TORTINI

For your delectation and delight, desultory dicta on the law of delicts.

Talc Litigation in Missouri – Show Me the Law and the Evidence

February 22nd, 2017

In New Jersey, where the courts are particularly plaintiff friendly but not beyond the persuasive force of evidence, lawsuit industry claims that talc causes ovarian cancer have not fared well. Last year, Judge Johnson, of Atlantic County, New Jersey, held that the plaintiffs’ causal claims failed to meet even the minimal New Jersey legal threshold of scientific validity.1 Meanwhile, in Missouri, juries have been returning large verdicts for plaintiffs on their claims that their use of talc products caused their ovarian cancers.2

What gives? Why is the outcome of similar litigation so different in New Jersey from that in Missouri? One might mistakenly think that courts in Missouri would be skeptical of scientifically dubious claims. After all, Missouri is the “Show Me” state; right? Many people understand the state’s nickname to mean that Missourians are not gullible.3

The reality of the origins of the Missouri nickname may well be different. The most cited account reports that a congressman from Missouri, Willard Duncan Vandiver, used the phrase in an 1899 speech:

I come from a state that raises corn and cotton and cockleburs and Democrats, and frothy eloquence neither convinces nor satisfies me. I am from Missouri. You have got to show me.”

Basically, according to Vandiver, Missourians are “show me” simple folks because they do not read or understand eloquent language. Vandiver might have thought that scientific language was beyond his neighbors’ ken as well. Of course, things have changed since 1899. Missouri is no longer a state populated by Democrats. In the 2016 general election, Donald Drumpf received 56.8% of the Missouri votes cast. Hilary Clinton received 38.1%.4  Inquiring minds will want to know whether “Show Me” connotes incredulity or illiteracy.

One relevant difference between Missouri and many other states, and all the federal courts, is that some courts in Missouri engage in a particularly edentulous form of judicial gatekeeping of expert witness opinion testimony. The talc claims that resulted in large verdicts in Missouri never got off the dime (or got a dime) in New Jersey because plaintiffs’ expert witnesses’ opinions were excluded from courtrooms in the Garden State.

The resulting trials in Missouri have showcased some curious, doubtful rhetoric from legal counsel for the lawsuit industry. In his closing argument in Giannecchini v. Johnson & Johnson, the plaintiff’s lawyer accused Johnson & Johnson of having “rigged” regulatory agencies to ignore the dangers of talc.5 The argument was apparently effective and it has been repeated in another Missouri trial, in Swann v. Johnson & Johnson6, now underway. The plaintiffs’ opening “statement” in Swann was marked by overwrought, hyperbolic rhetoric.7

And the first trial days in Swann were dedicated by plaintiff’s counsel to showing, not that talc actually causes ovarian cancer, but to showing that the defendants engaged in lobbying with respect to the carcinogenic classification of talc by regulatory agencies.8 According to the coverage in legal news media, the first testimony offered was offered to show that after the National Toxicology Program (NTP) nominated talc for inclusion in its list of potential carcinogens, industry trade groups, such as the Cosmetic, Toiletry and Fragrance Association, “shut down serious regulator concerns through intensive lobbying efforts.”9

This is a remarkable digression from the truth finding function of an American jury trial for several reasons. First, the “shutting down” of regulator concern was not, in the media reports, associated with any fraudulent misrepresentations of the scientific record. By casting the lobbying in an unflattering light, the plaintiff was able to undermine the truth value of agencies’ refusal to characterize talc as an ovarian carcinogen. The media coverage did not suggest that the lobbying involved the presentation of sham evidence or arguments that might have misled agencies about the correctness of their position.

Second, if the industry lobbying had badly misled the National Toxicology Program, or other government body, then there would no doubt be a conclusive case for causation today. The fact of the matter, however, is that there is no conclusive case for the claim that talc causes ovarian cancer. Late last year, the “Sister Study,” which explored whether there was any association between perineal talc use and ovarian cancer, was published in Epidemiology.10 The Sister Study (2003–2009) followed a cohort of 50,884 women whose sisters had been diagnosed with breast cancer. Talc use was ascertained at baseline, before diagnosis of subsequent disease and before any chance for selective recall. The cohort was followed for a median of 6.6 years, in which time there were 154 cases of ovarian cancer, available for analysis using Cox’s proportional hazards model. Perineal talc use at baseline was not associated with later ovarian cancer. The authors reported a hazard ratio of 0.73, less than expected, with a 95% confidence interval of 0.44, 1.2. Such a powerful study, showing the absence of any large or even modest association, would hardly be feasible if the science were so clear in the year 2000 that no reasonable scientist would have advocated against the NTP’s proposed classification.

Third, the lawsuit industry’s focus on lobbying activities in the Giannecchini and the Swann cases raises serious issues of infringing upon the defendants’ first amendment rights. The defendants’ advocacy for non-sham, non-fraudulent scientific positions is protected by the federal constitution, under what has come to be known as the Noerr-Pennington doctrine.

