There’s an old joke about a guy who keeps banging his head against the wall. When asked why, he says: Because it feels so good when I stop. That’s roughly how we feel about watching defendants in medical device cases make preemption arguments under the shadow of Medtronic, Inc. v. Lohr, 518 U.S. 470 (1996). The head-banging is painful. It never really stops. But somebody has to keep doing it, because the wall eventually has to come down.

The latest occasion for this exercise is Dyer v. Coloplast Corp., 2026 WL 1906706 (M.D. Fla. July 2, 2026) — a pelvic mesh case in which the defendant lost its preemption arguments on summary judgment. We are not happy about it. We are, however, not surprised. And we think the defendant was right to try, and should keep trying, because Lohr is a flawed decision and we’ve said so here for years.

Plaintiff underwent pelvic floor reconstructive surgery during which a surgeon implanted defendant’s mesh — an §510(k)-cleared Class II medical device — to treat stress urinary incontinence. Dyer, 2026 WL 1906706, at *1. Plaintiff developed pelvic, groin, and lower extremity pain, was diagnosed with neuralgias, and underwent two additional surgeries. Id. She sued the manufacturer for negligence, strict liability design defect, and strict liability failure to warn, and also sought punitive damages under Florida law. Id.

The defendant moved for summary judgment on several grounds—most notably that plaintiff’s design defect claims were impliedly preempted because, pursuant to FDA regulations it could not unilaterally change the mesh’s design without going through the §510(k) process. Id. Before deciding the impossibility preemption issue, the court noted the elephant that has been in the room for thirty years — the Supreme Court’s ruling in Lohr that §360k does not expressly preempt state-law design defect claims against §510(k)-cleared devices. Id. at *2. Lohr has been in our crosshairs for a long time, and not just because it goes against defendants. It is a jurisprudential mess. A fragmented plurality opinion whose central reasoning has been chipped away, overruled, or repudiated by nearly every subsequent Supreme Court decision to touch it.

We’ve laid out the case against Lohr in detail several times.  The short version: Lohr‘s anti-preemptive reasoning rested on (a) the presumption against preemption in express preemption cases; (b) deference to an FDA regulation (21 C.F.R. § 808.1(d)) restricting the preemptive scope of §360k(a); and (c) the Court’s gestalt view that Congress could not have intended to bar most tort relief for injured patients. Every one of those foundations has since been kicked out from under it.

So yes, Lohr controls on the express preemption issue for §510(k)-cleared devices. But that is a wall that needs to be brought down. The way to bring it down is to get a well-built record before a court willing to send the issue up, and to keep the issue alive and preserved in every case until the Supreme Court takes it.

In the meantime, the implied preemption argument is valid. It just needs to be paired with a common-law argument that the court in Dyer never addressed. The court rejected defendant’s implied preemption argument finding that the manufacturer “had complete and independent control over the [mesh’s] design before it sought §510(k) clearance for the device,” and therefore “[i]t was not impossible to simultaneously comply with federal and state law.” Id. at *3 (emphasis added, citations omitted).

But generally speaking state product liability law requires a finding that a product is defective when it leaves the defendant’s possession, not at some earlier time. The design defect theory the Dyer court allowed to survive is premised not on what the mesh was when it left defendant’s control post-clearance, but on what it should have looked like before it was ever submitted to the FDA for clearance. In other words — you could have designed it differently before you ever went to the FDA. Therefore, no impossibility. Therefore, no preemption.

But that reasoning has a common-law problem. The product liability claim is not against the defendant’s hypothetical, never-manufactured alternative design from years before FDA submission. It is against the device as actually sold. At the time of sale — indeed, at the time of submission — the device’s design was subject to federal clearance requirements. A state tort judgment holding that the design was defective is a judgment that the defendant should have done something different. What it should have done, according to the plaintiff’s theory, is submit a different design to the FDA. That is the very essence of a “stop selling” or “never start selling” claim that the Supreme Court held preempted in Mutual Pharmaceutical Co. v. Bartlett, 570 U.S. 472 (2013). Bartlett‘s logic applies here with full force. If the only way to comply with the state-law duty is not to have marketed the product in its FDA-cleared form, then complying with state law would have required federal regulatory action — the submission of a different §510(k) — that the defendant could not independently accomplish.

None of this analysis was in the opinion, but it should be in the next motion, and the next, and the one after that.

While the preemption ruling is probably our biggest problem with Dyer, it isn’t our only one. The failure to warn ruling illustrates a recurring problem. The IFU warned of pain, neuromuscular symptoms, chronic complications, and the need for revision surgery – the very injuries plaintiff claims. Dyer at *4. But the court let the claim proceed anyway, because plaintiff’s expert said the warning should also have included comparative risk data comparing transobturator slings with retropubic slings. Id. at *5.

This is the classic no-win warning game. No matter how comprehensive the IFU, some expert can always find something more that allegedly should have been said. The information that plaintiff says was missing here — comparative complication rates between sling types — has never been held necessary to an adequate warning by any Florida appellate court. The two cases the court cited are federal district court cases, neither of which is binding. Yet a Florida product liability defendant is now going to trial on a warning claim premised on a duty that Florida’s own courts have never recognized.

This is a textbook Erie problem. Federal courts sitting in diversity are supposed to apply existing state tort law, not expand it to include duties no state appellate court has ever imposed. Telling a physician that a transobturator sling carries a different risk profile than a retropubic sling is not a warning about potential harmful consequences — it is a comparative marketing disclosure. Florida has never required that.

One piece of good news. The court granted summary judgment against punitive damages. Id. at *7-8. Under Florida law, a plaintiff seeking punitive damages must show by clear and convincing evidence that the defendant was personally guilty of intentional misconduct or gross negligence — the latter defined as conduct “so reckless or wanting in care that it constituted a conscious disregard or indifference to the life, safety, or rights of persons exposed to such conduct.” Fla. Stat. § 768.72(2)(b). The court quoted the observation from Tesla, Inc. v. Banner, 411 So. 3d 1, 5 (Fla. 4th DCA 2025), that “[i]t would appear that the Florida Supreme Court has all but eliminated punitive damage awards in products liability cases.” Dyer, at *7. Plaintiff, whose response brief simply recycled her design defect and failure to warn arguments, never came close to meeting that standard. Summary judgment on punitives granted. Id. at *8.

We will take our wins where we can get them.

The defendant in Dyer lost the battles that almost every defendant in a §510(k) medical device case loses. That is the current state of play under Lohr‘s long shadow. But the shadow is not permanent. The jurisprudential ground beneath Lohr has been eroding for years. The express preemption argument, the properly-framed implied preemption argument paired with a state tort law defect-at-sale rule and Bartlett‘s stop-selling logic, and the Erie objection to court-created comparative warning duties — these are all worth pursuing, on the right record, in the right court. Keep banging. The wall will come down.