|Steve C. Gold
It is universally acknowledged that the Human Genome Project and related advances have immeasurably increased our understanding of the genetic factors that contribute to human disease. But when plaintiffs claim that their illness resulted from a wrongful exposure to a toxic substance, has advanced scientific understanding made it easier for the legal system to establish definitive proof of toxic causation in individual cases? Not really, says Associate Professor Steve C. Gold; in fact, in many respects it has made the job even more complex. In the forthcoming article “When Certainty Dissolves Into Probability: A Legal Vision of Toxic Causation for the Post-Genomic Era,” he argues that “genomic and molecular understanding, instead of sounding the death knell for proposals to reform toxic tort causation law, will strengthen the argument for those reforms.” Gold, who first wrote about toxic torts in a much-cited Note while he was a law student and then served as a trial attorney and senior attorney in the Environmental Enforcement Section of the U.S. Department of Justice, proposes that a probabilistic causal contribution model replace the traditional deterministic model of causation in toxic torts.
You write that the “deterministic model of causation has performed poorly when confronted with the mechanistic opacity of toxic torts.” Can you give an example from a recent case?
There are many, many examples. Two cases recently decided by the Tenth Circuit Court of Appeals are really good illustrations. The plaintiffs were people living in company mining towns who had been diagnosed with cancer after being exposed to radiation released during uranium mining and milling activities. There was no question that the radiation exposure increased their cancer risk, but the panel majorities held that under the deterministic model of causation each of the plaintiffs had failed to provide sufficient evidence that his or her cancer specifically was caused by the tortious exposure as opposed to other radiation or some other cause.
Another example is a 2010 decision by the Vermont Supreme Court in a worker’s compensation case, Estate of George v. Vermont League of Cities and Towns. Mr. George died of non-Hodgkins’ lymphoma after a long career as a firefighter. Although several epidemiologic studies linked this disease to firefighting, the court affirmed the denial of compensation, essentially because the epidemiology did not unambiguously show a doubling of the risk. The court held that without such a showing, the claimant had not proven that Mr. George “more likely than not” contracted the disease as a result of his employment. Some courts have demanded that plaintiffs prove that their disease has some characteristic that distinguishes illness caused by toxic exposure from illness caused by other factors. That has been impossible, except for a few “signature diseases” like asbestosis.
In the post-genomic era, why is the probabilistic model of causation more appropriate to toxic torts than the deterministic model?
The deterministic model assumes that it is possible to reach a conclusion about whether or not the plaintiff’s injury would have occurred “but for” the tortious behavior – whether we are talking about the failure to clear ice from a building’s entrance stairs where someone fell or driving above the speed limit before an automobile collision or performing surgery with improperly sterilized instruments on a patient who suffers an infection at the incision. In toxic tort cases, courts appropriately demand scientific evidence to support a causal conclusion, but the evidence science provides often does not well fit the legal model. Typically the best the science can do is demonstrate that the toxic exposure is a risk factor for the plaintiff’s disease – that it sometimes causes the disease. Usually there is no way to prove, scientifically, that the particular case of disease would not have occurred absent the toxic exposure.
This disjunction between the legal model of causation and the scientific evidence of causation has existed for as long as there have been toxic torts. It is not unique to the post-genomic era. What courts and legal scholars must understand, however, is that – contrary to their intuition and hopes – advances in genomics have not fundamentally changed this situation. In fact, genomics and related modes of scientific inquiry will likely increase the number of possible competing causes for many diseases by increasing our understanding of risk factors and their interactions. This would only exacerbate the uncertainties in attempting to prove exactly what deterministically caused any one case of disease.
Has genomic science begun to appear in toxic tort jurisprudence?
So far, only a little. In a few cases defendants have argued that the absence of certain molecular markers of toxic exposure is evidence that the plaintiff’s disease has some other cause. Defendants have also argued that the plaintiff’s own genes are to blame for the illness. Plaintiffs too are beginning to use molecular evidence to substitute for, or explain away the absence of, traditional epidemiologic proof of causation. More and more use of genomics and related sciences in toxic torts seems inevitable, however. To apply this evidence in the best possible way, courts should apply a probabilistic model of causation. If they won’t do that, however, in the post-genomic era courts should at least avoid making the same mistakes they made before genomics. I described some of those pitfalls in an earlier article.
Assuming further advances in our understanding of the interaction of environmental exposures and human genes, do you foresee a day in which conclusive evidence as to what made a particular plaintiff sick can be presented in court?
I would never say “never.” Even before genomics there were signature diseases. And science has made such incredible discoveries! It’s entirely plausible that for some disease and some toxic substance researchers will find a diagnostic marker that is always present when the exposure caused the disease and always absent when it did not. But that day is probably a long way off, and I question whether “conclusive evidence” will ever be typical. It seems to me that right now scientists are unveiling more and more complexity rather than finding sources of simply conclusive evidence. So long as the data are fundamentally about risks and probabilities, it would make sense for legal doctrine to think the same way.