TORTINI

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

Archives »

Reference Manual’s Chapter on Expert Witness Testimony Admissibility – Part 4

March 5th, 2026

In the district court, Judge George O’Toole conducted a pre-trial hearing over four days, and heard testimony from Smith and Cranor, as well as from defense expert witnesses. Judge O’Toole’s published opinion carefully and accurately stated the facts, the applicable law, and presented a well-reasoned judgment as to why Smith’s opinion was not admissible under Rule 702. Without admissible opinions on general causation to support Milward’s case, Judge O’Toole granted summary judgment to the defendants.

Milward appealed the judgment. A panel of judges in the First Circuit heard argument, and reversed in an opinion that is riddled with serious errors.[1] In reviewing the district court’s application of Rule 702, the panel, in an opinion written by Chief Judge Lynch, credulously accepted most of Smith’s and Cranor’s arguments that an ill-defined WOE approach is acceptable method of guiding scientific judgment. Cranor equated WOE, as used by Smith, to the approach that Sir Austin Bradford Hill described, in 1965, for identifying causal associations from epidemiologic data.[2] Chief Judge Lynch’s opinion tracked accurately Cranor’s and Milward’s lawyers’ misrepresentations about Sir Austin’s paper:

“Dr. Smith’s opinion was based on a ‘‘weight of the evidence’’ methodology in which he followed the guidelines articulated by world-renowned epidemiologist Sir Arthur [sic] Bradford Hill in his seminal methodological article on inferences of causality. See Arthur [sic] Bradford Hill, The Environment and Disease: Association or Causation?, 58 Proc. Royal Soc’y Med. 295 (1965).

Hill’s article explains that one should not conclude that an observed association between a disease and a feature of the environment (e.g., a chemical) is causal without first considering a variety of ‘viewpoints’ on the issue.”[3]

The quoted language from the First Circuit opinion, which twice refers to “Arthur Bradford Hill,” rather than Austin Bradford Hill, may suggest that neither Chief Judge Lynch nor his judicial colleagues and their law clerks read the classic paper. An even stronger indicator that the appellate court did not actually read this paper is evidenced in the court’s equating WOE to Bradford Hill viewpoints, without consideration of the necessary predicate for those nine viewpoints. In his short paper, Sir Austin clearly spelled out that there was a foundation needed before parsing the nine viewpoints:

“Disregarding then any such problem in semantics we have this situation. Our 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?”[4]

Whatever Sir Arthur had to say about the matter, Sir Austin defined the starting point of causal analysis as an association free of invalidating bias and random error. The Milward decision ignored this all important predicate for assessing the various considerations that might allow for a valid association to be considered a causal association.[5] The resulting abridgement was a failure of scientific due process that distorted the Bradford Hill paper.

The First Circuit amplified its error when it asserted that from the nine considerations “no one type of evidence must be present before causality may be inferred.”[6] Although Sir Austin said something similar, one of the considerations he noted was “temporality,” in which the putative cause must come before the effect.  Most scientists would consider this consideration to be essential, unless they were observing events that were moving faster than the speed of light. The other eight considerations are more dependent upon context of the exposures and outcomes of interest, but surely strength and consistency of the clear-cut association across multiple studies is an extremely important consideration.

The First Circuit proceeds from misreading Sir Austin’s paper to misunderstanding another paper invoked by Cranor and by Milward’s lawyers. Carelessly tracking Cranor, the appellate court suggested that there was no “hierarchy of evidence”:

“For example, when a group from the National Cancer Institute was asked to rank the different types of evidence, it concluded that ‘‘[t]here should be no such hierarchy.’’ Michele Carbon [sic] et al., Modern Criteria to Establish Human Cancer Etiology, 64 Cancer Res. 5518, 5522 (2004); see also Sheldon Krimsky, The Weight of Scientific Evidence in Policy and Law, 95 Am. J. Pub. Health S129, S130 (2005).”[7]

This quoted language from the Milward opinion shows how slavishly and credulously the court adopted and regurgitated plaintiff’s argument. Sheldon Krimky was actively involved with SKAPP, and his article was presented at the SKAPP-funded Coronado Conference, discussed earlier in this series. Krimsky actually acknowledged that although “the term [WOE] is applied quite liberally in the regulatory literature, the methodology behind it is rarely explicated.”

As for the article by Carbon [sic], this publication never rejected a hierarchy of evidence. The court’s language, quoted above, follows immediately after the court’s discussion of Sir Austin’s nine types of corroborating evidence that would support the causal interpretation of an association. As such, the court seems to imply, incorrectly, that there was no hierarchy of these considerations.[8]

The court’s language also suggests that the quoted language came from the National Cancer Institute (NCI), but its provenance is quite different. The cited article’s lead author, Michele Carbone (not Carbon), was reporting on a workshop hosted by the NCI at an NCI building; it was not an official NCI event or publication. The NCI did sponsor or conduct the meeting, and Carbone’s paper was not an official statement of the NCI. Carbone’s paper was styled “Meeting Report,” and published as a paid advertisement in Cancer Research, not in the Journal of the National Cancer Institute as a scholarly article.

The discipline of epidemiology was not strongly represented at the meeting; most of the chairpersons and scientists in attendance were pathologists, cell biologists, virologists, and toxicologists. The authors of the meeting report reflect the interests and focus of the scientists in attendance. The lead author, Michele Carbone, a pathologist at the University of Hawaii, was an enthusiastic proponent of Simian Virus 40 as a cause of mesothelioma, a hypothesis that has not fared terribly well in the crucible of epidemiologic science.

The cited article did report some suggestions for modifying Bradford Hill’s criteria in the light of modern molecular biology, as well as a sense of the group that there was no “hierarchy” in which epidemiology was at the top of disciplines.  The group definitely did not address the established concept that some types of epidemiologic studies are analytically more powerful to support inferences of causality than others — the hierarchy of epidemiologic evidence. The group also did not address or reject a ranking of importance of Bradford Hill’s nine viewpoints. There was nothing remarkable about the tumor biologists’ statement that in some cases causality can be determined by careful identification of genetic inheritance or molecular biological pathways. There was no evidence of this sort in the Milward case, and the citation by Cranor and Milward’s lawyers was nothing more than hand waving.

Carbone’s meeting report summarizes informal discussion sessions at the 2003 meeting.  Those in attendance broke out into two groups, one chaired by Brook Mossman, a pathologist, and the other group chaired by Dr. Harald zur Hausen, a virologist. The meeting report included a narrative of how the two groups responded to twelve questions. Drawing from plaintiff’s (and Cranor’s) argument, the court’s citation to this meeting report is based upon one sentence in Carbone’s report, about one of twelve questions:

“6. What is the hierarchy of state-of-the-art approaches needed for confirmation criteria, and which bioassays are critical for decisions: epidemiology, animal testing, cell culture, genomics, and so forth?

There should be no such hierarchy. Epidemiology, animal, tissue culture and molecular pathology should be seen as integrating evidences in the determination of human carcinogenicity.”[9]

Considering the fuller context of the meeting, there is nothing particularly surprising about this statement.  The full question and answer in the meeting report does not even remotely support the weight given to it by the court. There was quite a bit of disagreement among meeting participants over criteria for different kinds of carcinogens, as seen the report on another question:

“2. Should the criteria be the same for different agents (viruses, chemicals, physical agents, promoting agents versus initiating DNA-damaging agents)?

There were different opinions. Group 1 debated this issue and concluded that the current listing of criteria should remain the same because we lack sufficient evidence to develop a separate classification. Group 2 strongly supported the view that it is useful to separate the biological or infectious agents from chemical and physical carcinogens due to their frequently entirely different mode of action.”[10]

Carbone and the other authors of the meeting report noted the importance to epidemiology for general causation, while acknowledging its limitations for determining specific causation:

“Concerning the respective roles of epidemiology and molecular pathology, it was noted that epidemiology allows the determination of the overall effect of a given carcinogen in the human population (e.g., hepatitis B virus and hepatocellular carcinoma) but cannot prove causality in the individual tumor patient.”[11]

Clearly, the report was not disavowing the necessity for epidemiology to confirm carcinogenicity in humans. Specific causation of Mr. Milward’s APML was irrelevant to his first appeal to the First Circuit. Carbone’s report emphasized the need to integrate epidemiologic findings with molecular biology; it did not suggest that epidemiology was not necessary or urge that epidemiology be ignored or disregarded:

“A general consensus was often reached on several topics such as the need to integrate molecular pathology and epidemiology for a more accurate and rapid identification of human carcinogens.”[12]

                 * * * * *

“Ideally, before labeling an agent as a human carcinogen, it is important to have epidemiological, experimental animals, and mechanistic evidence (molecular pathology).”[13]

The court’s implication that there was “no hierarchy of evidence” is unsupported by the meeting report. The suggestion that WOE allows some loosey-goosey, ad hoc, unstructured assessment of diverse lines of evidence is rejected in the meeting report with a careful admonition about the lack of validity of some animal models and mechanistic research:

“Moreover, carcinogens and anticarcinogens can have different effects in different situations. As shown by the example of addition of β-carotene in the diet, β- carotene has chemopreventive effects in many experimental systems, yet it appears to have increased the incidence of lung cancer in heavy smokers. Animal experiments can be very useful in predicting the carcinogenicity of a given chemical. However, there are significant differences in susceptibility among species and within organs in the same species, and differences in the metabolic pathway of a given chemical among human and animals could lead to error.”[14]

Inference to the Best Explanation

The First Circuit asserted that “no serious argument can be made that the weight of the evidence approach is inherently unreliable.”[15] As discussed above, this assertion is demonstrably false. In his testimony at the Rule 702 pre-trial hearing, Cranor classified WOE as based upon “inference to the best explanation,” and the First Circuit obsequiously accepted this claim. In articulating and accepting Cranor’s reduction of scientific method to IBE, the appellate court seemed unaware that IBE as an epistemic theory has been roundly criticized. In a very general sense, IBE draws on Charles Pierce’s description of abduction as a mode of reasoning, although many writers have been eager to distinguish abduction from IBE. Bas van Fraassen criticized IBE as lacking merit as a mode of argument in a way germane to Cranor’s presentation of the notion, and the First Circuit’s uncritical acceptance:

“As long as the pattern of Inference to the Best Explanation—henceforth, IBE—is left vague, it seems to fit much rational activity. But when we scrutinize its credentials, we find it seriously wanting.”[16]

The IBE approach raises thorny problems of knowing how to discern the best explanation, or how to tell whether an explanation is simply the best of a bad lot. Other philosophers of science have questioned why explanatoriness should matter as opposed to predictive ability and resistance to falsification upon severe or robust testing.

In the hands of Smith and Cranor, these philosophical quandries become largely beside the point. For Smith and Cranor IBE becomes telling just so stories, which transform “but for” causation into “could be” causation. Drawing directly from Cranor, the Circuit Court explained that an inference to the best explanation involves six general steps for scientists:

“(1) identify an association between an exposure and a disease,

(2) consider a range of plausible explanations for the association,

(3) rank the rival explanations according to their plausibility,

(4) seek additional evidence to separate the more plausible from the less plausible explanations,

(5) consider all of the relevant available evidence, and

(6) integrate the evidence  using professional judgment to come to a conclusion about the best explanation.”[17]

Of course assessing causation requires judgment, but Cranor and Smith radically abridge the process of judging by eliminating:

  • the robust testing of, and attempts to falsify, hypotheses,
  • the weighting of study designs,
  • the pre-specification of kinds of studies to be included or excluded, the assignment of weights to different kinds and qualities of studies, and
  • the pre-specification of criteria of study validity, experimental design, consistency, and exposure-response.

The vague, contentless IBE and WOE, in the hands of Smith, operates just as van Fraassen anticipated. With Cranor’s “philosophizing,” IBE creates a permission structure to reach any desired conclusion. Indeed, Cranor’s approach makes no allowance for when careful scientists withhold judgment because the evidence is inadequate to the task. Furthermore, Cranor’s approach and the Milward decision would cheerily approve cherry picking of studies and data within studies, post hoc weighing of evidence, and even fabricating and rejiggering of evidence, all of which was on display in Smith’s for-litigation opinion.

The First Circuit uttered its mantra of approval of Smith’s scientific delicts in language that became the target of the revision of Rule 702 in 2023:

“the alleged flaws identified by the [district] court go to the weight of Dr. Smith’s opinion, not its admissibility. There is an important difference between what is unreliable support and what a trier of fact may conclude is insufficient support for an expert’s conclusion.”[18]

Earlier in its opinion, the appellate court quoted from the version of Rule 702 in effect when it heard the appeal:

“if (1) the testimony is based upon sufficient facts or data, (2) the testimony is the product of reliable principles and methods, and (3) the witness has applied the principles and methods reliably to the facts of the case.”[19]

Sufficiency, reliability, and validity were all preliminary questions to be decided by the court as part of its gatekeeping responsibility.  The appellate court simply ignored the law in its decision to green light Smith’s testimony.

                    (to be continued)

[1] Milward v. Acuity Specialty Products Group, Inc., 639 F.3d 11 (1st Cir. 2011), cert. denied sub nom., U.S. Steel Corp. v. Milward, 565 U.S. 1111 (2012).

[2] Austin Bradford Hill, The Environment and Disease: Association or Causation?, 58 PROC. ROYAL SOC’Y MED. 295 (1965).

[3] Milward, 639 F.3d at 17.

[4] Id. at 295.

[5] See Frank C. Woodside, III & Allison G. Davis, The Bradford Hill Criteria: The Forgotten Predicate, 35 THOMAS JEFFERSON L. REV. 103 (2013).

[6] Milward, 639 F.3d at 17.

[7] Id. (internal citations omitted).

[8] The Reference Manual chapter on medical testimony carefully discusses the hierarchy of evidence as it factors into the assessment of medical causation. John B. Wong, Lawrence O. Gostin & Oscar A. Cabrera, Reference Guide on Medical Testimony, in National Academies of Sciences, Engineering and Medicine & Federal Judicial Center, REFERENCE MANUAL ON SCIENTIFIC EVIDENCE 687, 723 -24 (2011); John B. Wong, Lawrence O. Gostin, & Oscar A. Cabrera, Reference Guide on Medical Testimony, in National Academies of Sciences, Engineering and Medicine & Federal Judicial Center, REFERENCE MANUAL ON SCIENTIFIC EVIDENCE 1105, 1150-52 (4th ed. 2025). Interestingly, the chapter on epidemiology in the third edition of the Reference Manual cited to the Carbone workshop with apparent approval, but the same chapter in the fourth edition has dropped the reference. Compare Michael D. Green, D. Michal Freedman & Leon Gordis, Reference Guide on Epidemiology, in National Academies of Sciences, Engineering and Medicine & Federal Judicial Center, REFERENCE MANUAL ON SCIENTIFIC EVIDENCE 549, 564 n.48 (3rd ed. 2011) with Steve C. Gold, Michael D. Green, Jonathan Chevrier, & Brenda Eskenazi, Reference Guide on Epidemiology, in National Academies of Sciences, Engineering and Medicine & Federal Judicial Center, REFERENCE MANUAL ON SCIENTIFIC EVIDENCE 897 (4th ed. 2025).

