Photo of Stephen McConnell

We can hardly think of a more useful recent law review article than D. Capra & L. Richter, “The Difference a Year Makes: The Admissibility of Expert Opinion Testimony under the 2023 Amendment to Federal Rule of Evidence 702,” 111 Cornell L. Rev. 127 (2026). Here is an online link. The authors are affiliated with the Advisory Committee on Evidence Rules, so their qualifications are not in doubt. Neither is their methodology. The article offers “an insider’s account of the process and thinking that generated the 2023 amendment to Rule 702.” Interestingly, we learn that a catalyst behind the amendment was a broad concern that forensic expert opinions in criminal cases (e.g., feature comparisons, such as fingerprints) were being presented with a degree of certainty that outstretched scientific support. Then the project became both sharper and broader. 

The “long journey to the 2023 amendment” included some false starts and blind alleys. The Advisory Committee considered directly calling out court decisions that had misapplied Rule 702, but ultimately decided against it. As the article points out, it is hardly surprising that so many courts misapplied the gatekeeping role, because Daubert “offered litigants a potpourri of pronouncements regarding a trial court’s evaluation of expert opinion testimony. There is language in Daubert for everyone.” In any event, we eventually got the amendment, which provided necessary clarification of the proponent’s burden of proving that the expert’s testimony more likely than not met the 702 requirements, and that the trial court must gatekeep against insufficient factual bases and errant applications of the purported methodology. 

The article then reviews every Rule 702 decision in the first year after the 2023 amendments.  This is the part of the article you will want to keep close at hand in your office the next time you are briefing a Rule 702 motion. 

While finding some progress, the article laments that litigants and courts continue to rely on the “antiquated” Daubert opinion (now superseded by two Rule amendments) and/or Rule 702 opinions that predate and are not compatible with the 2023 amendments.  Some courts “remain mired in the mixed messages of the now-dated Daubert opinion and risk admitting unreliable expert opinion testimony that threatens to alter outcomes of criminal and civil actions in federal court.”

The article contains a detailed examination and critique of hundreds of Rule 702 opinions. A pie chart shows 801 cases sorted into helpful categories. The article offers an interesting taxonomy of judicial attitudes toward amended Rule 702, including “Who, Me? Courts that Need No Correction,” “Old Habits Die Screaming: Federal Courts that Cannot Let Go of the Erroneous Tropes about ‘Liberal’ Admissibility,” and “Ignorance is not Bliss: Federal Courts that have Failed to Recognize the 2023 Amendment.” We feel we have encountered entirely too many courts in those last two categories. Too many courts still rely on “tropes” like “liberal thrust,” “presumption” of admissibility, and “questions of weight not admissibility,” “shaky opinions” and cross-examination that are no longer part of the rule.  

The article makes the key point that Rule 702 is not “synonymous” with Daubert.  Too many courts have failed to acknowledge the 2023 amendments at all.  Fifty-six opinions did not even cite the current language of Rule 702, but only the “defunct” prior version.  The authors express regret that the Advisory Committee failed to reject these “tropes” specifically.  The Committee “incorrectly” thought that courts would reach the proper conclusions themselves.  But some courts did arrive at proper conclusions.  Here is one of our favorite examples (not surprising given our Pennsylvania/Third Circuit focus): “The trial court in Knight v. Avco Corp. noted that Third Circuit precedent instructing that the Rule 702 standard is ‘not that high’ and that ‘the reliability and believability of expert testimony … is exclusively for the jury to decide’ was inconsistent with the preponderance requirement subsequently added to the Rule in 2023.”  Just so. 

A tone of exasperation creeps into the article. “The potential futility of painstaking efforts to lead horses to water is the rule makers’ lament. And the multitude of federal opinions that have failed to recognize the 2023 amendment to Rule 702, or that have flagrantly ignored its clear mandate, is disheartening to say the least.” More and better judicial education is recommended. It would also be helpful to get the citation services to red-flag Daubert as being superseded by Rule 702 amendments.  Counsel also need to get rid of outdated precedent in their motions and briefs. “[T]here were a stunning 316 cases in the year following the amendment to Rule 702 in which both the lawyers and the court were seemingly in the dark about the important amendment.”  

With the help of this article, and your briefing, that darkness can be dispelled. 

Photo of Michelle Yeary

This blogger spent the weekend in Morgantown, and West Virginia did everything right except the football score. The leaves were doing their early turn, that first crisp snap was in the air, and the drive in on 68 delivered exactly the scenery you hope for — ridge after ridge of Appalachian foothills rolling out in every direction, gold and green under a sky so blue it looked photoshopped. There to visit a sophomore daughter and catch a Mountaineers game, and while the football gods did not cooperate this particular Saturday, the vibe could not have been better.

Morgantown itself is a hilly little marvel. Every walk from campus to downtown, downtown to campus, or to literally anywhere, is somehow uphill. Both directions. This blogger is fairly certain the entire town was designed by someone who has never heard of a flat surface, and yet nobody there seems to mind, possibly because the view from the top of every hill makes the climb worth it. And yes, at some point during the weekend, in a car, on a porch, somewhere, John Denver’s Take Me Home, Country Roads made an appearance, because it is basically illegal to spend forty-eight hours in that state without it playing at least once. It is the state anthem in every way that matters, official designation or not.

Unfortunately, this blogger also had to bring some work along, in the form of a new decision out of the Southern District of West Virginia. And here is where the good vibes take a detour, because the ruling we are about to discuss does not treat federal pleading standards nearly as gently as West Virginia treated us this weekend. If Morgantown was almost heaven, the TwIqbal analysis in Davis v. Sagent Pharmaceuticals, Inc., 2026 WL 2797358 (S.D.W. Va. Sept. 18, 2026), is more like almost pleading. Let’s take the drive through it.

This is a generic drug product liability case arising from a recalled injectable steroid. Plaintiff alleged she received an injection of a methylprednisolone acetate suspension that was later subject to a recall due to contamination, and that she developed a serious infection leading to an amputation. Id. at *1. She sued the generic manufacturer and the distributor, asserting manufacturing defect, failure to warn, negligence, negligence per se, and breach of express and implied warranty claims. Id. at *2. Both defendants moved to dismiss.

Defendants opened with a shotgun pleading argument, pointing out that the complaint repeatedly referred to the defendants collectively without much differentiation. Id. at *3. The court rejected it, finding the complaint adequate because it identified, early on, which defendant was the manufacturer and which was the distributor, which the court found sufficient to put each defendant on notice of its role. Id. This one is defensible enough on its own terms. It is a low bar, but shotgun pleading doctrine has always been a low bar, aimed at truly indecipherable complaints rather than ones that merely lump defendants together in some paragraphs while differentiating their roles elsewhere.

