The case for legal regulation of physicians' off-label prescribing.

• Author: Rosoff, Philip M.

INTRODUCTION

Prescription drugs and biologics (1) can be used on- and off-label. On-label signifies that the particular use to which the drug is being put has been vetted by the Food and Drug Administration (FDA) through a series of trials or studies designed to establish the use's safety and efficacy. (2) Off-label use signifies that the particular use to which the drug is being put has not been formally approved by the FDA (3) and thus, unless it has been otherwise tested, its safety and efficacy for that use has not been established. (4) Typical off-label uses (OLU) include promoting, prescribing, and ingesting substances for conditions other than those for which they were approved, in higher-or lower-than-indicated dosages, and in populations other than those in which they were tested. (5)

Pharmaceutical companies are largely prohibited from promoting their products' OLU. (6) However, OLU are otherwise legal. That is, notwithstanding a lack of evidence of safety and efficacy, physicians may lawfully prescribe and patients may lawfully take OLU. (7) As a result, it is estimated that OLU account for up to twenty-one percent of drugs and biologics used annually in the United States; this figure is significantly higher in certain specialized settings such as oncology and pediatrics. (8)

It is likely the case that most people are unaware of the distinction between on- and off-label uses. (9) It is also likely that when physicians prescribe drugs for an off-label use, their patients are unaware of this fact or (correspondingly) that they may not be safe or effective. (10) Patients' lack of awareness is due at least in part to standard medical practice, which is not to disclose the OLU status of drugs to patients. (11) Depending on the physician, this practice has at least two explanations: some physicians believe it is unnecessary for patients to know of their drugs' OLU status, (12) while others are--like their patients--ignorant of that status and thus of the safety and efficacy of the drugs they prescribe. (13)

Previous commentators have established up to four categories of OLU depending upon the evidentiary basis supporting their safety and efficacy: OLU justified by high-quality evidence, OLU justified by some but not high-quality evidence, OLU justified by the need or desire to innovate, and unjustified OLU. (14) OLU are justified by high-quality evidence of safety and efficacy using clinical investigation techniques not dissimilar to those employed by pharmaceutical companies applying for FDA approval. (15)

OLU whose safety and efficacy have not been established by such clinical investigation techniques (problematic OLU) can cause significant harms: They can cause physical harm (including death) to patients both directly, if they turn out to be unsafe, (16) and indirectly, when they are ineffective and used in lieu of an existing, established alternative. (17) Patients also suffer harm when their physicians prescribe problematic OLU without disclosing in advance the relevant facts about the drug's off-label status and the dearth of evidence concerning safety and efficacy. Two related patient rights are violated in this setting.

First, patients have the right to decide for themselves (as part of the informed consent process) whether or not they will accept particular medical treatment; long gone are the days when the law and medical ethics permitted physicians to make treatment decisions for their patients without ensuring in advance that they knowingly and voluntarily understood and accepted the associated risks and benefits. (18) Modern physicians have a corresponding legal and ethical duty to disclose all of the facts that are material to their patients' treatment decisions. (19) Although the argument to the contrary has been strangely successful, (20) it is difficult to imagine that there is not a more material fact than that a proposed treatment's--in this case, an OLU's--safety and efficacy have not been established.

Second, people have the right not to be made the subjects of medical experiments--in this case, not to be prescribed untested OLU--unless they know about and have the capacity to acquiesce to their inherent risks. Protestations to the contrary notwithstanding, (21) this right, well-established internationally since the Nuremberg Trials, (22) is broader than that which applies to "research subjects" as these are currently defined by federal law. (23) That is, federal law does not consider unsupported OLU to be "research" when they are the result of a treating physician's prescription, (24) whereas the same OLU are "research" when they are the result of a principal investigator's prescription. (25) Leaving this federal regulatory wordplay aside, there can be no doubt that problematic OLU are experimental in the general sense of that word, and thus that patients are subject to experimental treatment without their knowledge when their physicians fail to disclose the relevant facts in advance. (26) Indeed, proponents of unregulated OLU are otherwise generally forthcoming in describing their motivation as assuring that this important avenue for "innovation" remains unfettered. (27)

