We spent the weekend in the Boston area, visiting the Drug and Device Law Rock Climber for “parents’ weekend.” Saturday’s schedule was replete with activities:  a keynote speech by the university president, lectures on topics academic and not, an informative session on the school’s “semester abroad” program (the DDLRC has informed us that she hopes to spend next spring in Australia), and dozens of presentations by student arts groups (theatre, improv, orchestra, a cappella).    

We didn’t get to do any of that.  By mid-morning, the DDLRC had tired of being trailed around campus by her parents.  All were shepherded to the “T,” where, several stops down the line, our procession (and its credit cards) descended on the downtown shopping district.  But before we left campus, we insisted on attending one lecture, a sparkling presentation by the head of the theatre department.  The subject was a famous circus mogul who was one of the university’s founding trustees and whose best-remembered donation was the carcass of his equally-famous circus elephant, who has since lent his name to all of the university’s sports teams.  According to legend, the elephant died while pushing a baby elephant out of the way of a speeding train.  The truth was a less-heroic, but equally fatal, elephant vs. train collision. In any event, he was duly stuffed and sent to live in a university museum.  Years later, he was incinerated in a building fire.  Some of his ashes were scooped into a container and displayed in the rebuilt museum, but a grander tribute awaited. Just this year, a massive bronze statue of the elephant was erected outside the building bearing his patron’s name.  We love this campus attraction, and we wax irrationally sentimental about the whole story.  

We are trying to think of a smooth transition to today’s discussion, but this is the best we can do:  like legendary elephants, Daubertand Frye standards for admissibility of expert testimony, as applied by modern courts, frequently defy segregation into discrete containers.  Notwithstanding a hybrid analysis squeezed into aFrye box, the conclusion was clear: plaintiff’s causation experts in a Philadelphia Zoloft case did not meet any relevant admissibility standards, and both were properly excluded. 

In a pair of decisions in Porter v. SmithKline Beecham Corp., et al., 2015 WL 5970639 (Pa. Com. Pl. Oct. 5, 2015), the Court of Common Pleas considered defendant Pfizer’s motions to  exclude the testimony of plaintiffs’ causation experts.  Pennsylvania is one of the minority of states that have not adopted Daubert and that ostensibly pay homage to  the District of Columbia Circuit’s 1923Frye decision (Frye v. United States, 293 F. 1013, D.D. Cir. 1923).  The Porter court began by explaining that, “although both the Frye and Daubert standards relate to methodology and not conclusions, the differences are dramatic.”  Porter, 2015 WL 5970630 at *2.  Frye is “an evidentiary standard for determining whether the methodology employed by a proposed witness is considered scientific by others in a relevant scientific field.”  Id.at *2.  In other words, a Frye inquiry is limited to whether an expert’s methodology has “gained general acceptance in the particular field in which it belongs.”  Id.  In contrast, “[t]heDaubert standard requires the court to make an independent judicial scientific judgment whether the methodology is scientifically sound even if a scientific consensus of propriety exists.”  As quickly as it explained these distinctions, the Portercourt rendered them meaningless, conducting its own analysis of the reliability of the experts’ methodologies a la Daubert, albeit punctuated with occasional repetition of the phrase “generally accepted.” 

Porter involved an infant born with a serious birth defect known as omphalocele – the abdominal organs formed outside the body because the abdominal wall did not close properly – allegedly caused by the mother’s ingestion of Zoloft while she was pregnant.  The court made short work of the first expert, a Ph.D. of unnamed specialty proffered “to testify as to general and specific causation of birth defects” of the infant plaintiff.  Id. at *3.  The court explained that the expert devoted twelve pages of his report to animal studies linking SSRIs to birth defects, and that the newest cited study was 29 years old.  Moreover, the court commented that “animal studies are of limited utility where a significant body of studies in human populations exists in the literature,” a limitation that the expert failed to acknowledge. Id.   In addition, the court concluded that the mechanism the expert identified – “alteration of serotonin signaling [that] can impact embryonic development resulting in . . . congenital malformations” – amounted to “nothing but speculation without support since [the expert had] no information as to the baseline level of serotonin causing signaling or the changed level of serotonin” caused by SSRIs.  Id.   As such, the expert was precluded from testifying.

Plaintiffs’ second expert, a “forensic epidemiologist” with an M.D. from Sweden and a Ph.D. in Public Health (and a former chiropractor) testified that “forensic epidemiology” was “the use of epidemiology in a legal setting for evaluating causation.”  Id. at *6.  The expert first ran afoul of the court by testifying that “the legal standard of ‘more probable than not’ [was] his definition of ‘reasonable degree of scientific certainty’ and that this standard varie[d] by jurisdiction.”  Id. at *6.  The court begged to differ: “Although [the expert] believes the appropriate standard for a scientific conclusion in the field of ‘forensic’ epidemiology depends on ‘jurisdiction,’ only a legal standard depends on jurisdiction and proper scientific methodology and conclusions do not vary whether testified to on the Pennsylvania side or the New Jersey side of the Benjamin Franklin Bridge.”  Id.  LOLs. 

The court went on to explain that proper epidemiologic methodology “requires that one not fall victim to the ‘Class Effect Fallacy’ or the ‘Lumping Fallacy’ – [that] the causation conclusion must be drug specific . . . [and that] [d]ifferent birth defects should not be grouped together unless they are part of the same body system, share a common pathogenesis, or there is a specific valid justification or necessity for an association . . . .”  Id, at *7. Nevertheless, the expert “improperly conflates three types of data:  Zoloft and omphalocele, SSRIs generally and omphalocele, and SSRIs and gastrointestinal and abdominal malformations.”  Id.  He also relied almost exclusively on one study, notwithstanding its small sample and broad confidence interval and its authors’ inability to rule out the possibility that the apparent association between Zoloft and omphalocele was the result of chance.  Id.

The expert “defer[red] to other experts” on the issue of the “temporal relationship between the exposure and disease” because he wasn’t “a teratologist or embryologist.”  Id. at *11. Specifically, despite the fact that omphalocele results from “incomplete folding of the abdominal wall during the third to fifth weeks of pregnancy” and the fact that the mother did not begin taking Zoloft until the seventh week of her pregnancy, the expert was willing to testify that Zoloft caused the infant plaintiff’s omphalocele.  Id.  Finally, the court noted that the expert, who was not a clinician and did not profess to perform a “clinical differential diagnosis,” had failed to rule out any other potential causes of omphalocele, including genetic causes accounting for a large percentage of cases.  In sum, he “made no analysis at all, only unwarranted assumptions.”  The court concluded, “Defendants claim that plaintiff’s expert . . . utilizes an improper methodology which requires preclusion.  For the foregoing reasons this Court concludes that the methodology is improper . . . .”

Recall that the court prefaced its opinion by explaining that it was not its role, under Frye, to determine whether an expert’s methodology was proper – only whether it was “generally accepted.”   No matter – experts spouting opinions supported only by their own ipse dixit should be excluded, whatever you want to call the analysis. Any other result would be a train wreck of elephantine proportions.