The Noerr-Pennington Doctrine of Immunity

One of the first agenda items for the first United States Congress was the drafting of a “Bill of Rights” to be submitted to the individual States for ratification. The First amendment (originally the third until the first two were dropped) sets forth a basic “right of the people to peaceably assemble, and to petition the government for a redress of grievances.”11 In the context of lobbying legislatures and regulatory agencies, the Supreme Court has long regarded lobbying and advocacy for and against legislation and regulation as core political speech that is protected by the right to petition the government.12

Part of this constitutional guarantee is a freedom to associate with others to lobby for redress.13 The constitutional protection is not lost by an economic or self-interested motivation in the lobbying or advocacy.14  This constitutional protection of advocacy positions results in an immunity from civil liability for speech, association, and conduct undertaken to advance advocacy positions before legislatures, agencies, and courts.15 This immunity, over half a century old, has come to be known as the Noerr-Pennington doctrine.

Although the original Noerr-Pennington doctrine cases specifically addressed claims of antitrust liability, later cases have held that the immunity applies with equal force in tort cases. State courts, regardless of their state constitutions, are of course obliged to grant and protect the federal Noerr-Pennington immunity.16

The unconstitutional infringement of defendants’ first amendment rights is hardly an innovation in Giannecchini and Swann cases. For decades, the lawsuit industry, which jealously guards its own first amendment rights, has overzealously pressed conspiracy and tort claims against manufacturing industry for trying to influence legislation and regulation. In Senart v. Mobay Chem. Corp., 597 F. Supp. 502 (D. Minn. 1984), plaintiffs alleged that they were harmed by exposure to toluene diisocyanate (TDI), a feedstock chemical used in making polyurethane foam. The plaintiffs sued TDI manufacturers, on conspiracy claims that the manufacturers had jointly influenced the Occupational and Safety Health Administration (OSHA) to reject a recommendation from the National Institute for Occupational Safety and Health (NIOSH) for lower permissible exposure standards for TDI. Senart, 597 F. Supp. at 504. The plaintiffs’ conspiracy complaint was based upon allegations that the manufacturing defendants knew of a body of scientific evidence which suggested that workers could suffer harm at exposure levels below the prevailing … standard,” and and that they “conspired to ‘obfuscate and confuse’ scientific findings which supported a more stringent standard.” Id. Plaintiffs also alleged that the TDI manufacturers knew that a more stringent TDI exposure standard would harm their businesses. Id.

The trial court dismissed the conspiracy count in Senart. “[E]ven accepting plaintiffs’ allegations as true, defendants concerted action sought only permissible ends and acted through permissible means.” Id. at 505-6 (footnote omitted). The defendants work in concert through their trade association to persuade OSHA to reject the NIOSH proposal was clearly protected by the first amendment. Id. at 506 (internal citations omitted).

Following Senart, federal courts in later products cases have applied he Noerr-Pennington doctrine to bar tort claims. In a 1996 class action, a district court held that the immunity barred a class action filed by relatives of gunshot victims against gun manufacturers. Hamilton v. ACCU-TEK 935 F. Supp. 1307 (E.D.N.Y. 1996). The court, in Hamilton, found the plaintiffs’ negligence and product liability claims untenable:

Defendants’ efforts to affect federal firearm policies through lobbying activities are prime examples of the types of activity the First Amendment, through its rights of free speech and petition, sought to protect… . A core principle of the Noerr-Pennington doctrine is that lobbying alone cannot form the basis of liability… .”

Id. at 1321. The court in Hamilton dismissed the product liability claims. See also Tuosto v. Philip Morris USA Inc., No. 05 Civ. 9384 (PKL), 2007 WL 2398507, at *5 (S.D.N.Y. Aug. 21, 2007) (noting that the immunity “applied to bar liability in state common law tort claims, including negligence and products liability claims, for statements made in the course of petitioning the government”).

The lawsuit industry is one of the largest rent-seeking groups in the United States. Our courts need to apply constitutional standards in a symmetrical fashion, with an understanding that what is spoken in the halls of legislatures and agencies is protected at least as much as speech in the courtroom, and that the constitutional rights of manufacturing industry should not be subordinated to the rights of the lawsuit industry. Maybe lawyers need to figure out how to “show” the constitution in pictograms, without all the 18th century eloquence.


1 Carl v. Johnson & Johnson, No. ATL-L-6546-14, 2016 WL 4580145 (N.J. Super. Ct. Law Div., Atl. Cty., Sept. 2, 2016).See New Jersey Kemps Ovarian Cancer – Talc Cases” (Sept. 16, 2016).

2Talc Litigation – Stop the Madness” (Nov. 10, 2016) (describing large verdict for plaintiff in Giannecchini v. Johnson & Johnson); see also Myron Levin, “Johnson & Johnson Hammered Again in Talc-Ovarian Cancer Verdict of $70 Million,” Law360 (Oct. 27, 2016); Brandon Lowrey, “J & J, Talc Co. Hit With $70M Baby Powder Cancer Verdict,” Law360 (Oct. 2016).

3 SeeThe Show-Me State,” last visited Feb. 21, 2017.

4 SeeState of Missouri – 2016 General Election – November 8, 2016,” last visited Feb. 21, 2017. I leave it to the reader to assess whether the state nickname describes incredulity or illiteracy.

5 Myron Levin, “Johnson & Johnson Hammered Again in Talc-Ovarian Cancer Verdict of $70 Million,” Law360 (Oct. 27, 2016); Brandon Lowrey, “J & J, Talc Co. Hit With $70M Baby Powder Cancer Verdict,” Law360 (Oct. 2016).