[9] Carbone at 5522.

[10] Carbone at 5521.

[11] Carbone at 5518 (emphasis added).

[12] Carbone at 5518.

[13] Carbone at 5519.

[14] Carbone at 5521.

[15] Milward, 639 F.3d at 18-19.

[16] Bas van Fraassen, LAWS AND SYMMETRY 131 (1989).

[17] Milward, 639 F.3d at 18.

[18] Milward, 639 F.3d at 22.

[19] Milward, 639 F.3d at 14.

Posted in Causation, Expert Witnesses, Reference Manual on Scientific Evidence, Rule 702, Scientific Evidence | No Comments »

Reference Manual’s Chapter on Expert Witness Testimony Admissibility – Part 3

March 2nd, 2026

Richter and Capra treat WOE in Justice Steven’s lone dissenting opinion in Joiner as if it were the law. Of course, it was not; nor was it a particularly insightful analysis into scientific method, Rule 702, or the law of expert witnesses. The Manual authors elevate WOE by their complete failure to offer any criticisms or by citing to the scientific and legal scholars who have criticized WOE.

Richter and Capra do cite to a couple of cases that are skeptical of expert witnesses who had offered WOE opinions, but they fail to cite to any cases that disparage WOE itself.[1] In aggravation of their misplaced focus on the Joiner dissent, Richter and Capra proceed to spend two full pages on the Milward case, which had posthumously appeared in Professor Berger’s version of the law chapter in the 2011, third edition of the Reference Manual. The attention given to Milward in the fourth edition is greater than to any other non-Supreme Court case, including Frye. Richter and Capra offer no commentary or analysis critical of the case, although many legal commentators have criticized the Milward opinion on WOE.[2]

Richter and Capra’s chapter fails to note that a dark cloud hangs over the Milward case due to the unethical non-disclosure of CERT’s amicus brief filed in support of reversing the exclusion of CERT’s founders, Carl Cranor and Martyn Smith,[3] or CERT’s funding Smith’s research, or CERT’s involvement in shaking down corporations in California for Prop 65 bounties.

In their extensive coverage of the 2011 Milward decision, Richter and Capra failed to report that after the First Circuit reversed and remanded, the trial court again excluded plaintiffs’ expert witnesses for failing to give a valid opinion on specific causation. On the second appeal, the First Circuit affirmed the exclusion of specific causation expert witness testimony and the entry of final judgment for defendants.[4] Given that the first appellate decision was no longer necessary to the final disposition of the case, it is questionable whether there is any holding with respect to general causation in the case.

The most salient aspect of Richter and Capra’s uncritical coverage of the Milward case is their complete failure to identify the legal errors made by the First Circuit in its decision on Rule 702 and general causation. As the Reporter to the Rules Advisory Committee, Professor Capra was intimately involved in many meetings and memoranda that addressed the failings of courts to engage properly in gatekeeping. These failings were the gravamen of the basis for the 2023 amendments to Rule 702. The Milward decision in 2011 managed to check almost every box for bad decision making: the appellate panel ignored the text of Rule 702, disregarded Supreme Court precedent in the Joiner case, relied upon over-ruled, obsolete, pre-Daubert decisions, ignored the policy considerations urged by the Supreme Court, bungled basic scientific concepts, and egregiously and credulously endorsed WOE as advocated as a scientific methodology. Professor David E. Bernstein has pointed to the 2011 Milward decision, as “the most notorious,” and “[t]he most prominent example of such judicial truculence” in resisting following the requirements of Rule 702, as it existed in 2011.[5]

Milward is an important case, much as the Berenstain Bears stories are important and helpful in teaching children what not to do. Unfortunately, Richter and Capra discuss Milward in a way that might lead readers to believe that the case represents a reasonable or proper treatment of the science involved in the case. To correct this biased coverage of Milward, readers will have to roll up their sleeves and actually look at what the court did and did not do, and what scientific methodology issues were involved.

Perhaps the best place to begin is the beginning. Brian Milward filed a lawsuit in which he claimed that he was exposed to benzene as a refrigerator technician.[6] He developed acute promyelocytic leukeumia (APML), and claimed that he had been exposed to benzene from having used products made or sold by roughly two dozen companies. APML is a rare disease, type M3 of acute myeloid leukemia (AML), defined by specific chromosomal abnormalities that are necessary but not sufficient to result in APML. APML has an incidence of fewer than five cases per million per year. APML occurs with equal frequency in both sexes; there are no known environmental or occupational causes of APML.[7] APML occurs in the general population without benzene exposure, and its occurrence in all populations is sparse. There are no biomarkers that suggest that some putative benzene-related mechanism is involved in some APML cases, which biomarker would identify the rarity of benzene involvement in causation.

Milward’s General Causation Expert Witness, Martyn T. Smith

Milward did not serve a report from an epidemiologist, or anyone with significant expertise in epidemiology. His only general causation expert witness was Martyn Smith, a toxicologist, who testified that the “weight of the evidence” supported his opinion that benzene exposure causes APML.[8] As noted above, Smith is a member of the advocacy group, the Collegium Ramazzini; and for over 30 years, he has been a frequent testifier for plaintiffs in chemical exposure cases.[9]

Despite the low but widespread prevalence of APML in the general population, with no sex specificity, and the absence of any identifying biomarker of supposed benzene-related etiology in individual cases, Smith maintained that epidemiology was not necessary to reach a causal opinion about benzene and APML. The principal thrust of Smith’s proffered testimony is that APML is a plausible outcome of benzene exposure, because benzene can cause other varieties of AML, by structurally altering chromosomes (clastogenic) by breaking them and causing re-arrangements.[10]

The trial court found that Smith’s extrapolations were problematic and lacking in supporting evidence. The clear differences among AML subtypes made the extrapolation to APML, a unique clinical entity, inappropriate. The characteristic translocation in APML is absent from other varieties of AML, and APML, unlike other AML varieties, is treatable with all-trans retinoic acid.[11]

Smith advanced speculation that benzene targeted cells in the pathway of  leukemic transformation to APML, but the state of science was clearly devoid of sufficient evidence to show that benzene was involved in the APML translocations. Although the parties agreed that mechanistic evidence showed that benzene can effectuate chromosome damage that are characteristic of some AML subtypes other than APML, the trial court found that:

“[n]o evidence has been published making a similar connection between benzene exposure and the t(15;17) translocation, characteristic of APL [APML].”[12]

The trial court assessed Smith’s extrapolation from benzene’s clastogenic effect in breaking and rearranging chromosomes to induce some types of AML to its causing the specific APML t(15;17) translocation, as a

“bull in the china shop generalization: since the bull smashes the teacups, it must also smash the crystal. Whether that is so, of course, would depend on the bull having equal access to both teacups and crystal. If the teacups were easily knocked over, but the crystal securely stored away, a reason would exist to question, if not to reject, the proposition that the crystal was in as much danger as the teacups.”[13]

The trial judge clearly saw that Smith’s plausibility proved too much, and would support attributing virtually any disease to benzene through a putative mechanism of breaking chromosomes.

Lacking the courage of his convictions, Smith, non-epidemiologist, proceeded to offer opinions about the epidemiology of benzene and APML, some of them quite fanciful. No published or unpublished study showed a statistically significant increase in APML among benzene-exposed workers. The most Smith could draw from the published epidemiologic studies on benzene was one Chinese study that found a small risk ratio, without even nominal statistical significance: a crude odds ratio of 1.42 for benzene exposure and APML. Despite Smith’s hand waving about lack of power,[14] this Chinese study suggested that chloramphenicol was a risk factor for APML (M3), and it was able to identify a nominally statistically significant association between benzene and another sub-type of AML (M2a), with an odds ratio of 1.54.[15]

Smith offered no meta-analysis to show that the available studies collectively established a summary estimate of increased risk for APML among benzene workers. Undaunted, Smith set about to re-jigger the numbers in published studies to make something out of nothing. Neither physician nor epidemiologist, Smith altered diagnoses and exposure status as reported in published papers so that his reclassified cases and controls would yield, where none existed. These re-analyses were done speculatively, inconsistently, and incompetently, and were driven by the motivation to make something out of nothing. His approach was unsupported, unprincipled, and lacking in any reasonable methodology. The proffered re-analyses were never published, never presented at a professional society meeting, and never could comply with the standards used by epidemiologists used in their non-litigation activities. As a toxicologist, Smith did not have any non-litigation epidemiologic activities of note.

Smith’s representation of the relevant epidemiologic methods and studies was misleading and contained numerous errors that cumulatively led to erroneous conclusions; his own re-jiggering was carried out to reach a preferred conclusion to support plaintiff’s litigation case.[16]

One of the epidemiologic studies relied upon by Smith was Golumb (1982).[17] This study did not explore associations with benzene; it was a study of insecticides, chemicals and solvents, and petroleum. Crude oil contains very little benzene, typically about 0.1 percent.[18] Smith, without any evidentiary support, assumed that petroleum exposure equated to benzene exposure.

There were eight cases of cases of leukemia with petroleum exposure; one of those cases was APML. The authors of Golumb (1982) reported that this particular case with APML was actually a crane operator.[19]

In analyzing published epidemiologic studies, Smith insisted that he could re-classify APML cases to non-APML in control subjects, in studies, when the karyotype was normal. Karyotype analysis identifies the defining translocations of specific chromosomes in APML, and is found in virtually all such cases. The obvious result of Smith’s ad hoc reclassifications were to increase risk ratios for APML among benzene-exposed subjects. His arbitrary reclassifications of data allowed him to create the result he desired. In reviewing other published studies, Smith insisted that normal karyotype did not require reclassifying cases out of the APML category, when this approach would yield a risk ratio above one. 

Taking data from the Golumb 1982 paper, Smith attempted to inflate his calculation of an odds ratio, which would support his causation opinion. He arbitrarily discarded two APML from the non-exposed cases, and he discarded eight non-APML cases from the exposed subjects. He did not report p-values or confidence intervals for his reanalyses. At the hearing, the defense epidemiologist showed that Smith’s rejiggered odds ratio (1.51) had a p-value of 0.72, and a 95 percent confidence interval of 0.15 – 14.91. Not only was the result not statistically significant, the confidence interval shows that there was a range of alternative hypotheses over an order of magnitude in range, with none of them being rejected based upon the sample data at an alpha of 0.05. Without the rejiggering of exposed and unexposed cases, the odds ratio would have been 0.71, p = 0.76. All results, both as reported in the published article and as rejiggered by Smith were highly compatible with no association whatsoever.

In discussing other studies, Smith repeated his re-labeling of leukemia cases as APML, in the absence of karyotyping, to support his claims that there were more APML cases observed than expect on general population rates.[20] Smith also cited studies improvidently in supposed support of his opinion (Rinsky 1981; updated in 1994), where there was no association at all. Even workers heavily exposed to benzene in these studies did not develop APML.[21]  Similarly, in support of his opinion, Smith cited another Chinese study, which actually declared that:

“Acute promyelocytic leukemia has been reported infrequently in benzene-exposed groups as well as in t-ANLL. Although ANLL-M3 occurred in at least 4 patients in this series, its general representation among the subtypes of ANLL was similar in its distribution in de novo ANLL in China.”[22]

Smith’s methodological improprieties were the subject of a four day pre-trial hearing before Judge O’Toole. In the course of the hearings, Smith attempted to defends his methods, but like Donny Kerabatsos, in the Big Lebowski, Smith was out of his depth. The trial court found that Dr. Smith’s arbitrary creating and choosing data to support his beliefs was unreliable and not in accordance with generally accepted scientific methodology in the fields of medicine or epidemiology. Smith was simply fabricating data to fit his made-for-litigation beliefs.

Carl Cranor’s Attempt to Bolster Smith

Milward also submitted a report from Carl Forest Cranor, Smith’s business partner in founding the Prop 65 bounty-hunting CERT, and a fellow member of the advocacy group Collegium Ramazzini. Cranor has no expertise in toxicology or epidemiology, and he has never published on the cause of APML. As a professor of philosophy, Cranor has written about scientific methodology, including WOE and “inference to the best explanation (IBE).” Cranor’s publications are riddled with basic misunderstandings of statistical concepts.[23] Essentially, Cranor testified at the Rule 702 hearing, as a cheerleader for Smith, and to advocate for open admissions of dodgy scientific conclusions as acceptable with a methodology he described as WOE or IBE. Cranor stretched to resurrect Justice Stevens’ use of WOE, and attempted to pass it off as a generally accepted scientific mode of reasoning.

The trial court carefully reviewed the proffered opinion testimony in a four day pre-trial hearing. The trial court found that Smith had shown that his hypothesis was plausible and possible, but not that it was “scientific knowledge,” as required by Rule 702. Lacking sufficient scientific methodological validity and support, Smith’s opinions failed to satisfy the requirements of Rule 702, and were thus inadmissible. As a result of excluding plaintiff’s sole general causation expert witness, the trial court granted summary judgment to the defendants.[24]

(to be continued)

[1] See, e.g., Allen v. Pennsylvania Eng’g Corp., 102 F.3d 194, 197-98 (5th Cir. 1996) (“We are also unpersuaded that the ‘weight of the evidence’ methodology these experts use is scientifically acceptable for demonstrating a medical link between Allen’s EtO [ethylene oxide] exposure and brain cancer.”); Magistrini v. One Hour Martinizing Dry Cleaning, 180 F. Supp. 2d 584, 601-02 (D.N.J. 2002) (excluding David Ozonoff, whose WOE analysis of whether perchloroethylene causes acute myelomonocytic leukemia was criticized by court-appointed technical advisor), aff’d, 68 F. App’x 356 (3d Cir. 2003).

[2] See Eric Lasker, Manning the Daubert Gate: A Defense Primer in Response to Milward v. Acuity Specialty Products, 79 DEF. COUNS. J. 128, 128 (2012); David E. Bernstein, The Misbegotten Judicial Resistance to the Daubert Revolution, 89 NOTRE DAME L. REV. 27, 29, 53-58 (2013); David E. Bernstein & Eric G. Lasker, Defending Daubert: It’s Time to Amend Federal Rule of Evidence 702, 57 WM. & MARY L. REV. 1, 33 (2015); Richard Collin Mangrum, Comment on the Proposed Revision of Federal Rule 702: “Clarifying” the Court’s Gatekeeping Responsibility over Expert Testimony, 56 CREIGHTON LAW REVIEW 97, 106 & n.45 (2022); Thomas D. Schroeder, Toward a More Apparent Approach to Considering the Admission of Expert Testimony, 95 NOTRE DAME L. REV. 2039, 2045 (2020); Lawrence A. Kogan, Weight of the Evidence: A Lower Expert Evidence Standard Metastasizes in Federal Court, Washington Legal Foundation Critical Legal Issues WORKING PAPER Series no. 215 (Mar. 2020); Note, Judicial Conference Amends Rule 702. — Federal Rule of Evidence 702, 138 HARV. L. REV. 899, 903 (2025); Nathan A. Schachtman, Desultory Thoughts on Milward v. Acuity Specialty Products, DOI: 10.13140/RG.2.1.5011.5285 (Oct. 2015), available at https://www.researchgate.net/publication/282816421_Desultory_Thoughts_on_Milward_v_Acuity_Specialty_Products .