The manufacturing defect claim is where things really start to get bumpy. Plaintiff’s own alleged evidence of exposure was a letter from her doctor stating only that she “may have been given one of the affected injections.” Id. at *5 (emphasis added). That is about as wishy-washy as causation allegations get, yet the court found it plausible that she received a contaminated unit and that the contamination caused her injury, relying on the temporal proximity between the injection and her symptoms and the fact that her prior injections apparently did not cause the same reaction. Id.

We have said it before and we will keep saying it — Twombly and Iqbal were supposed to mean something. Plausibility is supposed to require more than a “sheer possibility that a defendant has acted unlawfully.” Ashcroft v. Iqbal, 556 U.S. 662, 678 (2009). A complaint alleging that a plaintiff might have received a contaminated product, without any allegation actually linking her specific unit to the recalled lots, ought to be closer to that sheer possibility than to the plausible side of the line. We’ve seen sturdier bridges built out of duct tape. The court’s reasoning that a bad outcome following an injection, plus a recall notice phrased in the conditional, equals a plausible manufacturing defect claim is exactly the kind of res ipsa-by-complaint reasoning that TwIqbal was designed to prevent.

West Virginia’s innocent seller statute fared no better. It generally shields distributors from product liability suits unless a specific statutory exception applies. W. Va. Code § 55-7-31(b); Davis, at *4. Plaintiff invoked the actual knowledge exception, alleging that the distributor “knew” the product was contaminated. Id. at *5. No supporting facts. No allegations about how, when, or from whom the distributor supposedly gained this knowledge. Just the word “knew,” dropped into the complaint like a landmine plaintiffs hope the court will not notice is basically a legal conclusion masquerading as a fact.

The court noticed, and let it through anyway, accepting the bare assertion as sufficient to plead the exception plausibly and denying the motion to dismiss the manufacturing defect claim against the distributor. Id. This is Iqbal read backward. “Threadbare recitals of the elements of a cause of action, supported by mere conclusory statements, do not suffice.” Iqbal, 556 U.S. at 678. A single word alleging a state of mind, with zero factual enhancement, is about as threadbare as recitals get. If this is what plausibility requires, the pleading standard has been effectively repealed for one-word conclusions dressed up as knowledge allegations.

The one bright spot in the decision was failure to warn. Credit where due, because the court did get the preemption analysis right here. Relying on Mensing, the court held that failure to warn claims against the generic manufacturer were preempted because federal law requires generic drug labeling to remain identical to the corresponding brand-name label, making it impossible for the manufacturer to unilaterally add warnings. Davis, at *6. More notably, the court extended that preemption to the distributor as well, holding that state law warning claims are preempted whenever the defendant “lacked the ability to alter the drug’s label,” and that distributors, lacking any authority over labeling content, fall into that category regardless of the fact that Mensing technically addressed manufacturers. Id. The court rejected plaintiff’s argument that Mensing’s impossibility rationale should not extend to intermediate sellers, reasoning that a distributor cannot be liable for failing to do something the manufacturer itself is legally forbidden to do. Id.

The court also dismissed the negligence per se count outright, because West Virginia simply does not recognize an independent cause of action for negligence per se. A statutory violation is, at most, prima facie evidence of negligence. Id. at *8. No surprises here, and no complaints from us. This is settled West Virginia law, applied correctly.

Last stop on this tour of pleading standards gone soft — the express warranty claim survived even though plaintiff never identified any specific affirmation, promise, or representation the defendants supposedly made about the product. Id. at *9. The court found generic allegations that the defendants warranted the product would be “sterile, safe, and reasonably fit and suitable for its intended purpose” sufficient at the pleading stage, reasoning that plaintiff need not produce evidence of the warranty language this early in the case. Id.

Except an express warranty claim requires, at minimum, identifying the affirmation of fact that formed the basis of the bargain. W. Va. Code § 46-2-313(1)(a). Plaintiff did not quote any express language because there apparently was not any pled with particularity, just a generalized assertion that a warranty existed. Letting that survive a motion to dismiss is not accepting factual allegations as true, it is accepting a conclusion as true. Between this and the actual knowledge holding, the opinion is having a rough day for anyone hoping Rule 8 still has teeth in the Southern District of West Virginia.

This is a genuinely mixed bag, and not the good kind of mixed. The failure to warn preemption analysis, including the extension to the distributor, is a nicely reasoned application of Mensing that defense lawyers in generic drug cases should keep in their back pocket. But the rest of the opinion reads like plausibility pleading was optional homework. A causation theory resting on a doctor’s letter that only said the plaintiff “may have” gotten the recalled product, a knowledge exception satisfied by the word “knew” standing alone, and an express warranty claim that never identifies an express warranty, are not close calls under TwIqbal. They are the fact patterns those cases exist to weed out.

So we will take the preemption win, gladly, and file the rest of this one under “TwIqbal, we hardly knew ye.”

West Virginia gave this blogger perfect weather, a scenic drive, a promising football team, a very hilly campus, a much-needed daughter hug, and an unavoidable dose of Country Roads. It did not, unfortunately, give us a court willing to hold the line on federal pleading standards. Almost heaven only gets you so far when the complaint is held together with “may have” and “knew.” Country roads, take us home, but maybe drop the causation allegations off first.

Photo of Bexis

Last month Bexis filed his 134th, and probably last, amicus brief in the Pennsylvania Superior Court in a non-drug/device product liability case involving firearms.  Now edging into semi-retirement, Bexis is no longer in the business of drafting documents with footnotes.  Bexis final won/loss total is:

            Wins – 65       Losses – 38     Other – 29      Pending 2

            Wins Reversed On Further Appeal – 2      Losses Reversed On Further Appeal – 4

            Unpublished Wins – 11                                Unpublished Losses – 10

            Final, Published Wins – 52                        Final, Published Losses – 24

Anyone interested can see the complete list of Bexis’ 134 amicus briefs, dating back to 1992, with parties represented and issues briefed, here.

Here are some principles that Bexis has learned in over thirty years of writing and filing amicus briefs in significant pieces of litigation:

Number one:  Make sure to say something that is likely to contribute materially to winning the case.  Too many amicus briefs are filed by organizations and their counsel that do little more than wave the flag.  The presence of “me too” amicus briefs only devalues all amicus briefs, so would-be “me too” amici should simply sign onto amicus briefs written by someone else rather than file separately.

Number two:  Don’t expect to make any money directly from filing amicus briefs.  Writing an impactful amicus brief is often expensive, so amicus counsel should enter into their representations with their eyes open – realization rates for meaningful amicus briefs are likely to be very low.  Bexis rarely topped 25% for his amicus filings, that is, when he was paid at all.  Amicus briefs are loss leaders, with their value to their authors and their firms measured in gratitude from supporting industries – such as drug/device manufacturers – as a whole, and whatever publicity comes with helping win important cases.