6 Swann v. Johnson & Johnson, case number 1422-CC09326-01, in the 22nd Judicial Circuit of Missouri.

7 Cara Salvatore, “J&J Hid Talc Risk For ‘Love Of Money’, Jury Hears,” Law360 (Feb. 9, 2017).

8 Cara Salvatore, “Talc Lobbyists Stymied Carcinogen Classification, Jury Hears,” Law360 (Feb. 10, 2017).

9 Id.

10 Nicole L. Gonzalez, Katie M. O’Brien, Aimee A. D’Aloisio, Dale P. Sandler, and Clarice R. Weinberg, “Douching, Talc Use, and Risk of Ovarian Cancer,” 27 Epidemiology 797 (2016).

11 U.S. Const. amend. I.

12 California Motor Transp. Co. v. Trucking Unlimited, 404 U.S. 508, 510 (1972) (disallowing a cause of action “predicated upon mere attempts to influence the Legislative branch for the passage of laws or the Executive branch for their enforcement.”); United Mine Workers of Am. v. Ill. State Bar Ass’n, 389 U.S. 217, 222 (1967) (characterizing the right to petition as “among the most precious of the liberties safeguarded by the Bill of Rights”). United Mine Workers of Am. v. Pennington, 381 U.S. 657, 669-70 (1965); Doe v. McMillan, 566 F.2d 713, 718 (D.C.Cir. 1977), cert. denied, 435 U.S. 969 (1978) (holding that the first amendment constitutional right to petition the legislature “extends to administrative agencies and the courts”).

13 N.A.A.C.P. v. Button, 371 U.S. 415, 430 (1963) (protecting the right “to engage in association for the advancement of beliefs and ideas”); N.A.A.C.P. v. Alabama ex rel. Patterson, 357 U.S. 449, 460 (1958) (“[e]ffective advocacy of both public and private points of view, particularly controversial ones, is undeniably enhanced by group association … .”). The right of association to further lobbying activities has been described as having a “preferred place” along with other first amendment freedoms, such that the Court will not tolerate “dubious intrusions.” Thomas v. Collins, 323 U.S. 516, 530 (1945).

14 Virginia State Bd. of Pharmacy v. Virginia Citizens Consumer Council, 425 U.S. 748, 762 (1976); Sawyer v. Sandstrom, 615 F.2d 311, 316 (5th Cir. 1980) (“The right to freely associate is not limited to those associations which are ‘political in the customary sense’, but includes those which ‘pertain to the social, legal, and economic benefit of the members’.”) (citing Griswold v. Connecticut, 381 U.S. 479, 483 (1965)); International Union v. National Right to Work Legal Defense & Education Foundation, Inc., 590 F.2d 1139, 1148 (D.C. Cir. 1978) (“Even economically motivated expression or association is not disqualified from protection under the first amendment.”); Greminger v. Seaborne, 584 F.2d 275, 278 (8th Cir. 1978) (observing that the constitutionally protected [f]reedom of association includes membership in unions or other organizations concerned with ‘business and economic causes’.”); Senart v. Mobay Chem. Corp., 597 F. Supp. 502, 506 (D.Minn. 1984) (“Selfish motivations do not lessen one’s right to present views to the government.”).

15 Eastern Railroad Presidents Conference v. Noerr Motor Freight, Inc., 365 U.S. 127 (1961); United Mine Workers v. Pennington, 381 U.S. 657 (1965).

16 Fraser v. Bovino, 317 N.J. Super 23, 37 (App. Div. 1998) (recognizing “the fundamental values that undergird a citizen’s right to communicate on issues of public import”); Village Supermarket, Inc. v. Mayfair, 269 N.J. Super. 224, 229-32 (Law Div. 1995) (refusing to interpret New Jersey tort law to permit claims based on lobbying activity protected by the First Amendment); ARTS4ALL Ltd. v. Hancock, 810 N.Y.S.2d 15, 16 (App. Div. 2006) (denying employee’s motion for summary judgment on claim for breach of no-disparagement clause in severance agreement, holding that employer’s statements to government officials were protected by Noerr-Pennington doctrine); Concourse Nursing Home v. Engelstein, 692 N.Y.S. 2d 888, 891 (Sup. Ct. 1999) (holding law firm was immune from business tort claims for successful lobbying efforts); I.G. Second Generation Partners v. Reade, 793 N.Y.S.2d 379, 381 (App. Div. 2005) (holding that NoerrPennington immunity barred claim for tortious interference); Diaz v. Southwest Wheel, 736 S.W.2d 770, 771 (Tx. Ct. App. 1987) (holding that Noerr-Pennington immunity barred conspiracy claims against tire manufacturer, which as a member of a trade association, opposed the recall on defective tire rims and restrictions on multi-piece wheels).

Quackers & Cheese – Trump Picks Kennedy to Study Vaccine Safety

January 11th, 2017

Science necessarily involves a willingness to follow evidence to whatever conclusions are warranted, if conclusions properly can be had. When it comes to vaccination conspiracies, Democrats have it in their political DNA to distrust pharmaceutical companies that research, develop, and manufacture vaccines. The current Republican party, which has been commandeered by theocrats and populists, see vaccination as federal government aggrandizement, and resist vaccination policy as contrary to God’s will. Science is often the loser in the cross-fire.

And so we now have the public spectacle of watching the left and the right join in similar scientific apostasies. Consider how both McCain and Obama both suggested that vaccines and autism were related in the 2008 election. (Although both candidates were to some extent slippery in their suggestions, which might have been appropriate given how little they knew about the controversies.) And consider Michelle Bachmann was converted to a similar view about the HPV vaccine on the basis of a woman’s anecdote about her child. And then on the far left, you have the uplifting story of Robert F. Kennedy Jr, and his brief on how thimerosal supposedly causes autism.

