Supreme Court Botches Preemption Case

The Supreme Court handed down its decision last month in the case of Wyeth v. Levine, ruling that federal law did not bar plaintiff Diana Levine from suing pharmaceutical maker Wyeth over allegedly insufficient drug safety warnings, even though the warnings had been approved by the Food and Drug Administration (FDA). This decision establishes the troubling precedent that a sympathetic jury can now supersede the expert opinions of the FDA on what qualifies as adequate safety labeling.

Ms. Levine lost an arm to gangrene after a physician’s assistant injected the Wyeth drug Phenergan in such a way that it came into contact with oxygen-rich arterial blood. Although the drug’s label explicitly warned that doing so poses a high risk of tissue damage, Levine claimed that the label should have instructed physicians not to use this intravenous “IV-push” method at all. A Vermont jury agreed, and awarded Levine $7.4 million, which the court reduced to $6.7 million to account for an earlier settlement with the physician’s assistant and supervising doctor.

On appeal, Wyeth argued that the FDA’s extensive regulation of drug labeling should preclude claims of negligent failure-to-warn, but the Supreme Court rejected that argument by a 6-3 majority. According to Justice John Paul Stevens’s majority opinion, “The history of the [Food Drug and Cosmetics Act] shows that Congress did not intend to pre-empt state-law failure-to-warn actions.” Although true in a general sense, the Court failed to recognize that this is not a typical failure-to-warn case, and that permitting the suit here would interfere with the FDA’s authority to regulate drug labels.

As Justice Samuel Alito’s dissenting opinion notes, Ms. Levine conceded that, in 1988, Wyeth proposed a label change that “if followed, would have prevented the inadvertent administration of Phenergan into an artery,” but the FDA rejected that language. Nevertheless, Ms. Levine alleged not only that the warning on Phenergan’s label wasn’t strong enough, but that Phenergan was “not reasonably safe for intravenous administration” and that the label should have said so. The latter claim, however, poses a question regarding the FDA’s approval of the product for that use, not Wyeth’s alleged negligence in drafting the label.

What Happened With Phenergan

The FDA first approved Phenergan in 1955. As the risk of arterial injection became apparent over the following years, the agency approved several changes to the drug’s label, which now contains six statements (two in all capital letters and bold face type) warning doctors about the exact nature of that risk.

The doctor and physician’s assistant who treated Ms. Levine nevertheless administered a dose twice as high as indicated, injected Phenergan into the inner crook of her arm where the risk of accidental arterial injection is very high, and continued to push in the syringe’s plunger despite Ms. Levine’s protestations of pain, each in direct contravention of explicit label warnings. It was this negligent administration that caused the massive tissue damage that led to the amputation of Levine’s arm.

Ms. Levine settled a claim against the physician’s assistant and the prescribing doctor, then turned her sights on Wyeth. At trial, Levine’s attorney argued that, despite the FDA’s approval of Phenergan’s label, the drug firm acted negligently by failing to add even more stern warnings or changing the label to rule out intravenous injection altogether. The Court observed that, under the agency’s “changes being effected” (CBE) regulation, a drug manufacturer may legally update a product’s label prior to receiving FDA approval, provided the manufacturer submits a supplemental application with the agency seeking approval of the exact label change.

The Ruling

The majority correctly noted that the FDA cannot keep track of all safety issues that arise after a drug is approved, but incorrectly claimed that the agency never specifically made a determination regarding the safety of IV-push administration. Consequently, the Court argued, Wyeth could have changed its label without the FDA’s pre-approval upon receiving information regarding the risk of arterial injection of Phenergan. That assertion is largely mistaken.

The case record shows that the FDA was fully aware of the risk that intra-arterial injection of Phenergan poses. The agency repeatedly and intensively investigated this exact question and determined that, notwithstanding those risks, IV-push injection provided important medical benefits. Furthermore, there are no allegations that Wyeth hid any information about the risks of IV injection, nor has any new information regarding the risks arisen that would call FDA’s decision into question. Only by relying on the fact that the FDA’s Phenergan labeling decisions were made without giving states and the general public an opportunity to intervene through a notice-and-comment rulemaking was the majority able to make an otherwise preposterous claim that the agency did not explicitly strike a balance between the risks and benefits of IV-push injection. The record clearly demonstrates that the FDA did exactly that.

With this in mind, Wyeth surely thought its chances of victory were high, given the general rule regarding implied conflict pre-emption of state laws, elucidated most recently in Geier v. American Honda. As Justice Alito’s dissent makes clear, “the ordinary principles of conflict pre-emption turn solely on whether a State has upset the regulatory balance struck by the federal agency.” Here, the FDA considered removing IV-push injection as an available option, but decided against doing so. The agency decided that IV-push injection provides greater benefits than could be achieved with the alternatives of deep tissue injection or IV-drip administration alone, and that those benefits outweighed the risks. Permitting state tort law to over-ride that determination upsets the regulatory balance struck by the FDA.

As though recognizing that observers would easily see through such a transparent argument, the majority also suggested that the trial jury would not inherently have to countermand the FDA’s balancing decision in order to find Wyeth negligent. The Court concluded that, in some undefined no-man’s land between the six existing label warnings on the one hand and ruling out IV-push injection altogether on the other, there was some other conceivable way for Wyeth to have ramped up its warning in such a way as to have prevented Ms. Levine’s injury. Therefore, no conflict arose, and no preemption was necessary.


The dual purposes of tort law are to require that tortfeasors compensate injured parties for the damage caused by wrongful acts, and to deter future acts of negligence or intentional wrongdoing. Here, however, the Court offers no guidance describing what kind of behavior Wyeth or similarly situated drug manufacturers could or should engage in that would not be considered negligent. Today, in the wake of the Wyeth decision, pharmaceutical companies and the makers of many other regulated products now have little or no idea which acts are clearly lawful and which are not. As a consequence, manufacturers will be deterred not only from engaging in bad acts, but also many good ones.

Of course, as Justice Stevens’s opinion makes clear, when new information about the risks of an approved drug arises, or when there are serious allegations that a drug maker concealed or misrepresented relevant information, there is no good reason for relying on a past FDA decision as the basis for preempting state tort law. That did not happen here, however. It appears then that the Court went much too far out of its way to ensure that acts reasonably considered negligent would not be shielded from the reach of tort claims. By doing so, the Court exposed an entire industry to liability for behavior the vast majority of Americans would consider responsible.

The Court could have and should have held in Wyeth’s favor with a narrowly tailored opinion confined to the facts of this case. Doing so would not have insulated wrongdoers from punishment, but would have recognized that Congress gave the FDA statutory authority over questions of safety and efficacy because it believed a federal expert body could most effectively balance the benefits and risks of new medicines keeping bad ones off the market and ensuring that beneficial ones reach patients in a timely fashion. Empowering lay juries to override those decisions means that fewer patients will benefit from important medicines in the future. Not only is the majority’s decision bad law, it is very bad for patients.