Plaintiffs in Milward v. Acuity Specialty Products Group, Inc., 2011 U.S. App. LEXIS 5727 (1st Cir. Mass. Mar. 22, 2011), sued three chemical companies in the United States District Court for the District of Massachusetts, claiming workplace exposure to their benzene-containing products had caused the plaintiff husband to develop acute promyelocytic leukemia (“APL”). After a four-day evidentiary hearing, the district court excluded plaintiffs’ expert’s “general causation” opinion that benzene is capable of causing APL in humans generally, ruling that plaintiffs had not demonstrated the testimony to be reliable as required by Daubert v. Merrill Dow Pharmaceuticals, Inc., 509 U.S. 579 (1993). Based on the lack of essential expert testimony, the court entered judgment for defendants.
The United States Court of Appeals for the First Circuit reversed. The court held that the expert had indeed applied a reliable methodology, namely making a scientific judgment that the “weight of the evidence,” considering generally accepted criteria enumerated years ago by the British epidemiologist Sir Arthur Bradford Hill as well as additional similar criteria, “supported the inference that the association between benzene exposure and APL is genuine and causal.” The Bradford Hill criteria include such factors as the temporal relationship between exposure and disease, strength of the association between the two, presence of a dose-response relationship and biological plausibility of causation in light of existing scientific knowledge.
At the outset, the court acknowledged that APL was a relatively rare subtype of acute myeloid leukemia (“AML”), which was itself one of four broad leukemia types, and that while there was a scientific consensus that APL was in part caused by a particular genetic mutation there was no consensus as to the cause(s) of that mutation. In addition, while there was epidemiologic evidence of a statistically significant association between benzene exposure and AML generally, there was no such evidence specifically with respect to APL.
The district court had found the expert’s causation opinion unreliable because, among other things, it relied upon data concerning the genetic mutations involved in other AML subtypes, as well as animal and in vitro studies about the possible role of benzene metabolites in causing those mutations, to support the conclusion that benzene caused APL even though it involved a different mutation. The appeals court held, however, that in so doing the district court had “placed undue weight on the lack of general acceptance of [the expert]’s conclusions and crossed the boundary between gatekeeper and trier of fact” by making its own “evaluation of the weight of the evidence.” The appellate court also suggested the trial judge had misunderstood the Bradford Hill/weight of the evidence methodology as requiring that each criterion analyzed by the expert by itself reliably demonstrate causation, rather than merely that the criteria cumulatively do so.
It appears from the court’s opinion, however, that the court committed precisely the error of which it accused the trial court— namely, of misunderstanding the Bradford Hill methodology, which is only intended to guide the determination of causality where epidemiological studies have demonstrated a statistically significant association between the exposure and disease at issue. Indeed, many of the Bradford Hill criteria themselves— such as the strength of the association and the presence of a dose-response relationship—only underscore this fact. Here, there was no such association, and the court did not purport to analyze whether there was any methodology that supports drawing scientifically reliable conclusions, as opposed to merely offering hypotheses for further investigation, about the causation of one disease or genetic mutation from evidence about another.