[3] See David DeMatteo & Kellie Wiltsie, When Amicus Curiae Briefs are Inimicus Curiae Briefs: Amicus Curiae Briefs and the Bypassing of Admissibility Standards, 72 AM. UNIV. L. REV. 1871 (2022) (noting that amicus briefs often include “unvetted and potentially inaccurate, misleading, or mischaracterized expert information,” without the procedural safeguards in place for vetting expert witnesses at trial).

[4] Milward v. Acuity Specialty Prods. Group, Inc., 969 F. Supp. 2d 101, 109 (D. Mass. 2013), aff’d sub. nom., Milward v. Rust-Oleum Corp., 820 F.3d 469, 471, 477 (1st Cir. 2016).

[5] David E. Bernstein, The Misbegotten Judicial Resistance to the Daubert Revolution, 89 NOTRE DAME L. REV. 27, 53, 29 (2013).

[6] Milward v. Acuity Specialty Products Group, Inc., 664 F. Supp. 2d 137 (D. Mass. 2009) (O’Toole, J.), rev’d, 639 F.3d 11 (1st Cir. 2011), cert. denied, U.S. Steel Corp. v. Milward, 565 U.S. 1111 (2012).

[7] Andrew Y. Li, et al., Clustered incidence of adult acute promyelocytic leukemia in the vicinity of Baltimore, 61 LEUKEMIA & LYMPHOMA 2743 (2021); Hassan Ali, et al., Epidemiology and Survival Outcomes of Acute Promyelocytic Leukemia in Adults: A SEER Database Analysis, 144 BLOOD 5942 S1 (2024).

[8] Milward, 664 F. Supp. 2d at 142.

[9] See, e.g., PPG Industries, Inc. v. Wells, No. 21-0232 (Feb. 10, 2023 W.Va.S.Ct.); Hall v. ConocoPhillips, 248 F. Supp. 3d 1177 (W.D. Okla. 2017); In re Levaquin Prods. Liab. Litig., 739 F.3d 401 (8th Cir. 2014); Jacoby v. Rite Aid Corp., No. 1508 EDA 2012 (Dec. 9, 2013 Pa. Super.); Harris v. CSX Transp., Inc., 232 W.Va. 617, 753 S.E.2d 275 (2013); In re Baycol Prods. Litig., 495 F. Supp. 2d 977 (D. Minn. 2007); In re Rezulin Prods. Liab. Litig., MDL 1348, 441 F.Supp.2d 567 (S.D.N.Y. 2006) (advocating mythological “silent injury”); Perry v. Novartis, 564 F.Supp.2d 452 (E.D. Pa. 2008); Dodge v. Cotter Corp., 328 F.3d 1212 (10th Cir. 2003); Sutera v. The Perrier Group of America Inc., 986 F. Supp. 655 (D. Mass. 1997); Redland Soccer Club, Inc. v. Dep’t of Army, 835 F.Supp. 803 (M.D. Pa. 1993).

[10] Milward, 664 F.Supp. 2d at 143-44.

[11] Milward, 664 F.Supp. 2d at 144.

[12] Id. at 146

[13] Id.

[14] The claim that a study lacks power is meaningless without a specification of the alternative hypothesis, the risk ratio the researcher thinks is the population parameter, at a specified level of alpha (typically p < 0.05), and a specified probability model. While virtually all studies would have reasonable statistical power (say 80 percent probability) to reject an alternative hypothesis that the risk ratio exceeded 10,000, no study would have power to detect a risk ratio of 1.0001, at a high level of probability.

[15] Yi Zhongguo, et al. (National Investigative Group for the Survey of Leukemia & Aplastic Anemia), Countrywide Analysis of Risk Factors for Leukemia and Aplastic Anemia, 14 ACTA ACADEMIAE MEDICINAE SINICAE 185 (1992).

[16] Milward, 664 F. Supp. 2d at 148-49.

[17] Harvey M. Golomb, et al., Correlation of Occupation and Karyotype in Adults With Acute Nonlymphocytic Leukemia, 60 BLOOD 404 (1982).

[18] Bo Holmberg, Per Lundberg, Benzene: standards, occurrence, and exposure, 7 AM. J. INDUS. MED. 375 (1985).

[19] Golumb, supra at note 17, at 407.

[20] See, e.g., Song-Nian Yin, et al., A cohort study of cancer among benzene-exposed workers in China: overall results, 29 AM. J. INDUS. MED. 227 (1996).

[21] Robert A. Rinsky, et al., Leukemia in Benzene Workers, 2 AM. J. INDUS. MED. 217 (1981); Mary B. Paxton, et al., Leukemia Risk Associated with Benzene Exposure in the Pliofilm Cohort: I. Mortality Update and Exposure Distribution, 14 RISK ANALYSIS 147 (1994); Mary B. Paxton, et al., Leukemia Risk Associated with Benzene Exposure in the Pliofilm Cohort II. Risk Estimates, 14 RISK ANALYSIS 155 (1994).

[22] Lois B. Travis, et al., Hematopoietic Malignancies and Related Disorders Among Benzene-Exposed Workers in China, 14 LEUKEMIA & LYMPHOMA 91, 99 (1994).

[23] See, e.g., Carl F. Cranor, REGULATING TOXIC SUBSTANCES: A PHILOSOPHY OF SCIENCE AND THE LAW at 33-34(1993) (conflating random error with posterior probabilities: “One can think of α, β (the chances of type I and type II errors, respectively) and 1- β as measures of the “risk of error” or “standards of proof.”); id. at 44, 47, 55, 72-76.

[24] 664 F. Supp. 2d at 140, 149.

Posted in Causation, Expert Witnesses, Reference Manual on Scientific Evidence, Rule 702, Scientific Misconduct, Statistical Power | No Comments »

The First Daubert Motion

February 20th, 2026

As every school child knows, or at least every law student in the United States knows, Daubert was a Bendectin case. The plaintiff claimed that his mother’s use of Bendectin, a prescription anti-nausea medication, during pregnancy caused him to be born with a major limb reduction defect.

Filed in 1984, the Daubert case was pending, in summer 1989, before Judge Earl Ben Gilliam, in the Southern District of California. A trial date was approaching, and a deadline for motions for summary judgment. The first Daubert motion was filed in August 1989, in Daubert v. Merrell Dow Pharmaceuticals, Inc.[1] It was a motion for summary judgment, not a motion specifically to exclude plaintiffs’ expert witness’s proffered testimony.

By the time of the first Daubert motion, the plaintiff was relying upon the anticipated testimony of John Davis Palmer, M.D. For the time, John Davis Palmer was not an unlikely expert witness. Although Palmer practiced internal medicine, he had a doctorate in pharmacology. Palmer, however, had no experience studying Bendectin, and no real expertise in epidemiology. He had never designed or published an epidemiologic study, and he had never done any kind of research on Bendectin. The standard for qualifying an expert witness, even in federal court, has always been very low, and thus not an effective way to police the quality of scientific evidence.

Palmer was a rather late substitute for expert witnesses previously listed by the plaintiff. Alan Kimball Done, a pediatrician, had been the main warhorse of the Bendectin plaintiffs, but he was withdrawn by plaintiff’s counsel after he was found to have committed perjury about his academic credentials in another Bendectin case.[2]

Plaintiff also needed to drop another expert witness, William Griffith McBride, who had been a star in plaintiff’s counsel’s stable. McBride helped show the teratogenicity of thalidomide in the early 1960s,[3] and his work in the Bendectin litigation gave these dodgy cases some patina of respectability. In 1988, however, McBride was accused of fraud, for which he would eventually lose his medical license.[4] McBride also chose, rather improvidently, to sue journalists, journals, and Merrell Dow executives, for reporting his rather extensive fees, only to lose that litigation.[5] When plaintiff’s counsel withdrew McBride, plaintiff was left with only Dr. Palmer to serve as plaintiff’s sole expert witness on both general and specific causation.

At the time that the first Daubert motion was filed, manufacturer Merrell Dow had voluntarily withdrawn Bendectin from the market, without any suggestion from the FDA that this action was necessary or in the public interest. The manufacturer had also enjoyed considerable success in court. The company had tried a case that consolidated the general causation claims of over 800 plaintiffs, to a defense jury verdict, in 1985, before Chief Judge Carl Rubin, of the Southern District of Ohio.[6] Despite some isolated trial losses, the company was vindicated in three federal circuits at the time its lawyers filed the “Daubert” motion.[7] The First, Fifth, and District of Columbia Circuits of the United States Court of Appeals, had all held that the plaintiffs’ case was legally insufficient to sustain a verdict against the defendant, or that the expert testimony involved was inadmissible.

In the Daubert case, Merrell Dow Pharmaceuticals was represented by the law firm Dickson, Carlson & Campillo. The important task of drafting the motion for summary judgment landed on the desk of a first year associate, Pamela Yates, who is now a partner at Arnold & Porter. Given that Merrell Dow had succeeded in other appellate courts, the task may have seemed straight forward, but the legal theories were actually all over the map.

The first Daubert motion was not styled as a motion to exclude expert witness opinion testimony, but rather as a motion for summary judgment, pursuant to Federal Rule of Civil Procedure 56, on the issue of causation. Merrell Dow’s supporting brief did not clearly invoke the distinction between general and specific causation, which distinction was not widely drawn until later in the 1990s. The supporting brief implicitly addressed both general and specific causation.

At the time that the first Daubert motion was made, there was no clear consensus of precedent that identified the source of support for a trial court’s ruling peremptorily on a weak evidentiary display on the causation issue. The evidence supporting the defense expert witnesses’ opinion that Bendectin had not been shown to cause birth defects generally and limb reduction defects specifically was strong. For all major congenital defects, there had been no change in overall incidence for the years in which Bendectin was marketed. Such an ecological argument usually has no validity, but in the case of Bendectin, for several years, roughly half of all pregnant women used the medication. When the medication was abruptly withdrawn,[8] not because of the science but because of the cost of the litigation, the rate of birth defects remained unaffected. The great majority of birth defects have no known cause, and there was no scientific consensus that Bendectin caused birth defects; indeed by 1989, the nearly universal consensus was that Bendectin did not cause birth defects.[9]

There were also many analytical epidemiologic studies, which both individually or in combination failed to support a conclusion of causation.

In the face of the defense’s affirmative evidence, the plaintiff relied upon a potpourri of evidence:

1) chemical structure activity analysis;

2) in vitro (test tube) studies;

3 ) in vivo studies (animal teratology) studies; and

4) reanalysis of epidemiology studies.

Plaintiff’s lead counsel Barry Nace[10] had concocted this potpourri approach, which he called “mosaic theory,” and which might more aptly be called the tsemish or the shmegegge theory.[11] Whatever Nace called it, he fed it to his expert witness to argue that:

“Like the pieces of a mosaic, the individual studies showed little or nothing when viewed separately from one another, but they combined to produce a whole that was greater than the sum of its parts: a foundation for Dr. Done’s opinion that Bendectin caused appellant’s birth defects.”[12]

Although philosopher Harry Frankfurt had not yet written his seminal treatise on the subject, most courts saw that this was bullshit, which tends to result “whenever a person’s obligations or opportunities to speak about some topic exceed his knowledge of the facts that are relevant to that topic.”[13]

In addition to favorable opinions from the First, Fifth, and District of Columbia Circuits, Merrell Dow had a favorable Zeitgeist working in its favor. The plaintiff-friendly influential judge, Judge Jack Weinstein, had rolled up his sleeves and taken a hard look at the plaintiffs’ scientific evidence in the Agent Orange litigation. Judge Weinstein found that evidence wanting in an important opinion in 1985.[14] Although the alleged causal agent in Agent Orange was not Bendectin, Judge Weinstein recognized that epidemiological studies were, in a similar medico-legal context, “the only useful studies having any bearing on causation.[15] Judge Weinstein relied heavily upon Federal Rule of Evidence Rule 703, which governed what inadmissible studies expert witnesses could rely upon, to whittle down the reliance list of plaintiffs’ expert witnesses before declaring their opinions too fragile to support a reasonable jury’s verdict in favor of plaintiffs.

More generally, discerning members of the legal system were reaching the end of their tolerance for the common law laissez-faire approach to expert witness evidence. In 1986, the Department of Justice issued a report that explicitly called for meaningful judicial gatekeeping of expert witnesses.[16] And in that same year, 1986, Judge Patrick Higginbotham wrote an influential opinion, in which he warned that expert witness opinion testimony was out of control, with expert witnesses becoming mouth pieces for the lawyers and advocates of policy beyond their proper role. Judge Higginbotham observed that trial judges (with support from appellate court judges) had a duty to address the problem by policing the soundness of opinions proffered in litigation, and to reject the system’s reliance upon expert witnesses simply because they “say[] it is so:”[17]

“we recognize the temptation to answer objections to receipt of expert testimony with the short hand remark that the jury will give it ‘the weight it deserves’. This nigh reflective explanation may be sound in some cases, but in others it can mask a failure by the trial judge to come to grips with an important trial decision. Trial judges must be sensitive to the qualifications of persons claiming to be experts … . Our message to our able trial colleagues: It is time to take hold of expert testimony in federal trials.”[18]

Although Merrell Dow had a substantial tailwind behind its motion for summary judgment, there was no one clear theory upon which it could rely. Some of the Bendectin appellate court opinions were based upon the insufficiency of the plaintiffs’ expert witness evidence, on the basis of the entire record after trial. The evidence in Daubert was virtually the same if not more restricted than what was of record in some of those appellate court cases. The ecological evidence was clear.

Some of the judgments relied upon by Merrell Dow were based upon the Frye test, and some were based upon Rule 703, which addresses what kinds of otherwise inadmissible evidence expert witnesses may rely upon in formulating their opinions. Finally, some courts, such as Fifth Circuit in In Re Air Crash Disaster at New Orleans, were beginning to see Rule 702 as the source of their authority to control wayward expert witness opinion testimony.

