In 2010, almost 30 years after Ferebee was decided, the Solicitor General cited the case in an amicus brief before the Supreme Court case, in Matrixx Iniatives, Inc. v. Siracusano. The case was a securities fraud class action, which was dismissed initially by the trial court on consideration of the defendant’s motion that the complaint failed to allege causation supported by statistically significant studies. The Supreme Court would go on unanimously to reject causation as a criterion for establishing a prima facie case of securities fraud, which made statistical significance irrelevant. Because the FDA could (and later did) require the company to recall its product upon a showing that material evidence suggested that there might be a possible causally induced harm, the Court held that the plaintiff class did not have to allege causation.[1] The harm to the shareholders came in the form of management’s bullish financial projections for a product that was later recalled for safety concerns, even if the recalled product never was shown to cause any harm. The Solicitor General’s amicus brief advanced the Ferebee case as an example of a causal relationship that could be established “through consideration of multiple factors independent of statistical significance.”[2] Although the government’s amicus brief correctly discerned that the causal connection between paraquat exposure and pulmonary fibrosis was established without analytical epidemiologic studies, and the necessary tools of statistical analysis for such studies, the brief mistakenly placed the allegations that Zycam caused anosmia in the same conceptual framework as paraquat. Unlike paraquat toxicity, millions of people used Zycam for relief from cold and flu symptoms, and the alleged harm, anosmia, commonly occurs in the aftermath of colds and flu. The Zycam personal injury claims fared poorly in litigation because of the dearth of supportive evidence that was appropriate to support causation, as opposed to materiality in securities law.[3]
The Ferebee case correctly observed that epidemiology was not necessary to establish the causal claim involving paraquat dermal exposure and lung toxicity and fibrosis. At the time that Mr. Ferebee sustained extensive paraquat exposure as a result of his governmental employer’s extreme negligence, the scientific community fully accepted that paraquat exposure, by ingestion, inhalation, or dermal exposure caused systemic toxicity and deleterious lung effects. This “general causation” had been established by case reports and case series, along with studies of paraquat’s metabolic fate and distribution in humans and non-human animals (including non-human primates), and assessment of mechanistic effects in cells and tissues of the target organs affected by paraquat when it became systemically distributed in the human body.
About the time of the Ferebee litigation, a textbook on agricultural chemicals described the toxic effects of dipyridyl compounds in humans, including paraquat:
“Human Toxicology Experience: A considerable amount of clinical experience has been reported in the literature with over 100 cases of illness and/or death. The chemical is unique in the sense that there is not only an acute toxicity syndrome but, in addition, it has the ability to produce a delayed fibroblastic response in the lungs. The latter is usually the principal mechanism of death.
For industrial workers, paraquat is not considered very dangerous. Inhalation hazard is extremely low due to the low vapor pressure of the chemical. Nevertheless, protective respiratory equipment should be used particularly when other atmospheric contamination might occur. * * * On no occasion should an applicator be allowed to walk through drifting spray.” [4]
This textbook cited studies that suggested that dermal and respiratory exposure to paraquat did not appear to be a hazard to field applicators, despite the demonstration of absorption, as long as precautions against overexposure are taken.[5] The premise of the textbook discussion, that appropriate, well-known safety measures and protective gear are employed, was an important part of its analysis.
This early textbook discussion also flagged delayed lung fibrosis as the main problem caused by all modes of paraquat exposure, including dermal absorption:
“Although the acute symptoms of paraquat intoxication are of concern and are dangerous, the principal problem relates to the unique delayed manifestations of this chemical’s ability to produce a fibroblastic change in the lung which begins a number of days after absorption. * * * Experiments then found that it was possible to induce respiratory failure as a result of both dermal and aerosol routes of absorption (Newhouse, 1978).”[6]
An early review by the World Health Organization also emphasized that paraquat exposure was not expected to pose a health risk as long as safe work practices are followed:
“Occupational exposure to paraquat does not pose a health risk if the recommendations for use are followed and there is adherence to safe working practices.
