Skip to Content

Experts Flunk Reliability Test in BMW Case

March 15, 2016 in  News

New York Law Journal

This column reports on an appellate ruling issued in February that focused on reliability issues posed by proposed expert testimony. Sean R. v. BMW of North America,1 is an important decision by the New York Court of Appeals in a case claiming that a youngster’s mental and physical disabilities were caused by in utero exposure to unleaded gasoline vapor attributable to a defective gas hose in the pregnant mother’s BMW. New York’s highest court provided the state’s trial bench and bar with yet more guidance about what is needed in the way of reliable expert testimony in a toxic tort claim. BMW, thus, is another link in a chain of decisions by the court offering a road map regarding how expert proofs may suffice as opposed to those that do not.

Our prior columns often elaborated well-established New York law on judicial screening or “gatekeeping” of expert testimony.2 The Court of Appeals made it clear in Parker v. Mobil Oil Corp.3 that questions regarding reliability of experts’ testimony required diligent policing by courts. The basic rules seem simple enough. If “novel” scientific evidence is involved, the court applies the Frye “general acceptance” test to determine “whether the accepted techniques, when properly performed, generate results accepted as reliable within the scientific community generally.” If the answer is “no,” the testimony has flunked and must be precluded or excluded. If the answer is “yes,” the proponent of the novel scientific testimony has survived the threshold test but still has to get past the admissibility “gate.” The evidence must meet a second, “foundational reliability” inquiry.

This second admissibility standard applies to all expert testimony, not just “novel” scientific evidence. In Parker, the court said the Frye inquiry is “separate and distinct from the admissibility question applied to all evidence—whether there is a proper foundation—to determine whether the accepted methods were appropriately employed in a particular case.” Parker declared that the focus moves “from the general reliability concerns of Frye to the specific reliability of the procedures followed to generate the evidence proffered and whether they establish a foundation for the reception of evidence at trial.” These points were confirmed by the court in its Cornell v. 360 W. 51st Realtydecision in 2014, a claim that toxic mold caused illness.4

In BMW the plaintiff was born in May 1992 with severe mental and physical disabilities. The timeline underlying the claim that the child’s deficits were caused by a defective fuel hose emitting gas vapors that damaged the baby in utero was as follows. In May 1989, the father purchased a new BMW 525i for his wife, the car’s exclusive driver. In the spring of 1991 she began to notice a smell of gasoline that “came and went.” When the windows were down, for example during summer, she could tolerate the smell. But at other times the strong odor caused headaches, dizziness and throat irritation. Family members who rode in the car noticed the odor; one said it made her nauseous and dizzy.

The dealer couldn’t identify any problem and made no repairs. In July or August 1991, the wife became pregnant with plaintiff. The husband took the car to the dealer, and a split fuel hose was discovered. Plaintiff was born in May 1992. Subsequent testing revealed the child had, among other things, spastic quadriparesis (a form of cerebral palsy), developmental delays, ventricular asymmetry, aortic stenosis and impaired vision function. Two years later, BMW recalled the vehicles due to defects in the feed fuel hoses. The recall report noted that customers had associated the defect with a “conspicuous fuel odor.”

Causation Experts

The lawsuit against BMW and others was commenced in January 2008. The claim was that failure to discover and fix the defective hose exposed plaintiff in utero to toxic gasoline vapor. Plaintiff’s primary causation experts were Dr. Linda Frazier and Dr. Shira Kramer. They were prepared to testify that the in utero exposure to gasoline vapor proximately caused the birth defects. Dr. Frazier concluded that plaintiff’s mother inhaled 1,000 parts per million (ppm) of gasoline vapor based on the mother’s symptoms of “acute toxicity” during exposure, such as suffering headaches, nausea and throat irritations.

After ruling out other possible causes, Frazier concluded that the mother’s “high peak exposure” to the vapors during the first trimester of pregnancy was the “most likely cause of plaintiff’s injuries.” Dr. Kramer reached similar conclusions using a “different analysis.” Specifically, she said that toluene and benzene, constituent elements of gasoline vapor are “causally related to an elevated risk of birth defects.”5

BMW moved to preclude these causation experts from testifying at trial. Alternatively, defendant asked for a Frye hearing to determine whether Drs. Frazier and Kramer were reaching novel conclusions and not using generally accepted principles and methodologies.6 After reviewing lengthy submissions and supplemental expert reports, the trial court precluded the causation experts. The Appellate Division affirmed but certified the question to the Court of Appeals.

The state’s highest court held that the experts should be precluded from testifying. The opinion for the court, by Judge Eugene F. Pigott, Jr., explained that in toxic tort cases an expert opinion on causation must set forth (1) a plaintiff’s exposure to a toxin, (2) that the toxin is capable of causing plaintiff’s particular injuries (general causation) and (3) that the plaintiff was exposed to sufficient levels of the toxin to cause such injuries (specific causation).7 The expert must establish such exposure levels “through methods ‘found to be generally accepted as reliable in the scientific community.'” This “general acceptance” requirement, also known as the Frye test, “governs the admissibility of expert testimony in New York.”8

In a footnote, the court explained that its analysis and holding were limited to the Frye “general acceptance” inquiry due to the procedural posture of the case. Thus, the court expressed no opinion on the “separate and distinct” question of “whether there was a proper foundation for their opinions.”9  The significance of this clarification by the court should not be overlooked. As mentioned above, after the “novel” science Frye test is passed, there is a second “foundational reliability” test for admissibility of the expert opinion, as stated in the Parker case. That reliability test applies in all non-novel science cases as well. In the BMW case, the Frye analysis was dispositive.

