While many cases alleging a claim for serious injury might involve zealous advocacy, cases that involve a cancer allegedly caused by exposure to a carcinogen can be particularly adversarial because those cases often revolve around the disputed basis for any causation opinion and its admissibility. Plaintiffs and defendants in toxic tort actions are regularly confronted with issues of admissibility and utility of governmental regulatory standards, governmental assessments of the toxicity of particular chemicals, and a variety of federal and state statutes and regulations governing industries' conduct in the environmental, occupational, and consumer contexts. These toxic tort cases typically involve assessments by toxicologists, occupational health physicians, environmental experts, epidemiologists, and industrial hygienists, who relate estimates or calculations of the alleged exposures to various levels, including those levels instituted by government agencies.
Both plaintiff and defense lawyers may rely on governmental levels as an important tool in constructing their arguments, whether that argument is that the alleged exposure exceeded the level or complied with the standards. If the agent at issue is at levels below a governmental standard, a defendant will assert compliance with the government's standard, perhaps even arguing that the exposure was de minimis and not causative. However, if an agent exceeded a regulatory level, a plaintiff might offer that as evidence of an exposure to dangerous levels, sufficient to cause the alleged disease, while the defendant maintains that a governmental standard is the result of a different process with different goals and assumptions and based on different evidence.
There are no clear rules of law governing the admissibility and utility of government data or regulatory actions in the courtroom and the actual admissibility and impact of governmental regulations will likely vary from jurisdiction to jurisdiction, from judge to judge, and from case to case. Thus, parties in toxic tort cases must have an understanding of the data potentially relevant to their cases in order to persuade the court that the scientific data at issue in that case are or are not reliable and admissible and are or are not relevant to the issue of proximate causation.
However, with Daubert v. Merrell Dow Pharmaceuticals, Inc., 509 U.S. 579 (1993), the U.S. Supreme Court outlined a context in which to analyze the admissibility of scientific evidence and changed the state of evidence law that had existed for generations. In the wake of Daubert, courts are allowed to admit only that evidence and those opinions that are reliable. And while there is a lot about the science of cancer and carcinogens that remains unknown, what is known and accepted and what opinions are accepted and reliable are often at the heart of an argument about whether an agent is a carcinogen and whether it caused a plaintiff's cancer. So, logically, the decision of an administrative agency, such as the EPA or the fact that evidence or an expert's conclusion is consistent with that decision is obviously reliable and admissible in a court of law, right? Wrong. Any evidence, conclusions, or opinions that form the basis of an agency decision, as well as the agency decision itself, must still meet these same Daubert and Rule 702 reliability and admissibility requirements. Careful consideration must be given to the goal and purpose of that agency in reaching that decision and the methodology it employed.
Regulatory agencies often engage in causal analysis without any statistical information and take actions based more on precautionary principles than on the type of reliable causation evidence necessary for a determination in tort cases such as epidemiology. Therefore, these regulatory processes and any resulting classifications are insufficient in and of themselves to allow valid and admissible scientific conclusions to be drawn. Further, often the bases for those classifications are not independently admissible.
For example, the EPA relies heavily on animal studies and modeling in reaching its classifications. Scientists have been unable to come to a consensus about whether data from animal studies are an appropriate predictor of human cancer and, if so, in what circumstances and at what dose. Generally, scientists assume that agents that are carcinogenic in animals are also carcinogenic in humans, though there are a number of concerns with this assumption and in particular with regard to high-dose exposure animal studies. In addition, a major limitation of animal studies is that they can suggest risks that reflect very high and perhaps unrealistic exposure levels that are not at all comparable to any such exposure in humans. And the extrapolation of data from animal tests to human populations relies on specific assumptions about the similarities between the studied animal and human response. Although these assumptions may be appropriate in formulating regulatory policy given the charge and goal of that agency, the conclusions they yield are not necessarily adequate for establishing scientific causation. Using animal studies as a basis for a cause-and-effect conclusion, in the absence of epidemiological studies, the EPA cannot exclude bias or other confounding factors as a possible cause for any observed effect.