* * *
In the small number of reported cases of paraquat poisoning allegedly resulting from occupational exposure, the cause can be identified as one or a combination of a number of factors, viz contamination of the skin with concentrated products, use of inadequately diluted solutions, use of faulty equipment, misuse of equipment (e.g., blowing blocked spray jets) or failure to take action in the event of contamination of skin or clothing.”[7]
Cases of dermal exposure to undiluted paraquat (20%), especially when exposure involved dermal exposure to the scrotum, can produce serious systematic toxicity.[8]
Ferebee was not given an appropriate respirator even when exposed to intense atmospheric contamination. He was drenched in paraquat spray, and remained drenched for hours. The gross negligence of his employer ensured that there would not be many similar cases, and that the tools of analytical epidemiology would not be available.[9] Indeed, epidemiology was never involved in determining general causation of paraquat exposure and lung fibrosis. Given that the outcome of interest would likely occur only in the context of negligent or intentional over-exposure, epidemiology will never be available.
Revisiting the Ferebee decision and the unique facts of the case place the decision in a better perspective for judging how courts continue to cite the case. The facts of the case readily distinguish the case, the claimed harm, and the manner of showing causation, from the facts in cancer, birth defects, and other cases where epidemiology is essential. The principle of charity would require the frequently quoted language on expert witness admissibility to be taken as a statement of the appellate standard of review for the jury’s determination of medical causation. Most of the glib characterizations of the Ferebee turn out to be wrong on close inspection of the case.
a. Ferebee was not a precedent under the Federal Rules of Evidence
Chevron’s evidentiary arguments were posed under Maryland law. Neither Rule 702[10] nor Rule 703[11] was ever mentioned in the district court or the Court of Appeals decisions.
b. Ferebee does not support a false distinction between scientific and legal causation.
In a later Bendectin birth defects case, Richardson v. Richardson-Merrell, Inc., the Court of Appeals struggled to distinguish Ferebee and its holdings based upon the ample epidemiologic evidence involving Bendectin. The Court mischaracterized Ferebee as not pertinent because it was on the “frontier of current medical and epidemiological inquiry.”[12] The Richardson court was impressed by the 20 years of research on Bendectin, including multiple epidemiologic studies. Unfortunately, the court was apparently ignorant of the irrelevance of epidemiology to the Ferebee case, and the extensive research base for determining the lung toxicity of paraquat. This ignorance seems to have resulted from Judge Mikva’s generalizations and overstatements of the paucity of evidence in Ferebee.
c. Ferebee does not support the false distinction between scientific and legal certainty.
Courts and commentators have attempted to explain the result in Ferebee by invoking what is largely a false distinction between scientific and legal certainty (or sufficiency). The Ferebee decision itself provided the ammunition by asserting a distinction between the requirements of scientific and legal decision making:
“In a courtroom, the test for allowing a plaintiff to recover in a tort suit of this type is not scientific certainly but legal sufficiency.”[13]
This was a common approach in distinguishing Ferebee, in the Bendectin litigation,[14] but it was picked up and promulgated by scientists and legal commentators.[15] Invoking this alleged distinction has become a common rhetorical move to excuse inadequate or insufficient evidence to support an expert witness’s causation opinion in litigation. The generalization from the facts of Ferebee to all scientific and legal questions of causation was wrong from the inception, and citations to a single case, Ferebee, cannot make those generalizations true.[16]
d. Ferebee did not establish the irrelevance or the dispensability of epidemiologic evidence in cancer or birth defects cases.
The Ferebee case observed that “a cause-effect relationship need not be clearly established by animal or epidemiological studies before a doctor can testify that in his opinion such a relationship exists.”[17] The bit about animal studies certainly cannot be part of the holding because the plaintiffs’ expert witnesses relied extensively on animal studies, along with human case reports, and human clinical studies of the metabolic fate and distribution of paraquat in both animals and humans.
The Ferebee case does properly stand for the proposition that there is a subset of all health effect cases for which epidemiologic evidence is unavailable and unnecessary for an expert witness to have for a valid conclusion of general and specific causation. The case illustrates how the ill effects of paraquat were observed shortly after exposure, and were sufficiently unique to not have a meaningful base- or background- rate. Adding the studies of absorption, metabolic fate and distribution, and mechanism of action, the plaintiff’s expert witnesses had an ample scientific, and legal, basis to assess causality.