Merrell Dow advanced multiple lines of analyses to show that plaintiffs cannot establish causation based upon the then current scientific record. The first Daubert motion had no clear line of authority, and so, understandably, it cast a wide net on all available potential legal rules and doctrines to oppose the plantiff’s potpourri Bendectin causation theory. The motion harnessed precedents based upon sufficiency of the plaintiffs’ proffered expert witness, Federal Rules 702 and 703, as well as the 1923 Frye case.[19]

The cases that invoked Frye doctrine presented several interpretative problems. Frye was a criminal case that prohibited expert witnesses from testifying about their interpretations of the output of a mechanical device. The Frye case’s insistence upon general acceptance, when imported into a causation dispute in a tort case, was ambiguous as to what exactly had to be generally accepted: the specific causal claim, or the method used to reach the causal claim, or the method used as applied to the facts of the case. Furthermore, Frye’s requirement of general acceptance was not explicitly incorporated into either Rule 702 or 703, when promulgated in 1975.[20]

Merrell Dow had ample evidence that there was no general acceptance of the plaintiff’s causal claim, but its counsel also showed that by applying generally accepted methodology, scientists could not reach the plaintiff’s causal conclusion, and no scientist outside of the litigation had done so. In particular, there was general acceptance of the propositions that non-human in vivo and in vitro teratology experiments have little if any predictive ability for human outcomes. Because randomized controlled trials were never an option for testing human teratogenicity, observational epidemiology was required, and the available studies were largely exonerative. Only by post-publication data dredging and manipulation was plaintiffs’ expert witness Palmer (following what Shann Swan had done in previous cases) able to raise questions about possible associations. Plaintiff’s expert witness Palmer could not show that these manipulations were a generally accepted method for interpreting or re-analyzing published studies.

In its last point, the first Daubert motion also maintained that the standard for medical causation required that the relevant relative risk exceed two.[21] As noted, the brief did not distinguish general from specific causation, a distinction that had not entered the legal lexicon fully in 1989. The brief’s citation to swine-flu cases, however, clarifies the nature of Merrell Dow’s argument. In the swine-flu litigation, the United States government assumed liability for adverse effects of a vaccine for swine flu. The government recognized that within a certain time window after vaccination, patients had more than a doubled risk of Guillain-Barré syndrome (GBS), an autoimmune neurological condition. The government refused compensation for claimants outside that window. Merrell Dow relied heavily upon one swine flu case, Cook v. United States, which articulated and applied the principle:

“Wherever the relative risk to vaccinated persons is greater than two times the risk to unvaccinated persons, there is a greater than 50% chance that a given GBS case among vaccinees of that latency period is attributable to vaccination, thus sustaining plaintiff’s burden of proof on causation.”[22]

In other words, the government had conceded that the swine-flu vaccine could cause GBS in some temporal situations, but not others. The magnitude of the causal association had been quantified in relative risk terms by epidemiologic studies. Only for those claimants vaccinated in time windows with relative risks greater than two could courts conclude that GBS was, more likely than not, caused by vaccination.

Unlike the federal government in the swine-flu GBS litigation, Merrell Dow was not, however, conceding general causation for any exposure scenarios. The first Daubert motion can only be read to deny general causation, but to explain further that even if the court were to assume, arguendo, that Bendectin causes limb reduction deficits based upon Palmer’s schmegegge and Swan’s re-jiggered risk ratios, that there would still be no proper inference that Bendectin more likely than not caused Jason Daubert’s birth defects.

In response to these arguments, the plaintiff’s counsel argued their mosaic, potpourri, schmegegge theories. Although plaintiffs were down to Dr. Palmer, they filed transcripts and affidavits from a host of other expert witnesses, from previous Bendectin cases.

As for the legal rules of decision, Barry Nace, on behalf of plaintiffs, argued that Rule 703 had “absorbed” the Frye rule. Having been shown to be qualified under the minimal standard of Rule 702, these expert witnesses then satisfied Rule 703 by relying upon “scientific evidence” of the sort that experts in their field rely upon, even if other scientists would not rely upon such evidence in support of a conclusion. Otherwise those expert witnesses were unrestrained by the law, and they were free to assess their relied upon facts and data as sufficient to show that Bendectin probably causes birth defects and that Bendectin caused Jason Daubert’s birth defects. Nace argued that as long as expert witnesses, properly qualified, offered relevant opinions, based upon “things of science,” they could opine that the earth was flat, and it was for the jury to sort out whether to believe them.

Judge Earl Gilliam found Nace’s position untenable, and granted summary judgment later in 1989.[23] Interestingly Judge Gilliam’s opinion in the district court never cited Federal Rule of Evidence 702. Instead, the opinion pointed to Rule 703, as restricting evidence, even if “science,” unless the proponent showed that the underlying principle had gained general acceptance in the relevant field.[24] Opinions not based upon facts or data “of a type reasonably relied upon by experts in the particular field” would be confusing, misleading, and unhelpful, and thus inadmissible. The reference to helpfulness might perhaps be taken as an implicit invocation of Rule 702.

Judge Gilliam had the benefit of the Circuit decisions in Brock, Richardson, and Lynch, with their various holdings of insufficiency or inadmissibility of plaintiffs’ expert witness evidence. In particular, Judge Gilliam cited Brock for the proposition that trial courts must “critically evaluate the reasoning process by which the experts connect data to their conclusions in order for courts to consistently and rationally resolve the disputes before them.”[25] Following Judge Weinstein on Agent Orange, and the previous federal decisions on Bendectin, Judge Gilliam observed that causation in the Bendectin cases could be established, under the circumstances of plaintiffs’ evidentiary display, only through reliance upon epidemiologic evidence. Dr. Done’s schmeggege, concocted as it was by Barry Nace, would not get plaintiffs to a jury.

Judge Gilliam went further to point out that some of plaintiffs’ proffer did not even purport to claim causation. Shanna Swan’s prior testimony asserted that Bendectin was “associated” with limb reduction. Jay Glasser, a specialist in biostatistics, epidemiology and biometry had opined that “Bendectin is within a reasonable degree of epidemiological certainty associated with congenital disorders, including limb defects.” Dr. Johannes Thiersch, a specialist in pathology and pharmacology, proclaimed that “structure analysis” was “of great interest.”[26] In other words, there was a good deal of true, true, but immaterial opinion in what Mr. Nace had thrown over the transom, in opposition to the motion for summary judgment.

Nace appealed, and the Daubert case was argued to the Ninth Circuit in 1991. In a short opinion by Judge Kozinski, the appellate court affirmed the judgment below.[27] The affirmance did not mention Rule 702; rather it relied upon the decisions of other Circuits, in which the plaintiffs’ evidentiary display had been found insufficient to sustain a reasonable jury verdict.

Judge Kozinski’s opinion tilted towards Rule 703 and the Frye standard in citing to cases that stated, based upon Frye, that expert witnesses must use generally accepted techniques from the scientific community. As a legal determination, the determination of general acceptance vel non was a legal determination reviewable de novo. For its de novo decision on general acceptance, the Ninth Circuit relied upon the cases coming from the First, Fifth, and District of Columbia Circuits,[28] and of course, the record below.

By 1991, another Circuit, the Third, had weighed in on the same evidentiary display, when it reversed summary judgment for Merrell Dow, and remanded for reconsideration under the Third Circuit’s approach to Rule 702. Judge Kozinski declared that the Third Circuit’s approach was not followed in the Ninth Circuit, and proceeded to ignore the DeLuca case.[29]

Judge Kozinski treated the insufficiency and the invalidity of the Nace/Done schmeggege theory as legal precedent, and thus the court’s opinion gave very little attention by way of expository description or explanation of the problems with the four factors (in vitro, in vivo, structure analysis, and re-analysis of epidemiologic studies). As Judge Kozinski put the matter:

 “For the convincing reasons articulated by our sister circuits, we agree with the district court that the available animal and chemical studies, together with plaintiffs’ expert reanalysis of epidemiological studies, provide insufficient foundation to allow admission of expert testimony to the effect that Bendectin caused plaintiffs’ injuries.”[30]

And thus, summary judgment was proper in Daubert. Judge Kozinksi, like Judge Gilliam in the district court, never reached the specific causation argument that involved risk ratios less than two.

Some of the Circuit court cases relied upon by Judge Kozinski delved into the invalidity of these methods for determining the causes of human birth defects. The Lynch decision explored in some detail the Shanna Swan made-for-litigation rejiggering of a study based upon data from the Metropolitan Atlanta Congenital Defects Program, which included a challenge to whether it could be reasonably relied upon (Rule 703), as well as its pretense to support a scientific conclusion (Rule 702).[31] Later commentators would skirt the validity issue by asserting that re-analysis, in the abstract, is not impermissible or invalid, without addressing the specific issues discussed in the reported decisions. Other commentators have misrepresented Swan’s re-analysis as a meta-analysis, which it was not.

Some commentators have complained that the defense in Daubert made too much of the lack of statistical significance. Their complaint, in the abstract, might have some salience. In some contexts, an isolated and elevated risk ratio greater or less than one may well have important information, even if the p-value is a bit above 0.05. The lack of statistical significance at the conventional five percent, however, conveys important information about the finding’s imprecision, especially when there was a large dataset to evaluate. In 1994, a meta-analysis was published that found a summary estimate of all birth defects in the available epidemiologic studies to be an odds ratio of 0.95 (95% C.I., 0.88-1.04), and the summary estimate for limb reduction defects to be an odds ratio 1.12 (95% C.I., 0.83-1.48).[32]

[1] Defendant’s Memorandum of Points and Authorities in Support of Its Motion for Summary Judgment on the Issue of Causation, Daubert v. Merrell Dow Pharms., Inc., Case No. 84-2013-G(I) (S.D. Cal. Aug. 2, 1989). The motion was made in a companion case before Judge Gilliam as well, Schuller v. Merrell Dow Pharms., Inc., Case No. 84-2929-G(I). The first Daubert motion may not have been the first one drafted. The linked brief is the first one as filed.

[2] See Oxendine v. Merrell Dow Pharms., Inc., 563 A.2d 330 (D.C. Ct. App. 1989).

[3] William Griffith McBride, Thalidomide and Congenital Abnormalities, 278 LANCET 1358 (1961).

[4] William Griffith McBride, McBride criticizes inquiry, 336 NATURE 614 (1988); Norman Swan, Disciplinary tribunal for McBride, 299 BRIT. MED. J. 1360 (1989); G. F. Humphrey, Scientific fraud: the McBride case, 32 MED. SCI. LAW 199 (1992); Mark Lawson, McBride found guilty of fraud, 361 NATURE 673 (1993); Leigh Dayton, Thalidomide hero found guilty of scientific fraud, NEW SCI. (Feb.27, 1993); William McBride: alerted the world to the dangers of thalidomide in fetal development, 362 BRIT. MED. J. k3415 (2018).

[5] McBride v. Merrell Dow & Pharms., Inc., 800 F.2d 1208 (D.C. Ct. App. 1986). McBride ultimately failed against all his litigation targets.

[6] See In Re Richardson-Merrell. Inc. Bendectin Prods. Liab. Litig., 624 F.Supp. 1212 (S.D. Ohio 1985); aff’d sub nom. In re Bendectin Litig., 857 F.2d 290 (6th Cir. 1988); cert. denied, 488 US 1006 (1989).

[7] Brock v. Merrell Dow Pharmaceuticals Inc., 874 F.2d 307 (5th Cir. 1989); Richardson y. Richardson-Merrell, 857 F.2d 823 (D.C. Cir. 1988); Lynch v. Merrell-National Labs., 830 F.2d 1190 (1st Cir. 1987) (affirming grant of summary judgment).

[8] US Food & Drug Admin., Determination That Bendectin Was Not Withdrawn from Sale for Reasons of Safety or Effectiveness, 64 FED. REG. 43190–1 (1999).

[9] Brief at 3-4.

[10] Barry Nace was one of the lead plaintiffs’ counsel in the Bendectin litigation, and he represented the Daubert family. Nace was also formerly President of the lawsuit industry’s principal lobbying organization, the American Trial Lawyers Association (now the AAJ). See also In re Barry J. Nace, A Member of the Bar of the District of Columbia Court of Appeals (Bar Registration No. 130724), No. 13–BG–1439, Slip op. (Sept. 4, 2014), available at <https://www.dccourts.gov/sites/default/files/pdf-opinions/13-BG-1439.pdf>, last visited on Feb. 8, 2026.

[11] See Michael D. Green, Pessimism about Milward, 3 WAKE FOREST J. L & POL’Y 41, 63 (2013) (paraphrasing Nace as describing the mosaic theory as “[d]amn brilliant, and I was the one who thought of it and fed it to Alan [Done].”).

[12] Id. at 61 (2013) (citing Oxendine v. Merrell Dow Pharm., Inc., 506 A.2d 1100, 1110 (D.C. 1986).

[13] Harry Frankfurt, ON BULLSHIT 63 (2005).

[14] In re “Agent Orange” Prod. Liab. Litig., 611 F. Supp. 1223 (E.D.N.Y. 1985), aff’d, 818 F.2d 187 (2d Cir. 1987), cert. denied, 487 U.S. 1234 (1988).

[15] Id. at p. 1231.

[16] United States Dep’t of Justice, Tort Policy Working Group, Report of the Tort Policy Working Group on the causes, extent and policy implications of the current crisis in insurance availability and affordability at 35 (Report No. 027-000-01251-5) (Wash. DC 1986), available at https://archive.org/details/micro_IA41152903_0369.

[17] In Re Air Crash Disaster at New Orleans, 795 F.2d 1230, 1233-34 (5th Cir. 1986).

[18] Id. at 1233-34.

[19] The Brief, at 2, cited United States v. Kilgus, 571 F.2d 508, 510 (9th Cir. 1987) (citing Frye).

[20] An Act to Establish Rules of Evidence for Certain Courts and Proceedings. Pub. L. 93–595, 88 Stat. 1926 (1975).

[21] Brief at 17.

[22] 545 F.Supp. 306, 308 (N.D. Cal. 1982). See generally Richard E. Neustadt & Harvey V. Fineberg, THE SWINE FLU AFFAIR: DECISION-MAKING ON A SLIPPERY DISEASE (Nat’l Acad. Sci. 1978).

[23] Daubert v. Merrell Dow Pharms., Inc., 727 F.Supp. 570 (S.D. Cal. 1989).

[24] Id. at 571, citing United States v. Kilgus, 571 F.2d 508, 510 (9th Cir.1978).

[25] Id. at 572 (citing Brock, 874 F.2d at 310).

[26] Id. at 574. The use of “association” was at best ambiguous, because it begged the question whether it as an association that was “clear cut” (reasonably free from bias and confounding), and beyond that which we would care to attribute to chance.

[27] Daubert v. Merrell Dow Pharms., Inc., 951 F.2d 1128 (9th Cir. 1991).

[28] Brock v. Merrell Dow Pharms., Inc., 874 F.2d 307, modified, 884 F.2d 166 (5th Cir.1989), cert. denied, 494 U.S. 1046 (1990); Richardson v. Richardson–Merrell, Inc., 857 F.2d 823 (D.C.Cir.1988), cert. denied, 493 U.S. 882 (1989); Lynch v. Merrell–National Labs., 830 F.2d 1190 (1st Cir.1987).

[29] DeLuca v. Merrell Dow Pharmaceuticals, Inc., 131 F.R.D. 71 (D.N.J.) (granting summary judgment), rev’d and remanded, 911 F.2d 941 (3d Cir.1990). On remand, the district court entered summary judgment on the alternative reasoning of Rule 702, as interpreted by the Third Circuit. DeLuca v. Merrell Dow Pharms., Inc., 791 F.Supp. 1042, 1048 (D.N.J. 1992) (re-entering summary judgment after considering Rule 702), aff’d, 6 F.3d 778 (3d Cir.1993) (per curiam), cert. denied, 510 U.S. 1044 (1994).