The Frye “general acceptance” test asks “whether the expert’s techniques, when properly performed, generate results accepted as reliable within the scientific community generally.” Although unanimity is not required, the proponent must show “consensus in the scientific community as to the [methodology’s] reliability.”10 In the BMW case, the experts failed to make that showing. Here are highlights of the experts’ shortcomings. Drs. Frazier and Kramer concluded from occasional nausea, dizziness and throat irritations experienced by the mother that a “sufficient amount of gasoline vapor” caused the child’s injuries. But they cited no text, scholarly article or study that approves of or applies this type of methodology, let alone a “consensus” as to its reliability.

‘Reliability’ Shortcomings

None of the reports or studies advanced by the experts established that Frazier’s methodology, “when properly performed, generate results accepted as reliable within the scientific community generally.”11 They merely supported her own “conclusion” that there is a dose-response relationship between exposure to the chemical constituents of gasoline and symptoms of toxicity. Those studies did not support her approach of “working backwards from reported symptoms to divine an otherwise unknown concentration of gasoline vapor.” Her methodology was not supported by the studies.12

Although “odor threshold” analysis has been admitted in certain settings, the odor threshold of a substance can be “far below” its toxicity. Indeed, one of plaintiff’s sources, an article in the American Journal of Industrial Medicine, explained: “Smelling organic solvents is not indicative of a significant exposure, as the olfactory nerve can detect levels as low as several parts per million, which is not necessarily associated with toxicity. As an example, the odor threshold of toluene is 0.8 parts per million, whereas the [threshold limit value] is 100 parts per million.”13

Since unleaded gasoline in the early 1990s had a “very low” odor threshold of between 0.50 and 0.76 ppm, a person would have been able to detect the odor of gasoline vapor at less than 1 ppm. The only expert conclusion that could be reached, using odor threshold methodology, was that plaintiff was exposed to at least 1 ppm of unleaded gasoline—the minimum level at which gasoline is detectable by human smell. Instead, Dr. Frazier concluded that there is a minimum vapor threshold below which individuals do not experience headache, nausea or dizziness. Because the mother experienced those symptoms, Frazier concluded she must have been exposed to at least that concentration. But it was not shown that Frazier’s approach “has been generally accepted in the scientific community.”14

Although it is sometimes difficult, if not impossible, to quantify a plaintiff’s past exposure to a substance, it remains a requirement of the causation expert in a toxic tort case to show, via generally accepted methodologies, that a plaintiff was exposed to a sufficient amount of a toxin to have caused his injuries. Here the plaintiff’s experts failed to meet their burden to show that their methodology was “generally accepted in the scientific community.”


Toxic tort cases often present statistical literature noting “associations” between a toxin and disease. Such murky “connections” do not reliably establish either general or specific causation. The BMW holding confirms the courts’ task to gatekeep expert opinions for reliability. The court’s analysis of the proposed expert testimony offers litigators lots of practical guidance. Put BMW on your reading list.


  1.  2016 N.Y. LEXIS 134 (N.Y. Ct. App. Feb. 11, 2016).
  2.  See M. Hoenig “‘Cornell’ Ruling on Mold Clarifies ‘Reliability’ Needed From Experts,” New York Law Journal, April 14, 2014, p. 3; “Experts ‘Frye’d’ on Tylenol-Cirrhosis Link,” NYLJ, Dec. 12, 2011, p. 3; “Testifying Experts and Scientific Articles: Reliability Concerns,” NYLJ, Sept. 16, 2011, p. 3 (citing prior articles on experts’ use of unreliable hearsay, scientific papers questioning the reliability of biomedical articles, and reporting serious shortcomings even in those that were peer reviewed); “‘Parker,’ ‘Frye’ and Gatekeeping of Experts: an Update,” NYLJ, June 17, 2009, p. 3; “Judicial Gatekeeping: ‘Frye,’ ‘Foundational Reliability,'” NYLJ, Feb. 11, 2008, p. 3; “‘Gatekeeping’ of Experts and Unreliable Literature,” NYLJ, Sept. 12, 2005, p. 3. The articles are also available on LEXIS.
  3.  7 N.Y. 3d 434 (2006).
  4.  Cornell v. 360 W. 51st Realty, 22 N.Y. 3d 762 (2014).
  5.  BMW, 2016 N.Y. LEXIS 134, at *4—*6.
  6.  A Frye hearing is so-named after the decision in Frye v. United States, 293 F. 1013 (DC Cir. 1923).
  7.  BMW, LEXIS at *7 (citing Parker v. Mobil Oil Corp., 7 N.Y. 3d 434, 448 (2006)).
  8.  Ibid.
  9.  Id. LEXIS at *8 n.1.
  10.  Id. LEXIS at *7—*8.
  11.  Id. LEXIS at *9 (citing People v. Wesley, 83 N.Y. 2d 417, 423 (1994)).
  12.  Id. LEXIS at *10.
  13.  Id. LEXIS at *11 (quoting from McMartin, et al, “Pregnancy Outcome Following Maternal Organic Solvent Exposure: A Meta-Analysis of Epidemiologic Studies,” 34 Am. J. of Indust. Med. 288—292, 289 (1998)).
  14.  Id. LEXIS at *12.