Daubert and Suicide Risk of Antidepressants in Children

Facts and History of the Case

In the spring of 1997, Matthew Miller, 13 years of age, disclosed to his teacher and peers his thoughts of suicide. The school contacted Matthew’s parents and urged them to seek mental health treatment. On June 30, 1997, the Millers took their son to Dr. Douglas Geenens, a child and adolescent psychiatrist who diagnosed Matthew as having “depression not otherwise specified.” After three weeks, in which Matthew showed little improvement, Dr. Geenens prescribed sertraline, a selective serotonin reuptake inhibitor (SSRI) marketed as Zoloft by Pfizer, Inc. One week later, Matthew hanged himself.

The Millers sought to hold Pfizer liable for their son’s death, claiming both that the company had influenced Dr. Geenens to prescribe Zoloft through aggressive marketing practices that misrepresented the medication, and that it had failed to test the drug adequately and to warn of its potential to induce akathisia and suicide. The Miller’s case required demonstrating both general causation linking Zoloft and suicide and specific causation in the case of Matthew Miller.

The details of the case’s procedural history provide the context for the appellate court’s eventual decision. The Millers filed the initial civil complaint on July 27, 1999. Shortly thereafter, they submitted a preliminary report prepared by their expert witness on causation, Dr. David Healy, a neuropsychopharmacologist and vocal critic of the pharmaceutical industry. In his report, Dr. Healy opined that Zoloft and other SSRIs cause akathisia, which in turn can lead to suicide in some patients. The Millers then filed a motion to appoint independent experts in October, in anticipation of challenges from Pfizer.

The U.S District Court for the Twelfth District (trial court) set initial deadlines for the Millers to provide Pfizer with disclosures in accordance with Federal Rule of Civil Procedure 26 governing discovery of evidence and to provide disclosures of rebuttal experts in December 1999 and February 2000, respectively. Pfizer’s motion to limit revisions to expert opinions was granted in part in January 2000, so that the Millers were limited to filing a final supplemental expert report by March 7 and final rebuttal expert disclosures by March 28.

After Dr. Healy was deposed on March 27–28, 2000, the Millers provided Pfizer with supplemental responses explaining a statistical analysis Dr. Healy had used in forming his opinion. In April, Pfizer filed an emergency motion, arguing that the responses were untimely and that Dr. Healy should be barred from supplementing his opinion. The district court denied this motion the following day, reasoning that the new information did not substantially depart from the previous disclosures. On the same day, Pfizer filed to exclude Dr. Healy’s testimony as failing to meet the Daubert standard for expert testimony.

Of note, Pfizer also moved to exclude other plaintiff’s experts under Daubert. It sought to exclude Dr. Morton Silverman, a suicidologist who was to testify that Zoloft can cause akathisia and that Matthew Miller may have had akathisia, and Dr. Donald Marks, an expert on the pharmaceutical industry who was to testify that Pfizer had a duty to warn about suicide risk and conduct further testing on the matter. In June, the court denied both of Pfizer’s motions to exclude the testimony altogether, but ruled to exclude Dr. Silverman’s opinion about Pfizer’s duty to warn.

Pfizer’s counsel became ill, delaying proceedings for several months. The district court issued an order in August staying all proceedings and directing the parties to show cause why independent experts should not be appointed to assist the court in determining the admissibility of Dr. Healy’s testimony. Then, in December 2000, the court ordered the parties to submit either a joint nomination for an independent expert or separate lists, each proposing three experts.

Finally, in April 2001, the court appointed Drs. John Concato and John M. Davis as independent experts. They were charged with determining the quality of Dr. Healy’s methodology and whether its application to the question of general and specific causation represented valid, scientifically reliable reasoning. The independent experts submitted their report in September, and a Daubert hearing was held in November 2001.

Informed by the findings of the independent experts, the district court ruled that Dr. Healy’s testimony was inadmissible in February 2002. The court noted that Dr. Healy’s views on general causation were “a distinctly minority view” and that he relied on case reports and his own studies over randomized controlled trials (RCTs) and large-scale epidemiological studies. The court held that, though RCTs and epidemiological studies were not essential to the formation of an expert opinion, Dr. Healy’s calculation of relative risk of suicide for persons taking Zoloft could not be replicated by the independent experts and that his method had not been subject to peer review. In addition, the court ruled that the design and methodology of Dr. Healy’s own studies fell short of the Daubert standard.

As for specific causation, the trial court found that Dr. Healy had relied too narrowly on “preselected information from interested parties” and that this method in forming his opinion was not generally accepted practice. It also deemed inadmissible Dr. Healy’s opinions on suicidology, warning labels, and Food and Drug Administration (FDA) regulations.

Given that the opinions of the plaintiff’s expert on causation had been deemed inadmissible, the court granted Pfizer’s motion for summary judgment on February 11, 2002. The Millers appealed this decision to the 10th Circuit Court of Appeals, claiming that the trial court had not given them a fair opportunity to make a proper record supporting Dr. Healy’s conclusions and had abused its discretion in evaluating expert testimony.