[30] Daubert v. Merrell Dow Pharms., Inc., 951 F.2d 1128, 1131 (9th Cir. 1991).

[31] Lynch v. Merrell–National Labs., 830 F.2d 1190, 1194-95 (1st Cir.1987).

[32] Paul M. McKeigue, Steven H. Lamm, Shai Linn & Jeffrey S. Kutcher, Bendectin and Birth Defects: I. A Meta-Analysis of the Epidemiologic Studies, 50 TERATOLOGY 27 (1994). This meta-analysis made no correction for multiple comparisons in examining many different types of birth defects.

Posted in Causation, Expert Witnesses, Frye, Meta-analysis, Peer Review, Rule 702, Rule 703, Scientific Evidence | No Comments »

The 4th Reference Manual’s Treatment of Genetic Causes of Disease

January 23rd, 2026

After checking to see whether the new Reference Manual on Scientific Evidence[1] attended to some long overdue corrections, I turned my attention to the substance of the chapter on epidemiology. A cursory comparison between the third[2] and fourth[3] editions of the epidemiology chapter in the Reference Manual a lot of carry over from the third edition, some change in authorship, and at least one interesting change.

The two lawyer authors, Steve Gold and Michael Green, remain, but the authors with reasonable pretense to subject-matter expertise have changed. Gold and Green are both law professors with a long history of commenting on American tort and evidence law. Both are aligned with the lawsuit industry. Previous epidemiology authors, Daryl Michal Freedman and Leon Gordis are now gone from the chapter. Leon Gordis, who had been a chairman of the department of epidemiology, in the Bloomberg School of Public Health, Johns Hopkins University, died in September 2015, after the third edition was published. Daryl Michal Freedman, who been the other subject-matter expert on the third edition’s chapter on epidemiology, has been an epidemiologist with the Biostatistics Branch of the National Cancer Institute, for many years. It is not clear why he left the project.

Replacing Gordis and Freedman are Jonathan Chevrier and Brenda Eskenazi. Chevrier is an associate professor on the faculty of medicine, in the department of epidemiology, in McGill University. The focus of his work is on “common environmental contaminants,” and the role in the development and health of children. Brenda Eskenazi is professor emerita, in the University of California Berkeley School of Public Health, where she is the Director of the Center for Environmental Research and Children’s Health. Eskenazi is a member of a dodgy group known as the Collegium Ramazzini, which was responsible for staging an ex parte presentation of plaintiffs’ expert witnesses to judges presiding in asbestos litigation.[4] Eskenazi was not, however, a member of the Collegium at the time the group conspired with the late Irving Selikoff to pervert the course of justice in American asbestos litigation.

The second significant change is substantive; the fourth edition has added a new subsection to the epidemiology chapter. Comparing the texts of the third and fourth editions of this chapter reveals a new subheading in the new edition:[5]

Genetic and Molecular Epidemiologic Studies

Alas, there is not as much substance to the new subsection, which is less than four pages. Lawyers in the trenches might well have hoped for more substantive treatment of genetic epidemiology, and genetic causation. The chapter’s authors explain their abbreviated treatment with the comment:

“Although commentators have long forecast that the output of genetic and molecular epidemiology would revolutionize causal proof, as of this writing few judicial opinions have addressed these types of studies, and it is far from clear that a revolution is in the offing.”[6] 

The chapter authors are correct that some authors in the past proffered unrealistic predictions of how genetics would supplant correlational studies. Nonetheless, this area has not been as quiescent as the authors’ parsimonious treatment would suggest.

On the question of how prevalent are genetic causation issues, whether raised by plaintiffs or defendants, the chapter might have benefitted from the contributions of a practicing lawyer. Genetic issues come up with some frequency in the litigation of cases involving mesothelioma. The days of plaintiffs who had 30 years of amphibole asbestos exposure in the workplace are largely over. Today’s cases involve little to no exposure, and it stands to reason that the origins of the recently diagnosed cases are different from those diagnosed in the 1970s and 1980s.[7] Genetic cause of mesothelioma is a salient current issue that is passed over in this new Reference Manual.

The authors acknowledge a single birth defects case in which genetic causation was litigated,[8] which was already old news when the last edition of the Manual was published. There are now many more reported cases that cry out for discussion in this under-covered area of the Manual.[9] There are also many cases not reported that have turned on genetic issues. For instance, in some cancer and birth defect cases, the existence of a highly penetrant genetic mutation that could explain the occurrence of a disease completely raises a serious question whether the plaintiff who fails to test for the mutation can possibly have carried his burden of proof.[10] And then there are myriad cases in which the parties have engaged in motion practice, sometimes extended, over access to genetic testing materials.

Genetic issues have arisen in the litigation of high-profile general causation disputes. For instance, the failure to control for genetic effects in epidemiologic studies was a significant issue in the acetaminophen-autism litigation, with both sides presenting geneticists to explain whether the relevant studies were undermined by failure to control for genetic effects.[11]

In the Manual’s epidemiology chapter’s new section on genetics, the authors describe some basic terms and explain that genetic epidemiology may provide evidence for, or against, claims of health effects. The authors’ views come through most clearly in the following short passage:

“Alternatively, genetic epidemiology may reveal associations between genetic variations and a plaintiff’s disease, raising the issue of whether or not a genetic variation may be a competing cause of the disease. This requires assessment of whether the gene–disease association is causal in a general sense, whether it acts independently of the exposure, and whether it is a competing cause in the plaintiff’s specific instance. The extreme, though not typical, example would be a health outcome or disease entirely determined by genetics, 55 as is the case with sickle cell anemia.56”[12]

The authors never explain or defend their claim that cases involving diseases caused entirely by genetics are “extreme” and “not typical.” At several points, the authors emphasize that gene-environment interactions are the more prevalent determinants of diseases.[13] If we were to catalog the currently known genetic determinants of diseases, the authors may be correct on a percentage basis, but the issue in any given case is whether the disease or harm claimed by the plaintiff is one of the “extreme” cases of complete genetic causation, or an instance of genetic susceptibility. The authors’ generalization, even if it were correct, would not be very helpful or informative for any specific case.

Perhaps even more important for lawyers, there is a substantive issue on which the new chapter manages to provide confusing guidance. The epidemiology chapter appears to create a false dichotomy between rare, highly penetrant genetic mutations that are uncommon causes of certain diseases, and the more prevalent genetic mutations and polymorphisms that leave persons more susceptible to the deleterious effects of exogenous exposures to toxic chemicals.[14] There is, however, another scenario omitted in the chapter’s discussion of genetic causation. Genetic mutations and polymorphisms may leave persons susceptible to normal, endogenous chemicals, stochastic cellular events, and biological processes that result in diseases such as cancers. In other words, the knee-jerk reflex to invoke exogenous, external toxic chemical exposures promotes a false dichotomy and obscures the obvious implication that susceptibility mutations and polymorphisms may lead to cancer without environmental exposures to harmful chemicals.[15]

The number of endogenous events leading to DNA alterations is enormous, and requires us to rethink the mantra that attributes chronic diseases to gene-environment interaction. At the very least, we need to stop thinking of “environment” as chemical exposures from without ourselves. The epidemiology chapter authors, like many writers, point to external chemical exposures as the culprits in gene-environment reactions, but they ignore the normal, endogenous events that lead to DNA damage, for which genetic susceptibility may be relevant. Mutations that result in increased susceptibility to cancer may affect DNA alterations from both endogenous and metabolic factors as well as from exposures to external chemicals.

Ignorance is never a good thing, and the chapter does the bar and bench a disservice in not adequately exploring genetic susceptibility in view of both exogenous and endogenous exposures that may be responsible for chronic diseases, such as cancers.

[1] National Academies of Sciences, Engineering, and Medicine & Federal Judicial Center, REFERENCE MANUAL ON SCIENTIFIC EVIDENCE (4th ed. 2025) (cited as RMSE 4th ed.)

[2] Michael D. Green, D. Michal Freedman & Leon Gordis, Reference Guide on Epidemiology, 549, in RMSE 3rd ed.

[3] Steve C. Gold, Michael D. Green, Jonathan Chevrier, & Brenda Eskenazi, Reference Guide on Epidemiology, in RMSE 4th ed.

[4] See In re School Asbestos Litigation, 977 F.2d 764 (3d Cir. 1992). See also Cathleen M. Devlin, Disqualification of Federal Judges – Third Circuit Orders District Judge James McGirr Kelly to Disqualify Himself So As To Preserve ‘The Appearance of Justice’ Under 28 U.S.C. § 455 – In re School Asbestos Litigation (1992), 38 VILL. L. REV. 1219 (1993); Bruce A. Green, May Judges Attend Privately Funded Educational Programs? Should Judicial Education Be Privatized?:  Questions of Judicial Ethics and Policy, 29 FORDHAM URB. L. J. 941, 996-98 (2002).

[5] Steve Gold, et al., Reference Guide on Epidemiology, at 914, in RMSE 4th ed.

[6] Id. at 916.

[7] ToxicoGenomica, The Litigator’s Guide to Using Genomics in a Toxic Tort Case (2018).

[8] Id. at 917 & n.55 (citing Bowen v. E.I. Du Pont de Nemours & Co., No. CIV.A. 97C-06-194 CH, 2005 WL 1952859 (Del. Super. Ct. June 23, 2005), aff’d, 906 A.2d 787 (Del. 2006) (discussing the importance of a test for a genetic mutation, which was the defense’s alternative causation theory to plaintiff’s claim that a toxic exposure caused the birth defect at issue). The authors fail to mention that the Bowen case was actually dismissed.

[9] See, e.g., Oliver v. Sec’y Health & Human Servs., 900 F.3d 1357 (Fed. Cir. 2018); Ortega v. United States, 2021 WL 4477896, 2021 U.S. Dist. LEXIS 188969 (N.D.Ill. Sept. 30, 2021); Vanslembrouck ex rel. Braverman v. Halperin, 2014 WL 5462596 (Mich. App. 2014).

[10] See, e.g., Halter v. Boehringer Ingelheim Pharms. Inc., no. 2023-L-001382, Cir. Ct. Cook Cty., Illinois, jury verdict (Aug. 27, 2025) (defense verdict in colorectal cancer case in which plaintiff failed to test for genetic mutation); see also Lauraann Wood, Boehringer Wins Another Zantac Cancer Trial In Illinois, LAW360, Chicago (Aug. 27, 2025).

[11] See, e.g., In re Acetaminophen – ASD-ADHD Prods. Liab. Litig., 707 F.Supp.3d 309, 320  (S.D.N.Y. 2023).

[12] Id. at 916-17 (emphasis added).

[13] Id. at 915.

[14] See, e.g., id. at 967n.190, citing McMillan v. Dep’t of Veterans Affairs, 294 F. Supp. 2d 305, 312 (E.D.N.Y. 2003) (“It is generally accepted that genetic susceptibility plays a key role in determining the adverse effects of environmental chemicals. . . . [I]f polymorphisms of the gene encoding the AhR [protein] exist in humans as they do in laboratory animals, some people would be at greater risk or at lesser risk for the toxic and carcinogenic effects of TCDD [dioxin].”).

[15] See Edward J. Calabrese, Changing the paradigm: The biggest polluter and threat to your health is your body, J. OCCUP. & ENVT’L HYG. (2025), published on-line, ahead of print.

Posted in Causation, Reference Manual on Scientific Evidence | No Comments »

Signature Diseases in the New Reference Manual

January 16th, 2026

The third edition of the Reference Manual on Scientific Evidence[1] had some problems with its discussion of so-called signature diseases.[2] There was a distinct need for the  epidemiology chapter in particular to improve in its fourth edition[3] on the issue of so-called signature diseases, diseases caused by only a single cause. The third edition carved out a limited exception to its questionable generalization that epidemiology had nothing useful to say about specific causation by stating that some diseases do not occur without exposure to a specific chemical or substance.

The new, fourth edition carries forward its assertion that “[t]here are some diseases that do not occur without exposure to an agent; these are known as signature diseases.”[4] And in a footnote, the authors of the epidemiology chapter, fourth edition, attempt to explain:

“There are, however, some diseases that do not occur without exposure to a given toxic agent. This is the same as saying that the toxic agent is a necessary cause for the disease, and the disease is sometimes referred to as a signature disease (also, the agent is pathognomonic) because the existence of the disease necessarily implies the causal role of the agent. Two examples are asbestosis, which is a signature disease for asbestos, and vaginal adenocarcinoma (in young adult women), which is a signature disease for in utero DES exposure. See Kenneth S. Abraham & Richard A. Merrill, Scientific Uncertainty in the Courts, in Issues Sci. & Tech. 93, 101 (1986).”[5]

Much of this language in the footnote is repeated from the third edition, as is the citation to the article by Abraham and Merrill. That article was written by lawyers, not scientists, and is now 40 years old, inaccurate and out of date.

With respect to asbestosis, the epidemiology chapter is correct, at least in part. By definition, only asbestos can cause asbestosis, but asbestosis presents clinically in ways that are indistinguishable in many cases from idiopathic pulmonary fibrosis and other interstitial fibrotic diseases of the lungs. Over the years, the diagnostic criteria for asbestosis have changed, but these criteria have always had a specificity and sensitivity less than 100%. Saying that a case of asbestosis must have been caused by asbestos begs the clinical question whether the case really is asbestosis. The situation might be clearer for a pathology diagnosis of asbestosis, but even then there is often the problem of coincidental findings of asbestos bodies in the presence of interstitial fibrosis from another cause.

On the other hand, the chapter’s characterization of vaginal adenocarcinoma as a signature disease of in utero DES exposure is clearly not correct.  Although this cancer in young women is rather rare, there is a baseline risk that allows the calculation of relative risks for young women exposed in utero.[6] In older women, the relative risks are lower because the baseline risks are higher, and because the effect of DES is diminished for older onset cases.[7] The disease, however, was known before the use of DES in pregnant women, which began after World War II,[8] and thus not an apt or accurate example of signature disease.

The Reference Manual should really not weigh in on controversies that may arise in courtroom litigations, unless it has a very solid basis. Here the chapter on epidemiology cited to a decades old article, by lawyers, on a technical topic. The proposition about DES was readily falsified by a wee bit of research in PubMed.

[1] National Academies of Sciences, Engineering, and Medicine & Federal Judicial Center, REFERENCE MANUAL ON SCIENTIFIC EVIDENCE (3rd ed. 2011) (cited as RMSE 3rd ed.)

[2] See Schachtman, Reference Manual – Desiderata for 4th Edition – Part I – Signature Diseases, TORTINI (Jan. 30, 2023); see also Reference Manual on Scientific Evidence v4.0 (Feb. 28, 2021); Reference Manual on Scientific Evidence – 3rd Edition is Past Its Expiry (Oct. 17, 2021).

