One of the high points of 2013 for me was the Practicing Law Institute seminar on expert witness gatekeeping. I am indebted to my co-chair David Cohen, and the talented participants, Dr. David Garabrant, Joe Cecil, Mary Wells, and Mike Williams. I learned a great deal from all of them, even when we disagreed.
At the seminar, which addressed the slippage in the federal judiciary’s fidelity to Rule 702 text and principles, Joe Cecil kindly reminded me of Professor Gastwirth’s contribution on the Wells v. Ortho Pharmaceutical Corp., 615 F. Supp. 262, 298 (N.D. Ga. 1985), aff’d and rev’d in part on other grounds, 788 F.2d 741 (11th Cir.), cert. denied, 479 U.S.950 (1986). Joseph L. Gastwirth, “The need for careful evaluation of epidemiological evidence in product liablility cases: a reexamination of Wells v. Ortho and Key Pharmaceuticals,” 2 Law, Probability and Risk 151 (2003).
The Supreme Court recently reinvigorated the Wells case, and like Frankenstein’s monster, it walks again. Matrixx Initiatives, Inc. v. Siracusano, 131 S. Ct. 1309 (2011). See also “Matrixx Unloaded.” As a result of the Matrixx decision, I revisited the untoward Wells decision several times this year. See “Wells v. Ortho Pharmaceutical Corp. Reconsidered – Part 1”; and here, here, here, here, and here. Joe’s reminder, however, provoked me to look one more time at Wells, through the lens of Gastwirth’s review.
Gastwirth is a well-known statistician, not a lawyer. His article undertakes some interesting sensitivity analyses of the potential role of one known confounder in the epidemiologic studies of spermicides and birth defects, cited by witnesses in Wells. Not surprisingly, Gastwirth confuses important legal concepts, when he contends that science and law have different goals because science has the luxury of waiting for additional studies to resolve indeterminate datasets. Gastwirth at 154. Gastwirth suggests that in law, causality determinations turn on fairness, but in science, they turn on the truth. Gastwirth offers no support for his assertion, and there is much law to the contrary. Federal Rule of Evidence 102, for instance, explains the “Purpose” of the Federal Rules and how they should be construed:
“These rules should be construed so as to administer every proceeding fairly, eliminate unjustifiable expense and delay, and promote the development of evidence law, to the end of ascertaining the truth and securing a just determination.”
Truth does count, at least in some courts, and it is hard to imagine a case in which the goals of truth and justice are bifurcated. Gastwirth sees law courts as unable to avoid judgment when the evidentiary display is weakly developed or not replicated, but he is wrong. Courts can, do, and should avoid permitting judgments based upon indeterminate datasets.
Gastwirth also contends that the duty to warn in tort is based upon a possibility of causation, and he opines that this is perfectly satisfactory. Gastwirth at 155. Gastwirth cites the Restatement (Third) a/Torts for his view, but the Restatement does not couch the duty to warn as predicated upon merely possible risk:
“A defendant will not be liable under an implied warranty of merchantability for failure to warn about risks that were not reasonably foreseeable at the time of sale or could not have been discovered by way of reasonable testing prior to marketing the product. A manufacturer will be held to the standard of knowledge of an expert in the appropriate field, and will remain subject to a continuing duty to warn of risks discovered following the sale of the product at issue.”
Id. (emphasis added). Gastwirth fails to explain why the epistemic standard of the Restatement should be diluted to include merely possible risks; nor does he explain why the law should impose a duty because a minority of so-called experts claim knowledge of the existence of a risk, whether potential or known.
Risks are based upon known causal relationships, viewed ex ante rather than post hoc. Gastwirth attempts to defend the Wells decision on the basis that the trial judge, Judge Shoob, could have correctly found a possible risk, but this was not then or now the standard for imposing a duty to warn. Even if possibilities were the basis for a duty to warn, the courts in Wells had to find sufficient evidence of cause in fact between the spermicidal jelly and the birth defect present in the infant plaintiff. No merely possible risk could have satisfied this burden, and Gastwirth avoids this important issue in his review. Ultimately, Gastwirth fails to come to grips with how the Wells courts artificially distinguished two causation standards – warning-harm and product-harm. Gastwirth at 184.