In addition, regulatory agencies are increasingly relying on physiologically based pharmacokinetic (PBPK) modeling in risk assessments, such as the EPA's recent assessment of TCE. These risk assessments are now more frequently relying on dosimetry estimates in animals and humans obtained from PBPK modeling to reach conclusions on carcinogenicity and exposure levels. These PBPK models describe the relationship between those external measures of exposure and internal measures of a toxicologically relevant dose. However, the selection of proper dose metrics depends not only on the reliability of PBPK data and models but also on the understanding and application of the mode of action for a particular toxic effect. These PBPK models and the agency risk assessments that rely on them also increasingly require the characterization of uncertainty and variability. PBPK models, and the assumptions that must be made in them, may create even more potential pitfalls for the practitioner who intends to rely on those conclusions.
Also, many regulatory assessments assume that there are no thresholds for carcinogens and that risk is proportionate to dose, despite studies that demonstrate trivial risk at very low exposures. That assumption, in turn, rests on a number of other assumptions, each of which is controversial when examined in the context of legal causation. Further, there is agreement in scientific and regulatory communities that risk assessments based on animal studies represent worst-case scenarios rather than best-case estimates of the potential incidence of cancer. Consequently, reliance on animal studies alone to support an opinion on causation in humans, while perhaps acceptable in the regulatory setting and in furtherance of governmental policy, does not likely meet the burden imposed by Daubert and Rule 702. Further, proponents have been unable to convince courts regularly and repeatedly that animal studies alone are sufficient to establish causation under Rule 702 and Daubert.
Not only are the bases for an administrative classification very different from a judicial ruling on admissibility, but also the very goals of the two analyses differ greatly. For example, the goal of the EPA classification of carcinogens is to protect people from potential harm, whereas the goal of a Daubert or Rule 702 challenge is to prevent the admissibility of evidence that is unreliable. Further, the EPA classifications are based on an assessment of risk and are, by design, particularly cautious. That approach is a glaring juxtaposition to the courtroom, where only evidence that is demonstrated to be reliable may be offered and the judge serves as the gatekeeper to prevent experts from offering unsupported opinions. The EPA often engages in a "weight of the evidence" approach with regard to classifying agents as carcinogens, which is different than the reliability assessment required by Daubert and Rule 702. Thus, there is no reason for a court to allow the admission of evidence it would normally bar merely because the EPA relied on that evidence in reaching its conclusion or classification.
It is important to distinguish between a governmental entity setting a level for exposure to a particular agent in the regulatory framework and a scientifically reliable evaluation of whether an agent can and did cause an injury in the context of a tort claim. Regulatory standards and levels are often set based on the methodology of quantitative risk assessment. Risk assessment has been defined by the Federal Judicial Center as "the use of scientific evidence to estimate the likelihood of adverse effects on the health of individuals or populations from exposure to hazardous materials and conditions." Fed. Judicial Ctr., Reference Manual on Scientific Evidence (1994). This risk assessment differs from the methodologies used by physicians in diagnosing a patient's condition or scientists in studying the cause of disease and were not designed or intended to be dispositive in proximate causation. Instead, this risk assessment is a calculation for regulatory policy formulation that is based in large part on economics and considering what is the "acceptable" number of cancers that might arise from exposure to a certain agent at a particular level.
In addition, there has been some controversy over the admissibility of expert testimony premised on the weight-of-the-evidence methodology used by the EPA to classify a chemical as a carcinogen. This weight-of-the-evidence assessment of scientific validity is not to be confused with a conclusion based on the weight of the factual evidence at trial. The latter is a function of the jury, the finder of fact, whereas the former is a function of the gatekeeper, the judge. "Weight of the evidence" analysis was used to classify the chemical ethylene oxide as a carcinogen and was discussed in Allen v. Pennsylvania Engineering Corp., 102 F.3d 194 (1996). In Allen, the Fifth Circuit held that while regulatory agencies use the method to assess the carcinogenicity of substances, the threshold of proof is lower in the regulatory setting because the agencies are charged with protecting public health, while the tort system imposes a higher burden on the plaintiff to prove it is more likely than not that the agent caused the alleged injury. The Allen court rejected the weight-of-the-evidence methodology.