[3] National Academies of Sciences, Engineering, and Medicine & Federal Judicial Center, REFERENCE MANUAL ON SCIENTIFIC EVIDENCE (4th ed. 2025) (cited as RMSE 4th ed.).

[4] RMSE 4th ed. at 927-28 n.90.

[5] RMSE 4th at 990 n.274, citing Kenneth S. Abraham & Richard A. Merrill, Scientific Uncertainty in the Courts, 2 ISSUES SCI. & TECH. 93, 101 (Winter 1986). Thankfully, the new epidemiology chapter did not put its finger on the scale about the now discredited view that mesothelioma is a signature disease of asbestos exposure. See Michele Carbone, Harvey Pass, et al., “Medical and Surgical Care of Patients With Mesothelioma and Their Relatives Carrying Germline BAP1 Mutations,” 17 J. THORACIC ONCOL. 873 (2022). See also Mitchell Cheung, et al., Novel LRRK2 mutations and other rare, non-BAP1-related candidate tumor predisposition gene variants in high-risk cancer families with mesothelioma and other tumors, 30 HUMAN MOL. GENETICS 1750 (2021); Thomas Wiesner, et al., “Toward an Improved Definition of the Tumor Spectrum Associated With BAP1 Germline Mutations,” 30 J. CLIN. ONCOL. e337 (2012); Alexandra M. Haugh, et al., Genotypic and Phenotypic Features of BAP1 Cancer Syndrome: A Report of 8 New Families and Review of Cases in the Literature, 153 J.AM. MED. ASS’N DERMATOL. 999 (2017).

[6] See, e.g., Kadir Güzin, et al., “Primary clear cell carcinoma of the vagina that is not related to in utero diethylstilbestrol use,” 3 GYNECOL. SURG. 281 (2006).

[7] Janneke Verloop, et al., Cancer risk in DES daughters, 21 CANCER CAUSES & CONTROL 999 (2010).

[8] See Risk Factors for Vaginal Cancer, American Cancer Soc’y website (last visited Jan. 16, 2026).

Posted in Causation, Reference Manual on Scientific Evidence | No Comments »

A New Year, A New Reference Manual

January 5th, 2026

The fourth edition of the Reference Manual on Scientific Evidence was quietly released in the waning hours of 2025, in the twilight of American democracy.[1] The Manual had been slated to be published in 2023, but that date slid to 2024, and then to 2025.  Perhaps the change in directorship of the Federal Judicial Center slowed things up. (Judge Robin Rosenberg of Zantac fame is now the Director)

The new volume is available for download at:

https://www.nationalacademies.org/publications/26919

Although I was a reviewer of one chapter of the Manual, I am just seeing this new edition for the first time today. The basic structure of the volume has not changed, although it has now grown to over 1,600 pages. Many of the key chapters on statistics, epidemiology, toxicology, and medical testimony are carried over from previous editions, with some new authors added and some previous authors no longer participating. In addition, there are some new chapters on exposure science, artificial intelligence, climate science, mental health, neuroscience, and eyewitness identification.

The individual chapters and authors in the new edition of the Manual are:

Liesa L. Richter & Daniel J. Capra, The Admissibility of Expert Testimony, at 1.

Michael Weisberg & Anastasia Thanukos, How Science Works, at 47

Valena E. Beety, Jane Campbell Moriarty, & Andrea L. Roth, Reference Guide on Forensic Feature Comparison Evidence, at 113

David H. Kaye, Reference Guide on Human DNA Identification Evidence, at 207

Thomas D. Albright & Brandon L. Garrett, Reference Guide on Eyewitness Identification, at 361

David H. Kaye & Hal S. Stern, Reference Guide on Statistics and Research Methods, at 463

Daniel L. Rubinfeld & David Card, Reference Guide on Multiple Regression and Advanced Statistical Models, at 577

Shari Seidman Diamond, Matthew Kugler, & James N. Druckman, Reference Guide on Survey Research, at 681

Mark A. Allen, Carlos Brain, & Filipe Lacerda, Reference Guide on Estimation of Economic Damages, at 749

Prologue to the Reference Guide on Exposure Science and Exposure Assessment, the Reference Guide on Epidemiology, and the Reference Guide on Toxicology, at 829i

Elizabeth Marder & Joseph V. Rodricks, Reference Guide on Exposure Science and Exposure Assessment, at 831

Steve C. Gold, Michael D. Green, Jonathan Chevrier, & Brenda Eskenazi, Reference Guide on Epidemiology, at 897

David L. Eaton, Bernard D. Goldstein, & Mary Sue Henifin, Reference Guide on Toxicology, at 1027

John B. Wong, Lawrence O. Gostin, & Oscar A. Cabrera, Reference Guide on Medical Testimony, at 1105

Henry T. Greely & Nita A. Farahany, Reference Guide on Neuroscience, at 1185

Kirk Heilbrun, David DeMatteo, & Paul S. Appelbaum, Reference Guide on Mental Health Evidence, at 1269

Chaouki T. Abdallah, Bert Black, & Edl Schamiloglu, Reference Guide on Engineering, at 1353

Brian N. Levine, Joanne Pasquarelli, & Clay Shields, Reference Guide on Computer Science, at 1409

James E. Baker & Laurie N. Hobart, Reference Guide on Artificial Intelligence, at 1481

Jessica Wentz & Radley Horton, Reference Guide on Climate Science, at 1561

Some quick comments on changes in authorship in some of the chapters. Bernard Goldstein, a member of the dodgy Collegium Ramazzini, remains an author of the toxicology chapter in the new edition. David Eaton, however, has been added. Professor Eaton was the president of the Society of Toxicology for many years, and perhaps he has brought some balance to the new edition’s work on toxicology.

An author of the statistics chapter, David Kaye, is also the sole author of the chapter on DNA evidence. Professor Kaye is a distinguished scholar of DNA evidence with serious statistical expertise. David Freedman had been a co-author of the statistics chapter in the third edition, but sadly Professor Freedman died before the third edition was published. Freedman is replaced by Hal Stern, an accomplished statistician from the University of California.

The chapter on epidemiology lost Leon Gordis, who died in 2015. The chapter in the fourth edition has the return of law professors Steve C. Gold and Michael D. Green, whose pro-plaintiff biases are well known, along with two new authors, epidemiology professors Jonathan Chevrier, & Brenda Eskenazi. Like Goldstein, Eskenazi is a fellow of the Collegium Ramazzini.

The Reference Manual, for better or worse, has had substantial influence on the litigation of scientific and technical issues in federal court, and in some state courts as well. I hope to write more substantively about the new edition in 2026.

[1] National Academies of Sciences, Engineering, and Medicine & Federal Judicial Center, Reference Manual on Scientific Evidence (4th ed. 2025).

Posted in Causation, Expert Witnesses, Reference Manual on Scientific Evidence, Rule 702 | 1 Comment »

Prada – Fashionable, But Unreliable Review on Acetaminophen and Autism

September 30th, 2025

Back in the first week of this month, I posted about a paper (Prada 2025),[1]  which featured a so-called navigation-guide systematic review of the scientific evidence on the issue whether pregnant women’s ingestion of acetaminophen causes their children to develop autism.[2] The focus of my post was on some dodgy aspects of the Prada review, such as its anemic disclosures of interest, and its squirrely claim to have been “NIH funded.”

Since posting, the Prada review has been very much in the news. Last week, President Trump held a news conference, where we learned that he cannot pronounce acetaminophen and that he has a strongly held opinion that acetaminophen causes autism.[3] Trump was surrounded by officials in his administration, including plaintiffs’ lawyer Robert Kennedy, Jr., and three physicians, Drs. Oz, Makary, and Bhattacharya, who looked on in apparent approval. Once upon a time, a risk communication such as this one about acetaminophen, would have come out from a non-political FDA employee, such as Janet Woodcock, who was head of Drug Safety, and for many years the Director of Center for Drug Evaluation and Research. Over her tenure, Dr. Woodcock weighed in on many pharmaceutical safety issues. Those of us who have been involved in litigation of those safety issues remember that Dr. Woodcock chose her language very carefully. She did not just give opinions; she marshalled facts.

Admittedly, Trump’s autism press conference was not as deranged as his 2020 press conference at which he suggested that injecting sodium hypochlorite (bleach) into patients would cure Covid-19 infections. Still, most of the world was left with the impression that Trump was replacing (DOGE-ing) scientific research and replacing it with irrational speculation. Trump’s press conference on acetaminophen and vaccines was widely met with skepticism and disbelief. Medical ethicist Dr. Arthur Caplan, who is not given to hyperbole, called the conference “the saddest display of a lack of evidence, rumors, recycling old myths, lousy advice, outright lies, and dangerous advice I have ever witnessed by anyone in authority.”[4]

When the administration physicians communicated with the public, they said something very different from Trump’s presentation. In her press release, Press Secretary Karoline Leavitt used the meaningless locution, “suggested link,” and cited the Prada review, which eschewed causal conclusions:[5]

“Andrea Baccarelli, M.D., Ph.D., Dean of the Faculty, Harvard T.H. Chan School of Public Health: “Colleagues and I recently conducted a rigorous review, funded by a grant from the National Institutes of Health (NIH), of the potential risks of acetaminophen use during pregnancy… We found evidence of an association between exposure to acetaminophen during pregnancy and increased incidence of neurodevelopmental disorders in children.

Harvard University: Using acetaminophen during pregnancy may increase children’s autism and ADHD risk.”

Of course, saying that something “may increase risk” is not even close to saying that something causes the outcome in question. And Baccarelli’s description of his paper, Prada review, as funded by the National Institutes of Health is misleading at best.[6]

Leavitt went on to declare that “[t]he Trump Administration does not believe popping more pills is always the answer for better health.” Unless of course, it is Propecia for Mr. Trump, testosterone for Mr. Kennedy, or ketamine for Mr. Musk.

FDA Commissioner Martin A. Makary issued a Notice, the same day, in which he declared:

“In recent years, evidence has accumulated suggesting that the use of acetaminophen by pregnant women may be associated with an increased risk of neurological conditions such as autism and ADHD in children.

* * *

To be clear, while an association between acetaminophen and autism has been described in many studies, a causal relationship has not been established and there are contrary studies in the scientific literature.”[7]

So the FDA is clearly not declaring that acetaminophen causes autism.

Dr. Mehmet Oz, former surgeon and television talking head, who stood mute by Trump’s side at the infamous press conference, found his voice later in the week, when he acknowledged that pregnant women of course should take acetaminophen when physicians direct them to do so.

In Europe, where pharmaceutical regulation is typically more precautionary than in the United States, both the European Medicines Agency and the U.K.’s Medicines and Healthcare Products Regulatory Agency announced that using acetaminophen during pregnancy was safe with no showing that it causes autism in offspring.[8] Steffen Thirstrup, the EMA’s Chief Medical Officer, announced a day after the Trump bungle, that:

“Paracetamol [acetaminophen] remains an important option to treat pain or fever in pregnant women. Our advice is based on a rigorous assessment of the available scientific data and we have found no evidence that taking paracetamol during pregnancy causes autism in children.”

Most medical organizations were appalled at the administration’s sloppy messaging. The day after the press conference, the American College of Medical Toxicology (ACMT) issued a statement in response, to affirm the safety of acetaminophen in pregnancy.[9] The ACMT noted that its position was in agreement with the American College of Obstetrics and Gynecologists, the Society for Maternal-Fetal Medicine, the American Academy of Pediatrics, and the Society for Developmental and Behavioral Pediatrics.

The acetaminophen kerfuffle seems always to come back to the Prada “navigation guide” systematic review and its authors, including the Harvard Dean, Andrea Baccarelli, who was the well-paid member of the plaintiffs’ expert witness team in acetaminophen litigation.[10] Why did Dr. Andrea Baccarelli in the Prada review use this curious, arcane, and infrequently used method of review? Why did Baccarelli and his co-authors publish this review in Environmental Health, which is dedicated to publishing “manuscripts on important aspects of environmental and occupational medicine,” which places maternal ingestion of a licensed pharmaceutical outside its stated competence? Why did Baccarelli offer a litigation opinion that acetaminophen causes autism, but retreat to “association” when writing for the scientific community? And why did Baccarelli and his co-authors not disclose that Baccarelli had submitted essentially the same navigation guide systematic review as his proffered expert witness testimony, and that a federal court had rejected his opinion as not “the product of reliable principles and methods,” and not “a reliable application of the principles and methods to the facts of the case”[11]? Perhaps the answers are obvious to most observers, but candid disclosures certainly would have provided important context, and saved some people the embarrassment of relying upon the Prada review.

In digging deeper into the history of the navigation guide method itself, the earliest citation I could find to such systematic reviews was in 2009, in a conference paper that discussed this approach as a proposal.[12] The authors that made up the Navigating the Scientific Evidence to Improve Prevention Workshop Organizing Committee were not particularly well known or distinguished in the field of research synthesis. Still, there must be other reasons that “navigation guide” reviews are not more prevalent if the Organizing Committee had been truly on to something important.

The Committee never identified a rationale for a new systematic review approach. When the Organizing Committee outlined its approach in 2009, there were well over three decades of experience with systematic reviews,[13] with well-regarded full-length textbook treatment by experts in the field.[14]

In addition to the lack of experience among its authors and the preemption of the subject by comprehensive treatments elsewhere, there were three additional curious take aways from a cursory reading of the Organizing Committee’s 2009 manuscript. First, Committee emphasized the alleged need for a review methodology for environmental exposures. This emphasis was never accompanied by a showing that well-described methodologies long in use were somehow inadequate or inappropriate for environmental exposures.

Second, the authors urged the need for precautionary assessments, which might make their method fine where syntheses for precautionary pronouncements were called for. In the United States, regulatory assessments vary depending up the governing statutes that create the regulatory mandate.  In personal injury litigation, the precautionary principle is nothing less than an end run around the burden of proof on the party claiming harm and suing in tort. The designated subject matter of environmental exposures for the proposed systematic review technique offers an insight into why these authors believed that they had to propose a new fangled systematic review methodology. Previously described methods interfered with authors’ ability to elevate “iffy” associations into conclusions of causality in the name of the precautionary principle.

The third curiosity in the 2009 manuscript is that the authors never described the need for a pre-specified protocol. Later articles on this proposed methodology similarly failed to describe the need for such a protocol,[15] although by 2014, authors from the original Organizing Committee reversed course to add a pre-specified protocol to the requirements for a navigation guide systematic review.[16]

A recent article defines a systematic review essentially in terms of a protocol:

“Systematic review (SR) is a rigorous, protocol-driven approach designed to minimise error and bias when summarising the body of research evidence relevant to a specific scientific question.”[17]

The purpose of a protocol may be obvious to anyone who has been paying attention to the replication crisis in biomedical literature, but the same article offers a helpful description of its rationale:

“The purposes of the protocol are to discourage ad-hoc changes to methodology during the review process which may introduce bias, to allow any justifiable methodological changes to be tracked, and also to allow peer-review of the work that it is proposed, to help ensure the utility and validity of its objectives and methods.”[18]

Systematic reviews vary widely in quality, methodological rigor, and validity, but one of the key determinants of their validity is whether they were preceded by pre-specified protocol. Although systematic reviews are often described the “gold standard” for evidence synthesis, their methodological rigor vary widely. Reviews that lack a pre-specified protocol are decidedly less rigorous than those reviews that employ a protocol.[19] The absence of a protocol is thus an important tell that a systematic review may be untrustworthy.

