We are watching the post-MDL-remand Bard IVC filter litigation with interest. It bears some resemblance to the Bone Screw litigation of the late 1990s/early 2000s (except in Bone Screw, there was no MDL settlement) in that the targeted defendant is engaged skirmishes across the country. In mass torts, the plaintiffs have the benefit of numbers – knock out one and 999 remain. But defendants have some advantages, too: better coordination, and stare decisis.
The latter is primarily how the Bone Screw litigation (and later Aredia/Zometa and (outside the MDL context) Infuse) mass tort was won. The enduring importance of defense-friendly precedent is one counter our side has to the plaintiffs’ raw numbers. Good precedent will weaken all plaintiffs’ positions – now and in the future. From Buckman Co. v. Plaintiffs Legal Committee, 531 U.S. 341 (2001), down to dozens of trial court decisions recognizing the learned intermediary rule’s applicability to medical devices or rejecting FDCA-based negligence per se, plaintiffs are still losing cases due to Bone Screw precedent twenty-plus years later. Not wanting to lose like this creates favorable incentives to resolve litigation.
We offer Nelson v. C.R. Bard, Inc., ___ F. Supp.3d ___, 2021 WL 3578874 (S.D. Miss. Aug. 6, 2021), as Exhibit A.
Inferior vena cava (where the “IVC” comes from in this litigation) filters function something like internal nets to catch blood clots forming in the legs (deep vein thrombosis) before they can migrate to the lungs and kill or cripple people via pulmonary embolism. The netting function is accomplished by long, thin metal “struts” anchored to the vein interior that catch the clots. Metal being metal, these struts can bend or break under the constant pressure of blood flow, particularly as successfully caught blood clots increase overall flow resistance. This can cause the filter itself to tilt, or even come loose. Flesh being flesh, it can be penetrated by the metal legs, or by pieces of broken legs, which can cause the filter to tilt and get stuck.
All of these risks are well known to the experienced surgeons who implant these devices, and these devices’ FDA-regulated instructions for use explicitly warn about all of these risks.
In Nelson the filter “had fractured,” and “[s]ome of the struts of the filter had penetrated through the inferior vena cava wall and some had migrated to other parts of [plaintiff’s] body.” 2021 WL 3578874, at *2. Since the basic risks of these devices are well-known and understood in the relevant medical community, plaintiffs in IVC cases have to invent other information that they can then claim was not included in these product’s warnings.
So plaintiffs mine the FDA’s adverse event reporting system for whatever they can find to show that any given IVC filter has “comparatively” more reports than others. Id. at *3 (“rate of fracture . . . was higher compared to other filters on the market”). This effort, of course, is often self-fulfilling, since every filed lawsuit against a targeted device model generates a report − inflating its “reported” adverse reaction rate as against other devices that aren’t targeted. Publicity generates more reports.
Our posts here and here describes the various biases inherent in FDA reporting. So does the FDA’s webpage for the “MAUDE” medical device reporting database. “MAUDE data is not intended to be used either to evaluate rates of adverse events or to compare adverse event occurrence rates across devices.” Which is precisely what the plaintiffs misuse MAUDE data to claim. MAUDE also comes with this disclaimer:
A report or other information submitted by a reporting entity under this part, and any release by FDA of that report or information, does not necessarily reflect a conclusion by the party submitting the report or by FDA that the report or information constitutes an admission that the device, or the reporting entity or its employees, caused or contributed to the reportable event.
Quoting 21 C.F.R. §803.16.
That kind of statistically invalid “comparison” is exactly what the plaintiff in Nelson tried to pull, quite unsuccessfully. Summary judgment was entered on the comparative claim under state law (Mississippi has an exclusive product liability statute), and the reasoning is instructive. The statute includes a state of the art provision (plaintiff must prove that the defendant “should have known about the danger” in question) and a definition of adequacy (information “that a reasonably prudent person . . . would have provided with respect to the danger”). Nelson, 2021 WL 3578874, at *6 (quoting statute). Mississippi also follows the learned intermediary rule. Id.
Thus, from a plain reading of the statute, to establish liability on the part of a medical device manufacturer on a failure to warn claim, a plaintiff must show that (1) a product was unreasonably dangerous because it failed to contain adequate warnings to the professional prescribing the product (2) of a danger that caused the damage for which plaintiff seeks recovery, (3) of which danger the manufacturer knew, or should have known.
As to the risks that befell this plaintiff – filter fracture, perforation, and migration – these were “specifically warned against,” id. at *10 (quoting extensive warnings) so that the defendant’s instructions for use were adequate as a matter of law:
These warnings are clear and explicit and were sufficient to warn [the implanter] of these possible complications. [Plaintiff’s] filter fractured; fragments migrated; fragments perforated the vena cava wall; fragments embolized; and these complications resulted in medical intervention. . . . Because Defendant’s warnings expressly warned [implanter], as the learned intermediary, of these very complications, under Mississippi law, Defendants’ warnings . . . were adequate as a matter of law.
