’Then you should say what you mean’, the March Hare went on.
‘I do’, Alice hastily replied; ‘at least–at least I mean what I say–that’s the same thing, you know’.
‘Not the same thing a bit!’ said the Hatter. ‘You might just as well say that “I see what I eat” is the same thing as “I eat what I see!”’

Lewis Carroll, Alice’s Adventures in Wonderland, Chapter VII (1865)

Anick Bérard is an epidemiologist at the Université de Montréal. Most of her publications involve birth outcomes and maternal medication use, but Dr. Bérard’s advocacy also involves social media (Facebook, YouTube) and expert witnessing in litigation against the pharmaceutical industry.

When the FDA issued its alert about cardiac malformations in children born to women who took Paxil (paroxetine) in their first trimesters of pregnancy, the agency characterized its assessment of the “early results of new studies for Paxil” as “suggesting that the drug increases the risk for birth defects, particularly heart defects, when women take it during the first three months of pregnancy.”¹ The agency also disclaimed any conclusion of “class effect” among the other selective serotonin reuptake inhibitors (SSRIs), such as Zoloft (sertraline), Celexa (citalopram), and Prozac (fluoxetine). Indeed, the FDA requested the manufacturer of paroxetine to undertake additional research to look at teratogenicity of paroxetine, as well as the possibility of class effects. That research never showed an SSRI teratogenicity class effect.

A “suggestion” from the FDA of an adverse effect is sufficient to launch a thousand litigation complaints, which were duly filed against GlaxoSmithKline. The plaintiffs’ counsel recruited Dr. Bérard to serve as an expert witness in support of a wide array of birth defects in Paxil cases. In her hands, the agency’s “suggestion” of causation became a conclusion. The defense challenged Bérard’s opinions, but the federal court motion to exclude her causal opinions were taken under advisement, without decision. Hayes v. SmithKline Beecham Corp., 2009 WL 4912178 (N.D. Okla. Dec. 14, 2009). One case in state court went to trial, with a verdict for plaintiffs.

Despite Dr. Bérard;s zealous advocacy for a causal association between Paxil and birth defects, she declined to assert any association between maternal use of the other, non-paroxetine SSRIs and birth defects. Here is an excerpt from her Rule 26 report in a paroxetine case:

“Taken together, the available scientific evidence makes it clear that Paxil use during the first trimester of pregnancy is an independent risk factor that at least doubles the risk of cardiovascular malformations in newborns at all commonly used doses. This risk has been consistent and was further reinforced by repeated observational study findings as well as meta-analyses results. No such associations were found with other types of SSRI exposures during gestation.”²

In her sworn testimony, Dr. Bérard made clear that she really meant what she had written in her report, about exculpating the non-paroxetine SSRIs of any association with birth defects:

“Q. Is it fair to say that you will not be offering an opinion that SSRIs as a class, or individual SSRIs other than Paxil increased the risk of cardiovascular malformations in newborns?

A. This is not what I was asked to do.

Q. But in fact you actually write in your report that you don’t believe there’s sufficient data to reach any conclusion about other SSRIs, true?

A. Correct.”³

In 2010, Dr. Bérard, along with two professional colleagues, published what they called a systematic review of antidepressant use in pregnancy and birth outcomes.⁴ In this review, Bérard specifically advised that paroxetine should be avoided by women of childbearing age, but she and her colleagaues affirmatively encouraged use of other SSRIs, such as fluoxetine, sertraline, and citalopram:

“Clinical Approach: A Brief Overview

“For women planning a pregnancy or when a treatment initiation during pregnancy is deemed necessary, the decision should rely not only on drug safety data but also on other factors such as the patient’s condition, previous response to other antidepressants, comorbidities, expected adverse effects and potential interactions with other current pharmacological treatments. Since there is a more extensive clinical experience with SSRIs such as fluoxetine, sertraline, and citalopram, these agents should be used as first-line therapies. Whenever possible, one should refrain from prescribing paroxetine to women of childbearing potential or planning a pregnancy. However, antenatal screening such as fetal echocardiography should be considered in a woman exposed prior to finding out about her pregnancy.”⁵

When Bérard wrote and published her systematic review, she was still actively involved as an expert witness for plaintiffs in lawsuits against the manufacturers of paroxetine. In her 2010 review, Dr. Bérard gave no acknowledgment of monies earned in her capacity as an expert witness, and her disclosure of potential conflicts of interest was limited to noting that she was “a consultant for a plaintiff in the litigation involving Paxil.”⁶ In fact, Bérard had submitted multiple reports, testified at deposition, and had been listed as a testifying expert witness in many cases involving Paxil or paroxetine.