The Prada paper put together by Baccarelli’s team has no protocol. It may satisfy the Trump administration’s Fool’s Gold Standard for Science, but that is far short of the requirements of Federal Rule of Evidence 702. Given Baccarelli’s abridgement of scientific method, we should not be overly surprised by Judge Cote’s judgment of the failures of Baccarelli’s and the other plaintiffs’ expert witnesses’ proffered opinions in the acetaminophen litigation:

“their analyses have not served to enlighten but to obfuscate the weakness of the evidence on which they purport to rely and the contradictions in the research. As performed by the plaintiffs’ experts, their transdiagnostic analysis has obscured instead of informing the inquiry on causation.”[20]

Judge Cote carefully reviewed Baccarelli’s proffered testimony and found it replete with cursory analyses, cherry-picked data, and result-driven assessments of studies.[21] Her Honor’s findings would seem to apply with equal measure to the Prada review.

[1] Diddier Prada, Beate Ritz, Ann Z. Bauer and Andrea A. Baccarelli, “Evaluation of the evidence on acetaminophen use and neurodevelopmental disorders using the Navigation Guide methodology,” 24 Envt’l Health 56 (2025).

[2] See Schachtman, “Acetaminophen & Autism – Prada Review Misleadingly Claims to Be NIH Funded,” Tortini (Sept. 9, 2025).

[3] Jeff Mason, Ahmed Aboulenein, and Julie Steenhuysen, “Trump Links Autism to Tylenol and Vaccines, Claims Not Backed by Science,” Reuters (Sept. 22, 2025); Brianna Abbott & Andrea Petersen, “The Trump administration said acetaminophen could cause autism. Doctors maintain it is safe during pregnancy,” Wall St. J. (Sept. 22, 2025) (“Studies looking at a link [sic] between acetaminophen and autism are inconclusive.”); Will Weissert, “Dr. Trump? The president reprises his COVID era, this time sharing unproven medical advice on autism,” Wash. Post (Sept. 23, 2025).

[4] Ali Swenson & Lauran Neergaard, “Trump makes unfounded claims about Tylenol and repeats discredited link between vaccines and autism,” Assoc. Press (Sept. 23, 2025).

[5] Leavitt, “FACT: Evidence Suggests Link Between Acetaminophen, Autism,” The White House (Sept. 22, 2025).

[6] See Schachtman, “Acetaminophen & Autism – Prada Review Misleadingly Claims to Be NIH Funded,” Tortini (Sept. 9, 2025). The referenced grants had nothing to do with acetaminophen and autism, or even autism generally. The NIEHS granted Dr. Baccarelli money to study air pollution and brain aging. The exposure of interest was not acetaminophen, and the outcome of interest was not autism. By claiming that his research was “NIH funded,” Baccarelli was attempting to boost the prestige of the research even though his acetaminophen review was done for litigation, not for the federal government. Apparently the NIEHS acquiesces in this charade because it suggests to the uninitiated that its research grants result in more published papers, even though the topics of those papers are unrelated to the funded research proposal, and the unrelated topics never receiving committee peer review.

[7] Martin A. Makary, “Notice to Physicians on the Use of Acetaminophen During Pregnancy,” (Sept. 22, 2025).

[8] E.M.A., “Use of paracetamol during pregnancy unchanged in the EU,” (Sept. 23, 2025).

[9] ACMT Supports the Safe Use of Acetaminophen in Pregnancy (Sept. 23, 2025).

[10] Rebecca Robbins & Azeen Ghorayshi, “Harvard Dean Was Paid $150,000 as an Expert Witness in Tylenol Lawsuits,” N.Y. Times (Sept. 23, 2025).

[11] Fed. R. Evid. 702.

[12] Patrice Sutton, Heather Sarantis, Julia Quint, Mark Miller, Michele Ondeck, Rivka Gordon, and Tracey Woodruff, “Navigating the Scientific Evidence to Improve Prevention: A Proposal to Develop A Transparent and Systematic Methodology to Sort the Scientific Evidence Linking Environmental Exposures to Reproductive Health Outcomes,”  (July 29, 2009).

[13] See Quan Nha Hong & Pierre Pluye, “Systematic reviews: A brief historical overview,” 34 Education for Information 261, 261 (2018) (describing the evolution of systematic reviews as made up of a “foundation period 1970-1989,” an “institutionalization period 1990-2000, and a “diversification period” from 2001 forward.)

[14] Matthias Egger, Julian P. T. Higgins, and George Davey Smith, Systematic Reviews in Health Research: Meta-Analysis in Context (3rd ed. 2022). The first edition of this text was published in 1995.

[15] Tracey J. Woodruff, Patrice Sutton, and The Navigation Guide Work Group, “An Evidence-Based Medicine Methodology To Bridge The Gap Between Clinical And Environmental Health Sciences,” 30 Health Affairs 931 (2011); Julia R. Barrett, “The Navigation Guide Systematic Review for the Environmental Health Sciences,” 122 Envt’l Health Persp. A283 (2014).

[16] Tracey J. Woodruff & Patrice Sutton, “The Navigation Guide Systematic Review Methodology: A Rigorous and Transparent Method for Translating Environmental Health Science into Better Health Outcomes,” 122 Environ Health Perspect. 1007 (2014).

[17] Paul Whaley, Crispin Halsall, Marlene Ågerstrand, Elisa Aiassa, Diane Benford, Gary Bilotta, David Coggon, Chris Collins, Ciara Dempsey, Raquel Duarte-Davidson, Rex Fitzgerald, Malyka Galay-Burgos, David Gee, Sebastian Hoffmann, Juleen Lam, Toby Lasserson, Len Levy, Steven Lipworth, Sarah Mackenzie Ross, Olwenn Martin, Catherine Meads, Monika Meyer-Baron, James Miller, Camilla Pease, Andrew Rooney, Alison Sapiets, Gavin Stewart, and David Taylor, “Implementing systematic review techniques in chemical risk assessment: Challenges, opportunities and recommendations,” 92-93 Env’t Internat’l 556 (2016).

[18] Id. at 560.

[19] Julia Menon, Fréderique Struijs & Paul Whaley, “The methodological rigour of systematic reviews in environmental health,” 52 Critical Rev Toxicol. 167 (2022).

[20] In re Acetaminophen ASD-ADHD Prods. Liab. Litig., 707 F. Supp. 3d 309, 334, 2023 WL 8711617 (S.D.N.Y. 2023) (Cote, J.).

[21] Id. at 354-56.

Posted in Causation, Expert Witness Hall of Shame, Rule 702, Scientific Evidence | No Comments »

Specific Causation – The Process of Elimination

September 24th, 2025

Specific causation causes some courts to become costive, and sometimes, courts overuse so-called differential etiology as a laxative. The phrase “differential etiology” is an analogy to differential diagnosis, the reasoning process by which clinicians assess the identity of a disease or disorder. Differential etiology, like laxatives, can be overused and misused.

Last month, the Ninth Circuit affirmed a district court’s summary judgment in a glyphosate case. Engilis v. Monsanto Co., No. 23-4201, D.C. No. 3:19-cv-07859-VC (9th Cir. August 12, 2025). The trial court found that plaintiff’s expert witness’s differential etiology was unreliable because the putative expert witness acknowledged that obesity could be a cause of plaintiff’s disease, but then failed reliably to rule out obesity as a differential etiology. Instead, the excluded expert witness glibly inferred that glyphosate was a cause of plaintiff’s cancer. The trial and appellate courts were faced with a great example of invalid, motivated reasoning, or the lack of reasoning.

The Ninth Circuit’s affirmance was significant because it clearly acknowledged that there was no presumption of admissibility, and that the district court was well within its discretion to find that the proffered expert witness opinion had failed to meet the requirements of Rule 702.[1]

The decision in Engilis was simple and straightforward; it was based upon specific or individual causation or its absence. In cases involving diseases with multiple potential causes, none of which is necessary for the outcome, an exposure or lifestyle factor may be capable of causing a particular disease, but that factor may not have played a causal role in everyone who experienced the exposure or lifestyle factor and who developed the disease. (Not everyone who smoked cigarettes develops lung cancer, and not all lung cancer patients smoked.) Courts and litigants are thus left with the puzzle of individual causation.

In a case such as Engilis, courts can basically assume, arguendo, that glyphosate can cause the claimed outcome (Non-Hodgkin’s Lymphoma or NHL), but then insist that there is competent and sufficient evidence to show that the claimant’s specific case of NHL was caused by the claimed exposure.

Some courts and commentators have suggested that a process of “differential etiology, by analogy to differential diagnosis, can get a claimant to the finish line. This attempted solution assumes arguendo that glyphosate can cause NHL, but then it still must resolve whether this specific case of NHL (or whatever claimed) was caused by the claimed exposure.

As suggested above, differential etiology is something like constipation, which is resolved by the process of elimination. Formally, the reasoning process is an “iterative disjunctive syllogism.” We start with an exhaustive listing of the possible established general causes of the claimed disease:

A or B or C (exhausting the possible general causes of the claimed disease).

Because the diseases may multifactorial, the set of disjuncts may be more complex:

A or B or C or A*B or B*C or A*C or A*B*C.

But if the claimant had never been exposed to A, we can deduce:

B or C or B*C.

And if the claimant had never been exposed to B, we can infer that:

C.

And if C is the tortogen under investigation, for which general causation was established, the claimant would have an unequivocal submissible case to the jury.

Of course many diseases have unknown causes, so-called idiopathic or sporadic cases.  In such instances, any proper differential etiology must include a disjunct D, for idiopathic cause. We can see that the iterative disjunctive syllogism in such cases leaves us with uneliminated D in some of the remaining disjuncts, and the claimant cannot reach an unequivocal conclusion in support of his claim.

There may perhaps be a solution to this problem that turns on the effect size, and the probability of attribution associated with each uneliminated disjunct, but that is a story for another day.

[1] See Paul Driessen, “Nation’s most liberal court rejects plaintiff expert’s claims that glyphosate caused couple’s cancer,” Eurasia Review (Sept. 23, 2025).

Posted in Causation, Expert Witnesses, Rule 702 | No Comments »

Acetaminophen & Autism – Prada Review Misleadingly Claims to Be NIH Funded

September 9th, 2025

A few weeks ago, four scientists published what they called a “navigation guide” systematic review on acetaminophen use and autism.[1] The last named author, Andrea A. Baccarelli, is an environmental epidemiologist, who has been an expert witness for plaintiffs’ counsel in lawsuits against the manufacturers and sellers of acetaminophen. Another author, Beate Ritz, frequently testifies for the lawsuit industry in cases against various manufacturing industries. A third author, Ann Z. Bauer, was the lead author of a [faux] “consensus statement” that invoked the precautionary principle to call for limits on the use of acetaminophen (N-acetyl-p-aminophenol or APAP) by pregnant women, on grounds that such use may increase the risks of neurodevelopmental (including autism), reproductive and urogenital disorders.[2] The lead author was Diddier Prada, who works in Manhattan, at the Icahn School of Medicine at Mount Sinai, in the environmental and climate science department, within the Institute for Health Equity Research. The Mount Sinai website describes Dr. Diddier Prada as an environmental and molecular epidemiologist who focuses on the role of environmental toxicants in age-related conditions

Curious readers might wonder how someone whose interest is in environmental issues and “health equity” became involved in a review of pharmaco-epidemiology and teratology. The flavor of systematic review deployed in the paper, “navigation guide,” originated and has had limited use in the field of environmental issues. To my knowledge, so-called navigation guides have never been used previously in pharmaco-epidemiologic or teratologic controversies.[3]

The Prada paper and its deployment of a “navigation guide” systematic review deserve greater critical scrutiny.  In this post, however, I want to address some peripheral issues, such as “competing interests” and misleading claims about the paper’s having been NIH funded.

Only Dr. Baccarelli disclosed a potential conflict of interest, in a statement that many would judge to be anemic:

“Dr. Baccarelli served as an expert witness for the plaintiff’s legal team on matters of general causation involving acetaminophen use during pregnancy and its potential links to neurodevelopmental disorders. This involvement may be perceived as a conflict of interest regarding the information presented in this paper on acetaminophen and neurodevelopmental outcomes. Dr. Baccarelli has made every effort to ensure that this current work—like his past work as an expert witness on this matter—was conducted with the highest standards of scientific integrity and objectivity.”

The disclosure fails to mention whether Dr. Baccarelli was compensated for his playing on the “plaintiff’s legal team,” and if so, then how much. Using the passive voice, he suggests that this work might be perceived as a conflict of interest, when surely he knows that it is a serious issue. If industry scientists working on the relevant issue had published, they surely would be accused of having had a conflict.

Dr. Baccarelli self-servingly, falsely, and with epistemic arrogance, asserts that he made every effort in this paper, and in his past work as an expert witness, to conform to the “highest standards of scientific integrity and objectivity.” Despite his best efforts to be “scientific,” Baccarelli’s work failed critical scrutiny in the multi-district litigation that consolidated acetaminophen cases for pre-trial handling. In that litigation, the defense challenged Dr. Baccarelli’s opinions under Rule 702, for their lack of validity. In an extensive, closely reasoned opinion, federal district court judge Denise Cote ruled that Dr. Baccarelli’s proffered opinions failed to meet the relevance and reliability standards of federal law.[4]

The MDL court easily found that Dr. Baccarelli was qualified to provide an opinion on epidemiology, although the focus of his career has been on environmental issues. Baccarelli’s substantive problem was that he deviated from accepted and valid methods of causal inference by cherry picking different results and outcomes across multiple studies. Baccarelli’s sophistical trick was to advance a “transdiagnostic” analysis that lumps an already heterogenous autism spectrum disorder (ASD), with attention-deficit hyperactivity disorder (ADHD), and a grab bag of “other neurodevelopmental disorders.” If a study found a putative association with only one of the three end points, Baccarelli would claim success on all three. Baccarelli avoided conducting separate ASD and ADHD analyses, and he cherry picked the end points that supported his pre-determined conclusions.

Judge Cote found that the transdiagnostic analyses advanced by plaintiffs’ expert witnesses, including Baccarelli, obscured and obfuscated more than they informed the causal inquiry.[5] The court’s analysis casts considerable shade upon Baccarelli’s self-serving claim to have used “the highest standards of scientific integrity and objectivity.” Judge Cote barred Baccarelli and the other members of the plaintiffs’ “expert team” from testifying.