So it should be no surprise that Donald Trump, a Birther, a Mirther, a mid-night Twitterer, should embrace the anti-vaccination movement. Trump has made it clear that he rejects evidence-based policy, and so no one should expect him to embrace a scientific policy that is driven by high-quality scientific evidence. According to Kennedy, Trump wants Kennedy to head up a “commission on vaccine safety and scientific integrity.” Michael D. Shear, Maggie Haberman & Pam Belluckjan, “Anti-Vaccine Activist Says Trump Wants Him to Lead Panel on Immunization Safety,” N.Y. Times (Jan. 10, 2017); Domenico Montanaro, “Despite The Facts, Trump Once Again Embraces Vaccine Skeptics,” National Public Radio (Jan. 10, 2017).

Who needs the National Academy of Medicine when you can put a yutzball lawyer in charge of a “commission”?

Some of the media refer to Robert F. Kennedy Jr. as a vaccine skeptic, but their terminology is grossly inaccurate and misleading. Kennedy is a vaccine denier; he has engaged in a vitriolic campaign against the safety and efficacy of vaccines. He has aligned himself with the most extreme deniers of science, medicine, and public safety, including the likes of Andrew Wakefield and Jenny McCarthy. Kennedy has not merely engaged hyperbolic rhetoric against vaccines, he has used his radio show on the lawsuit industry’s Ring of Fire, to advance his campaign against public health as well as to shill for the lawsuit industry on other issues. SeeRFK, Jr.: Science Shows That Autism — Mercury Link Exists – PT. ½,” Ring of Fire (Mar 8, 2011).

Kennedy should not be characterized as a skeptic, when he is a shrill ideologue, for whom science has no method that he is bound to respect. Back in July 2005, Kennedy published an article, “Deadly Immunity,” in both Rolling Stone and on Slate’s website. The article was a hateful screed against Big Pharma and government health agencies for an alleged conspiracy to hide the autism risks of thimerosal preservatives in vaccines. Several years later, on January 16, 2011, Salon retracted the article. Seehttps://en.wikipedia.org/wiki/Deadly_Immunity” entry in Wikipedia. See also Phil Plait, “Robert F. Kennedy Jr.: Anti-Vaxxer,” Slate (June 5 2013) (describing Kennedy as a full-blown anti-vaccination conspiracy theorist); Rahul K. Parikh, M.D., “Inside the vaccine-and-autism scare: A pediatrician traces the rise of the anti-vaccine movement that falsely linked thimerosal with autism and turned parents away from the most lifesaving medicine in history,” Salon (Sept. 22, 2008); Keith Kloor,Is Robert F. Kennedy Jr. Anti-Science?” Discover Magazine (June 1, 2013); Steven Novella, “RFK Jr.s Autism Conspiracy Theory,” (Jun 20 2007).

Back in 2008, President Obama apparently considered Robert Kennedy for a cabinet-level position, but on sober reflection, thought better of it. See Steven Novella, “Politics and Science – The RFK Jr. Test,” (Nov. 07 2008). The Wall Street Journal, joined by many others, are now urging Trump to think harder and better about the issue, perhaps with some evidence as well. See Alex Berezow & Hank Campbell, “Ignore Anti-Vaccine Hysteria, Mr. Trump: Robert F. Kennedy Jr.’s conspiracy theories have no place in the White House,” Wall Street J. (Jan. 10, 2017).

Kiker v. Smithkline Beecham & the Pathology of Judicial Gatekeeping

January 4th, 2017

There is no expedient to which a man will not go to avoid the labor of thinking.”                                                                                    Sir Joshua Reynolds

Medical students study pathology not only to understand the nature, course, and causation of disease, but also to understand better normal tissue and cellular function and structure. Similarly, lawyers can improve their understanding of judicial decision making, not only from studying well-reasoned judicial opinions, but from also studying pathological opinions, with clear, demonstrable errors that help illustrate both the pathogenesis of intellectual and judicial error, as well as the normal, proper function of judging.

At the end of each year, bloggers and pundits traditionally call attention to the best and the worst decisions, usually from a partisan perspective. One federal judicial decision on Rule 702, however, stands out for special treatment as a veritable Berenstain Bears’ manual on how not to adjudicate so-called Daubert motions. Kiker v. Smithkline Beecham Corp., 2:14-cv-02164-EAS-TPK, (S.D. Ohio, Dec. 15, 2016) (Sairgus, C.J.) [cited below as Kiker slip op.] The Kiker opinion is as worthy of dissection as a judicial opinion for lawyers, as is the dissection of a cadaver by medical students in their first-year course on clinical anatomy.

The Kiker plaintiffs claimed that maternal use of paroxetine (tradename Paxil) caused her child to develop a ventricular septal defect. The defendant, GlaxoSmithKline LLC (GSK), invoking Federal Rule of Evidence 702, moved to exclude opinion testimony of several of plaintiffs’ expert witnesses, including Laura M. Plunkett, Ph.D., Ra-id Abdulla, M.D. Kiker slip op. at 1. The gravaman of the plaintiffs’ case is that GSK did adequately warn physicians of the risk to offspring of women who took paroxetine in pregnancy until September 2005. At that time, GSK revised its labeling for Paxil to warn of the “increased risk for cardiovascular malformations.” Kiker slip op. at 3.