Number three:  Act quickly, particularly if supporting the appellant.  Briefing schedules are short and, unless aided by the party they support, amici are very unlikely to obtain any extensions.  A second reason to act quickly is that early circulation of a persuasive draft amicus brief is the best way for an amicus to influence the arguments that the party it is supporting will make.  In amicus work, plagiarism is the ultimate form of flattery.  Amici will never be as influential as parties, so if a supported party wants to lift something that an amicus drafted, by all means the amicus brief writer should let that happen.

Number four:  Make sure that your amicus’ arguments are appropriate for the amicus organization.  If the amicus party is a nationwide organization, it’s brief should offer a nationwide perspective on the issue(s) being briefed.  If the amicus party is a state or local organization, that party’s brief should select and focus on an appropriate state or local issue.  In any event, most state and local amicus parties, at least on the defense side, are not in the financial condition to undertake extensive research on their own, so see item 3.

Number five:  The amicus brief writer must thoroughly research whatever legal issues are the subject of the filing.  Amicus briefs are important mainly because they focus on a particular issue in more depth than the party being supported can give it in this era of strictly enforced word-counts.  A comprehensive argument on the targeted issue is job one for any amicus.  A prospective amicus writer needs to know from the outset exactly what issue(s) will be covered, and needs to coordinate with the supported party.  Amici must also be careful not to bite off more than they can chew.

Number six:  Amicus curiae research must be both thorough and professional.  AI slop has negative value, so stay away from any use of AI in actually drafting an amicus brief.  While AI can be useful in an early round of research, no AI program yet exists that is better than legal research – and writing – performed by an experienced and dedicated lawyer.  AI does not understand the difference between recent and old cases.  In amicus (and most other) legal argument, reliance only on a pre-2000 precedent often looks like a tacit admission that the proposition in question lacks recent support.

Number seven:  Be a team player.  The more important the case, the more amici it will attract.  Experienced defense counsel in major litigation will have developed an amicus strategy for the appeal, in order to maximize perceived amicus impact and to reduce overlap and repetitive briefing.  Amicus brief writers may be able to influence this strategy, particularly if they have experience in the area, but should in all but the most extreme situations, stay in their designated lanes.

Number eight:  Write decisively.  Use declarative sentences.  Avoid passive voice.  Do not say something is “clear” or “obvious” – show that with research and argument.  Get rid of unnecessary words.  “In order to” almost always can be shortened to “to.”  Avoid starting sentences with “and” or “but.”  Break up long sentences and paragraphs.  Once a sentence has more than forty words, or a paragraph has more than five sentences, it is a candidate for splitting in two.  Do not start successive sentences or successive paragraphs with the same word.  The same holds true for using the same relatively distinctive word more than once in the same sentence. 

Number nine:  Block quotes are perfectly proper, but should be used sparingly and kept as short as possible.  Remember that the longer block quotes are, the more likely the reader is to pass over them.  The ellipsis is a friend, so long as it is used to shorten, rather than change, the meaning of, the text.  Using an ellipsis to alter a quote’s meaning is asking for trouble.  An introductory sentence to a block quote, with a few key words in quotes, is always good idea, and that idea gets better the longer the quote.

Number ten:  Make sure the brief is, and looks, professional.  This is a blogpost, so it’s fine to use contractions.  Stay away from them in briefs.  Likewise, avoid legal slang.  Yes, the Blog helped coin the portmanteau “TwIqbal,” but it’s not appropriate for an amicus (or any other) brief.  Use a standard font, such as Times New Roman, or Century Schoolbook, except on the cover, where putting the name of the court in Old English Text MT helps make the brief look as if it were professionally published.  Ditto with full justification of the text of the brief.  Ragged right side margins look like ragged teeth.  Footnotes should be in the same size text as the rest of the brief and should also be fully justified.  A justified text brief with unjustified footnotes just looks weird. 

Number eleven:  We’re not in law school anymore, so to hell with blue-book citation style.  Keep abbreviations in case citations to a minimum – limit them to “Co.,” Inc.,” “Assn,” and terms of that ilk.  Do not abbreviate directions.  Do not abbreviate anything that creates uncertainty, such as “Com.” or “Con.” or “Elec.”  The only exception that comes to mind is the names of law reviews.  Also, the most worthless word in any citation is “Ct.” in the closing parenthetical.  Of course, it’s a court; otherwise the brief would not be citing it.  String citations are OK if, and only if:  (1) every citation contains a jump citation to the relevant page of the opinion, and (2) every citation includes a parenthetical describing or quoting the relevant holding.  In briefs that are presenting nationwide research, it is a good option to replace long string citations with state-specific bullet points.

Number twelve:  Look like a local, particularly in state court.  Respect local citation forms, such as only using state reporters in New Jersey and following California’s placement of citation parentheticals before the actual citation.  Some jurisdictions require docket numbers for Westlaw/Lexis citations. If you are representing an amicus in an unfamiliar jurisdiction, it is always helpful – and frequently required – to have local counsel for advice on peculiar court rules or customary practices that affect how an amicus brief is written.

Number Thirteen:  A professional’s proof-reading is essential, and should not be left to secretaries, and never, ever to computers.  Read through the brief aloud before finalizing it.  Orally stumbling over words in the text is a dead giveaway that, even if grammatical, the text of the brief is not sufficiently clear.  A professional should be responsible for the final review of the table of authorities, and should do so manually, going page by page and checking off every citation.  Computers can only do so much; unusual citations will be missed.  Computers often treat numeric typos in citations as if they involve a new case.  Bexis’ last brief included a 50-state collection of standard jury instructions.  The computer omitted every one of them.

Number Fourteen:  Finally, be electronically literate.  Every part of every filing should be word searchable, since in the computer age text searches have reduced the importance of the table of authorities.  If the amicus brief contains Internet links, make sure that they all work.  Efiling has improved, but remains glitchy in many jurisdictions.  Therefore, waiting until the last minute to file your brief is asking for trouble.

Photo of Eric Hudson

Today’s opinion, Digan v. Luxco, Inc., 2026 U.S. Dist. LEXIS 209391 (D. Mass. Sept. 16, 2026), arises from the bad combination of an open flame, a fraternity party, and Everclear alcohol. For those who’ve never encountered Everclear, it’s a 95% ethanol, 190-proof alcoholic beverage. Stiff and obviously flammable stuff. A frat-member poured a bottle of Everclear on a backyard fire during a party, creating a fireball that badly burned the plaintiff. Plaintiff sued the manufacturer of Everclear alleging a variety of claims, most of which would not make the opinion blog-worthy for us. But one aspect of the opinion caught our eye—the court’s rejection of “negligent undertaking” liability. 

We’ve blogged about some troubling aspects of negligent undertaking liability before, and we don’t want to see it creep into our practice area. Here the court held that marketing a product does not create a special duty to ensure the protection of its consumers, and that’s worth a post.

