Samuel L. Tarry Jr. is a partner in the Tyson’s Corner, VA and Washington, DC offices of McGuire Woods LLP and Davis M. Walsh is a partner in the firm’s Richmond, VA office. The authors represent the defendant in the litigation discussed below.
In Part 1 of this two-part commentary, we reviewed the Eastern District of Pennsylvania’s recent reasoning on general medical causation in Hoefling v. U.S. Smokeless Tobacco, which resulted in the court striking the plaintiffs’ expert witnesses under FRE 702 and granting summary judgment. __ F. Supp. 3d ___, No. CV 19-3847, 2021 WL 6051382 (E.D. Pa. Dec. 21, 2021). Here in Part 2, we discuss the court’s treatment of expert testimony regarding specific medical causation. General causation addresses “ruling in” potential causes for the injury generally. Specific causation requires both the ruling in of the most likely cause of the individual’s injury, as well as the “ruling out” other potential causes based on the facts of the case.
A handful of facts were most relevant to the question of specific causation. Plaintiff Hoefling, diagnosed with squamous cell carcinoma of the tonsil in his early 80s, dipped Skoal moist snuff and chewed Red Man loose leaf chewing tobacco for roughly 40 years. He claimed no significant history of alcohol consumption, a known risk factor. He also testified to no smoking history, another risk factor, but the medical records provided conflicting information.
Most importantly, Hoefling’s social history, which included four marriages, placed him at increased risk for HPV infection, which the experts agreed independently causes approximately 80% of all tonsil cancers. Due to limitations of the tissue taken at biopsy, Hoefling’s tumor was never tested for HPV, leaving that question open as a factual uncertainty. Hoefling’s well-credentialed medical experts initially claimed to “rule out” HPV by taking the position that there was “no evidence” of HPV infection, but retracted that stance at deposition based on the high prevalence of HPV generally. The experts then turned to a purported ranking methodology, claiming that it was more likely (by a slight amount) that smokeless tobacco caused Hoefling’s cancer as opposed to HPV.
The court rejected this methodology under FRE 702 based on the fact that the experts “could [not] rule out HPV” and their lack of “good grounds” for “believing smokeless tobacco use was the more likely cause.” The court’s rejection of the experts’ faulty “differential etiology” approach provides several important reminders about the threshold requirements of FRE 702.
- Specific Causation – All obvious alternative causes must be reliably ruled out.
The court followed the reasonable rule that “[e]xperts are not required to address all possible causes, but obvious alternative causes need to be ruled out.” To this end, “once a defendant points to a plausible alternative cause of the plaintiff’s illness, the expert must offer a good explanation as to why his or her conclusion remains reliable.” Again, the standard for excluding a potential cause is “good grounds” to do so.
Here, HPV undoubtfully qualified as an obvious cause, indeed the most obvious cause, given that all experts agreed it causes 80% of tonsil cancers diagnosed in the United States. The question then turned on whether Plaintiffs’ experts had good grounds to rule out HPV, which they did not.
- Specific Causation – Plaintiffs cannot claim an absence of evidence as evidence to rule out an established risk factor.
In light of the overwhelming percentage of tonsil cancers caused by HPV, it was necessary that the experts rule it out. The court rejected the experts’ attempt to reverse the burden of proof and claim that the absence of evidence of HPV was evidence HPV did not cause the cancer. Instead, the court held that an expert needs actual evidence.
In his report, Dr. Busse explains he could not “rule … in” HPV without a positive P-16 test.6 (Busse Rpt. at 4.) Dr. Chabner said he dismissed HPV as a possible cause because there was no “overwhelming evidence” Hoefling had the virus. (Chabner Dep. at 174:19–22.) Nothing short of “a wife with cervical cancer” or “a positive P-16” test would have led him to consider HPV as a cause. (Id. at 174:24–175:1.) This approach is flawed given both HPV’s pervasiveness and the low probability it would have been detected before Hoefling’s cancer diagnosis. As Dr. Busse explained, most men with HPV infections in the throat “would never know” they were infected. (Busse Dep. at 135:17–136:2.)
The court held “[a] proper differential etiology does not work like this.” Citing to Paoli, the court explained,
“[d]ifferential etiology is a reliable technique because it requires doctors to demonstrate, through a process of elimination, that their hypothesized cause is more likely than alternative causes. . . . The absence of a positive HPV test is not evidence HPV was absent. . . It was improper to treat HPV as implausible simply because no conclusive test was performed. Doing so spared Drs. Busse and Chabner’s preferred hypothesis from the rigors of scientific testing, and undermined the reliability of their differential etiologies.
- Specific Causation – Epidemiology informs specific causation as well as general causation because medical experts often must compare and weigh competing risk factors.
Throughout the court’s discussion of specific causation, it referenced the epidemiology on tonsil and oropharyngeal cancers—most notably the fact that HPV causes such a high percentage of these tumors.
The first problem with Drs. Busse’s and Chabner’s differential etiologies is that they resisted “ruling in” tonsil cancer’s predominant cause: HPV. HPV accounts for eighty percent of tonsil cancers. (Pinkerton SOMF ¶ 92.) Despite this statistic, both experts demanded affirmative evidence that Hoefling had an HPV infection before meaningfully considering it as a possible cause.
By citing to the proportion of tonsil cancers caused by HPV across the population, the court demonstrated the appropriate use of epidemiological evidence in specific causation analysis. Where, as here, there exists a dominant cause of a disease or cancer in most people, that risk factor remains the most probable explanation, statistically speaking, for any given individual until the physician reliably rules it out as inapplicable. Epidemiology thus informs specific causation analysis; because the most reliable way for an expert to demonstrate “that their hypothesized cause is more likely than alternative causes” is to cite to epidemiological studies showing that the potential cause carries high risks of the disease.
- Specific Causation – An “equal probability” of causation between two risk factors does not reach a reasonable degree of medical certainty.
Although the experts claimed to employ a ranking system for smokeless tobacco and HPV, they conceded that even in their minds, it was a close call. Indeed, one expert admitted there was “‘equal probability’ that HPV or smokeless tobacco caused the cancer, but [admitted] ‘unfortunately there’s no way to know” because the test that would have shown whether the cancer was HPV caused was not completed. In common-sense fashion, the court concluded that an “equal probability” does not equate to ruling out any cause.
Outside of the Daubert realm, the court also held that the defendants were entitled to summary judgment because Pennsylvania substantive law requires medical experts to testify to a “reasonable degree of medical certainty.” This is a substantive-law issue, specific to Pennsylvania. The court held that an expert’s testimony that there was a near equal chance that HPV caused the cancer, as opposed to smokeless tobacco, meant that the expert could not reliably testify about causation to a “reasonable degree of medical certainty.”