A recent ruling by the D.C. Circuit Court of Appeals in a seemingly obscure case has important implications for limits on the Food and Drug Administration’s regulatory reach and for the freedom of physician–patient relationships.
The case had an unusual background. The Judge Rotenberg Educational Center in Massachusetts treats people with severe mental disabilities including severe self-injurious or aggressive behaviors such as head-banging and self-biting. Some self-injurious patients do not respond to conventional therapies. The center, with oversight from state authorities, uses an electrical-stimulation device to briefly shock these patients to reduce or cease their behaviors. While electrical stimulators are used to treat a wide variety of conditions, the center is the only facility in the country that utilizes them for self-injurious behaviors.
In 2016 the FDA proposed banning electrical-stimulation devices to treat self-injurious behaviors, based on its determination that the device’s long-term effectiveness was uncertain for that indication and that it posed a risk of substantial and unreasonable harm to patients. The agency issued a final rule based on those findings in 2020.
Families and guardians of self-injurious patients disagreed with the FDA’s benefit–risk determination, stating that the devices had been the only effective means of deterring serious patient self-injury. Along with the center, they sued to reverse the FDA’s ban. The baseline legal question was: Does the FDA have authority to ban an otherwise legal device from a particular use while allowing it for other uses? The court answered, 2–1, in the negative.
All the parties agreed that the FDA has the power to ban a medical device altogether. One section of the Food, Drug & Cosmetics Act (21 U.S.C. §360f) gives the agency authority to ban a device that poses “an unreasonable and substantial risk of illness or injury.” But another section of the statute, entitled “Practice of medicine” (21 U.S.C. §396), prohibits the FDA from “limit[ing] or interfer[ing] with the authority of a health care practitioner to prescribe or administer any legally marketed device to a patient for any condition or disease within a legitimate health care practitioner–patient relationship.”
The FDA did not extend its ban to other uses of the electrical-stimulation devices. The devices were, for example, still approved to treat tobacco, alcohol, and drug addictions. The two-judge majority reasoned that if a device could be legally marketed for those conditions, then section 396 constrains the agency’s authority to ban a practitioner from prescribing it for another condition — self-abusive behavior.
The decision reaffirms that the FDA may not regulate the practice of medicine. Physicians must be free to prescribe approved drugs and devices for on- and off-label uses. Off-label drug use is common, especially in certain specialties such as oncology where off-label drug uses are often considered the standard of care and in pediatrics where scientific, ethical, and logistic concerns militate against conducting large trials for approval in children. Allowing off-label use enables physicians to assess their patients’ unique circumstances and the evolving state of scientific knowledge, as opposed to waiting for lengthy and costly FDA review and approval of each and every possible use.
The FDA argued that the statute does not limit its ability to ban devices for particular uses while maintaining them as approved devices for other uses. The agency claimed that “it would be a peculiar construction of the statute” to allow the agency to completely ban a device but to prohibit the “FDA from using its expertise to narrowly tailor a ban to those circumstances in which a device presents a uniquely substantial risk.”
The dissent by Judge Sri Srinivasan echoed this argument. If the agency approved a device for two intended uses and subsequently decides that the device has limited effectiveness and high risk for one of them, Srinivasan opined, it is “hard to perceive why Congress could want to deny the agency” the ability to ban a device for a risky use while preserving it for the other use that has an acceptable benefit–risk profile. “Why force the agency to make an all-or-nothing choice,” preserving or losing both uses?
Yet this, the majority found, is exactly what the congressionally enacted statute demanded. A natural reading of the language of section 360(f) “suggests a device either is banned or it is not.” No statutory language indicates that a device can be banned for some purposes yet allowed for others. A “legally marketed device” in section 396 is a device that can be lawfully sold, prescribed, or administered for any use. Nothing in the statute suggests that it “must be marketed for the particular use for which the practitioner wants to utilize the device.”
The court continued that allowing the FDA to allow a device for one use but ban it for another would “eviscerate” section 396’s protection of off-label uses. The plain language of the statute suggests that Congress wanted to preserve physicians’ freedom to make professional judgments about utilizing devices that are legally marketed for one indication for “any [other] condition or disease.”
The FDA has other avenues to communicate safety and efficacy information to physicians through labeling changes and public warnings. Highly trained and sophisticated physicians can evaluate this information and share it with patients and their families to make the most appropriate medical choices. But micromanaging medical practice by making an otherwise approved device unavailable to physicians for a particular use exceeds the FDA’s statutory authority. The agency is, of course, free to urge Congress to give it that authority.
The majority opinion is important because, rather than deferring to the agency’s statutory interpretation as the dissent urged and as so many courts do, it undertook its own, careful statutory analysis. Congress has chosen to preserve an important value: Physicians and patients determine medical care, not bureaucrats.