The plaintiffs threw in the kitchen sink with their allegations, which included specific averments that GSK should have informed the medical community about “significant” adverse event reporting and the meaning of claimed deaths among rat pups in high-dose maternal toxicity testing. Not content with a failure to warn case, plaintiffs ratcheted their allegations into a fraudulent misrepresentation case, as well. Kiker slip op. at 3-4. Laura Plunkett and Ra-id Abdulla were the principal expert witnesses relied upon by plaintiffs for their hyperbolic claims.

The Standard

Chief Judge Sargus started his description of the governing law by insisting that the standard for expert witness gatekeeping was “flexible”; that is, he would follow the “Gumby Rule,” which allows the trial judge maximal flexibility and stretch to admit dubious expert witness opinions. Chief Judge Sargus employed the usual reductionist criteria for assessing “reliability.” Citing Kumho Tire, he explained that the court’s role was to ascertain whether

an expert . . . employs in the courtroom the same level of intellectual rigor that characterizes the practice of an expert in the relevant field.”

Kumho Tire Co. v. Carmichael, 526 U.S. 137, 152, (1999). He also acknowledged that Daubert had provided some indicia of reliability in factors such as

testing, peer review, publication, error rates, the existence and maintenance of standards controlling the technique’s operation, and general acceptance in the relevant scientific community.”

Kiker slip op. at 7, quoting from United States v. Langan, 263 F.3d 613, 621 (6th Cir. 2001) (citing Daubert v. Merrell Dow Pharm, Inc., 509 US. 579, 593-94 (1993)).

Chief Judge Sargus was then quick to point out that the cited Daubert factors do not make up a definitive, dispositive test or checklist, which presumably gave him license to ignore these factors and their absence, all together. Nowhere later in his opinion on the contested reliability of plaintiffs’ expert witnesses’s causation opinions is there any discussion of the actual testing, its validity, its pre-publication and post-publication peer review, error rates, standards for assessing causation, or general acceptance of the claimed methodologies. And of course, the discretion permitted district judges in performing their gatekeeping function is not the discretion to abandon the gatekeeping function and to ignore relevant methodological criteria. See Kumho Tire, 526 U.S. 137, 158-59 (Scalia, J., concurring).

Semantic Legerdemain Substitutes for Demonstration of General Causation

Chief Judge Sargus acknowledged that there is a “specific methodology” used by scientists to assess a body of evidence for causation of birth defects, but then proceeded to ignore that methodology without bothering to describe or apply it. Kiker slip op. at 10. What gave the trial judge his argument for ignoring the “specific methodology” used by scientists, the Daubert factors, and indeed any and all factors for assessing the validity of a scientific claim and conclusion, was the language used by the Food and Drug Administration (FDA) and GSK, the NDA-holder, in various communications. Rather than engage in an intellectually challenging exploration and evaluation of the actual scientific evidence and analysis that underlay the plaintiffs’ expert witnesses’ causation opinions, Chief Judge Sargus pointed to the language used by the FDA in its original Public Health Advisory about the issue of congenital cardiac malformations in children of mothers who ingested paroxetine in their first trimester of pregnancy:

[t]he FDA has determined that exposure to paroxetine in the first trimester of pregnancy may increase the risk for congenital malformations, particularly cardiac malformations. At the FDA’s request, the manufacturer has changed paroxetine’s pregnancy category from C to D and added new data and recommendations to the WARNINGS section of paroxetine’s prescribing information. FDA is awaiting the final results of recent studies and accruing additional data related to the use of paroxetine in pregnancy in order to better characterize the risk for congenital malformations associated with paroxetine.”

Kiker slip op. at 10, quoting from FDA Public Health Advisory (Dec. 8, 2005), available at <http://www.fda.gov/Drugs/DrugSafety/PostmarketDrugSafetyInformationforPatientsandProviders/ucm051731.htm> (emphasis added).

Chief Judge Sargus apparently was oblivious to the difference between “X causes Y” and “X may increase the risk of Y.” As the trial judge, he also fixed on the FDA’s decision to change the pregnancy category labeling for paroxetine from Category C to Category D, with the latter category’s reflecting “positive evidence of human risk.” Kiker slip op. at 11. Again, the existence of evidence for risk is not, and never has been, the existence of evidence that would support a reasonable, reliable conclusion that paroxetine causes cardiac birth defects. Nothing can explain or justify this incredible reliance and misinterpretation of language, and Chief Judge Sargus makes no attempt to defend his linguistic contortions.

Chief Judge Sargus ends with an implied assertion that he, as trial judge, need not spend any time on assessing the quantity or quality of evidence for a conclusion of causality because GSK has admitted that paroxetine causes cardiac birth defects. The GSK Dear Healthcare Provider Letter, the FDA Safety Alert, along with the (preliminary) results of a single epidemiologic study

combine in this instance to constitute an admission that Paxil can cause injury, and is sufficient to create an issue of fact regarding causation.”

Kiker slip op. at at 15.

Whence comes this incredible reliance upon the language of a package insert?  Chief Judge Sargus points to Judge James Gwin’s decision in In re Meridia, and proceeds to provide two pages, single-spaced, of block quotation from the Meridia decision. Kiker slip op. at 13-15, quoting from In re Meridia Prods. Liab. Litig., 328 F. Supp. 2d 791, 800-01 (N.D. Ohio 2004).