Conspicuously absent from the conflict disclosure section of the Prada article was any mention of the litigation work of co-author Beate Ritz. In 2007, Ritz became a fellow of the Collegium Ramazzini, which functions in support of the lawsuit industry much as the scientists of the Tobacco Institute supported tobacco legal defense efforts in times past. Ritz’s fellowship in the Collegium makes her a full-fledged member of the Lobby and a supporter of the lawsuit industry.[6] Ritz has testified, for claimants, in cases involving claims of heavy metals in baby food, in cases involving claims that paraquat exposure caused Parkinson’s disease, and most notoriously for plaintiffs in glyphosate litigation, where her witnessing is often done for the Wisner Baum lawfirm that employs the son of Robert F. Kennedy, Jr.[7]

The conflict of interest disclosure statement is hardly the only misleading aspect of the Prada paper. At the end of the paper, the authors state, with respect to funding that their “study was supported by NIH (R35ES031688; U54CA267776).” Some people may incorrectly believe that the Prada review was directly sponsored and funded by the National Institutes of Health.  Nothing could be further from the truth.

The research grant referenced, R35ES031688, is a National Institute of Environmental Health Sciences (NIEHS) research grant. The curious reader might inquire what whether and why the NIEHS would be concerned about a pharmacological issue. The short answer is that the NIEHS is not, and that this grant has nothing to do with children’s neurological status in relation to their mother’s ingestion of acetaminophen.

The NIEHS award this research grant to Andrea Baccarelli, while he was at Columbia University, for his project “Extracellular Vesicles in Environmental Epidemiology Studies of Aging.” The research focuses on extracellular vesicles (EVs) and their role in environmental health, particularly as it relates to aging. What Baccarelli promised to do with this NIEHS grant was to study the effects of air pollution on accelerated brain aging, and disease states such as dementia. Baccarelli noted that his focus would be on intra-cellular communication enabled by extracellular vesicles, in reaction to air pollution. The described research would understandably be viewed as potentially relevant to the NIEHS mission statement, but it has nothing to do with autism among children of women who ingested acetaminophen during pregnancy.  The phrases “extracellular vesicles” and “air pollution” do not appear in the Prada review.

The second grant listed under funding for the Prada review was U54CA267776. The U54 designation marks this as a career award, not specific to a specific topic or this published work. Ironically, the grant is a diversity, equity, and inclusion grant to the Mount Sinai Icahn School of Medicine, in Manhattan. The Icahn School has long had one of the most ethically, racially, culturally diverse faculties of any medical school, and hardly needs financial incentives to hire minority physicians and scientists.

The NIH awarded grant U54CA267776 for “Cohort Cluster Hiring Initiative at Icahn School of Medicine at Mount Sinai.” The NIH describes the grant as aiming to reduce “[t]he barriers to research and career success for underrepresented groups in academic medicine.” The text of the U54 grant is written largely in bureaucratic jargon, which may require a degree in DEI to understand fully. What is abundantly clear is that nothing in this U54 grant, or in its stated criteria for evaluation, has anything to do with studying the teratologic potential of acetaminophen.

What so far has escaped the media’s attention is that Prada and colleagues did not have NIH (or NIEHS) support for their acetaminophen review. They had career-level support for DEI purposes, or perhaps general “walking-around” money for research on environmental pollution and brain aging, which has nothing to do with the subject of their navigation guide review. The authors of the Prada review never prepared a study proposal related to acetaminophen for evaluation by a funding committee at NIH. The authors never submitted a protocol to the NIH, and the NIH provided no peer review or guidance for the authors’ acetaminophen review. In short, there is nothing that marks the Prada review as an NIH work product other than the over-claiming of the authors with respect to funding sources.

The Prada review has attracted a lot of attention in the media and from the worm-brained Secretary of Health and Human Services. An article in the Washington Post described the Prada review as NIH funded, which tracks the paper’s misleading disclosure.[8] The media no doubt jumped on the publication of the Prada review last month because Secretary Kennedy promised to reveal the cause of autism by September. We can imagine that Kennedy will be tempted to embrace the Prada review because he can falsely mischaracterize it as an NIH-funded review.

Not only is the funding claim dodgy, but so is the suggestion that the review supports a conclusion of causation between maternal ingestion of acetaminophen and autism in children. The lead author, Dr. Diddier Prada, noted the frequent confusion between correlation and causation and explicitly stated the authors of the review “cannot answer the question about causation.”[9]

[1] Diddier Prada, Beate Ritz, Ann Z. Bauer and Andrea A. Baccarelli, “Evaluation of the evidence on acetaminophen use and neurodevelopmental disorders using the Navigation Guide methodology,” 24 Envt’l Health 56 (2025).

[2] Ann Z. Bauer et al., “Paracetamol Use During Pregnancy — A Call for Precautionary Action,” 17 Nature Rev. Endocrinology 757 (2021).

[3] See Tracey J. Woodruff, Patrice Sutton, and The Navigation Guide Work Group, “An Evidence-Based Medicine Methodology To Bridge The Gap Between Clinical And Environmental Health Sciences,” 30 Health Affairs 931 (May 2011).

[4] In re Acetaminophen ASD-ADHD Prods. Liab. Litig., 707 F. Supp. 3d 309, 2023 WL 8711617 (S.D.N.Y. 2023) (Cote, J.).

[5] Id. at 334.

[6] See F.D.K. Liddell, “Magic, Menace, Myth and Malice,” 41 Ann. Occup. Hyg. 3, 3 (1997).

[7] See, e.g., In re Roundup Prods. Liab. Litig., 390 F. Supp. 3d 1102 (2018); Barrera v. Monsanto Co., Del. Super. Ct. (May 31, 2019); Pilliod v. Monsanto Co., 67 Cal. App. 5th 591, 282 Cal. Rptr. 3d 679 (2021). See also Dan Charles, “Taking the stand: For scientists, going to court as an expert witness brings risks and rewards,” 383 Science 942 (Feb. 29, 2024) (quoting Ritz as suggesting that she was reluctant to get involved as an expert witnesses).

[8] Ariana Eunjung Cha, Caitlin Gilbert and Lauren Weber, “MAHA activists have been pushing for more investigation into use of the common pain killer during pregnancy,” Wash. Post (Sept. 5, 2025). See also Liz Essley Whyte & Nidhi Subbaraman, “RFK Jr., HHS to Link Autism to Tylenol Use in Pregnancy and Folate Deficiencies,” Wall St. J. (Sept. 5, 2025).

[9] Jess Steier, “Saturday Morning Thoughts on the Tylenol-Autism News: The public health whiplash continues as we play another round of ‘autism cause’ roulette,” Unbiased Science (substack) (Sept. 06, 2025).

Posted in Causation, Conflicts of Interest, Expert Witnesses, Rule 702, Scientific Evidence | 2 Comments »

AAAS Conference on Scientific Evidence and the Courts

September 8th, 2025

Back in September 2023, the American Association for the Advancement of Science (AAAS), with its Center for Scientific Responsibility and Justice, sponsored a two day meeting on Scientific Evidence and the Courts. If there were notices for this conference, I missed them. The meeting presentations are now available online. Judging from camera views of the audience, the conference did not appear to be well attended. Most of the material was forgettable, but some of the presentations are worth watching.

Jennifer L. Mnookin opened the conference with a keynote presentation on “Where Law and Science Meet.” Chancellor Mnookin presented a broad overview and some interesting insights on the development of the evidence law of expert witness testimony.

Following Mnookin, Professors Ronald Allen and Andrew Jurs presented on the “Unintended Impacts [sic] of the Daubert Standard.” The conference took place only a few months before amendment to Rule 702 became effective, and the reference to a “Daubert” standard was untoward. Allen’s comments followed the path of his previous articles. Jurs presented some empirical legal research, which seemed flawed for its assumption that the Frye standard was universally applied in federal court before the advent of Daubert. Assessing whether these standards lead to different outcomes when both standards have been applied heterogeneously, and one standard, Frye, is often not applied at all, and Daubert is often flyblown by judges hostile to the gatekeeping enterprise, Jurs’ empirical research seemed both invalid and very much beside the point. Both presenters missed the key point of Daubert, in which case plaintiff’s counsel advocated for no standard at all, beyond basic subject-matter qualification, for giving expert opinions in court.

A Session on “An International Perspective,” Scott Carlson discussed the efforts of the American Bar Association (ABA), and its Center of Global Programs, on supporting judges in foreign countries. Prateek Sibal discussed the history and work of the UNESCO Global Judges Initiative. My sincere wish is that the ABA would support judges more in the United States.

Panelists Valerie P. Hans, Emily Murphy, and Dr. Michael J. Saks presented on various jury issues, in a session “In the Minds of the Jury.” The presentations on how foreign countries process expert witness testimony were lacking any mention of how juries rarely if ever sit in civil cases that involve complex technical and scientific issues.

Two editors of scientific journals, Adriana Bankston and Valda Vinson, along with law professor Michael Sakes, spoke about peer review and publication, in  a session “As a Matter of Fact: ‘General Acceptance’ in Emerging vs. Established Science.” Their discussion on the publication process shed very little light on how courts and juries should assess the validity of specific papers, particularly in view of the lax practices at many journals. Towards the end of this session, a question from the audience proved to be very revealing of the prejudices of the law professor on the panel. The questioner rose to complain that after beginning research on a topic that has litigation relevance her research is now frequently questioned. She asked the panel how she might deal with the annoyance of being questioned. Some on the panel basically urged her to buck up, but the law professor invoked the spirit of agnothologist, and lawsuit industry expert witness, David Michael, to suggest that “manufacturing doubt” was just a corporate tactic in the face of scientific evidence. The prejudice against corporate speech is remarkable when the lawsuit industry has a long history of playing the ad hominem game in advancing its pecuniary interests.

The session that followed addressed how trustworthy science might best be put before courts. The organizers described this session, Utilizing Scientific and Technical Expertise, as going to the heart of the issues targeted by the conference. Joe S. Cecil, Deanne M. Ottaviano, and Shari Seidman Diamond discussed how scientific expertise enters into the evidentiary record in American courtrooms. Their presentations were interesting, but curiously no one mentioned that the primary avenue for expert witness opinion is through oral testimony!

Joe Cecil discussed methods judges have to obtain scientific and technical evidence to advance justice. (By this I hope he meant the truth, and not just the outcome preferred by social justice warriors.) As noted, Joe Cecil did not focus on the ordinary methods of direct and cross-examination of party expert witnesses, but rather, he identified other methods of introducing expertise into the courtroom for the benefit of the judge or the jury. Only one suggestion really affects jury comprehension, namely the appointment of non-party expert witnesses by the court. The other methods really only provide expertise to the trial judge, who perhaps is challenged to make a ruling under Federal Rule of Evidence 702. The federal courts have the inherent supervisory power to appoint technical advisors to act as special law clerks on issues. Similarly, appointed special masters can address technical implementation issues, subject to the district judges’ control. The judges are always free to read outside the briefs and testimony, but there are ethical and notice issues for such conduct. The Reference Manual on Scientific Evidence (RMSE) sits on the shelves on every federal judge’s bookshelf, even if in pristine, unused condition. Judges can at least read the RMSE on specific issues without having to disclose their extra-curricular research to the parties.  Of course, parties are well advised to consider any materials in the RMSE, which support or oppose their contentions.

In discussing the RMSE, Cecil noted that the fourth edition was in the works. He also mentioned that all the old chapter topics would be carried forward to the fourth edition, and that new topics would include eyewitness identification, computer science, artificial intelligence, and climate science. Sadly, there will be no chapter on genetic determination of disease, but perhaps the clinical medicine chapter will take on the subject in greater detail than previous editions. This conference took place two years ago, and yet the RMSE, fourth edition, is still not published. The National Academies website previously listed the project as completed, but the site now describes the work as “in progress.”

Joe Cecil’s analysis of the various extraordinary expert techniques was pretty much spot on, especially his assessment that “experiments” with court-appointed experts were often failures or at best modest successes. The discussion of Judge Pointer’s Rule 706 independent expert witnesses in the silicon [sic] breast implant litigation, MDL926, seemed to lack context. Cecil acknowledged that the court’s expert witnesses contributed some value to admissibility decisions, but Judge Pointer notoriously did not believe that he, as the MDL judge, had any responsibility for Rule 702 determinations, and he made none except in cases that he tried in the Northern District of Alabama. (And these decisions were before the Science Panel was appointed.) So the Rule 706 witnesses really could not have aided in admissibility decisions.

The real value – in my view – of the Science Panel was that it demonstrated that Judge Pointer was quite wrong in believing that both sides’ expert witnesses were simply “too extreme,” or too partisan, and that the truth was somehow in the middle. Indeed, Judge Pointer said so on many occasions, and he was judicially gobsmacked when all four of his experts roundly rejected the plaintiffs’ distortions of the science of immunology, epidemiology, toxicology, and rheumatology. The courts’ expert witnesses sat for discovery depositions, and then gave testimony de bene esse. To my knowledge, their testimony was never admitted in any of the subsequent trials.

Judge Jed Rakoff gave an interesting presentation, “Strengthening Cooperation Between the Scientific Enterprise and the Justice System,” on the intersection between scientific and legal expertise and the need for their better integration. Judge Rakoff focused on the astonishing lack of compliance of trial judges with the gatekeeping requirements of Rule 702 in addressing the admissibility of forensic evidence. Several subsequent panels also addressed forensic topics, including “A Texas Case Study in Accountability for Forensic Sciences,” “Innovations in Investigative Technologies Improvements and Drawbacks,” and “Artificial Intelligence and the Courts,” “Wrongful Convictions and Changed Science: Statutes,” and “Standing Up for Justice: When the Law and Science Work Hand-in-Hand.”

One of the more curious sessions was on “Statistical Modeling and Causation Science,” presented by the American Statistical Association along with the AAAS. Maria Cuellar, from the University of Pennsylvania, discussed the role of statistical thinking in causal assessment, with slides that referred to a nonparametric estimator for the probability of causation. Cuellar, however, never defined what an estimator was; nor did she differentiate nonparametric from parametric estimators. She displayed other equations, again without explaining their origin and meaning, or identifying symbols or meanings. Similarly, Rochelle E. Tractenberg, discussed the use of statistics as evidence and as part of inferring causal inference in litigation, in a model of unclarity. At one point, Tractenberg appeared to suggest that general causation could be taken from regulatory pronouncements. Her discussion of glyphosate implied that general causation was established, which may have led me to disregard her presentation.

Finally, the conference sported a discussion, “Toxic Tort 2.0: Emerging Trends in Climate Change Related Litigation,” The two presenters were Dr. L. Delta Merner, the “Lead Scientist” for the Science Hub for Climate Litigation, Union of Concerned Scientists, and Dr. Paul A. Hanle, Visiting Scholar and  Founder of the Climate Judiciary Project, Environmental Law Institute. The Science Hub actively promotes climate change litigation, which made me wonder whether its scientists are involved in that new chapter in the upcoming fourth edition of the Reference Manual.

Posted in Causation, Reference Manual on Scientific Evidence, Rule 702, Scientific Evidence, Scientific Publishing | No Comments »

« Older Entries