Continue Reading Dismissal of Negligent Undertaking Claim in Massachusetts
Photo of Lisa Baird

Having clerked for a federal district judge and cut our litigation teeth in MDLs, federal court practice has burrowed deep.  When we think of procedural issues, our brains reflexively think in FRCP terms, and sometimes in state court the mental shift takes a beat.

Most of the time, it is a pretty short beat.  Sometimes, say in Louisiana or Pennsylvania state court, when they start throwing around the “dilatory exceptions” and “praecipes”, that beat takes a tad longer.  Things are similar but not quite the same.

Now think about product liability litigation in the U.S.—the general substantive legal principles we work with daily, as well as the broader litigation framework entailing document discovery, depositions, dispositive motion practice, the whole enchilada. 

It would be a mistake to take a brain stewed in the U.S. product liability milieu over to the European Union and not pause it to think through whether E.U. discussions of “strict liability” or “disclosure” or even “presumptions” mean what we think they do. 

We bring this up now because the new E.U. Product Liability Directive (aka “E.U. PLD” or Directive (EU) 2024/2853) will replace the nearly 40-year-old Product Liability Directive 85/374/EEC and take effect in less than 3 months, on December 9, 2026. 

Once it takes effect, all of us in the U.S. may do well to take a beat and figure out if the Directive’s language and procedures really mean what one might assume on first glance.

As an initial matter, we talk about “the” Directive, but it isn’t really one thing.  We rely on the English version of the E.U. PLD because we are essentially monolingual, but the European Union’s Official Journal has published the Directive in all E.U. languages, so the Polish version might be your touchstone for the E.U. PLD, or the Portuguese.  It’s a fair guess that there may be significant differences amongst the various translations, and no one version is more authoritative than the others.  That alone presents a definite opportunity for definitional arbitrage.

Then there is transposition.  Each E.U. member nation is obliged to transpose the Directive into national law by December 9, 2026.  A few have done so, and those presumably were able to figure out where the Directive best fit into their existing national law (both substantive and procedural) and had time to think through and resolve potential conflicts and inconsistencies.  Others have just started, and yet other countries have not yet begun (see here for a tracker regarding implementation).  The procrastinators likely will have no time to attend to where the Directive may create ambiguities or problems with laws already in force.  The point being that the E.U. Product Liability Directive may be a pretty clunky fit with existing national law in any particular country, so particular member states’ transposition laws are another place to look for differences and distinctions that may work to a defendant’s advantage once litigation commences under the new regime.

Turning back to the language of the E.U. PLD itself (English version), we cannot stress this enough:  The E.U. set out to increase the number of lawsuits, increase the number of product liability claimant victories, and increase the monetary recoveries, and they designed legislation to accomplish those goals. 

But while we are confident that the E.U. PLD will increase the number of lawsuits, we also think that whether the E.U. PLD meets the rest of the drafter’s stated goals will turn on the vigorousness of the defense once litigation begins.  That means not giving in to quick glosses on, or AI summaries of, the various E.U. PLD provisions (in whatever language you are litigating in).  The Directive itself has inconsistencies and knowing these exist will allow the unfortunate first few defendants to argue for the best possible version of conflicting language.

For example, take the term “defect.”

If you are in the U.S., you hear the term defect and you immediately wonder, manufacturing, design, or warning?  Under the E.U. PLD, there is just “defect” and no sub-categories. 

Do good warnings help a defendant avoid a finding of defect under the E.U. PLD?  No.  Or, maybe. Yes, definitely? 

In one place, the drafters declare that “warnings or other information provided with a product cannot be considered sufficient to make an otherwise defective product safe.” See Whereas Para. 31.  On the other hand, Article 7(2) declares:  “In assessing the defectiveness of a product, all circumstances shall be taken into account, including: (a) the presentation and the characteristics of the product, including its labelling….”  Right there is one gap, and we’d fight for the Article precluding a finding a defect because of good labeling, and a rejection of the inconsistent (and incomprehensible) prefatory language.

The standard for “defect” presents other openings for the defense as well. 

Under Article 7(1):  “A product shall be considered defective does not provide the safety that a person is entitled to expect or that is required under Union or national law”.  (Set aside the first part for a minute, but let’s not assume it automatically signifies a U.S.-style consumer expectations test). 

Look at the second part, “safety…that is required under [E.U.] or national law.”  It is written in a one-way fashion—non-compliance equals defect. 

But there are cracks.  Article 7(2) makes “relevant product safety requirements, including safety-relevant cybersecurity requirements” one of the “all circumstances” taken into account in “assessing the defectiveness of a product”.  That’s good.  Article 10(2(b) makes defectiveness a presumption upon the claimant’s demonstration “that the product does not comply with mandatory safety requirements laid down in Union or national law that are intended to protect against the risk of the damage suffered by the injured person.”  That’s less good.

But which is it?  If we are representing the defendant facing this word salad (in whatever language, and through and with qualified local counsel of course) we might decide to take a run at just arguing that Article 7(1) contains an implicit, and automatic, regulatory compliance defense that Articles 7(2) and 10(2)(b) do not foreclose. 

The infamous E.U. PLD presumptions provide additional openings for defendants to try to take back some ground.  Even within the U.S., how presumptions work is always a complicated matter.  They might help a plaintiff survive a dispositive motion and then disappear for trial purposes, they might form part of a jury instruction, they might be rebuttable, or not, or something else. 

We are guessing here, but we suspect that most E.U. member nations do not have a well-established and developed practices regarding the effect and operation of presumptions, so defendants should have a real opportunity—country to country, judge-by-judge—to advocate for what these presumption parts of the E.U. PLD mean, and how they should function. 

And the presumption provisions (Article 10) are particularly messy and confusing.  

Article 10(2) says a product is presumed defective where:

  • (a) The defendant fails to comply with Article 9(1), newly requiring defendants to cough up evidence;
  • (b) The claimant demonstrates that the product does not comply with mandatory safety requirements under E.U. or national law (as noted above); or
  • (c) The claimant demonstrates that the damage was caused by an “obvious malfunction” of the product during reasonably foreseeable use or under ordinary circumstances—whatever that means. 

Article 10(3) says that a causal link between defectiveness and damage is presumed where the product has been established as defective and the damage is “of a kind typically consistent with the defect in question”.  Again, whatever that means.

Article 10(4) says that where, despite disclosure of evidence, the claimant faces “excessive difficulties, in particular due to technical or scientific complexity,” in proving defectiveness or the causal link, the court shall presume defectiveness, causation, or both — provided the claimant “demonstrates” it is likely that the product was defective or that there is a likely causal link.