When courts are required to work with an epistemic standard for evaluating opinions about causal claims, they will often confront sincere beliefs supported by “putative” justifications, but which turn out to be untrue. The obvious implication is that the causal opinions were not “true” when offered, and the asserted, putative justification was no justification at all. Twenty years post-Daubert should have made our courts more sensitive to the pathology of claiming, or pathoepistemology, as I call it.
Although Gastwirth is intent to show that the district and circuit court decisions in Wells were reasonable given the record evidence at the time of their decisions, Gastwirth’s defense is tepid at best. Gastwirth concedes that the plaintiffs’ evidence in the Wells case was not very convincing, and he is quick to point out that he does not conclude that spermicides cause limb reduction defects (although he does believe that the data were sufficient for his idiosyncratic diluted duty-to-warn legal standard:
“The epidemiological evidence in the Wells case surely was not very convincing and it is understandable that scientists and legal scholars might desire stronger scientific evidence before holding a defendant liable. What seems unfortunate is the lack of discussion of the part of the record most supportive of the legal decisions from the published criticisms.201 While it is reasonable to ask how much evidence should be required before a warning is required, is it fair to rely on studies that were published after the time of exposure to assess the producer’s duty at the time?”
Gastwirth at 184-85. The equivocation in this passage between assessing causation for a duty to warn and for product harm is typical of Gastwirth’s confusion throughout this article. As noted, Gastwirth appears to believe it is sufficient to show the plausibility of the causal relationship such that the trial court’s findings of a duty to warn, and a breach of that duty, were reasonably supported:
“It should be stressed that our examination of the studies does not conclude that exposure to spermicides causes limb reduction defects. 202 Indeed, the studies published in the 1980s are sensitive to the potential effect of unmeasured known potential risk factors that were not controlled for in the analysis. Many studies observed that it is difficult to rule out small but meaningful risks given their sample size or that more studies including potential confounding variables be carried out before a scientific conclusion can be reached.203”
Gastwirth at 185 & n.202. So small increased risks were difficult to rule in or rule out, but then how can a judgment for plaintiffs be sustained unless, as I have shown elsewhere, Judge Shoob improperly placed the burden of proof upon the defendant? The studies published in the 1980s may have been sensitive to the potential effect of unmeasured and uncontrolled known potential risks, but Judge Shoob clearly was not sensitive to the problem at all.
Other problems abound. The studies involved in the Wells case not only failed to account for the few known confounders at the time, but they could not and did not sort out the residual confounding. Gastwirth concedes the point in a footnote, in which he cites a paper by Professor Gary Shaw on the role of multivitamin use in preventing birth defects. See Gastwirth at 185 & n. 202 (citing Shaw et al., “Maternal Periconceptional Use of Multivitamins and Reduced Risk for Conotruncal Heart Defects and Limb Deficiencies Among Offspring,” 59 Am. J. Med. Genetics 536 (1995) (showing that children of mothers who had taken vitamins and folic acid had a 30-35% lower risk of limb defects). Gastwirth sees the discovery of causes not previously controlled for in the spermicide/birth defect studies of the 1980s as example of the differences between scientific and legal judgments. Another way to look at such discoveries is to impose some standard of intellectual modesty upon courts in basing their findings upon speculative causal claims.
Gastwirth’s casual dismissiveness of the need for replication in a legal standard of causation is difficult to square with his acknowledgment that early studies often prove misleading about the existence of a true risk. Gastwirth, for instance, acknowledges that later publications quite commonly do not support initial hypotheses. Gastwirth at 179, 184 (citing Michael B. Bracken, “Spermicidal Contraceptives and Poor Reproductive Outcomes: The Epidemiologic Evidence Against an Association,” 151 Am. J. Obstet. & Gyncecol. 552, 555 (1985)). Gastwirth channels the anti-manufacturer Zeitgeist for his claim that “firms” may not report studies that show risk, but he ignores the pervasive, prevalent practice of academic and advocacy scientists’ not publishing negative studies in the first place, and even later in the natural history of scientific controversies.
Ultimately, however, Gastwirth’s own judgment is a self-fulfilling prophecy of how the law and science should part company over false-positive assertions of causality. Fortunately, the law has evolved in a different direction, towards convergence of scientific and legal modes of assessing causal claims. Now judges must evolve as well.
