Safe Harbors—A Commentary by Peter Schuck
The following commentary was published in The American Lawyer on June 26, 2007.
By Peter Schuck
Consider this scenario. A drug manufacturer develops a new product, then spends a billion dollars or more to satisfy the Food and Drug Administration that the drug is safe and effective. The FDA spends many years analyzing this data, including a review by outside experts, and finally permits the company to market the drug under a carefully formulated label describing the drug's uses and risks. A consumer of the drug becomes ill and sues the manufacturer. She alleges that the drug is defective and its labeling inadequate and demands massive compensatory and punitive damages. Both the drug and the label comply with all of the FDA's rules, and no new risks have come to the attention of the agency or manufacturer, yet a jury finds the company liable. A month later, another jury in a different state dismisses an essentially identical claim, upholding both the drug and its labeling.
What is wrong with this picture, legally? In almost all states, nothing. State tort law lets juries consider or ignore the FDA's approval, substitute their own judgment that the product is unsafe and improperly labeled, and award unlimited tort damages. (Michigan is the only state where FDA approval pre-empts most tort liability; a handful of others offer limited protection to manufacturers.) Otherwise, even a drug company's full and continuing compliance with the FDA requirements provides no safe harbor against tort liability. Instead, the law casts the manufacturer onto the stormy sea of litigation where the jury may treat the FDA's expert judgment as entirely irrelevant.
This capricious state of affairs is the reason I think that the states or Congress should establish a qualified regulatory compliance defense to tort liability under which FDA approval would ordinarily, but not always, pre-empt such liability. In practice, this means that a drug manufacturer that receives FDA approval could obtain summary judgment against injured claimants, unless they could show that the manufacturer failed to fully inform the agency of risks about which it should have known.
This issue has gained fresh saliency for two reasons. First, the FDA is now issuing regulations whose preambles assert that state tort law is pre-empted. The courts will decide whether only Congress, not the agency, can do this -- an issue I do not address. Second, criticisms of FDA safety regulation have sharpened, most recently with reports that a popular diabetes drug, Avandia, increases heart problems. As I explain below, the remedy for the FDA's shortcomings is not to use tort law as a separate regulatory system but rather to strengthen its regulatory capacity, as Congress is about to do, and to limit tort to situations where manufacturers misled the agency.
New, improved drugs can save countless lives and improve the quality of life for millions. This social cornucopia, however, depends on the willingness of the pharmaceutical industry, operating in a fiercely competitive international market, to invest vast sums of money and many years of research and development in hopes of coming up with a blockbuster drug like Lipitor to make up for the far more numerous duds. But since a few adverse jury verdicts can destroy a product -- or even a company -- drug manufacturing is a high-stakes roll of the dice. To elicit the necessary large, long-term investments, some legal predictability is needed.
The FDA is responsible for balancing drug safety risks against therapeutic benefits. The agency operates through the work of specialist career scientists and policy-makers under the direction of a political appointee. Like all such agencies in technologically advanced countries, the FDA has received sharp criticism and suffered occasional scandals. Even so, the agency's technical requirements, review processes, and scientific quality are generally considered among the most rigorous in the world. But precisely because the agency's decisions are so important and inevitably controversial, the pivotal legal and regulatory policy question concerns institutional design: Who should resolve these controversies -- the agency, or a series of juries?
Even an imperfect FDA, I argue, is much better suited to make these decisions than lay juries, which are inherently far more flawed decision makers. Still, some leading analysts with no ideological axes to grind support the status quo. Robert Rabin, a pre-eminent torts scholar at Stanford Law School and the co-editor of a leading torts casebook, has published a prominent defense of the traditional rule. He also argues that products failing to comply with FDA regulations are per se defective.
Let me explain why I and some other torts scholars reject Rabin's characteristically careful analysis and instead favor a federal statutory qualified regulatory compliance defense. Rabin's first argument invokes the value of state autonomy in tort law. Absent express federal pre-emption, he contends, this value militates against any regulatory compliance defense. But because the market for pharmaceuticals is national and international, the interest in nationally uniform rules is correspondingly greater than in other tort areas, like medical malpractice or workplace injuries. When the FDA acts within its unified, comprehensive system of drug regulation, its approval should usually pre-empt state tort liability.
For this reason, Rabin's next argument -- that a regulatory compliance defense would leave a "compensation gap" for injured plaintiffs -- misses a fundamental point about tort law that he has actually emphasized in his other work comparing social insurance to tort liability. Congress has authorized the FDA to determine the socially optimal level of drug safety and information, taking other factors into account. If, as I maintain, a properly labeled, FDA-approved drug is thereby not defective, tort compensation -- as distinct from no-fault compensation for injured consumers that Congress might authorize, as it has for certain childhood vaccines -- is simply inappropriate.