Not long after the 2010 review article, Glaxo settled most of the pending paroxetine birth defect cases, and the plaintiffs’ bar pivoted to recast their expert witnesses’ opinions as causal teratogenic conclusions about the entire class of SSRIs. In 2012, the federal courts established a “multi-district litigation,” MDL 2342, for birth defect cases involving Zoloft (sertraline), in the Philadelphia courtroom of Judge Cynthia Rufe, in the Eastern District of Pennsylvania.

Notwithstanding her 2010 clinical advice that pregnant women with depression should use fluoxetine, sertraline, or citalopram, Dr. Bérard became actively involved in the new litigation against the other, non-Paxil SSRI manufacturers. By 2013, Dr. Bérard was on record as a party expert witness for plaintiffs, opining that setraline causes virtually every major congenital malformation.⁷

In the same year, 2013, Dr. Bérard published another review article on teratogens, but now she gave a more equivocal view of the other SSRIs, claiming that they were “known carcinogens,” but acknowledging in a footnote that teratogenicity of the SSRIs was “controversial.”⁸ Incredibly, this review article states that “Anick Bérard and Sonia Chaabane have no potential conflicts of interest to disclose.”⁹

Ultimately, Dr. Bérard could not straddle her own contradictory statements and remain upright, which encouraged the MDL court to examine her opinions closely for methodological shortcomings and failures. Although Bérard had evolved to claim a teratogenic “class effect” for all the SSRIs, the scientific support for her claim was somewhere between weak to absent.¹⁰ Perhaps even more distressing, many of the pending claims involving the other SSRIs arose from pregnancies and births that predated Bérard’s epiphany about class effect. Finding ample evidence of specious claiming, the federal court charged with oversight of the sertraline birth defect claims excluded Dr. Bérard’s causal opinions for failing to meet the requirements of Federal Rule of Evidence 702.¹¹

Plaintiffs sought to substitute Nicholas Jewell for Dr. Bérard, but Dr. Jewell fared no better, and was excluded for other methodological shenanigans.¹² Ultimately, a unanimous panel of the United States Court of Appeals, for the Third Circuit, upheld the expert witness exclusions.¹³

“This is the end
Beautiful friend

This is the end
My only friend, the end”

Jim Morrison, “The End” (c. 1966)

The General Electric Co. v. Joiner, 522 U.S. 136 (1997), case was based upon polychlorinated biphenyl exposures (PCB), only in part. The PCB part did not hold up well legally in the Supreme Court; nor was the PCB lung cancer claim vindicated by later scientific evidence. See “How Have Important Rule 702 Holdings Held Up With Time?” (Mar. 20, 2015).

The Supreme Court in Joiner reversed and remanded the case to the 11^th Circuit, which then remanded the case back to the district court to address claims that Mr. Joiner had been exposed to furans and dioxins, and that these other chemicals had caused, or contributed to, his lung cancer, as well. Joiner v. General Electric Co., 134 F.3d 1457 (11th Cir. 1998) (per curiam). Thus the dioxins were left in the case even after the Supreme Court ruled.

After the Supreme Court’s decision, Anthony Roisman argued that the Court had addressed an artificial question when asked about PCBs alone because the case was really about an alleged mixture of exposures, and he held out hope that the Joiners would do better on remand. Anthony Z. Roisman, “The Implications of G.E. v. Joiner for Admissibility of Expert Testimony,” 1 Res Communes 65 (1999).

Many Daubert observers (including me) are unaware of the legal fate of the Joiners’ claims on remand. In the only reference I could find, the commentator simply noted that the case resolved before trial.[1] I am indebted to Michael Risinger, and Joseph Cecil, for pointing me to documents from PACER, which shed some light upon the Joiner “endgame.”

In February 1998, Judge Orinda Evans, who had been the original trial judge, and who had sustained defendants’ Rule 702 challenges and granted their motions for summary judgments, received and reopened the case upon remand from the 11th Circuit. In March, Judge Evans directed the parties to submit a new pre-trial order by April 17, 1998. At a status conference in April 1998, Judge Evans permitted the plaintiffs additional discovery, to be completed by June 17, 1998. Five days before the expiration of their additional discovery period, the plaintiffs moved for additional time; defendants opposed the request. In July, Judge Evans granted the requested extension, and gave defendants until November 1, 1998, to file for summary judgment.