Id. (footnote omitted).
That’s an excellent result right there – a to-be-published decision finding warnings adequate as a matter of law. But even more interesting is Nelson’s denial of the plaintiff’s overreaching summary judgment motion based on his fall-back claim that the warning was defective for not adversely comparing the defendant’s own product’s adverse report rate with competing products. First, there was no recognized warning claim based on “failure to provide comparative risk information.” Id. at *7. Plaintiff’s claim that comparative risk information was something an “adequate” warning required was contrary to Johnson & Johnson, Inc. v. Fortenberry, 234 So.3d 381 (Miss. 2017), which has held a drug label adequate as a matter of law for stating all the relevant risks notwithstanding “testimony from an expert . . . who testified that the safest . . . medications were two other brands.” Id. (discussing Fortenberry, 234 So.3d at 393). The proper focus of a warning claim is the “actual” label, not “marketing materials and internal documentation.” Id. at *8. “Plaintiffs’ theory of liability utilizing Defendants’ internal documentation to argue that the warning was inadequate goes beyond the statutory scope and takes us far afield from a manufacturer’s duty under Mississippi law.” Id.
Second, it was improper to go beyond “decades of strict liability principles” and allow comparative warning claims even though a product “specifically and clearly warned of [the relevant] dangers/complications.” Id. (cleaned up). Plaintiffs claimed that “physicians and patients expect that kind of [comparative] information” was belied by the simple fact that the implanting surgeon “ha[d] never seen comparative risk information included in any IFU.” Id. Further, “subjective expectations are not the standard for liability . . . regardless of the reasonableness of such expectations.” Id. Nelson thus provides two excellent responses – one factual and one legal – to plaintiffs questioning a prescriber/implanter about what s/he “expects” from a product label.
Third, Nelson recognized that comparative labeling claims generally are bad policy, since they would impose a difficult and self-defeating warning standard. That theory’s first problem is it “effectively charges manufacturers and sellers with having to know their competitors’ products’ failure rates.” Id. at *9. Product warnings are properly limited to what a manufacturer can control − the characteristics of its own products. The second problem is what we discussed above, comparative claims:
Ignore the informational biases associated with the latency with which manufacturers or sellers receive complaints for new products as compared to established products, and create new liability in every instance where one product is alleged to have higher risk of complication than another.
The third problem is medical causation. Comparative warning claims seek to impose liability “regardless of whether any such difference [in reporting rates] has clinical significance.” Id. Such claims embark on a “slippery slope”:
At present, a manufacturer’s duty is to adequately warn of known adverse effects. To go beyond that and find that, even though a manufacturer warned of a known danger, it can be liable because it failed to give the frequency of the danger or the percentages of danger compared with other products makes legal compliance nearly impossible and potential liability wholly unpredictable . . . [since] rate of failure statistics would change on a regular basis.
Id. Requiring manufacturers to warn of “the ‘degree’ of the danger” in addition to the fact that it is a known significant risk would allow liability despite these being “complications that are encountered with all IVC filters.” Id. Moreover, “[t]he fact that a danger exists without a proper warning of its existence is what makes a product potentially unreasonably dangerous, not whether the danger is more likely than not to occur.” Id.
Finally, Nelson also rejects the use of comparative risk evidence in design defect claims. Plaintiffs claimed that “other IVC filter designs exist” that are “safer,” based on the same allegations of comparative risk. That alternative design argument also failed:
First, Plaintiffs cite no legal authority holding, or even discussing, that a predecessor design of a product that carried the same risks but less chance of occurrence would legally suffice as a feasible design alternative. Second, regardless of any difference between [purported alternative designs], it is Plaintiffs’ burden to show that a feasible design alternative would have to a reasonable probability prevented the harm. Simply reducing the risk is not preventing it.
Id. at *13. Anytime a supposed “alternative” design carries with it the same risks as the challenged design, the plaintiff must prove causation – that the “particular alternative design would have prevented the harm” to this particular plaintiff. Id. In Nelson plaintiff’s expert’s opinions failed “to show that these design elements would have, to a reasonable probability, prevented the harm without impairing the utility, usefulness, practicality, or desirability of the” product. Id. (emphasis original).
In sum, Nelson contains one of the most extensive and well-reasoned refutations of the concept of comparative product risk claims that we’ve ever encountered. It’s precedent that we expect defendants to use whenever this type of claim crops up against any number of different products for years to come. As such, its stare decisis effect will impose the type risk on plaintiffs and their counsel – both now and in the future − that can result in favorable resolution of this particular mass tort litigation.