Rabin doubts that tort liability in such situations would deter manufacturers. Tell that to Merrill Dow Pharmaceuticals, whose safe and effective drug for treating morning sickness, Bendectin, was driven from the market by a few jury awards wholly inconsistent with the findings of objective scientists -- and indeed of most other Bendectin juries. Tell it to the breast implant manufacturers that paid more than $4 billion to settle lawsuits based on specious claims that this widely used product caused autoimmune disease. (A decade later, the FDA is allowing the product to be remarketed.) Tell it to Merck & Co. Inc., which withdrew Vioxx, the best nonsteroidal pain reliever for many (but not all) arthritis sufferers, partly out of fear of tens of thousands of tort claims alleging that it increased the risk of heart failure. So far, most Vioxx juries have found it safe and the labeling adequate, but some have not. In early March, a New Jersey jury awarded $20 million (plus possible punitive damages) to the family of a heart attack victim who had used Vioxx for two months. A few weeks later, an Illinois jury exonerated Vioxx in the case of a longtime user. Because the plaintiffs had different risk factors, these verdicts are not necessarily inconsistent -- but they could be. We'll never know what each of these juries was thinking. My larger point, however, is that this is no way to regulate drugs.
Rabin also observes, correctly, that many drug-related risks become apparent only when adverse drug reactions come to the FDA's attention -- after it has approved a drug -- and also notes that manufacturers might knowingly submit false or incomplete information to the agency. But these troubling possibilities, far from refuting the case for a carefully limited regulatory compliance defense, suggest only what those limits should be.
First, the defense should apply only if the agency has approved the very same risk that the plaintiff says makes the drug defective. Second, the agency's rule must be one that sets an optimal safety level, not merely a floor above which a finding of defect is still appropriate. FDA drug regulation is precisely this kind of risk-optimizing scheme; it prescribes what the FDA deems the socially desirable safety level, considering cost, the risk to consumers of not using the drug, the efficacy of the specific labeling and other factors. This expert balancing judgment, reached after careful analysis, should be binding on the jury.
Third and most important, the manufacturer cannot enter the defense's safe harbor if it failed to provide the agency, in timely fashion, with all safety and efficacy information required by Congress and the FDA. This includes any post-approval information of which the company knew or should have known, and any information that it has about off-label uses, misprescribing by physicians and so forth. Thus, the company would risk tort liability if (1) it failed to keep the agency fully informed about any facts relevant to the continuing rationale for its approval, or (2) the company's advertising or other communications misled physicians or consumers about the drug's safety or efficacy. (Indeed, one who knowingly did so would invite criminal prosecution.)
A regulatory compliance defense (including its exceptions) that expressly pre-empted inconsistent state laws would make a difference in certain kinds of cases. Where symptoms are so common in the general population (e.g., cardiovascular events among Vioxx users) that the drug's causal responsibility cannot be detected until post-approval reports have accumulated, the defense's availability would depend on how promptly the company learned of and reported this data. In the more usual situation where serious risks are so rare that they simply cannot be detected by pre-approval tests but can only be seen after large-scale consumer use, the defense would also depend on how well the company monitors side effects and how quickly it discloses them to consumers, physicians and the FDA.
The key rationale for this limited regulatory compliance defense is the superior institutional competence of the FDA compared with a lay jury. Juries have many virtues, but making technocratic decisions about drug design and labeling is not one of them. (Indeed, savvy lawyers would probably use voir dire to exclude any potential juror who knew anything about these subjects.) But the FDA's comparative advantage in balancing risk, benefit and cost is not just technocratic. It is also politically accountable for its decisions through constant congressional hearings and discipline, Government Accountability Office audits, media reports, peer review and professional oversight, judicial review and even internal whistle-blowers. (A House committee, for example, is now studying how the FDA regulates anemia drugs.) A jury, in contrast, has all the responsibility of a one-night stand. Even the smartest, most conscientious juror cannot comprehend and take responsibility for the larger context of the FDA's continuous, complex policy development. By design, the jury is a black box. It lacks even the minimal discipline of having to give reasons for its decisions, thus leaving little or nothing for a court (or anyone else) to review.
The FDA, of course, is very far from being perfect. Egregiously slow to improve its post-approval adverse effects reporting system, the agency has left many consumers exposed to newly discovered drug risks. Critics accuse it of being all too accountable -- but to Big Pharma, not to the public interest. This "agency capture" notion, however, is simplistic. Many factors and interests influence agency decisions, and money does not always call the tune -- if only because there is lots of money on every side of the kinds of complex regulatory issues that the FDA handles. Indeed, other critics accuse the FDA of being too risk-averse, often sacrificing large potential health benefits in order to prevent very small risks of harm to relatively few consumers. The FDA, in this view, is so obsessed with avoiding a Thalidomide-type disaster that it delays or denies approval of some drugs with high benefit-risk ratios.
Still, there are legitimate concerns about FDA sluggishness, coziness with industry, error and risk-aversion. The solution, however, lies in improving how it makes its decisions, not in allowing a lay jury, which is less accountable, more error-prone and subject to its own biases, to override, ignore or misapprehend those decisions. If the agency needs more funds, better information and transparency, greater authority to conduct post-approval surveillance and limit direct advertising to consumers, and more insulation from industry pressures and balanced incentives, then Congress should unquestionably provide them -- as pending bipartisan reform bills promise to do.
Finding the regulatory sweet spot -- the ideal compromise among the goals of drug safety and availability, reasonable cost, timely agency decisions and consumer information and choice -- is an enormous challenge that a limited regulatory compliance defense would help to meet.
Peter Schuck, a law professor at Yale, is the author most recently of "Targeting in Social Programs "(Brookings) and "Meditations of a Militant Moderate" (Rowman & Littlefield).