Prescription Requirements and Patient Autonomy

Abstract

Although many drugs are reasonably available only by prescription, for many others—such as oral contraceptives and statins—this paternalistic gatekeeping cannot be justified. To better support valid claims of patient autonomy, the U.S. Food and Drug Administration should change its default approach to favor over-the-counter status for new and established drugs.

In the United States, patients can have over-the-counter (OTC) access to a wide array of drug products, including acne medications, antacids, cold and allergy remedies, sleep aids, smoking cessation products, weight loss drugs, and various pain relievers. But for many drugs, especially those that are newly approved, patients must secure a prescription from a licensed clinician. When the prescription requirement is cast against increasing attention to patient autonomy, growing rejection of medical paternalism, and the broader democratization of access to information and medical products via the Internet, one question becomes paramount: in what circumstances is prescription gatekeeping ethically justified?

Today’s patients are increasingly able to control their access to a variety of health-related services.¹ Consider the recent proliferation of online prescription services. Companies such as Roman, Hers, and Hims offer access to treatments for erectile dysfunction, hair loss, genital herpes, and oral birth control, all without an in-person doctor visit.² Instead, the patient completes a brief health questionnaire, which is reviewed by a remote physician who has no prior relationship to the patient and who then prescribes the requested product if they determine it to be appropriate. In many cases, the physician never engages directly with the patient, even virtually, although there are mechanisms for follow-up communication where needed. These services have been touted both as improving convenience for patients, especially those with stigmatizing conditions, and as improving the efficiency of the health care system by avoiding time-consuming, expensive, and medically unnecessary office visits. Critics, however, suggest that the practice could harm patients.³ They worry that important medical needs might be left unidentified and unaddressed if prescribing occurs outside of a doctor-patient relationship, without routine communication between doctor and patient and without the option for an in-person physical exam—all of which are features of traditional telemedicine.

Although online prescribing services retain the physician gatekeeper, albeit in a diminished form, other recent approaches go further toward reducing physician authority over prescription access. For example, a 2019 California law permits specially trained pharmacists to provide, without a physician’s prescription, a sixty-day supply of preexposure prophylaxis (PrEP) for HIV, as well as doses of postexposure prophylaxis (PEP), to patients who have tested negative for HIV within the past seven days.⁴ The law is intended to help patients avoid stigma and address the difficulty some may face in securing a doctor’s appointment during the critical window for avoiding infection. While the law retains some gatekeeping for drug access in the form of pharmacist control, that gatekeeping is substantially reduced compared to the status quo in other states where a physician’s prescription is still required. Similar reasoning supports efforts to allow pharmacists to legally dispense Naloxone, an opioid antagonist that can reverse overdose, without an individual prescription⁵ and to move some emergency contraceptives from prescription to behind-the-counter or OTC status.⁶

These and other developments raise questions about whether, when, and why we need prescription gatekeeping at all. If the drugs being offered are truly low risk, or if they have high potential for net benefit under certain circumstances, then perhaps they should simply be available OTC with appropriate warning labels. And even if they are potentially risky, perhaps competent adult patients should nevertheless be allowed to make these kinds of treatment decisions for themselves. Overall, should we be satisfied with where the dividing line between prescription and OTC status has been drawn? We think change is in order.

Under the U.S. Food and Drug Administration’s current system, new drug approvals are nearly always for prescription use only. Since the mid-1980s, there have been fewer than five instances in which a new molecular entity was approved first for OTC use.⁷ If sponsors seek OTC approval, either initially or after a period of prescription status, they are expected to provide additional information to justify that approach. Thus, rather than assuming that approvable new products will be appropriate for OTC use unless demonstrated otherwise, the default assumption is precisely the opposite.⁸ When FDA approval is required for new drugs, OTC status currently stands as the exception rather than the rule, with direct-to-OTC approvals and prescription-to-OTC switches being relatively infrequent.

The prescription default has substantial paternalistic tendencies, often directed at protecting competent adults from harms related to the use of pharmaceutical products.⁹ Although paternalism can sometimes be justified, uncritical acceptance of it violates respect for individual autonomy, running contrary to a core tenet of bioethics and a long-standing American value. To give the FDA’s regulatory approach a stronger normative foundation in light of considerations related to respect for patient autonomy, we propose a reversal of its default: the presumption should favor OTC status. Thus, rather than assuming that new drugs can be used safely only with a prescription, the FDA would approve new drugs for OTC use unless a sponsor demonstrates, or the FDA otherwise determines, that either or both of the following conditions are met: (a) patients are likely to use a drug in a manner that would seriously harm third parties, or (b) a significant proportion of the general adult population would have difficulty safely using a drug without clinician support.

Because we anticipate that many drugs will satisfy one or both of these conditions, changing the default approach to favor OTC status may not dramatically affect the distribution of drug approvals across the prescription and OTC categories. However, a presumption in favor of OTC status would be substantially more respectful of patient autonomy than the status quo because it would demand justification for government interference with that autonomy. Moreover, application of these conditions suggests that some categories of drugs currently designated for prescription only should be considered for a switch to OTC status. These include oral contraceptives for birth control, bronchodilators for asthma, epinephrine autoinjectors for severe allergies, and insulin (analogs) for diabetes.