Meanwhile, in June 1998, new counsel entered their appearances for plaintiffs – William Sims Stone, Kevin R. Dean, Thomas Craig Earnest, and Stanley L. Merritt. The docket does not reflect much of anything about the new discovery other than a request for a protective order for an unpublished study. But by October 6, 1998, the new counsel, Earnest, Dean, and Stone (but not Merritt) withdrew as attorneys for the Joiners, and by the end of October 1998, Judge Evans entered an order to dismiss the case, without prejudice.

A few months later, in February 1999, the parties filed a stipulation, approved by the Clerk, dismissing the action with prejudice, and with each party to bear its own coasts. Given the flight of plaintiffs’ counsel, the dismissals without and then with prejudice, a settlement seems never to have been involved in the resolution of the Joiner case. In the end, the Joiners’ case fizzled perhaps to avoid being Frye’d.

And what has happened since to the science of dioxins and lung cancer?

Not much.

In 2006, the National Research Council published a monograph on dioxin, which took the controversial approach of focusing on all cancer mortality rather than specific cancers that had been suggested as likely outcomes of interest. See David L. Eaton (Chairperson), Health Risks from Dioxin and Related Compounds – Evaluation of the EPA Reassessment (2006). The validity of this approach, and the committee’s conclusions, were challenged vigorously in subsequent publications. Paolo Boffetta, Kenneth A. Mundt, Hans-Olov Adami, Philip Cole, and Jack S. Mandel, “TCDD and cancer: A critical review of epidemiologic studies,” 41 Critical Rev. Toxicol. 622 (2011) (“In conclusion, recent epidemiological evidence falls far short of conclusively demonstrating a causal link between TCDD exposure and cancer risk in humans.”

In 2013, the Industrial Injuries Advisory Council (IIAC), an independent scientific advisory body in the United Kingdom, published a review of lung cancer and dioxin. The Council found the epidemiologic studies mixed, and declined to endorse the compensability of lung cancer for dioxin-exposed industrial workers. Industrial Injuries Advisory Council – Information Note on Lung cancer and Dioxin (December 2013). See also Mann v. CSX Transp., Inc., 2009 WL 3766056, 2009 U.S. Dist. LEXIS 106433 (N.D. Ohio 2009) (Polster, J.) (dioxin exposure case) (“Plaintiffs’ medical expert, Dr. James Kornberg, has opined that numerous organizations have classified dioxins as a known human carcinogen. However, it is not appropriate for one set of experts to bring the conclusions of another set of experts into the courtroom and then testify merely that they ‘agree’ with that conclusion.”), citing Thorndike v. DaimlerChrysler Corp., 266 F. Supp. 2d 172 (D. Me. 2003) (court excluded expert who was “parroting” other experts’ conclusions).

Last year, an industrial cohort, followed for two decades found no increased risk of lung cancer among workers exposed to dioxin. David I. McBride, James J. Collins, Thomas John Bender, Kenneth M Bodner, and Lesa L. Aylward, “Cohort study of workers at a New Zealand agrochemical plant to assess the effect of dioxin exposure on mortality,” 8 Brit. Med. J. Open e019243 (2018) (reporting SMR for lung cancer 0.95, 95%CI: 0.56 to 1.53)

[1] Morris S. Zedeck, Expert Witness in the Legal System: A Scientist’s Search for Justice 49 (2010) (noting that, after remand from the Supreme Court, Joiner v. General Electric resolved before trial)

CONFOUNDING¹

Back in 2000, several law professors wrote an essay, in which they detailed some of the problems courts experienced in expert witness gatekeeping. Their article noted that judges easily grasped the problem of generalizing from animal evidence to human experience, and thus they simplistically emphasized human (epidemiologic) data. But in their emphasis on the problems in toxicological evidence, the judges missed problems of internal validity, such as confounding, in epidemiologic studies:

“Why do courts have such a preference for human epidemiological studies over animal experiments? Probably because the problem of external validity (generalizability) is one of the most obvious aspects of research methodology, and therefore one that non-scientists (including judges) are able to discern with ease – and then give excessive weight to (because whether something generalizes or not is an empirical question; sometimes things do and other times they do not). But even very serious problems of internal validity are harder for the untrained to see and understand, so judges are slower to exclude inevitably confounded epidemiological studies (and give insufficient weight to that problem). Sophisticated students of empirical research see the varied weaknesses, want to see the varied data, and draw more nuanced conclusions.”²

I am not sure that the problems are dependent in the fashion suggested by the authors, but their assessment that judges may be reluctant to break the seal on the black box of epidemiology, and that judges frequently lack the ability to make nuanced evaluations of the studies on which expert witnesses rely seems fair enough. Judges continue to miss important validity issues, perhaps because the adversarial process levels all studies to debating points in litigation.³

The frequent existence of validity issues undermines the partisan suggestion that Rule 702 exclusions are merely about “sufficiency of the evidence.” Sometimes, there is just too much of nothing to rise even to a problem of insufficiency. Some studies are “not even wrong.”⁴ Similarly, validity issues are an embarrassment to those authors who argue that we must assemble all the evidence and consider the entirety under ethereal standards, such as “weight of the evidence,” or “inference to the best explanation.” Sometimes, some or much of the available evidence does not warrant inclusion in the data set at all, and any causal inference is unacceptable.

Threats to validity come in many forms, but confounding is a particularly dangerous one. In claims that substances such as diesel fume or crystalline silica cause lung cancer, confounding is a huge problem. The proponents of the claims suggest relative risks in the range of 1.1 to 1.6 for such substances, but tobacco smoking results in relative risks in excess of 20, and some claim that passive smoking at home or in the workplace results in relative risks of the same magnitude as the risk ratios claimed for diesel particulate or silica. Furthermore the studies behind these claims frequently involve exposures to other known or suspected lung carcinogens, such as arsenic, radon, dietary factors, asbestos, and others.

Definition of Confounding

Confounding results from the presence of a so-called confounding (or lurking) variable, helpfully defined in the chapter on statistics in the Reference Manual on Scientific Evidence:

“confounding variable; confounder. A confounder is correlated with the independent variable and the dependent variable. An association between the dependent and independent variables in an observational study may not be causal, but may instead be due to confounding. See controlled experiment; observational study.”⁵

This definition suggests that the confounder need not be known to cause the dependent variable/outcome; the confounder need be only correlated with the outcome and an independent variable, such as exposure. Furthermore, the confounder may be actually involved in such a way as to increase or decrease the estimated relationship between dependent and independent variables. A confounder that is known to be present typically is referred to as a an “actual” confounder, as opposed to one that may be at work, and known as a “potential” confounder. Furthermore, even after exhausting known and potential confounders, studies of may be affected by “residual” confounding, especially when the total array of causes of the outcome of interest is not understood, and these unknown causes are not randomly distributed between exposed and unexposed groups in epidemiologic studies. Litigation frequently involves diseases or outcomes with unknown causes, and so the reality of unidentified residual confounders is unavoidable.

In some instances, especially in studies pharmaceutical adverse outcomes, there is the danger that the hypothesized outcome is also a feature of the underlying disease being treated. This phenomenon is known as confounding by indication, or as indication bias.⁶

Kaye and Freedman’s statistics chapter notes that confounding is a particularly important consideration when evaluating observational studies. In randomized clinical trials, one goal of the randomization is the elimination of the role of bias and confounding by the random assignment of exposures:

“2. Randomized controlled experiments

In randomized controlled experiments, investigators assign subjects to treatment or control groups at random. The groups are therefore likely to be comparable, except for the treatment. This minimizes the role of confounding.”⁷

In observational studies, confounding may completely invalidate an association. Kaye and Freedman give an example from the epidemiologic literature:

“Confounding remains a problem to reckon with, even for the best observational research. For example, women with herpes are more likely to develop cervical cancer than other women. Some investigators concluded that herpes caused cancer: In other words, they thought the association was causal. Later research showed that the primary cause of cervical cancer was human papilloma virus (HPV). Herpes was a marker of sexual activity. Women who had multiple sexual partners were more likely to be exposed not only to herpes but also to HPV. The association between herpes and cervical cancer was due to other variables.”⁸

The problem identified as confounding by Freedman and Kaye cannot be dismissed as an issue that goes to the “weight” of the study issue; the confounding goes to the heart of the ability of the herpes studies to show an association that can be interpreted to be causal. Invalidity from confounding renders the studies “weightless” in any “weight of the evidence” approach. There are, of course, many ways to address confounding in studies: stratification, multivariate analyses, multiple regression, propensity scores, etc. Consideration of the propriety and efficacy of these methods is a whole other level of analysis, which does not arise unless and until the threshold question of confounding is addressed.