Frankly, we think these all are awful.  But nothing about these presumptions is self-apparent, and so every term, every phrase, presents an opportunity to imagine what they could mean, or should mean, under this entirely new regime.  Why litigate Article 10(2)’s “obvious malfunction” as some form of U.S.-style res ipsa loquitur when the Directive’s “obvious malfunction” is so undefined that it calls for some new definition of its own?  At most, the drafters gave “a glass bottle that explodes in the course of reasonably foreseeable use” (Whereas para. 46) by way of explanation, but that hardly answers the question of what “obvious malfunction” means.  Maybe “obvious malfunction” in the E.U. PLD means the claimant must come forward with proof that the product deviated from its manufacturing specifications in a way that caused it to work improperly and caused harm.  We could live with that.

There are more E.U. PLD inconsistencies, ambiguities, and internal tensions worth exploring and defining for the better.  Additional ones include:

  • The “excessive difficulties” presumption standard of Article 10(4), which is utterly vague.  The Directive states that a claimant who “demonstrates” excessive difficulties in proof gets a presumption of defect, causation, or both.  What does “demonstrate” mean?  Article 10(4) does not say, but the prefatory language suggests it won’t take much—elsewhere the Directive says the claimant “should provide arguments to demonstrate excessive difficulties” but that “proof of such difficulties should not be required.” (Whereas, para. 47).  We see no reason to roll over on the proof of requirement, particularly because the concept of “technical or scientific complexity” is left entirely to case-by-case judicial interpretation.
  • Article 3 mandates “maximum harmonization” in that E.U. countries “shall not” diverge from the Directive.  But at the same time, national law and each country’s judicial system provides an overlay that will help determine the standard of proof for the claimant’s prima facie case, rules on calculating compensation and damages (including “non material” damages like pain and suffering), how the disclosure requirements work, and procedural rules for bringing claims.
  • The “development risk defense”, which countries are expressly allowed to reject or modify pursuant to Article 18.

All of the above thoughts go to things to think about once the Directive takes effect and the first lawsuits are filed under it.  You have 76 days left until then, and the action items for now remain the same:

  1. Audit your supply chain to figure out who is an economic operator subject to liability the Directive;
  2. Review and update any indemnification, distribution, and software integration agreements;
  3. Review and update any product liability insurance coverage for E.U. activities;
  4. Audit regulatory compliance and systems;
  5. Prepare for new disclosure obligations by reviewing document and data creation and retention practices, and consider both the GDPR and the Directive’s provisions about “excessive difficulties” of proof; and
  6. Look closely at software components and digitally-connected products, including lifecycle management and expectations for updates under the Directive.
Photo of Stephen McConnell

Perez v. Apyx Medical Corp., 2026 WL 2691359 (N.D. Cal. Sept. 14, 2026), is a product liability suit alleging misrepresentations about off-label use of a company’s cosmetic surgery §510(k) device. That, at least, is how we read the complaint.  The court read the complaint a bit differently. The opinion was authored by the Judge Tigar, who possesses a burning bright intelligence, so we must allow for the possibility (unlikely as it might be) that we are wrong.  Even so, Judge Tigar dismissed the complaint (sparing us the need for despair or to pursue the Blake-ian inquiry of “What immortal hand or eye/Dare frame they fearful symmetry?”)

The plaintiff in Perez claimed that she suffered severe burns and scarring from the defendant’s medical device when it was used on her to perform skin resurfacing and body contouring procedures. The plaintiff’s causes of action included (1) strict liability under the risk-expectation theory, (2) strict liability under the consumer expectation theory, and (3) negligent design, manufacturing, marketing, and sale of the device. The defendant moved to dismiss based on preemption and failure to state a claim. 

The court held that the claims were not preempted.  It held that express preemption went nowhere because the device had gone through the 510(k) process, not premarket approval. We, along with at least one of our faithful readers, must grit our teeth at this semi-spurious distinction, but then we will move on to the implied preemption argument, which needed a bit more analysis.  The defendant argued that the complaint was preempted because it was based on allegations of misrepresentations about off-label use in violation of the Food,  Drug, and Cosmetics Act (FDCA), but the FDCA prohibits private enforcement.  The court rejected that argument, reasoning that the complaint travelled on traditional state claims, not disguised attempts to enforce the FDCA.  The court opined that the defendant “misreads Perez’s complaint. At heart, Perez’s claims are design defect and negligence claims, not off-label marketing claims, and therefore not preempted.” To our eyes, the court fell for the oft-used plaintiff lawyer trick of couching claims at a higher level of abstraction to avoid specific defenses.  That trick did not work in the recent SCOTUS Monsanto case, and it should not have worked here.  

But that high level of abstraction did end up costing the plaintiff in this case. The court dismissed the entire complaint, with leave to amend. The decision dismissed the design defect claims for failures to plead, but not because California law prohibits such claims against prescription medical products – because the defendant did not raise that argument in a timely fashion.  Rather, the court focused on how the allegations that physicians misused the product in a dangerous fashion did not plead that the product had any defect.  Foreseeable misuse is not enough to sustain a defect claim, whether design, warning, or manufacturing.  Because the plaintiff did not identify any product defect, the negligence claim also failed: “Perez does not identify what design defect affected Apyx’s Renuvion product and only contends in conclusory fashion that the foreseeable use or misuse of the product and the fact that it caused her injury sufficiently pleads the existence of a defect.” The complaint “does not identify the design defect that caused her injury.”  Goodbye, complaint (but try again). 

The defendant made a couple of other arguments, on causation and consumer expectation defect, but they were denied. The court thought the causation argument was more suitable for summary judgment, after factual development, and not so much as a way to test the pleadings. The Perez court held that the consumer expectation issue would not be ripe until jury instructions would be determined.

Of course, depending on whether the plaintiff can fix her complaint, there might not ever be jury instructions. 

Photo of Michelle Yeary

We’ve written a lot of benzene posts on this blog over the years, and most of them share a common thread. Plaintiffs find a chemical decomposition byproduct nobody put on a label, dress it up as a corporate conspiracy, and ask a court to do what years of FDA rulemaking declined to do. Every so often, though, a court doesn’t just reject the theory — it takes the theory apart piece by piece, cites every reason it fails, and locks the door on the way out. That’s In re L’Oréal USA Benzoyl Peroxide Products Litigation, 2026 WL 2719653 (S.D.N.Y. Sept. 15, 2026), where the court granted the defendant manufacturer’s motion to dismiss with prejudice, essentially ending this MDL. If you’ve been waiting for the decision that says everything the defense bar has been saying about OTC benzene litigation for the last two years, this is it.

The case involved three benzoyl peroxide (“BPO”) acne treatment products that plaintiffs allege contained or degraded into benzene, a known carcinogen. Plaintiffs allege they would not have bought the products, or would have paid less for them, had that been disclosed. The complaint pleaded four state-law causes of action — deceptive trade practices, breach of implied warranty, negligent misrepresentation/omission, and unjust enrichment — all resting on the claim that the products were “adulterated and misbranded.” Id. at *1.

