In trial or deposition I have been asked the question: “are you requiring scientific or legal certainty.” This question arises in the context of a causal assessment in, for example, a toxic tort or pharmaceutical liability claim. The question is designed to be a “gotcha.” Why? Because attorneys believe that scientific certainty is more rigorous (i.e. 95%) than legal certainty (more probable than not or 51%.) There is truth to that belief, but the question actually compares apples to oranges. Scientific certainty is needed for one aspect of an analysis. Legal certainty is required for a different aspect. Thus, far from a “gotcha,” this question reflects the attorney’s naivete regarding the application of probabilities to the respective issues–scientific versus legal.
Scientific “certainties” (actually scientists don’t use this legal term of art, but I shall explain it from an attorney’s point of view), is a reflection of the quality of the underlying data. In other words, are the studies that underlie the claim at issue statistically significant. Is there a scientifically-accepted likelihood that an observed relationship is not simply due to chance. That’s where the 95% number comes from. A p value of 0.05, by convention the cutoff between statistically significant and not, is that 95% likelihood. But that percentage applies only to the quality of the science used to assess causality in a claim. If, in fact, a study showed only a 51% likelihood of reflecting a true, rather than a chance relationship, no scientist, no regulatory body, no one who reviews scientific data would consider that study supportive of a causal relationship. In the words of legalese: “the relevant scientific community would consider the use of such a study methodologically improper.”
By contrast, the legal issue of more probable than not or reasonable medical/scientific certainty asks an entirely different question. First, it assumes that there is recognized and accepted scientific support for a position. In other words, it first requires that there are studies which meet the generally accepted scientific test of 95%. Then it asks, is it more probable than not that X caused Y in this case. It is the answer to that legal question–the more probable than not question–that requires only a 51% assurance.
As an example. Assume that chemical X has been strongly connected scientifically to disease Y. Scientific studies have shown (with a p value of 0.05 or 95% confidence) that X is connected to Y. I have assumed for this argument that there are also sufficient numbers of studies and that they have also shown meaningful attributable risks. Now assume that we have a claimant who has been exposed to X and has developed disease Y. In that case general causation–that X can cause Y–is scientifically satisfied. The next question–did it actually cause this person’s disease requires other elements. Sufficiency of dose, correct timing of exposure and others are among them. If those, too, are affirmatively satisfied, then an expert might state that it is more probable than not (greater than 51%) that this claimants disease Y was caused by agent X. On the other hand, if the studies linking X to Y were not available, or if they were not, by scientific standards (95%), positively connected, then one could not even approach the legal causation question, because there would be no scientific foundation upon which to base it. Said another way: if X has not been shown scientifically (using scientific measures of statistical significance) to be connected to disease Y then a claim that disease Y came from exposure to agent X, has neither scientific nor legal merit.
Therefore, scientific certainty (step 1) is, indeed, more rigorous than the ultimate test of legal certainty (step2). However one doesn’t get to step 2 without first satisfying step 1. Thus, both scientific and legal certainty are part of a causal assessment. And, asking a witness whether his test was one of legal or scientific certainty makes no sense. The proper answer is “both, each for a different component of the analysis.”
Ronald E. Gots, M.D., Ph.D.
International Center for Toxicology and Medicine