As explained by the Tenth Circuit Court of Appeals in Mitchell v. Gencorp Inc., 165 F.3d 778, 783 n.3 (10th Cir. 1999) (quoting Allen), the methodology used by a government agency "results from the preventive perspective that the agencies adopt in order to reduce public exposure to harmful substances." The court observed that "[t]he agencies' threshold of proof is reasonably lower than that appropriate in tort law, which traditionally makes more particularized inquiries into cause and effect and requires a plaintiff to prove it is more likely than not that another individual has caused him or her harm." The Fifth Circuit's holding, however, conflicts with the analysis and holding of the Third Circuit in In re Paoli Railroad PCB Litigation, 35 F.3d 717, 741–71 (3d Cir. 1994), which accepted expert causation opinions that employed the same weight-of-the-evidence methodology employed by the EPA.
Judges and the lawyers responsible for educating them about the scientific evidence in their cases need to know more than what makes good science; they need to understand how to identify insufficient or even bad science and explain what makes it so. Cases reflect judicial recognition that courts, in fulfilling their gatekeeping function, have an obligation to keep "junk science" out of the courtroom. While there are public policy reasons to allow broad discretion to admit evidence, there are also sound public policy reasons for requiring judges to assess expert testimony carefully to determine both its relevance and reliability before allowing its admissibility.
As the U.S. Supreme Court recognized in Daubert, 509 U.S. at 595, "expert evidence can be both powerful and quite misleading because of the difficulty in evaluating it." Justice Breyer noted in his concurring opinion in General Electric Co. v. Joiner:
And it may, therefore, prove particularly important to see that judges fulfill their Daubert gatekeeping function, so that they help assure that the powerful engine of tort liability, which can generate strong financial incentives to reduce, or to eliminate, production, points towards the right substances and does not destroy the wrong ones.
522 U.S. 136 at 148–49 (1997) (Breyer, J., concurring).
To be sure, the battle faced by anyone stricken with cancer is, at times, seemingly incomprehensible. Cancer is a horrible disease that afflicts too many, too often, and too frequently results in the ultimate consequence. However, courts and counsel must endeavor to keep any concerns for cancer patients separate from the question of causation and be sure that the jurors and juries do as well. And though it is a question scientists endeavor to answer, there is usually no simple answer to the question of whether an agent caused or even likely caused an individual's cancer. In our system, it is the plaintiff's burden to prove that an agent, in fact, did cause his or her cancer and if the plaintiff cannot do so, his or her case should not proceed to trial. Merely because an individual who was diagnosed with cancer suffered does not mean that an exposure to a carcinogen caused that disease, as numerous environmental agents can also cause cancer. The diagnosis also does not mean that we can determine the actual cause. Nevertheless, our legal system provides us with the means and responsibility to try to do so.
In this regard, the fact that a regulatory agency has classified an agent as a carcinogen or made a determination as to a level of exposure to that agent as causative or dangerous does not mean that if there is evidence of exposure to that agent in a toxic tort claim, causation has been determined or, for that matter, that such evidence is even admissible to demonstrate the probability or even the possibility of causation. As discussed above, there is a profound difference between the classifications from the regulatory process and the verdicts from the adjudicative process. For that reason, the determinations and conclusions of a regulatory agency do not and should not dictate any determination on the issue of proximate cause in tort litigation. Any evidence offered to support such a claim should be reliable and admissible in and of itself, whether or not it is consistent with any regulatory agency conclusion or employs the same weight-of-the-evidence methodology.
Keywords: agency action, EPA, IRIS, integrated risk, TCE, trichloroethylene, cause and proximate causation