Interspersed in the two pages of quotation from Meridia were citations to Ferebee and Wells, two of the most discredited, disreputable federal court decisions on biomedical causation, both of which were effectively overruled sub silentio by the Supreme Court in Daubert. Chief Judge Sargus argues that the Meridia decision held that “product inserts to both physicians and patients” constituted “admissions of Meridia’s potential to cause substantial increases in blood pressure in some patients. Meridia, 328 F. Supp. 2d at 810. Affirming the district court’s decision in Meridia, the Sixth Circuit specifically upheld the district court’s determination that the FDA warning label at issue in that case “constitutes an admission that Meridia can cause injury.” Meridia Prods. Liab. Litig. v. Abbott Labs, 447 F.3d 861, 866 (6th Cir. 2006).

This analytical shortcut has serious problems. First, as a first year law student might observe, the Meridia decision resulted in the exclusion of plaintiffs’ key expert witness and the grant of summary judgment to the defendant on adequacy of its warning, all of which the Sixth Circuit affirmed. Given that there was no liability, the comments about causation would seem to be dictum, not holding. Second, with respect to the issue of warnings as admissions, the Circuit agreed that the district court had construed the defendant’s package insert warning that the medication ‘‘substantially increases’’ blood pressure as an admission, but that such unequivocal language was quite different from warning language that states medication use ‘‘is associated with’’ an adverse event. 447 F.3d at 866. The FDA’s Public Health Advisory, the change to Category D, and GSK’s own sponsored study did not, individually or collectively, state a finding of anything more than an association, and that there “may be an increased risk.”

Of course, Chief Judge Sargus’s glib exercise eliminated all the difficult thought of evaluating actual scientific evidence. The indolent approach used in Kiker committed another blatant error. The approach not only relied incorrectly upon some language of the FDA and medication license holder, but it ignored all the contrary evidence, context, and analysis that kept the FDA from reaching a conclusion of causality in 2005, and most scientists to this very day. Furthermore, the Kiker approach conveniently ignored that over a decade of additional evidence, much of it exonerating paroxetine. Chief Judge Sargus has misidentified the weakest, incomplete, out-of-date, cherry-picked evidentiary display with reliable evidence that purports to support a causal conclusion.

Non-Specific Confusion on Specific Causation

Having announced that the court will not grant a hearing, or even an on-the-paper review of the actual evidence for plaintiffs’ causal claims, Chief Judge Sargus proceeded to make even shorter work of the issue of specific causation. The only support for specific causation in the case was in the proffered testimony of Dr. Ra-id Abdulla, a serial testifying expert witness in anti-depressant birth defects cases. Abdulla purported to conduct a differential diagnosis to discern the cause of the infant plaintiff’s birth defect, a ventricular septal defect. Kiker slip op. at 16.

The diagnosis of the infant Kiker’s birth defect, however, was never in doubt; rather it was the etiology of the septal defect, which was at issue. Abdulla claimed to have ruled out all other potential alternative causes. Kiker slip op. at 18. Even if Abdulla’s claim could be accepted for known causes of septal defects, he would still be faced with a situation in which there are baseline or background cases of septal defects, which occur in children with no known or even suspected risk factor. The court failed to explain how Abdulla ruled out such unknown, prevalent causes of septal defects in the Kiker plaintiff. To be sure, the court appeared to have fallen for the “treating physician” ruse, which suggests that treating a condition provides some magical insight into the cause of that condition. Kiker slip op. at 19-20.

No explanation was cited by the court for how Abdulla worked his magical clinical inference of specific causation. Sadly, there is no such magic, except in the form of the magic thinking evidenced here by Abdulla, and acquiesced in by Chief Judge Sargus. No biomarker of causal originst distinguishes the Kiker plaintiff’s septal defect from one caused by any other cause, whether or not established by current medical science. Moreover, Abdulla’s magical thinking cannot be swept under the Kumho Tire rug of appropriate level of rigor in the field. The Kiker court cited no evidence that pediatric cardiologists routinely and reliably make the specific causal attribution that Dr. Abdulla made in this case, as a paid, testifying expert witness. The court incredulously accepted Abdulla’s hand waving about the epistemic warrant of experience, education, training that has nothing to do with discerning individual causes.

GSK asked for oral argument, which may have been Chief Judge Sargus’s last clear chance to avoid these errors. Declaring that the record was fully developed, Judge Sargus denied the request for a hearing. Kiker slip op. at 1, 4. We are left with a profoundly flawed misunderstanding of scientific evidence and causal inference.

The Lawsuit Industry’s Cookbook for Bogus Epidemiology

November 15th, 2016

In 1965, Sir Austin Bradford Hill was appropriately elected to the President of the Royal Society of Medicine. Along with Sir Richard Doll and others, Hill had pioneered the use of epidemiologic and statistical methods, which in the 1950s, he had applied to the issue of tobacco smoking and lung cancer. By the late 1950s, Hill and Doll had embraced and urged a causal association between smoking and lung cancer, but they had formidable opponents in Joseph Berkson and Sir Ronald A. Fisher.

By 1964, Hill and Doll’s causal thesis had largely prevailed. On January 11, 1964, the Office of the Surgeon General, in the United States, issued a Committee report that reviewed the available evidence and concluded that the relationship between smoking and lung cancer was indeed causal. See Surgeon General’s Advisory Committee on Smoking and Health, Smoking and Health (Office of the Surgeon General, United States Public Health Service 1964). See alsoProfiles in Science – 1964 Report,” National Library of Medicine Website.