Ruling

The 10th Circuit Court affirmed the trial court’s summary judgment, holding that the trial court did not abuse its discretion in limiting the information provided to the independent experts and that the scope of its Daubert inquiry into Dr. Healy’s opinions was appropriate.

Reasoning

In its reasoning, the 10th Circuit Court rejected the Millers’ argument that the district court had misapplied Federal Rule of Civil Procedure 26. The Millers objected to the restrictions placed on their responses to the report of the independent experts, claiming that the trial court did not offer an adequate opportunity to present and defend Dr. Healy’s opinions. The court ruled to the contrary, delineating the procedural history of the case and highlighting the numerous opportunities Dr. Healy had to revise and supplement his opinion. The court acknowledged case law in which district courts had been found to have abused their discretion (Dodge v. Cotter Corp., 328 F.3d 1212 (10th Cir. 2003)) but ruled that such an abuse had not occurred in this case, citing the trial’s court allowance of Dr. Healy’s revision of his original opinion and his supplemental responses.

The circuit court addressed three specific concerns raised by the Millers. First, it took on the Millers’ assertion that the district court had improperly limited the information supplied to the independent experts and, in doing so, had allowed Pfizer to prevail based on a procedural technicality rather than on the science. The court recognized that revisions to an expert’s initial Rule 26 report are permitted and may be required, but ruled that Dr. Healy had ample opportunity to update his opinion and respond to critiques. Noting that “the orderly conduct of litigation demands that expert opinions reach closure,” the court ruled that the trial court did not abuse its discretion and had acted with “patience and concern for fairness.” Second, the Millers argued that the trial court had gone beyond its gatekeeper function, addressing questions relevant to the weight rather than the admissibility of Dr. Healy’s testimony. The circuit court disagreed, finding that the scope of the district court’s investigation was appropriate for a Daubert inquiry. It noted several examples of case law supporting “careful and meticulous review” of scientific evidence as within the scope of Daubert. Last, the Millers contended at oral argument on appeal that the trial court erred in granting summary judgment because it had sufficient evidence without Dr. Healy’s testimony to go forward. The circuit court declined to consider this argument, however, because the plaintiffs had not raised it at the trial court level or in appellate briefs.

Discussion

The trial court applied the Daubert standard to expert opinion on an issue that has come under increasing scientific and public scrutiny: antidepressants and suicide risk in children. The circuit court’s opinion deals with the problem of marshaling scientific debate in the legal arena. Specifically, it addresses the extent to which expert opinions can be refined and updated in an area of evolving research. The opinion emphasizes the differing nature of scientific and legal inquiry. While scientific knowledge is continually evolving, the legal arena demands a level of certainty and peer-reviewed acceptance that often preclude leading-edge science in rapidly evolving areas of research. The legitimacy of a technique or method may turn on the results of one study; what is hypothetical today may be textbook science in a month. Ironically, in this case, the science necessary for demonstration of general causation may have matured to meet Daubert standards of testability, peer review, and general acceptance over the procedural history of the case.

Recently, the body of evidence supporting a link between antidepressants and suicide risk in children has grown. Governmental agencies have responded with increased surveillance and warnings. In June 2003, the Medicine and Health Care Products Regulatory Agency of the British Department of Health warned against the off-label use of paroxetine in children. Subsequently, the British Committee on Safety of Medicines recommended that only fluoxetine be prescribed in pediatric populations. In the United States, the FDA issued a talk paper in October 2003, emphasizing the lack of certainty about the safety of antidepressants in children. In February 2004, the FDA asked manufacturers of 10 antidepressants to include a warning that patients be observed for treatment-emergent suicidality. In August 2004, the agency released results of a meta-analysis of pharmaceutical company clinical trials that showed an increased relative risk of suicidal behavior in children receiving the drugs. Such developments invite speculation about whether Dr. Healy’s opinions, supplemented with the above-mentioned studies, might have passed Daubert muster.

Controversy has also emerged over suppression of research data, recalling Miller’s contention that Pfizer had misrepresented the dangers of its drug and failed to test it adequately. In September 2004, a congressional panel examining antidepressant use in children criticized the FDA for silencing Dr. Andrew Mosholder, a researcher it had charged with investigating a link between the drugs and suicide. The panel found evidence that the FDA had suppressed Dr. Mosholder’s findings in February 2004 and that, as early as September 2003, data were available suggesting a link. The apparent intentional obfuscation of risk by a regulatory agency does little to reassure the public that a drug manufacturer might not do the same.

Would Miller v. Pfizer have had a different outcome had it been litigated today? Certainly, public outcry over the FDA’s failure to disclose promptly the risks of antidepressant use in children would have made for a different climate for the case. However, despite some evidence supporting a general causal link between antidepressants and suicidal behavior in children, showing specific causation in the case of a child with depression is complicated by the inherent risk of suicide due to the disorder itself. A more robust argument for specific causation might be made in a case involving a child’s suicide during antidepressant treatment for an indication with less intrinsic suicide risk, like obsessive-compulsive disorder or generalized anxiety disorder. However, since the purported increased relative risk of suicide with these agents has only been gleaned from studies of depressed children, the case for general causation might be hampered by a lack of research sufficient to meet the requirements of Daubert.