The Ferebee case is sometimes mistakenly thought of as a cancer case, which would have required epidemiologic evidence.[18] This mistake may well result from later courts citing Ferebee in cancer cases, for the proposition that epidemiologic evidence is unnecessary to support plaintiff’s causal claim between some exposure and some cancer. Perhaps the Illinois Supreme Court has provided the most egregious example of this sort of mistake. In Donaldson, a case involving plaintiff’s exposure to coal tar and his later development of neuroblastoma, the Court confusedly found Ferebee to be “of particular significance.”[19] The Donaldson court affirmed the trial court’s admission of plaintiffs’ expert witness testimony about “a causal link causal link despite the lack of a statistical number of others with neuroblastoma and a history of coal tar exposure.”[20] The court rambled on about how the plaintiffs were not required to prove general or specific causation “with 100% certainty that neuroblastoma.” This assertion was a common strawman argument that channels a misreading of Ferebee. The defense in Donaldson did not argue that 100% certainty was required, and such a level of posterior probability has never been required in law or in science.
The Ferebee case also did not establish that statistical significance was not required in epidemiologic studies for expert witnesses to be able reasonably to rely upon such studies. Because epidemiologic evidence was not at issue, there was no holding about epidemiologic studies or statistical significance as a criterion of the validity of such studies.
e. Ferebee did not establish that a clinician can opine about causation without sufficient facts and data.
One of plaintiff’s expert witnesses in Ferebee was Dr. Crystal, who was both a physician and a research scientist. His opinion as an expert witness was hardly without supporting facts and data, and the facts and data were of the exact kind that led to the scientific acceptance of the causal connection between some paraquat exposures and lung fibrosis. Crystal’s opinion was certainly not proffered without any evidentiary basis, as some have suggested.[21] Nor was Ferebee a case in which expert witnesses opined without facts and data to support unprecedented opinions on general and specific causation.[22]
This overwrought, over-extended interpretation of Ferebee as permitting causation opinions based upon only clinical observations of the patient appeared in the first edition of the Reference Manual on Scientific Evidence, but disappeared in all subsequent editions. In the chapter by evidence law professor Margaret Berger, the Manual reported that Ferebee was frequently cited for a “holding that causation can be established by the testimony of treating physicians.”[23] Berger’s observation about frequent citation is correct, but the observation does nothing to validate the opinion cited. Berger offered no comments or analysis in critique of the frequent miscitation of Ferebee, leaving the reader to believe that citing Ferebee for the sufficiency of treating physician opinion without data was somehow appropriate. The citations to which Berger referred were erroneous in 1994, and they remain erroneous today.
[1] Matrixx Iniatives, Inc. v. Siracusano, 563 U.S. 27, 131 S.Ct. 1309, 1320 (2011).
[2] Brief for the United States as Amicus Curiae, in Matrixx Iniatives, Inc. v. Siracusano, No. 09-1156, 2010 WL 4624148, at *15 (Nov. 2010).
[3] See, e.g., Benkwith v. Matrixx Initiatives, Inc., 467 F. Supp. 2d 1316, 1326, 1330, 1332 (M.D. Ala. 2006) (granting defendant’s motion to exclude testimony of an expert in the field of epidemiology regarding Zicam nasal spray’s causing plaintiff’s anosmia, because the opinions had not been tested and a rate of error could not be provided).
[4] Sheldon L. Wagner, CLINICAL TOXICOLOGY OF AGRICULTURAL CHEMICALS 198, 199-200 (1983).
[5] Id. at 200 (citing “[s]tudies by Staiff and co-workers (1975)” on occupationally exposed persons).
[6] Id. at 201. See also A. J. Gardiner, Pulmonary oedema in paraquat poisoning, 27 THORAX 132 (1972).
[7] WORLD HEALTH ORGANIZATION, ENVIRONMENTAL HEALTH CRITERIA 39: PARAQUAT AND DIQUAT at § 1.1.5. Effects on man (1984).