This litigation traces back to a March 2024 citizen petition asking the FDA to recall BPO products, followed by an FDA notice a year later confirming that the agency’s own testing found benzene contamination in some products, though less than third-party testing had suggested, and that a limited retail-level recall was warranted — a recall that, notably, did not instruct consumers to do anything with product already in their homes because, as the FDA itself said, the cancer risk from daily use over decades is very low. The manufacturer here agreed to a single voluntary recall of one lot of one product. No plaintiff alleged buying from that lot. Id. at *2.

The manufacturer moved to dismiss on the ground that the FDCA preempts all of it. The court agreed, and it did not take a narrow path to get there. It walked through the entire regulatory architecture and closed off every possible workaround plaintiffs tried to construct.

Start with the basics — OTC drugs like these get to market without individualized FDA review so long as they comply with the applicable monograph.  Here, the Acne Monograph, which functions as a regulatory recipe specifying the approved active ingredients and the labeling requirements for the category. The FDCA’s cosmetic labeling preemption provision, 21 U.S.C. § 379s(a), bars any state-law labeling requirement that is “different from or in addition to” the federal one. The Second Circuit has read that to mean that state law is preempted if it is not “exactly the same as” federal law. Id. at *4. A parallel provision, § 379r(a), does the same work for adulteration claims. Id.

The court examined the Acne Monograph and found that it says nothing about benzene. It permits BPO as an active ingredient, and it lists the required warnings, and benzene appears in none of them. That should have been the end of the misbranding theory. It was. But plaintiffs, being plaintiffs, did not stop there. They tried two workarounds, and the court shut down both.

First, plaintiffs argued that the Acne Monograph’s own “general conditions” incorporated 21 C.F.R. § 201.66’s requirement that OTC drugs list active and inactive ingredients on the label, and that benzene was an undisclosed ingredient that should have been listed. The court rejected this because benzene is not an “inactive ingredient” under the regulation, which defines a component as something a manufacturer “intended for use in the manufacture of a drug product.” Nobody alleged the manufacturer intended benzene to show up. It shows up, if at all, because BPO decomposes, not because anyone put it there. Id. at *7.

Second, plaintiffs invoked the FDCA’s general “misbranding” provision in chapter V, which prohibits false or misleading labeling, arguing it independently required a benzene warning. The court explained that chapter V is a backstop against affirmatively misleading statements — the “this product cures baldness” problem — not an engine for generating new disclosure obligations the FDA considered and declined to impose in a monograph it spent decades building. As the court put it, reading the general misbranding language to create a benzene-warning requirement the FDA never included would “directly contravene the Acne Monograph’s explicit labeling requirements.” Id. at *6. We’d go further than the court did and just call that reading absurd. The FDA looked at BPO for the better part of a century, decided the word “cancer” shouldn’t even appear on the label, and plaintiffs want a court to override that considered judgment using the FDCA’s general catch-all clause. No.

The court also made quick work of plaintiffs’ attempts to use two recent Second Circuit decisions — Rutledge v. Walgreen Co. and Yousefzadeh v. Johnson & Johnson Consumer Inc. — as escape hatches. Rutledge says manufacturers are permitted to add supplemental warnings; it does not say they are required to. And the dictum in Yousefzadeh about a monograph drug still being misbranded if it has “a dangerous effect on the user” was aimed at something like cyanide in a bottle, not a substance the FDA specifically evaluated and approved as an ingredient. Nice try, but citing a case for a principle it doesn’t actually support is not a strategy that wins in front of a court paying attention — and this court was paying attention. Id. at *8.

Next the court dismantled plaintiffs’ claimed “2 ppm federal benzene cap.” Plaintiffs alleged that a 2 parts-per-million cap on benzene existed under federal law, that the products exceeded it, and that this made the products “adulterated.” The court found no such cap exists anywhere in the FDCA or FDA regulations, and without it, the adulteration theory has nothing to stand on.

Plaintiffs tried three sources for this imaginary cap, and the court took each one out individually. The voluntary recall notice doesn’t establish it — a voluntary recall does not, by itself, establish a violation of federal requirements, and the notice itself never called the recalled product “adulterated.” Id. at *9. FDA guidance documents and warning letters don’t establish it either — they are expressly non-binding, don’t carry the force of law, and the FDA’s own Q3C Guidance says on its first page that it creates no rights and binds nobody. Id. at *9-10. And the USP doesn’t establish it. Chapter 467’s 2 ppm figure applies only to benzene as a “residual solvent,” meaning a chemical used or produced directly in manufacturing that isn’t fully removed by practical techniques. But plaintiffs’ own complaint described a multi-step decomposition process — BPO breaking down into benzoic acid radicals, which further decompose into benzene radicals, which then produce benzene — not a residual solvent. Plaintiffs’ own pleading defeated their theory. Id. at *10.

As a parting shot, the court noted that USP Chapter 467 repeatedly uses the word “should,” not “shall” — and under ordinary usage, “should” recommends, it doesn’t mandate. Even if plaintiffs had cleared every other hurdle, they were trying to build a federal requirement out of a document that speaks in suggestions. Id.

The court denied leave to amend and dismissed with prejudice, reasoning that plaintiffs had already amended once, had been given a further opportunity to amend before this motion was even fully briefed and declined to take it, offered no proposal for how a new complaint would fix anything, and — most importantly — were trying to plead around a preemption defect that no amount of additional facts can cure. Id. at *11. This isn’t a case getting narrowed for round two. It’s over.

This opinion is as complete a preemption win as you’ll see in the OTC benzene space, and it’s worth reading in full for the sheer discipline of it — every plaintiff argument gets a paragraph, and every paragraph ends the same way. But there is another part of this decision that we want to highlight. It’s the court’s recognition, running through the whole opinion, that the FDA already did the science here, and plaintiffs’ lawyers are trying to use an economic class action complaint to relitigate a regulatory judgment the agency reached after decades of study.

The court didn’t gloss over this. It noted that BPO’s tendency to decompose into benzene has been studied since 1936, that the FDA has been examining BPO’s safety since the early 1980s, that the agency concluded in 1982 that BPO at these concentrations is safe and effective. Further, when the FDA finalized the Acne Monograph in 2010, its own advisory committee unanimously decided the word “cancer” shouldn’t even appear on the label. That is not an agency that overlooked benzene. That is an agency that considered it, repeatedly, over the course of roughly ninety years, and made a judgment call. The court’s language on this point deserves to be quoted directly:

[W]here the FDA has constructed an elaborate labeling framework [in a monograph] specifying labeling requirements, and [a particular substance] appear[s] nowhere in that framework, the inference that the agency made a considered judgment about label content is substantially stronger than in cases where the regulatory scheme merely touches on the relevant product category.

Id. at *7 (quoting Brower v. Colgate-Palmolive Co., 2026 WL 1831253, at *5 (S.D.N.Y. June 25, 2026).

