April 1995 No. 5
Breast Implants: A Study in Phantom Risks
March 1, 1995, marked the final day women who believe they were harmed by silicone gel breast implants could register for a share of the $4.25 billion class-action settlement that plaintiffs' attorneys had won from implant manufacturers. Some 400,000 women had registered by the deadline.
Yet despite the vast sums of money involved, the danger of disease from silicone breast implants is a "phantom risk"—a risk whose very existence is unproven, probably nonexistent. No longterm scientific study has demonstrated that breast implants cause any serious disease. The defendants were nonetheless forced to pay billions of dollars, because courts imposed on them the scientifically impossible burden of proving that the implants were absolutely safe, that each dubious theory of risk was false.
Since the early 1980s, courts have often allowed plaintiffs to base their arguments on scientifically unsound theories, making it far easier for them to prevail in product liability cases. The Supreme Court's 1993 decision in Daubert v. Merrell Dow Pharmaceuticals signaled a change of course, imposing a stricter standard. The breast implant fiasco illustrates the dangers of allowing bad science into court and demonstrates that judges must be more vigilant in holding plaintiffs to the new standard set by the Supreme Court.
Litigation over injuries allegedly caused by silicone gel breast implants started in the late 1970s. But despite the best efforts of the Naderite group Public Citizen and trial lawyers, tort suits remained a minor nuisance for implant manufacturers until 1991. In December of that year, attorneys representing a woman claiming illness linked to breast implants won a $7.5 million jury award, including $6.5 million in punitive damages.
On the heels of this widely reported verdict, Food and Drug Administration Commissioner David Kessler banned silicone gel breast implants for cosmetic purposes. Kessler acknowledged that he was banning them only because manufacturers had failed to demonstrate their safety conclusively, as required by statute.
But plaintiffs' attorneys used the fear generated by the FDA decision to sign up tens of thousands of new clients. At the time no major longterm epidemiological studies had been conducted, leaving implant manufacturers especially vulnerable to a variety of scientifically unfounded accusations.
Initially, plaintiffs' attorneys had focused on the alleged risk of breast cancer, basing claims on animal studies and preliminary FDA calculations that a small risk might exist. Although the scientific community generally did not accept the link between silicone implants and cancer, one plaintiff won a multimillion-dollar verdict based on the alleged connection. In the spring of 1992, however, researchers published two epidemiological studies showing that breast implants were associated with a reduced rate of breast cancer. These studies discouraged attorneys from basing further claims on the risk of breast cancer.
Lawyers shifted their focus to claims that silicone breast implants cause immune system disease. Occasional articles had appeared in the scientific literature presenting anecdotal case reports of women who developed immunological disorders after receiving silicone implants. A few physicians and researchers began to believe that there might be a connection, but most of those who looked into the issue were highly skeptical.
This skepticism proved justified. In June 1994, the New England Journal of Medicine published a major epidemiological study that found "no association between implants and the connectivetissue diseases and other disorders that were studied." The study offered the first definitive scientific evidence that silicone breast implants did not cause the diseases alleged by plaintiffs' attorneys. Since then, several major epidemiological studies of women with breast implants have likewise shown no association between breast implants and immune system disease. But these findings came too late. Defendants, having lost tens of millions of dollars in several jury trials, had already agreed to the $4.25 billion classaction settlement.
Meanwhile, plaintiffs' lawyers—representing an estimated 11,000 women with implants who chose not to participate in the classaction suit—have come up with a new theory. They now argue that even if women with breast implants do not have unusual rates of recognized immune system disease, a disproportionate number of those women have "clusters" of symptoms that constitute "atypical" autoimmune disorders.