Almost a year to the day after the Surgeon General’s report was issued, Hill gave the President’s Address at the Royal Society of Medicine, in London. For Hill, by then Professor Emeritus of Medical Statistics in the University of London, the occasion was triumphant. Not only had he prevailed over the intellectual doubts of Berkson and Fisher, and the animadversions of the tobacco industry, but he had shown that causal relationships can be identified and established with statistical methods in population studies, even in the absence of demonstrated mechanisms or experimental randomization. Fittingly, his after-dinner speech outlined the methodology that had proved successful, and that speech was published in the Proceedings of the Royal Society of Medicine. Austin Bradford Hill, “The Environment and Disease: Association or Causation?” 58 Proc. Royal Soc’y Med. 295 (1965) [cited as Hill].

As a tribute to Hill, the publication was noteworthy, but Hill, if still alive, would be surprised, perhaps shocked, certainly annoyed, that the text of his Presidential Address was still being cited as the canonical statement of causal assessments more than half a century later. Epidemiologic science has progressed in many important ways since 1964, and science has refined and improved substantially upon Hill’s articulation of the epidemiologic method for assessing the causality of observed associations. Nonetheless, lawyers, on both sides of the bar, continue to publish analyses of Hill’s 1965 publication. A few years ago, Dr. Frank Woodside and Allison Davis published one such article.1 This month, a lawyer from the lawsuit industry published a plaintiffs’ vision of Hill’s methodology, in Trial, the trade journal of his industry. R. Jason Richards, “Reflecting on Hill’s Original Causation Factors,” 52 Trial 44 (Nov. 2016) [cited as Richards]Frankly, everyone would be better off if they simply read the original speech and understood it for what it was – a 50 year-old informal statement of a complex problem.

As do many defense lawyers, Richards treats the Hill factors as a canonical guide to causation. He states that “[t]he scientific community has generally accepted these viewpoints, and scientists regularly use them to assess causality between exposure and an outcome,” and then cites legal decisions only. Richards at 45.2

Some of Richards’ exegesis is a benign, helpful reminder to consider Hill’s recommendations in their original context, and that no one factor is typically dispositive. Richards at 49. What Richards fails to say is that a concordance of factors will often be needed to establish causation, as they were in showing that smoking caused lung cancer.

Richards tells us, plaintively and accurately, that “[m]any of Hill’s key insights about how to make decisions based on epidemiological evidence have been largely ignored or distorted over the years.” Richards at 46. Richards has in mind the indeterminacy of any given factor, which he casts as “no single interpretation is infallible.” Of course, if no single interpretation is infallible, then all are fallible, which Richards no doubts sees as immunizing any expert witness’s opinion against exclusionary gatekeeping.

Ultimately, Richards becomes yet another author who abridges and bastardizes Hill’s key insights. He further urges his readers to consider and invoke “seldom-cited but significant passages” from Hill (1965). One passage that Richards fails to consider and invoke himself is Hill’s important predicate for considering the nine factors in the first place. The starting point for Hill, as he clearly expressed in his President’s address, was that

[o]ur observations reveal an association between two variables, perfectly clear-cut and beyond what we would care to attribute to the play of chance. What aspects of that association should we especially consider before deciding that the most likely interpretation of it is causation?”

Hill at 295. The nine factors follow, with elaboration, but importantly, the nine factors answer the question Hill posed about the aspects of the association, which we have already seen to be “perfectly clear-cut and beyond what we care to attribute to the play of chance.” To be sure, and fair, Hill did not use the words valid and statistically significant, but his meaning was perfectly clear and is completely ignored by Richards and occasionally by trial courts led into error by lawyers who advocate for causal associations on weak, inconsistent evidence from invalid associations.

There is some mischief perpetuated by treating Hill’s language as legislating a decision procedure for demonstrating causality. First, Hill spoke informally, without the scholarly apparatus of footnotes or extensive research. So when Hill wrote that we should not dismiss a putative causal claim merely because the association is “slight,” he was writing in the context of known causal associations with relative risks in the 100s (for chimney sweeps and scrotal cancer) or twenty to thirty fold for smoking and lung cancer. A slight association might well be one that is a merely doubling or tripling of base rate for the outcome. And Hill was not urging that such slight associations were much in the way of evidence for causality, only that we should not dismiss the causal claim solely because of the small size of the association.

Second, some of what Hill said was wrong when he said it in 1965. For instance, he wrote that none of the factors is necessary. The temporality factor, which specifies that the putative cause come before the putative effect, however, is indeed necessary. Unless of course we have “spooky action at a distance” that permits simultaneous causality across the universe in biomedicine. Hill was probably not thinking about the philosophical problems invoked by quantum physics when he incorrectly branded temporality as not necessary.

Third, some of what Hill said is distorted in commentary, such as Richards’, by ignorance, or by design. Perhaps the distortions occur because Hill was speaking colloquially to fellow scientists. Much of what he said, and is written in his 1965 article, is nuanced and contextual. For instance, Hill wrote that:

[n]o formal tests of significance can answer those questions. Such tests can, and should, remind us of the effects that the play of chance can create, and they will instruct us in the likely magnitude of those effects. Beyond that they contribute nothing to the ‘proof’ of our hypothesis.”

