Using the Bradford Hill Criteria to analyze injury causation opinions of expert witnesses under Daubert
When litigating bodily injury claims, medical experts proffering causation opinions frequently cite to medical and epidemiological research to supporting causal links between various risk factors and a particular outcome or disease. Whether a hypothesis can be tested, and whether it has been tested, is one of the factors for courts to consider in evaluating the admissibility of expert testimony under Daubert v. Merrill Dow Pharmaceuticals. Attorneys challenging causation opinions in litigation must be prepared to critically analyze the scientific literature that supposedly supports the expert witness’s testimony. Attorneys must be prepared to challenge the scientific premises underlying the expert’s opinion at the deposition stage in order to develop testimony required to later challenge the expert’s opinions via a Daubert motion.
In 1965, British physician Sir Austin Bradford Hill proposed nine factors to consider in determining whether an observed association between two phenomena can establish a causal relationship. Epidemiologists refer to these as the “Bradford Hill Criteria.” They are:
- Strength of association. By what factor does exposure to the risk increase the likelihood of developing the disease? The higher the multiplication of risk, the more likely the risk factor causes the outcome.
- Consistency. Have multiple studies, incorporating a variety of confounding variables, demonstrated a consistent association between the risk factor and the disease?
- Specificity. Does the exposure case only one disease?
- Temporality. Does exposure precede the onset of disease?
- Biological gradient. Is there a dose-response relationship? Does exposure to more of the risk factor increase the likelihood of developing the disease?
- Plausibility. Is there a biological explanation of the mechanism by which the exposure to the risk factor causes the disease?
- Coherence. Does the cause-and-effect relationship make sense given all the other knowledge and information available to the expert?
- Experiment. Does the disease incidence change if exposure to the risk factor is reduced or eliminated?
- Analogy. Is the purported causal agent similar to another agent for which a causal link with the disease has already been established?
Focusing on the strength of one of the above criteria at the cost of others may lead the expert to incorrect conclusions about the validity of a causation opinion. In traumatic brain injury litigation, expert witnesses often latch onto one of these criteria as a sole basis for concluding a causal connection between exposure to a risk and the development of a disease or an outcome. In particular, experts often rely heavily on temporality as a basis for concluding causation—i.e., the outcome came after the accident, so the accident must have caused the outcome. This “temporal causation” theory—also called post hoc, ergo propter hoc, or “after this, therefore because of this” —has been soundly rejected as unscientific by state and federal courts applying the Daubert criteria. E.g., Terry v. Caputo, 115 Ohio St. 3d 351 (2007) (precluding expert from testifying that plaintiff’s mold exposure caused the plaintiff’s symptoms because the expert’s opinion “relied too heavily on the temporal proximity of the exposure to the onset of symptoms.”). As another example, just because a causal relationship is biologically plausible does not mean that a litigant’s disease was caused by the purported agent at issue in the case—especially where other factors like strength of association and specificity are weaker. Thus, focusing on one criteria at the expense of others may thus lead to invalid conclusions about causality. Experts—and attorneys critically evaluating expert opinions—must evaluate the validity of a causation opinion based on multiple factors and the totality of available information.
Under Daubert, expert witness testimony must be based upon the scientific method. Although, like the Daubert criteria, the Bradford Hill Criteria cannot be applied like an inflexible checklist in every case, the Criteria provide a helpful framework for attorneys to use in critically analyzing the epidemiological literature in preparing to depose an expert witness regarding causation opinions. Using the criteria to frame questions regarding both the expert’s own causation opinions and the validity of the medical or epidemiological literature underpinning those opinions can be an important strategy to prepare for an eventual Daubert motion to challenge the admissibility of the causation testimony.
The original paper describing the Bradford Hill Criteria, published in Austin Bradford Hill, “The Environment and Disease: Association or Causation?” Proceedings of the Royal Society of Medicine, 58 (1965), 295-300, is available here.