[8] See K. Tungsanga, S. Chusilp, S. Iarasena & V. Sitprija, Paraquat poisoning: evidence of systemic toxicity after dermal exposure, 59 POSTGRAD. MED. J. 338, 338 (1983).
[9] Cf. Zuchowicz v. United States, 140 F.3d 381 (2nd Cir. 1998) (analyzing causation in the context of defendants clear negligence that resulted in undisputed overexposure to prescription medication Danocrine). Unlike Zuchowicz, however, the Ferebee case did provide a strong evidentiary base for causation.
[10] See Kenneth J. Chesebro, Taking Daubert’s “Focus” Seriously: The Methodology/Conclusion Distinction, 15 CARDOZO L. REV. 1745, 1747, 1753 (1994) (misciting Ferebee as a Rule 702 case).
[11] See Alani Golanski, Judicial Scrutiny of Expert Testimony in Environmental Tort Litigation, 9 PACE ENVT’L L. REV. 399, 406-07 (1992) (misrepresenting Ferebee as a case under Rule 703; “this evidentiary issue was resolved through examination of the facts or data underlying the proffered expert opinion only to the extent necessary to make a Rule 703 determination on whether they are of the type reasonably relied upon by experts in the field.”). See also Michael C. McCarthy, “Helpful” or “Reasonably Reliable”? Analyzing the Expert Witness’s Methodology Under Federal Rules of Evidence 702 and 703, 77 CORNELL L. REV. 350, 373 (1992) (discussing Ferebee as a Rule 702 and 703 decision).
[12] Richardson v. Richardson-Merrell, Inc., 857 F.2d 823, 831-832 (D.C. Cir. 1988).
[13] Ferebee, 736 F.2d at 1536.
[14] Id.
[15] Louis Lasagna & Sheila R. Shulman, Bendectin and the Language of Causation, chap. 5, at 111, in Kenneth R. Foster, David E. Bernstein & Peter W. Huber, eds., PHANTOM RISK: SCIENTIFIC INFERENCE AND THE LAW (1993)
[16] See Michael C. McCarthy, “Helpful” or “Reasonably Reliable”? Analyzing the Expert Witness’s Methodology Under Federal Rules of Evidence 702 and 703, 77 CORNELL L. REV. 350, 373 (1992) (“Essentially, the Ferebee decision distinguished between the level of certainty required by a scientific discipline-and the level of certainty required by a court in drawing conclusions regarding causation”: and “Ferebee stands for the proposition that courts, in determining whether a given substance more likely than not caused a plaintiff’s injury, cannot always wait for the sciences.”).
[17] Ferebee, 736 F.2d 1529, 1535 (D.C. Cir. 1984).
[18] David E. Bernstein, The Misbegotten Judicial Resistance to the Daubert Revolution, 89 NOTRE DAME L. REV. 27, 36 (2013); David E. Bernstein, Expert Witnesses, Adversarial Bias, and the (Partial) Failure of the Daubert Revolution, 93 IOWA L. REV. 451, 465 (2008) (“Ferebee involved a claim that exposure to an herbicide caused an individuals’ cancer.”).
[19] Donaldson v. Central Illinois Public Service Co., 313 Ill. App.3d 1061, 730 N.E.2d 68, 79 (2000).
[20] Id.
[21] Lee Loevinger, Evidentiary Framework Margaret A. Berger Reference Manual on Scientific Evidence, 36 JURIMETRICS J. 149, 153 & n.21 (1996) (discussing the before (Daubert) times when “mere qualification and the facial relevance of an opinion might suffice to let an expert testify in some jurisdictions.”).
[22] See Kenneth J. Chesebro, Taking Daubert’s “Focus: Seriously: The Methodology/Conclusion Distinction, 15 CARDOZO L. REV. 1745, 1747 & n.20 (1994) (incorrectly arguing that Ferebee involved an expert witness offered an “unprecedented expert factual conclusion which no published literature supported.”
[23] See Margaret A. Berger, Evidentiary Framework, 39, 81 & n.164, in FEDERAL JUDICIAL CENTER, REFERENCE MANUAL ON SCIENTIFIC EVIDENCE (1st ed. 1994).
