Medical Examiner

Drug Dealing

Who should decide when a medication is safe?

Diana Levine

The medical events that culminated in last week’s dramatic U.S. Supreme Court decision Wyeth v. Levine began in 2000, when Vermont musician Diana Levine came down with a migraine and visited a small medical clinic. Her treatment led to a horrifying complication, and her right forearm—which she used to strum her guitar—had to be amputated. After she sued and received a settlement from her treating medical providers, Levine successfully hunted bigger game: the drug giant Wyeth, which makes the anti-nausea drug responsible for her injury. The story of Levine’s victory has narrative gusto: A one-armed children’s musician takes down a heartless multinational corporation. But that’s the wrong drama to highlight.

Wyeth’s failed case hinged on a legal doctrine called “implied pre-emption,” which would have allowed companies that comply with FDA regulations—as Wyeth did—to avoid later state lawsuits about the content of their drugs’ labels. A win for Wyeth would not (as is popularly believed) have immunized big drug companies from all litigation. A company that intentionally omits important safety data from FDA review (as some allege, for example, about Merck’s nondisclosure of the cardiovascular risks associated with Vioxx) would still not be protected. The company could still be sued under product-liability and fraud laws. Furthermore, Wyeth’s loss in court may not help avoid future catastrophes, and the ruling has the potential to undermine the centralized authority of the Food and Drug Administration—which, though far from perfect, is arguably the most effective public-health agency in the nation’s history.

So, what was the case really about? Let’s review what happened to Diana Levine: To treat her migraine, a health provider first injected a combination of Demerol (a narcotic pain reliever) and Phenergan (an antihistamine to reduce nausea) into her muscle. When Levine’s pain persisted, the provider concluded the medicine was being absorbed too slowly from the muscle to offer speedy relief, placed an intravenous line in Levine’s right arm, and quickly injected another dose directly into Levine’s bloodstream. The provider made a serious error by accidentally and unknowingly putting the IV into an artery (a high-pressure, pulsatile vessel going out to the arm) instead of a vein (a low-pressure vessel returning blood to the heart). Dispersed at high pressure throughout the distal forearm, Phenergan caused irreversible blood vessel damage resulting in gangrene and, ultimately, amputation.

There’s no question the provider fouled up and was deservedly sued. But how is that Wyeth’s fault? The package insert for Phenergan—which is worth reading—clearly warned about the risk of gangrene, explained the danger of accidentally putting the IV in an artery, recommending stopping injections if patients felt any pain, and prescribed a specific infusion rate and dose. According to court documents, the risk of gangrene is only 1 per 10 million doses. Even fool medical students graduating at the bottom of their classes know you’re never supposed to inject drugs into arteries.

Aaron Kesselheim, a Harvard researcher and intellectual-property lawyer, admitted to me that any reasonable physician knows not to inject drugs into arteries. But in his opinion, Wyeth still should have put a “black box” warning about push injections on the package insert decades ago, when the first reports of gangrene surfaced.

By that loose standard, however, every drug given intravenously probably needs one. (Stronger warnings might also be needed for over-the-counter drugs like infant Tylenol, which each year causes thousands of emergency-room visits and a handful of deaths in young children because parents regularly mess up the dosing despite the label’s instructions.) The Phenergan insert already has a “black box” warning not to give the drug to young children, an admonition not to give anti-nausea drugs of any kind to kids with “uncomplicated vomiting,” and, as a bonus, numerous dire warnings about glaucoma, peptic ulcers, prolonged exposure to the sun, permanent nerve damage, and dozens of other precautions.

The simple fact is that package inserts aren’t terribly useful because they’re too damn complicated, much like the laundry list of side effects at the end of television commercials for drugs. How can anybody make sense of labels that list dozens of horrible outcomes, even for the most commonly used drugs? In ruling against Wyeth—and concluding the FDA-approved labels may be insufficient—the Supreme Court has invited manufacturers to lard medications with even more useless warnings to head off lawsuits. It’s hard to see how this will prevent future harms to innocent patients, since adding yet another warning to the dozens already listed for Phenergan probably wouldn’t have saved Diana Levine’s arm.

Remarkably, some of the nation’s most respected physicians, such as a collection of editors of the New England Journal of Medicine (one of whom wrote a book assailing the ridiculously unscientific breast-implant litigation that bankrupted the Dow Corning Co.), supported Levine: They think personal injury attorneys can do a better job than the FDA in protecting Americans from complicated and sometimes dangerous medications on the market. In an editorial in last summer’s New England Journal of Medicine arguing against Wyeth’s position, several editors pointed out that Avandia, Vioxx, and Redux were all approved by the FDA and were later found to have important safety risks. They suggest that ruling in favor of Wyeth would “erase” a harmed patient’s “right to seek legal redress,” even though the truth is that outright lying wouldn’t be protected. And they also overlook their own contribution to our nation’s drug problems.

Consider the example of Vioxx, Merck’s pain medication that caused heart attacks. Merck likely was truthful in submitting its clinical trial data to the FDA during approval, including those suggesting an increased risk of heart attack. As nicely documented by physician John Abramson in Overdosed America, these data were easily accessible to the public and also emphasized by warning letters issued to Merck by the FDA. It wasn’t exactly a secret. In 2001, the Journal of the American Medical Association published a clear warning about the excess risks. Yet the New England Journal of Medicine published a review article at the same time minimizing the extent of the problem, uncritical health insurers added the drug to their formularies and paid for them, and many doctors happily wrote prescriptions without reviewing any data. If the drug companies should pay for resulting harm, shouldn’t the New England Journal of Medicine also be sued? How about health insurers? Or the pharmacists who fill the prescriptions?

Now, there’s no question Merck deserves punishment if the company blatantly lied about risks. But here’s the thing: The data are really, really complicated. Major medical journals published vastly different takes on it. And in the end, further studies on Vioxx within a brief period of time confirmed the preliminary risks, and the drug was withdrawn from the market—arguably a regulatory success story.

Ultimately, there is no such thing as a “safe” medication. Drugs each have a balance of risks and benefits best evaluated by highly qualified sources. And the truth is that sometimes drugs have unforeseen side effects. Complex nuances of medical practice rarely survive courtroom battles—which, unfortunately, is where many drug debates may continue to occur.