A ruling in one of the Cymbalta cases last week confirms that Daubert requires an expert’s testimony to be based on his particular area of expertise and on a methodology recognized by a community greater than one.

Plaintiff Erin Hexum is one of many who claim Eli Lilly and Company (“Lilly”) failed to warn her of the risk and severity of side effects upon ceasing to take Cymbalta, a prescription antidepressant medication. See Erin Hexum et al v. Eli Lilly and Co., No. 2:13-cv-02701 (C.D.Ca.). Lilly moved to exclude the testimony of two experts: Dr. Joseph Glenmullen, a clinical psychiatrist who professes to be a “national spokesperson for the appropriate, measured use of psychiatric medications,” and Dr. Louis Morris, a Ph.D. in social psychology and former Chief of FDA Marketing Practices and Communication Branch.

While the court did not grant Lilly’s motion to exclude Glenmullen’s testimony entirely for alleged lack of reliable methodology, “cherry-picking” through documents and data, and periodically providing supplemental reports explaining how “newly” discovered data supports his previously expressed opinions (although it acknowledged these practices could raise credibility questions), it did exclude proposed testimony on several specific issues.

For one, the judge agreed that Glenmullen’s proposed testimony regarding Cymbalta’s label’s adequacy, veracity, and tendency to mislead was inadmissible, because the history of Eli Lilly’s clinical studies and marketing efforts has nothing to do with his psychiatry expertise. Glenmullen’s proposed testimony regarding Lilly’s “motive” and “conduct,” specifically that the company concealed an article on Cymbalta’s risks in an “obscure” Dutch medical journal unlikely to be read by physicians prescribing the drug, was likewise inadmissible. The court found that Glenmullen failed to cite any relevant authority supporting his opinions regarding physicians’ general prescribing practices, reliance on the label, whether they would have been familiar with the article at issue, and whether Lilly intended to obscure clinical trials. Glenmullen was also precluded from simply recounting the history of Lilly’s clinical studies, marketing efforts, and publication of the article, because such testimony “is neither helpful to the trier of fact nor based on his psychiatry expertise.” Order at 8.

The court also ruled that Glenmullen’s opinion regarding the severity of discontinuation symptoms, aimed to support the claim that the labels do not adequately describe withdrawal symptoms, was inadmissible because he failed to establish that his definitions of “mild,” “moderate,” and “severe” were based on objective criteria: “[I]t appears that Glenmullen invents his own definitions and calls the Cymbalta label inadequate on that basis. The only authority that he cites for these definitions is his own book (which he cites generally without pointing to any specific pages).” Order at 7.

Finally, the court excluded some of Glenmullen’s proposed testimony on specific causation primarily on the grounds that as a psychiatrist, he is not qualified to opine regarding plaintiff’s need for medical treatment for fibromyalgia.

As for what remains of Glenmullen’s proposed testimony, the court ruled that he may opine that Cymbalta can cause discontinuation symptoms and may testify regarding the nature of those symptoms, the clinical data he examined, and his critique of the methodology of certain studies. He can opine that Cymbalta’s label does not adequately reflect the clinical data regarding Cymbalta and may testify about the alleged correlation between half-life length and the risk of discontinuation side effects.  Finally, he was permitted to opine that based on Cymbalta’s half-life, a reasonable drug manufacturer should have known that Cymbalta carried a significant risk of discontinuation side effects.

Lilly also moved to exclude in its entirety the testimony of Dr. Morris, whom plaintiffs offered to testify regarding the adequacy of Cymbalta’s label. The court found that plaintiffs failed to show that Morris’s opinion regarding the label’s adequacy was based on sufficient data regarding physician’s’ knowledge or was sufficiently related to Morris’s experience. To defeat Lilly’s argument, Morris claimed to have gained significant experience reviewing “promotional materials and the labeling of prescription drugs” while working at the FDA and that he arrived at his conclusions using the same methodology as the FDA. However, the court reasoned that even if Morris’s assertions regarding the FDA’s procedures were correct, he is not an expert with sufficient training and experience to determine whether the information in Cymbalta’s label is truthful and supported by sufficient data, because he is not a clinical psychologist and has never prescribed medication to patients. Order at 11. Morris was deemed similarly unqualified to opine whether Cymbalta’s label is inadequate as a matter of semantic and linguistic interpretation, because he was not offered as an English language expert, and his opinion regarding how a general “reader,” as opposed to a physician, would interpret the label, would not be helpful to the trier of fact. Order at 12.

Based on its findings, the court designated Morris as a rebuttal witness and limited his testimony to the FDA’s general procedures, practices, relevant regulations and Guidelines and his opinion regarding whether applicable laws and regulation would have permitted Lilly to modify Cymbalta’s label.

This decision represents a careful exercise of the court’s gatekeeping function to significantly narrow, without excluding altogether, the scope of plaintiffs’ proffered expert testimony.  Rightfully so, the court was scrupulous to prevent a clinical psychiatrist and social psychologist from testifying outside their respective areas of expertise to opine on issues relating to the adequacy of labels, corporate conduct and motives, and practices of prescribing physicians.