Beyond the individual patient, problematic OLU also cause harm to society in the form of unnecessary increases in health care spending. Drugs are prescribed with little or no evidence that they will work; the waste inherent in this effort is compounded if the built-in delays (in securing effective treatment) result in increased costs of care. (28) Drugs are also prescribed for pseudodiseases, illnesses that are poorly defined, thus causing unwarranted growth in pharmaceutical costs. (29) Finally, dangerous side effects causing physical harm and thus additional medical care are more common with OLU than with drugs used on-label. (30)

In this Article, we argue that regulation of physicians' OLU prescribing behavior is warranted to address these harms. Specifically, we argue that when OLU are not justified by high-quality evidence of safety and efficacy, they ought to be appropriately restricted. Notably, this argument excepts OLU that are evidence-based, because these are equivalent to FDA-approved uses. Many drug uses fit into this category and are, as a result, appropriately established as the relevant standard of care. (31)

The argument that problematic OLU ought to be regulated is not new or controversial. Against the consistent refrain that, in the interests of preserving this avenue of medical innovation, all OLU ought to remain unregulated, (32) scholars and policymakers have long urged regulation to protect patients' interests in this context. In the modern context, this latter view has been part of the more general movement toward "evidence-based medicine" and away from treatment approaches that lack evidentiary support. (33) The problem is that regulatory approaches to date have not and, without more, likely cannot solve the problem. Indeed, even if this is a matter of balancing the interest in innovation against the interest in protecting patients' welfare otherwise, a new approach is necessary.

The favorite target of regulators, restriction of pharmaceutical companies' promotion of problematic OLU, (34) has been only marginally effective. This is due, at least in part, to the fact that the companies' incentives are almost entirely in the other direction: their financial gains are directly related to the extent to which their products are used, experimentally or not, off-label. (35) The most recent twists on this old regulatory approach are unlikely to work because they do nothing to alter these incentives. Indeed, to the contrary, they were developed with the understanding that pharmaceutical companies would inevitably continue to find ways to market OLU, and with the hope that if this marketing could be controlled--specifically by permitting the marketing only of OLU with third-party evidence of safety and efficacy--patients' welfare would be better protected. (36) Recent revelations that pharmaceutical companies are paying third parties to claim authorship of articles submitted to peer-reviewed journals that are actually ghost written by company contractors (37) lay bare the fundamental flaw in this strategy: one can keep hoping and (more pragmatically) fining the companies, but so long as the profits to be made from unlawful marketing of OLU outweigh those fines, it is rational for the companies to write them off as part of the cost of doing business.

At least in part for this reason, commentators who continue to be concerned about the implications of problematic OLU have renewed older calls for nonbinding professional standards to govern physicians' prescribing behavior. (38) This latter step is particularly important, because it signals the recognition that a comprehensive strategy regulating all relevant actors, or else a targeted strategy focusing on the physician-patient relationship, is more likely to be effective than one focusing exclusively on manufacturers. Proposals to regulate physicians' OLU, even if only through nonbinding professional standards, also reflect the recognition that physicians share responsibility for their implications. The problem with this alternative approach is that nonbinding standards often go unheeded. This is perhaps especially true in circumstances where the audience is, as physicians are, inculcated in a culture that has long viewed professional independence as normative. (39)

It is for these reasons that we argue in this Article on behalf of binding legal regulation of physicians' related prescribing behavior. Such regulation is likely to be effective because physicians are the most immediate link between drugs and the patients who would take them; in legal parlance, unlike pharmaceutical companies, physicians are the "direct cause" of patients' OLU. Regulating physicians is also likely to be effective because their incentives are fundamentally...