Photo of Bexis

We’re calling this a “quasi” guest post because the author, Reed Smith‘s Dean Balaes, is actually trying out to join our blogging team.  This is his inaugural post, of what we hope will be many more.  This particular post addresses the causation aspects of a case, Gayle v. Pfizer, Inc., ___ F. Supp.3d ___, 2020 WL 1685313 (S.D.N.Y. April 7, 2020), that we mentioned in yesterday’s preemption post.  As we hope you’ll see shortly, the causation issues in Gayle are also interesting, and favorable to the right side of the “v.”  Without further ado, take it away Dean.


As is often the case in pharmaceutical product liability actions, another group of plaintiffs’ claims suffered a fatal deficiency:  establishing a causal association between the drug at issue and the alleged injury.  Gayle v. Pfizer, Inc., ___ F. Supp.3d ___, No. 19-CV-3451, 2020 WL 1685313 (S.D.N.Y. April 7, 2020), illustrates this debilitating deficiency when the other side alleged that Lipitor causes Type 2 diabetes by relying on anecdotal adverse event reports.  This blog will provide specific attention and commentary regarding causation fallacies as observed in Gayle, but note that one should also review Gayle for its favorable preemption ruling.

The other side tried to Febreze the air around this malodorous causation reality by relying on 6000 adverse event reports related to an earlier “changes being effected” (“CBE”) labeling change between the defendant manufacturer and the FDA.  The idea was to suggest that the mere existence of adverse event reports related to a drug (without evidence showing a causal association between the drug and the alleged injury) means that the manufacturer knew that the drug categorically caused the disease.  Thus, the Gayle plaintiffs nitpicked the drug’s label change – claiming it only warned about two conditions related to Type 2 diabetes, but should have warned that it causes Type 2 diabetes.

Opposing counsel’s legal gymnastics to hurtle over the fallacy in causation did not work.  According to the FDA, adverse reactions listed in the label are limited to “only those adverse events for which there is some basis to believe there is a causal relationship between the drug and the occurrence of the adverse event.”  21 C.F.R. § 201.57(c)(7).  After initial FDA-approval, a manufacturer can change labels through the CBE regulation to promote the safe use of the drug product only on the basis of “newly acquired information.”  21 C.F.R. § 314.70(c)(6)(iii).  As the blog’s readers know from scores of preemption posts, the FDA narrowly defines “newly acquired information” as:

[D]ata, analyses, or other information not previously submitted to the [FDA], which may include (but is not limited to) data derived from new clinical studies, reports of adverse events, or new analyses of previously submitted data (e.g., meta analyses) if the studies, events, or analyses reveal risks of a different type or greater severity or frequency than previously included in submissions to the FDA. 

21 C.F.R. § 314.3(b) (cited at 2020 WL 1685313, at *3) (emphasis added).  FDA imposes this limit for a reason.  “Over-disclosure dilutes warnings of more significant adverse reactions both by likelihood and severity of the reaction and can unjustifiably deter patients from a helpful drug or therapy.”  Id. at *3.

On a motion to dismiss, there is no weighing of evidence yet to prove causation, so Gayle highlights a well- (or at least over-) used plaintiffs’ strategy just to plead some purported evidence – here adverse event reports – and allege that such reports mean that the defendant knew that Lipitor causes Type 2 diabetes.  It is like Senator Hendrickson in the 1954 comic book hearings alleging that reading comic books causes juvenile delinquency.  With no actual evidence proving causality, plaintiffs rely on anecdotal, one-sided, and unverified hearsay (see the blog’s cheat sheet here) such as adverse event reports.  Throw mud against the wall and hope something sticks − either plaintiffs do not have a firm grasp of logical causation, or (more likely) they are hoping judges do not, when all they plead is that there are adverse event reports making the same allegations.

Thus, these plaintiffs erroneously alleged that the 6,000 adverse event reports constituted “newly acquired information.”  Gayle, 2020 WL 1685313, at * 5.  TwIqbal prevailed again, as plaintiffs’ claims were dismissed with prejudice and either barred by the state’s statute of limitations or preempted by federal law (see the blog’s preemption cheat sheet here).  Whether it was one, or 6000, the pleading of adverse event reports were simply vague causation assertions that did not amount to a plausible claim.  See Id. at *5.  As the blog has already discussed, an entire MDL raising the same purported injury against the same drug was dismissed a few years ago for – guess what – failure to establish causation.  Indeed, one would not be at all surprised if the purported 6000 adverse event reports include thousands of plaintiffs from that previous MDL whose claims have already been conclusively adjudicated non-causal.

The fundamentally flawed presumption in the plaintiffs’ causation reasoning is simple:  a phenomenon cannot be proven to occur on the basis of anecdotal adverse event reports.  It is like the child who believes that the song of birds makes the morning happen or the Monte Carlo Fallacy, which makes gamblers believe that a winning streak increases the chances of future winnings.  Alas for plaintiffs, coincidence is not causation.  These plaintiffs’ reliance on adverse event reports did not constitute “newly acquired information” because the reports merely described instances “where patients taking [the drug] were diagnosed with type 2 diabetes, but do not reach any conclusions regarding a causal association.”  Id. at * 6.  Thus, “under a plain reading of the regulations, adverse event reports, without any analysis indicating causality, cannot constitute ‘newly acquired information.’”  Id.

Gayle is worth the read because it is yet one more case (and this is yet one more post) reminding plaintiffs that the mere presence of anecdotal adverse event reports does not necessitate a causal connection based on coincidental relationships. Thankfully, Gayle also shows that judges do understand the fallacies of logical causation.