 702(b) was Crockett’s dismissal of the defendant’s challenges to the factual bases of certain opinions.  As the Rules Committee points out, “disputes concerning the data underlying” an expert opinion implicate one of the express Rule 702 prerequisites to admissibility, and thus cannot be left “for cross-examination.”  We criticized Crockett here.
9. KeraLink International, Inc. v. Geri-Care Pharmaceuticals Corp., 60 F.4th 175 (4th Cir. 2023).  As the only other published pro-plaintiff appellate decision of 2023, Keralink would have ranked higher (lower?) on our bottom ten, but for the issues it decided – the “sealed container” and “economic loss” defenses − being relatively uncommon in prescription medical product liability litigation.  It was a commercial product liability case, and summary judgment for the plaintiff was affirmed.  The plaintiff alleged economic loss because the defendant’s product, used to preserve human tissue, was contaminated and rendered some of plaintiff’s tissue unusable.  The state’s (Maryland) sealed container defense had an exception for any intermediate supplier that held itself out as the manufacturer.  Unfortunately, one defendant placed its logo on the material’s container and did not list anyone else as manufacturer on FDA filings.  Plaintiff won as a matter of law.  The other defendant labeled the product as “sterile,” which supposedly was enough to establish another exception, warranty.  KeraLink serves as a negative checklist – what you don’t want to do – on the sealed container issue.  KeraLink also affirmed summary judgment for plaintiff on the economic loss rule, albeit on the unusual fact pattern that the plaintiff did not technically own, and could not sell, human tissue, and thus was limited to lost service fees and employee time.  As a matter of law, the plaintiff’s “possessory rights to the donated tissue” were enough to avoid dismissal for seeking purely economic damages.  We kvetched about KeraLink here.
10. Estate of Cronin v. G4 Dental Enterprises, 2023 WL 2779206, 526 P.3d 1111 (table) (Nev. App. April 4, 2023).  Cronin is the only other adverse 2023 state appellate decision in our prescription medical product bailiwick, so we included it, even though it is not citable in Nevada.  It expanded strict liability to medical professionals – dentists – for the allegedly mis-manufactured customized dental implant they made in their office.  We’ve never seen a medical procedure considered to be a product defect before.  If allowed to stand, that imposition of strict liability could significantly complicate and deter the use of personalized medicine, as well as the medical use of 3D printing (Cronin doesn’t say, but its holding would cover 3D-printed bespoke medical devices).  Because the FDA does not regulate dentists (or other healthcare providers), and its authority over 3D printed medical devices is underdeveloped, the usual grounds for preemption become problematic.  Likewise, Nevada recognizes the learned intermediary rule, but how it would apply where the same defendant is both prescriber and product manufacturer is also unclear.  Cronin did not discuss any of this, but simply jumped off the cliff into the great unknown of ever more liability.  We contemplated Cronin here.

We also thought long and hard whether In re Zostavax (Zoster Vaccine Live) Products Liability Litigation, 2023 WL 5044944 (E.D. Pa. Aug 8, 2023) (here) should make the 2023 list.  But while we don’t like another federal rule, Rule 4, getting the MDL treatment, the issues in the other decisions at the bottom of our list – preemption, scope of liability, and expert admissibility under Fed. R. Evid. 702, are ultimately more consequential.  Also, given other recent rulings in Zostavax (2022+10; 2021+19), the remaining plaintiffs in Zostavax aren’t likely to profit from that mistake.

Good riddance.  Once we’ve recovered from our excessive exposure to toxic jurisprudence, we’ll start preparing for something we like a lot better − next week’s presentation of the top ten best drug/device decisions of 2023.