Reference Manual on Scientific Evidence

The epidemiology chapter of the Second Edition of the Manual stated that ruling out of confounding as an obligation of the expert witness who chooses to rely upon the study.⁹ Although the same chapter in the Third Edition occasionally waffles, its authors come down on the side of describing confounding as a threat to validity, which must be ruled out before the study can be relied upon. In one place, the authors indicate “care” is required, and that analysis for random error, confounding, bias “should be conducted”:

“Although relative risk is a straightforward concept, care must be taken in interpreting it. Whenever an association is uncovered, further analysis should be conducted to assess whether the association is real or a result of sampling error, confounding, or bias. These same sources of error may mask a true association, resulting in a study that erroneously finds no association.”¹⁰

^{Elsewhere in the same chapter, the authors note that “chance, bias, and confounding” must be looked at, but again, the authors stop short of noting that these threats to validity must be eliminated:}

^{“Three general categories of phenomena can result in an association found in a study to be erroneous: chance, bias, and confounding. Before any inferences about causation are drawn from a study, the possibility of these phenomena must be examined.”11}

                ^{*  *  *  *  *  *  *  *}

“To make a judgment about causation, a knowledgeable expert must consider the possibility of confounding factors.”¹²

Eventually, however, the epidemiology chapter takes a stand, and an important one:

“When researchers find an association between an agent and a disease, it is critical to determine whether the association is causal or the result of confounding.”¹³

Mandatory Not Precatory

The better reasoned cases decided under Federal Rule of Evidence 702, and state-court analogues, follow the Reference Manual in making clear that confounding factors must be carefully addressed and eliminated. Failure to rule out the role of confounding renders a conclusion of causation, reached in reliance upon confounded studies, invalid.¹⁴

The inescapable mandate of Rules 702 and 703 is to require judges to evaluate the bases of a challenged expert witness’s opinion. Threats to internal validity, such as confounding, in a study may make reliance upon any given study, or an entire set of studies, unreasonable, which thus implicates Rule 703. Importantly, stacking up more invalid studies does not overcome the problem by presenting a heap of evidence, incompetent to show anything.

Pre-Daubert

Before the Supreme Court decided Daubert, few federal or state courts were willing to roll up their sleeves to evaluate the internal validity of relied upon epidemiologic studies. Issues of bias and confounding were typically dismissed by courts as issues that went to “weight, not admissibility.”

Judge Weinstein’s handling of the Agent Orange litigation, in the mid-1980s, marked a milestone in judicial sophistication and willingness to think critically about the evidence that was being funneled into the courtroom.¹⁵ The Bendectin litigation also was an important proving ground in which the defendant pushed courts to keep their eyes and minds open to issues of random error, bias, and confounding, when evaluating scientific evidence, on both pre-trial and on post-trial motions.¹⁶

Post-Daubert

When the United States Supreme Court addressed the admissibility of plaintiffs’ expert witnesses in Daubert, its principal focus was on the continuing applicability of the so-called Frye rule after the enactment of the Federal Rules of Evidence. The Court left the details of applying the then newly clarified “Daubert” standard to the facts of the case on remand to the intermediate appellate court. The Ninth Circuit, upon reconsidering the case, re-affirmed the trial court’s previous grant of summary judgment, on grounds of the plaintiffs’ failure to show specific causation.

A few years later, the Supreme Court itself engaged with the actual evidentiary record on appeal, in a lung cancer claim, which had been dismissed by the district court. Confounding was one among several validity issues in the studies relied upon by plaintiffs” expert witnesses. The Court concluded that the plaintiffs’ expert witnesses’ bases did not individually or collectively support their conclusions of causation in a reliable way. With respect to one particular epidemiologic study, the Supreme Court observed that a study that looked at workers who “had been exposed to numerous potential carcinogens” could not show that PCBs cause lung cancer. General Elec. Co. v. Joiner, 522 U.S. 136, 146 (1997).¹⁷