That’s the whole ballgame, and it’s an argument we’ve been making for years across every mass tort that tries to use a courtroom as a second regulatory agency. Litigation should not lead science, and it especially should not lead science that the actual scientific regulator has already led. Plaintiffs’ lawyers didn’t bring new data. They didn’t bring a new toxicological study. They brought a reading of “should” as “shall,” a warning letter with no legal force, and a recall notice that says the opposite of what they needed it to say — and asked a federal court to override the FDA’s considered, decades-long judgment anyway. The court said no. Score one for letting the people who actually do the science keep doing the science. We’ll take it.

Photo of Bexis

Nearly a decade ago we published the blogpost “Pre-Approval Design Defects – No Such Thing.”  As we discussed, “pre-approval” design (later joined by warning) claims were a transparent preemption dodge.  No such claims had ever been advanced before the Supreme Court’s holdings in the generic drug preemption cases (Mensing/Bartlett) that actions requiring FDA prior approval – which includes all design changes affecting safety – were necessarily preempted because the defendants could not unilaterally make those changes purportedly required by state law.  Our prior post wasn’t really about preemption, but rather the incompatibility of “pre-approval” defect claims with accepted state law:

The reason that [the case we discussed] was “unable to conceive of” a state law pre-approval duty is because such duties do not exist.  Design defects under Restatement (Second) of Torts §402A (1965) . . . [are] limited to products that are defective at sale. “The rule stated in this Section applies only where the product is, at the time it leaves the seller’s hands, in a condition . . . which will be unreasonably dangerous” to the ultimate consumer.  Restatement (Second) of Torts §402A, comment g (1965).  In the Third Restatement, this time in the black letter, . . . all “categories of product defect” are likewise determined “at the time of sale or distribution.”  Restatement (Third) of Torts, Products Liability §2 (1998).

That “pre-approval” defect claims are contrary to state law (and therefore aren’t available ab initio as a preemption dodge) was strongly reinforced by the California Supreme Court’s recent decision in Gilead Tenofovir Cases, ___ P.3d ___, 2026 WL 2223748 (Cal. 2026) (which we discussed generally, here).  Since Brown v. Superior Court, 751 P.2d 470 (1988), California has not allowed strict liability design claims against prescription medical products, so plaintiffs had to pitch the “duty to innovate” claims that the Gilead court rejected as “negligence.”

Plaintiffs in Gilead had every opportunity to pursue standard defect allegations in that litigation, but they declined to do so.  Instead, they pursued their novel theory that, regardless of any defect, the defendants were negligent when they elected to delay pursuing FDA approval of an allegedly “safer” version of their breakthrough anti-HIV drug, in favor of successfully obtaining FDA approval of a related drug – the drug the undoubtedly saved most of these ingrate plaintiffs from slow and inevitable death from AIDS.  But for our purposes, Gilead also illustrates why the law unanimously restricts defect allegations, and any other kind of tort claim, to conditions existing at the time of product manufacture and sale, as opposed to speculative pre-FDA-approval allegations.

Gilead recognized that, in California as elsewhere, product-based negligence “claims still require proof that the manufacturer’s negligence resulted in a defective product.”  Id. at *5 (citations omitted).  Divorcing negligence from product defect (what the Gilead plaintiffs advocated) “would create substantial tension with that [product liability] body of law.”  Id. at *6.  That’s because “products liability law employs specific tests to determine whether a product is defective.”  Id.  Such specificity is particularly necessary when the plaintiff’s allegations attack prescription medical product development:

Unlike product defect claims, which are evaluated based on contemporaneous and relatively complete scientific evidence available at the time of the product’s distribution, evaluating the reasonableness of drug-development decisions would require retrospective assessment of inherently provisional judgments made based on evolving scientific data.

Id. (citation omitted) (emphasis added).  Such development is a “multistep, highly regulated process.”  Id. at  *7.  FDA approval matters.  It prevents speculative and conjectural claims.

The plaintiffs’ claims in Gilead were based “on a flawed premise; namely, that a drug can be sufficiently developed, and its safety and efficacy sufficiently known, before completion of phase III clinical trials and FDA approval.”  Id.  Similarly, “pre-approval” defect claims assert that same false premise – that “defects” can be determined prior to a product’s FDA approval. 

That premise is inconsistent with the statutory and regulatory framework governing drug [and medical device] development.  It also underscores the difficulty of imposing negligence liability based on preliminary and evolving safety assessments of an alternative [design] still under development, rather than on a defect in the [product] currently being sold.

Id.

“[D]efinitive judgments regarding” yet-to-be approved products’ “risks and benefits ordinarily cannot be made before completion of phase III trials and FDA approval.”  Id. at *8.  For that same reason, state law provides no basis for alleging “pre-approval” defects.  Instead, the common-law limits product liability defect claims to those in existence at the time that the product at issue was sold.  Gilead recognized that “[p]roducts liability law has long tied a manufacturer’s duty of care to defects in the product placed into the stream of commerce.”  Id.

In particular, Gilead held that sufficient knowledge of the risks and benefits of FDA-regulated products for those risks to be “foreseeable” does not exist “before completing phase III clinical trials and obtaining FDA approval.”  Id. at *9.  “[T]he difficulty of making comparative judgments, the structure of the [FDA]-approval process, and the governing statutory and regulatory framework demonstrate that such knowledge ordinarily cannot be established at earlier stages of development.”  Id.  “[O]nly about one-third of drugs evaluated in phase II clinical trials yield results promising enough to proceed to phase III testing,” and “fewer than one out of eight will obtain FDA approval.”  Id. at *10, 11. (citations omitted).  Thus, as a matter of law, Gilead held that “a manufacturer cannot truly know of a drug’s safety and efficacy before phase III testing and FDA approval,” making such risks unforeseeable.  Id. at *10.

Any such injury [to a plaintiff] would depend on speculation regarding the independent conduct and discretionary decisions of multiple third party actors, as well as the uncertain outcomes of additional clinical trials and scientific research required to develop, approve, and market the new drug.

Id. at *11.  No outside “assessment of a drug’s safety and efficacy” can “make FDA approval certain.”  Id. (citation omitted).  Pre-approval risk/benefit assessment is, as a practical matter, impossible:

[B]ecause of this high failure rate, . . . drug manufacturers must make complex strategic decisions about where to devote finite resources based on limited information.  These decisions are often provisional, iterative, and contingent on preliminary laboratory or clinical studies that may later prove incorrect.  As a result, they do not readily lend themselves to retrospective evaluations of reasonableness armed with the benefit of hindsight.

Id. at *16.

For these reasons, prescription medical product liability claims are not evaluated prior to FDA approval.  Rather, like all product defect claims, defect are only measured “at the time of manufacture and distribution.”