Hill at 299 (emphasis added). This passage, referenced by Richards, is a favorite of lawsuit industry lawyers but an astute reader will note that it opens with a reference to “those questions.” To understand what “those questions” are, one would sensibly look at the paragraph above. In that previous paragraph, Hill states that the “proof” of causality requires judgment, and that the nine factors cannot be taken as providing a quod erat demonstrandum; rather the factors structure the inquiry, which is essential, into how to explain the facts observed and to rule out explanations that do not turn on causality. Understandably, statistical analyses do not replace the judgment involved in synthesizing the evidence with respect to the available factors. Of course, formal tests of significance do answer specific questions, such as whether the association in one beyond that which we care to ascribe to chance, a consideration that must be made before delving into the nine factors. And even after crossing the threshold of statistically significant, valid association, an analyst would consider statistical tests in connection with consistency and exposure gradient. The former consideration today is often addressed by meta-analyses, tests of homogeneity, and p-curves, none of which was in common use in 1965, at the time of Hill’s Presidential address. Some simple tests for exposure gradient were available, but often the judgment was left to a visual assessment of the dose-response curve. Today, formal statistical analyses would indeed be invoked to determine whether the dose-curve was likely inconsistent with random variability.

Fourth, what Hill said in 1965 must be viewed in the light of over 50 years of scientific developments in field of epidemiology and statistics. Ultimately, what Richards presents as précis is a gross distortion of Hill in 1965, and and an even greater distortion of acceptable scientific method in 2016.

Just as it is helpful to study historical propaganda, Richards’ article should be studied to understand how the lawsuit industry will try to induce error in judicial judgments. Richards argues that all the nine factors are “inherently subjective,” and so it is understandable and inevitable that expert witnesses will disagree. Richards at 48. The fact is that some of the Hill factors are not the least subjective. Strength, consistency, and exposure gradient, for instance, are all objective, quantifiable variables that can be used to evaluate a body of available epidemiologic studies (after the the available studies have shown associations perfectly clear cut and beyond the play of chance).

Richards deploys a sophistical argument that because none of the factors is necessary, then causation can be inferred in the absence of any of the factors, or perhaps in the presence of only the most unimportant factors, such as analogy. Readers will be hard pressed to come up with an example of a generally accepted causal relationship evidenced solely by an analogy or by a subjective assessment of plausibility, but Richards argues for the validity of such an inference in the abstract, and in the absence of any real-world examples.

To some lawyers, all facts are created equal.” Felix Frankfurter3

To some lawyers, all epidemiologic studies are created equal, but they too are mistaken. Richard attempts to engage in extreme deconstuctionist analysis, verging on Daubert by Derrida. Richards avers that all studies are flawed, and one can always find always question a study’s validity, and “so it would be a stretch to establish any cause-and-effect relationship if statistically significant data were the only acceptable basis for asserting causation.” Richards at 48. Richards confuses validity with statistical significance, but worse, his argument ignores important qualitative and quantitative differences between and among studies in terms of their design, implementation, analysis, and interpretation. His argument is akin to saying that all human beings have flaws so we should do away with honors and awards, as well as prisons and penalties.

Finally, no article such as Richards’ would be complete without misstating the holding and dictum in Matrixx Initiatives:

The U.S. Supreme Court has finally recognized this as well, agreeing that statistical significance is not the touchstone of reliability under a Daubert analysis.”

Richards at 48 (citing Matrixx Initiatives, Inc. v. Siracusano, 131 S. Ct. 1309, 1319 (2011)). This is a remarkably misleading citation given that the Court, in Matrixx, noted that it was not considering whether “expert testimony was properly admitted,” and that it was not trying “to define here what constitutes reliable evidence of causation.” Matrixx, 131 S. Ct. at 1319.

In reaching, nay stretching, to address the statistical issue, the Supreme Court cited three cases, two of which involved differential etiology and contested specific causation, for which statistical analysis was absent and irrelevant. Id. (citing Best v. Lowe’s Home Centers, Inc., 563 F.3d 171, 178 (C.A.6 2009); Westberry v. Gislaved Gummi AB, 178 F.3d 257, 263-264 (C.A.4 1999)). The third case was relevant to statistical inference, but involved a case in which the plaintiff’s expert witness had cited at least one study with a nominally statistically significant result, which was vitiated by internal and external validity concerns. Wells v. Ortho Pharmaceutical Corp., 788 F.2d 741, 744-745 (11th 1986), cert. denied, 479 U.S.950 (1986).4

Richard asserts that “Hill recognized that scientists could efficiently use data and draw reasonable inferences about causation in the absence of statistically significant findings.” Richards at 48. Tellingly, Richards provides no citation for his assertion. There is none, but nothing is the hallmark of epistemic nihilism.


1 Frank C. Woodside, III & Allison G. Davis, “The Bradford Hill Criteria: The Forgotten Predicate,” 35 Thomas Jefferson L. Rev. 103 (2013).

2 citing In re Trasylol Prods. Liab. Litig., 2010 WL 1489734, at *8-9 (S.D. Fla. Mar. 8, 2010); In re Zoloft (Sertraline Hydrochloride) Prods. Liab. Litig., 26 F. Supp. 3d 449, 454 (E.D. Pa. 2014).

3 quoted in Comes v. Microsoft Corp., 709 N.W.2d 114, 116 (Iowa 2006) without source information.