In re Acetaminophen MDL Reiterates That Cherry-Picking Data And Exceeding Study Limitations Are Unreliable Expert Methodologies

The Southern District of New York’s recent opinion in the In re Acetaminophen MDL ^[1] establishes strong guardrails regarding how expert witnesses can rely on studies and data in support of their opinions. By way of background, In re Acetaminophen involves more than 600 plaintiffs who allege that the manufacturer and retailers of acetaminophen products failed to warn that children may develop autism spectrum disorder (ASD) or attention-deficit hyperactivity disorder (ADHD) from in utero exposure to the drug.^[2] In an opinion issued on December 18, 2023, Judge Denise Cote of the Southern District of New York excluded all five of the plaintiffs’ general causation experts as unreliable.^[3] 

The In re Acetaminophen decision reaffirms two reasons that general causation expert opinions may be excluded in product liability and mass tort proceedings: (1) if the expert cherry-picks scientific literature or isolated findings within a study; or (2) if the expert, in their reliance on a study, exceeds the limitations or conclusions placed by the study’s own authors. For other key takeaways from the In re Acetaminophen decision, please see our prior analyses here and here.

Experts cannot cherry-pick scientific literature or isolated findings within a study

The In re Acetaminophen court criticized the plaintiffs’ experts for cherry-picking data and failing to address the entire body of relevant scientific literature. As many other courts have observed, the court explained that “an expert must not cherry-pick from the scientific landscape and present the [c]ourt with what he believes the final picture looks like,” and that “exclusion of the proffered testimony is warranted where the expert fails to address evidence that is highly relevant to his or her conclusion.”^[4] Here, the court repeatedly noted examples of plaintiffs’ experts’ “cherrypick[ing] those findings that support [their] conclusions” and disregarding those that did not.^[5] Such examples include failing to even mention relevant studies, disregarding studies employing the same methodology as plaintiffs’ preferred studies but reaching unsupportive conclusions, and ignoring subsequent—and relevant—publications by the same authors of a cited study. ^[6] The court thus excluded plaintiffs’ experts for cherry-picking data—or explicitly ignoring the findings of a number of studies that were contrary to their opinions and choosing to only emphasize those that aligned with their theses.^[7]

The court also faulted the plaintiffs’ experts for relying on isolated positive findings within studies and failing to mention or address the contrary findings. The court noted that for many of the studies that plaintiffs’ experts relied on, “the devil is in the details” and “in reality those studies reported conflicting or unreplicated individual outcomes with varying relevance to either ADHD or ASD.”^[8] For example, the court criticized one of plaintiffs’ experts for identifying one study as supporting their causation opinion based on a single positive finding reported, while giving “no mention of the fifteen findings [in the same study] of no effect.”^[9] Another study relied on by plaintiffs’ experts “measured dozens of outcomes,” which were “much more mixed” than plaintiffs’ experts had represented.^[10] And yet another study found “a variety of potentially inconsistent results,” but nevertheless was referred to by one of plaintiffs’ experts as “clear evidence” in support of a causal association.^[11] Ultimately, the court determined that the “results-driven analysis” conducted by plaintiffs’ experts rendered their opinions unreliable, and thus inadmissible.^[12]

Experts cannot exceed the limitations or conclusions of a study’s own authors

The In re Acetaminophen court also heavily criticized the plaintiffs’ experts for ignoring—or even contradicting—the authors’ conclusions of the very studies they cited. The court faulted one of plaintiffs’ experts for his continued “willing[ness] to press conclusions that study authors are not willing to make,” explaining that this “willingness creates an ‘analytical gap’ between the conclusions reached by the authors and the conclusions he draws from their work.”^[13] As an example, the court pointed out that this same expert “repeatedly ignore[d] [study] authors’ cautions that familial or genetic confounding may explain, at least in part, the observed association” between acetaminophen and ADHD or ASD.^[14]

The court similarly faulted other experts for “dismiss[ing] the express limitations of study authors”^[15] and ignoring the authors’ conclusions that “additional investigation on this subject are needed.”^[16] And the court rejected yet another expert’s opinion regarding dose-response studies for ignoring the express “limitations that authors place on their own studies.”^[17] While the plaintiffs contended that “the limitations expressed by authors in their studies should be ignored as simply an overly conservative requirement that scientists impose on each other to get peer reviewed studies published,” the court rejected this position, noting that “[t]hese and more arguments like them” did not relieve the court of its obligation to scrutinize an expert’s methodology and ensure they are “sufficiently rigorous to pass muster by the standard established by [the expert’s] discipline, Rule 702, and Daubert.”^[18]

Strategic Takeaways

The recent In re Acetaminophen MDL decision underscores the importance of closely reviewing the scientific publications on which an expert relies and determining whether an expert is fairly characterizing both the findings of an individual study and the overall body of literature. If an expert cherry-picks only supportive findings or isolated findings within a given study, exclusion of that expert’s testimony may be proper.

---

^{[1] In re Acetaminophen – ASD-ADHD Prods. Liab. Litig., No. 22MD3043 (DLC), 2023 WL 8711617 (S.D.N.Y. Dec. 18, 2023).} 

^{[2] Id.}

^{[3] Id. at *1.}

^{[4] Id. at *18.} 

^{[5] Id. at *39; see also id. at *36 (noting one of plaintiff’s experts “ignore[d] inconsistent results,” rendering his methodology unreliable).}

^{[6] See id. at *26, *40, *44.}

^{[7] See, e.g., id. at *40.}

^{[8] Id. at *37.}

^{[9] Id. at *38.} 

^{[10] Id. at *39. Relatedly, the court observed that “[c]herry-picking of isolated findings is of particular concern” where studies “measure[] many markers” or “outcomes at once”—such as these—without correcting for multiple comparisons. Id. at *40.}

^{[11] Id.}

^{[12] See id. at *42.}

^{[13] Id. at *33.} 

^{[14] Id.}

^{[15] Id. at *36.}

^{[16] Id. at *42.}

^{[17] Id. at *49.} 

^{[18] Id. at *35.}