The beauty of the atypical disease theory—from the trial lawyers' perspective—is that it cannot be disproved. Because there is no clear definition of an "atypical" disease, it is impossible to design an epidemiological study to determine the existence or causes of such a condition. Hundreds of thousands of people suffer from a variety of symptoms—chronic fatigue, insomnia, depression, headaches, muscle and joint pain, and so on—that do not constitute a recognizable immunological disease. Based on chance alone, these symptoms will occur in women with breast implants as well as in women without implants. Yet lawyers, arguing that these symptoms are caused by silicone, have been able to turn a phony scientific claim into real dollars.
The breast implant cases reflect changes that have been under way in product liability law for more than a decade. Traditionally, courts placed the burden of proof in product liability lawsuits on the plaintiff, who was required to prove that the defendant's product caused his injury. Beginning in the early 1980s, however, some legal scholars began to argue that in "toxic tort" cases—those involving alleged injury from exposure to environmental pollutants or pharmaceutical products—the traditional standard is too strict. These reformers argued that plaintiffs should only be required to present some scientific evidence showing that the product could have caused the injury. This approach would shift the burden of proof to the defendant, who would have to disprove causation.
Many courts explicitly adopted this theory, allowing plaintiffs to proceed to trial based on weak, speculative evidence. Once a case reaches a jury—made up of ordinary citizens with no scientific expertise—science becomes a secondary concern. The case turns into a morality play, in which the plaintiff's lawyer presents his client as the sympathetic victim of rapacious companies. A former professor at Yale Law School has written, "Toxic tort cases are about good and evil, about corporate geed and indifference, and about the risk of the unknown. Above all, toxic tort cases are about redefining our public morality for a new era."
Plaintiffs' attorneys found they could boost their cases by presenting the jury with documents suggesting that the defendant had acted improperly when testing or marketing its product—even if the alleged impropriety had nothing to do with the claim of injury. Lawyers got help from sensationalistic media coverage, publicityseeking politicians, selfstyled public interest groups, and paid expert witnesses whose views were outside the scientific mainstream. Thus, they were able to create a public perception of grave risk from a myriad of products, based purely on speculation that most scientists rejected. This perception, in turn, led to still more litigation.
Even when the scientific evidence was overwhelmingly in the defendant's favor, as was the case with the morning sickness drug Bendectin, the massive cost of fighting such lawsuits frequently prompted companies to pull safe and effective products from the market.
The Daubert Case
All of this may change if courts follow the 1993 Daubert ruling, in which the U.S. Supreme Court held that lowercourt judges should serve as "gatekeepers," forbidding plaintiffs from presenting scientific evidence to a jury unless they can show that the evidence is "not only relevant but reliable." A key test of scientific reliability is "falsifiability": it must be possible in theory to demonstrate that a claim is false. A claim that is too vague to be falsifiable has no scientific force.
Some judges have been loath to exercise their gatekeeper role. In Hopkins v. Dow Corning, the Ninth Circuit Court of Appeals upheld a $7.3 million damage award to a California woman who claimed she had contracted a rare immune system disease from implants she received after breast cancer surgery. The court subjected the plaintiff's scientific evidence to only the most cursory review, failing to ensure that it met the standards promulgated in Daubert.
Most judges, however—including others on the Ninth Circuit—are adhering far more closely to Daubert's mandate. Appropriately, they are changing their focus from "redefining our public morality" to the more prosaic task of ensuring that multimillion dollar damage awards are supported by reliable scientific evidence.
The implant litigation is simply an extreme example of the way in which courts have been inundated with claims based on phantom risks. Such litigation has driven safe and effective products, such as Bendectin, off the market and has discouraged research and development into contraceptives, vaccines, and other pharmaceuticals. If courts enforce Daubert and put the burden on plaintiffs to show that their claims are based on scientifically valid evidence, they will go a long way toward making the justice system fairer and the market more responsive to consumers.
--David E. Bernstein
About the Author
David E. Berristein is the Mellon Foundation Research Fellow in the Julius Silver Program in Law, Science and Technology at the Columbia University School of Law. He edited, with Kenneth Foster and Peter Huber, PHANTOM RISK: Scientific Inference and the Law, published by MIT Press in 1993.