Product defect claims are judged based on contemporaneous and relatively complete scientific evidence available at the time of manufacture and distribution. . . .   [A] retrospective inquiry creates a substantial risk that hindsight will supplant the manufacturer’s contemporaneous judgment, even when its decisions were reasonable when made.

Id. at *17 (emphasis added).  Thus, the “novel theory” in Gilead that would have required a jury’s evaluation of a defendant’s pre-approval conduct and information was rejected:

That inquiry asks fact finders to reconstruct years of research and development decisions made in the face of evolving scientific evidence and uncertain outcomes, rather than to evaluate the safety of a product that actually reached the market.  Neither the plaintiffs nor the dissent have identified a single case wherein a fact finder has been tasked with making such determinations.

Id. (emphasis added).  Gilead emphasized that product liability is limited to those prescription products “that are already on the market,” not “decisions made long before a drug is approved or marketed.”  Id. at *20 (emphasis original), 21.

Although the plaintiffs stayed away from the term “defect,” Gilead involved a disguised “pre-approval” design defect claim.  To avoid preemption, such claims propose to shift the defect inquiry from the time that a drug or device was actually manufactured and sold, to some indeterminate time before FDA approval, when some alternative, purportedly safer, unapproved design was theoretically possible without FDA review.  Evaluating purported design defects prior to any FDA approval is a bad idea for all of the reasons Gilead discussed – which is why no state’s common law that we know of allegations of “defects” to any other time than traditional requirement of defect at sale.

And, as Gilead noted, plaintiffs in that high-stakes litigation were unable to cite “a single case” supporting a “defect” existing prior to FDA approval.

“Pre-approval” defect claims are figments of plaintiffs’ perfervid imagination, lacking any basis in existing tort law.  They are alleged solely for the purpose of avoiding preemption.  The doctrine of constitutional avoidance counsels against reaching constitutional arguments when non-constitutional issues are dispositive.  Defense counsel should respond to such made-up claims by establishing that they have no legal support in the relevant jurisdiction.

Photo of Bexis

Time for another shameless plug. The 31st Annual ACI Drug & Medical Device Litigation Conference takes place December 2–3, 2026 at the New York Marriott Marquis, and several of our Reed Smith bloggers are planning to be there. The full agenda is here.

We’re especially pleased that our own Lisa Baird is speaking on “Putting the Offense in the Preemption Defense” – exploring whether the time has come to go on the offense with preemption, including leveraging the Supreme Court’s FIFRA opinion in Monsanto v. Durnell and whether the reasoning in Medtronic v. Lohr has finally been undermined enough for courts to take notice.

The conference features 75+ speakers, including in-house leaders from Medtronic, Bristol-Myers Squibb, Novo Nordisk, Johnson & Johnson, Gilead Sciences, and many more, alongside top defense firms and members of the judiciary. The conversations you have here can genuinely shift your practice.

ACI has once again invited the blog to serve as a media sponsor, and they’re offering our readers a special 10% discount on registration. Just be sure to use the code D10-999-DDLB27 when you sign up.

You can register here. We hope to see you in New York!

Photo of Steven Boranian

No, your DDL bloggers are not offering a money-back guarantee—although we suppose offering one would not cost us anything, since reading the blog is free.  For enterprises that actually sell products though, the story is sometimes different, and many product sellers do guarantee satisfaction—or your money back.  Moreover, when those guarantees are unrestricted, we have a difficult time seeing how consumers could ever file lawsuits claiming they were overcharged.  All they had to do was ask for refunds, without the need for litigation or exorbitant claims for attorneys’ fees.  Case closed.

That is what happened a little while back in a short-but-smart order from a New York trial court.  In Scott v. Walmart, Inc., No. EF012051-2025, 2026 WL 2561736 (N.Y. Sup. Ct. July 28, 2026), the court dismissed a consumer class action brought by the purchaser of an over-the-counter sleep aid.  Because the seller offered a money-back guarantee, the purchaser/plaintiff had no standing to sue.  And, federal law preempted the claims to boot.

In Scott, the OTC product’s label said the product was “non-habit forming.”  The plaintiff alleged that statement was misleading because the product’s active ingredient allegedly could elevate the user’s mood, which could cause the user to want to use the product more.  Now mind you, this plaintiff did not allege that he actually experienced these sensations, or experienced any alleged injury at all.  He alleged instead that he paid more for the product because of the “misleading representations” and thus sought a partial refund for himself and his proposed class.

The court was having none of it.  First, the product seller offered an unrestricted money-back guarantee under which purchases could return the package for a full refund.  Sure, if there were strings attached, perhaps the purchaser could allege an injury in fact under New York’s consumer protection statutes.  But here the guarantee was unrestricted.  As a result, the court concluded that “Plaintiff herein would likewise be fully compensated by virtue of [the defendant’s] unrestricted money-back guarantee for the Product and therefore, having suffered no injury on fact, he lacks standing.”  Scott, at *4.

Second, even if the plaintiff had alleged a cognizable injury, federal law preempted his claims in any event.  As the court explained, Congress has created a national and uniform scheme to protect consumers from fraud or misrepresentations in the sale of food, drugs, and cosmetics.  Moreover, the FDCA contains an express preemption clause that bars states from establishing any requirement for OTC drugs that is “different from or in addition to, or that is otherwise not identical with” federal requirements.

Here, the FDA regulates the OTC product at issue through an FDA Monograph, which sets forth federal requirements for product labeling.  Because the plaintiff’s lawsuit would require the product seller to include different or additional information on a federally approved drug label, the plaintiff’s state-law claims were preempted.  In so holding, the court rejected precedent from our home district, the Northern District of California, which ruled that similar claims were not preempted because the Monograph did not explicitly address whether the “non-habit-forming” statement is misleading.  The New York court correctly saw it differently, and it held that if the FDA regulates a given subject matter—here, dependency on a certain active ingredient—then federal law “preempts all non-identical state laws within that subject matter.”  Scott, at *5.  Regulation of particular or specific statements within that subject matter is not required. 

The court closed with some choice words for the plaintiff’s counsel, who evidently favors exaggerated punctuation (lamely echoed in the title to this blogpost!!).  The court did not appreciate counsel’s attempt to denigrate defense counsel, nor counsel’s “specious attempt to cast aspersions on his adversary and to detract from the issues raised in the motion, especially in light of the disjointed and excessively punctuated submissions by Plaintiff’s counsel.”  Scott, at *5.  We will accept the court’s admonition on behalf of law bloggers everywhere.  Will our analysis be disjointed?  Usually not.  Will we excessively punctuate our pithy observations?  Never!!!!  Guaranteed, or your money back. 

Drug & Device Resources

Bexis’ Book

FDA-regulated products now account for an estimated one-fifth of overall economic activity in the U.S. They have also been the focus of a litigation explosion. This timely guide covers all aspects of litigation involving drugs, medical devices, vaccines and other FDA-regulated prescription products.
Learn more about this book