Regulatory Foundations

To clarify, our concern is about potentially unjustified paternalism related to decisions that bear on prescription-only approval—it is not about the FDA’s premarket approval requirement itself. The requirement that certain pharmaceutical products receive FDA approval prior to marketing is essential to the development and validation of critical information that all patients need to make autonomous decisions about their care, with or without clinician gatekeeping.¹⁰

By demanding adequate evidence of safety and efficacy before allowing new drugs to be marketed, the approval requirement provides an important incentive for manufacturers to conduct robust clinical trials that assess not only a drug’s benefits but also its risks. As the dietary supplement industry demonstrates, market forces alone may not be enough to ensure that this evidence is produced. Even if manufacturers were willing to conduct clinical testing in a timely fashion while marketing their drugs, patients might not want to participate in the testing if they were able to access the products outside of a study, making it difficult to conduct rigorous trials at this stage.¹¹ Finally, although it is possible that nongovernment entities could help clinicians and patients understand and translate the data produced or could make recommendations about whether and how a product should be used on the basis of those data (through a certifying function akin to the “Good Housekeeping” seal of approval or Consumer Reports recommendations), only the FDA has the authority to access and review all clinical trial data and to require that balanced data be produced prior to product marketing for a given indication (a validating function).¹² Accordingly, the FDA has legitimate authority both to determine which products are allowed on the market in the first place and to regulate permissible claims for marketed products. Our focus is, therefore, on critically evaluating the standards for deciding whether a drug that meets the standard for marketing approval should be available OTC or by prescription.

At the start of the twentieth century, only narcotics required a prescription, with all other drugs available for self-medication.¹³ The Pure Food and Drug Act of 1906 focused on regulating pharmaceutical products based on their labeling, prohibiting false and misleading claims, and requiring disclosure of the presence and amount of certain dangerous ingredients.¹⁴ The act imposed no requirements for clinician prescribing, however. Over three decades later, in 1938, the Federal Food, Drug, and Cosmetic Act imposed a new requirement mandating FDA preapproval of drugs based on a manufacturer’s demonstration of safety.¹⁵ The law relied on labeling to promote safe use, including adequate directions and warnings to protect users.¹⁶ Like the 1906 act, the 1938 act said nothing about prescription gatekeeping, and scholarly analysis suggests that Congress’s emphasis was on providing patients with the information needed to safely self-medicate.¹⁷ The regulations the FDA promulgated to implement the statute, however, required that labeling bear adequate directions for use, specifically by laypersons. If no such directions could be provided, then the regulations allowed manufacturers to use alternative labeling: “Caution: To be used only by or on the prescription of a physician.” With this label, a product could be sold only with a prescription; otherwise, it would be deemed misbranded.¹⁸ In 1944, the FDA amended its regulations to place more specific parameters around prescription labeling, including its appropriate use, to promote consistency among manufacturers. Congress codified these regulations into statutory law in 1951.¹⁹ Finally, in 1962, Congress added proof of efficacy to the premarket approval requirement.²⁰

Today, the law states that if a drug, “because of its toxicity or other potentiality for harmful effect, or the method of its use, or the collateral measures necessary to its use, is not safe for use except under the supervision of a practitioner licensed by law to administer such drug,” then it “shall be dispensed only” with a prescription.²¹ Most nonprescription drugs have been on the market since the 1970s and satisfy one of the FDA’s OTC “monographs” for specific therapeutic classes. OTC monograph drugs do not require preapproval because their active ingredients are generally recognized as safe and effective for a specific OTC use, with regard to both dosage strength and form, alongside precisely specified labeling requirements about indications, directions for use, and warnings.²² Drugs that fall outside the monograph process are generally considered new drugs, which must secure premarket approval from the FDA. Although new drugs technically can be approved for “direct-to-OTC,” they are almost always first approved for prescription only.²³ The FDA may subsequently effectuate an “Rx-to-OTC switch” if it becomes clear that prescription gatekeeping is “not necessary for the protection of the public health.”²⁴ A switch may occur either following application of the sponsor or via FDA rulemaking, but the process can be complex.

The FDA’s regulations exempt prescription drugs from consumer labeling requirements, whereas OTC drugs must bear “adequate directions for use” under which a layperson “can use a drug safely and for the purposes for which it was intended.”²⁵ In particular, OTC labeling is to be provided in terms “likely to be read and understood by the ordinary individual, including individuals of low comprehension, under customary conditions of purchase and use.”²⁶ When determining whether a product should be available OTC, the FDA also considers a drug’s overall safety and whether the condition being treated can be self-diagnosed, self-treated, and self-managed.²⁷ In addition to traditional clinical investigations mandated for all new drugs, the FDA may require sponsors of drugs that are intended to be made available OTC to conduct various consumer studies to assess users’ ability to comprehend proposed labeling and to appropriately self-select OTC products, as well as studies of actual use.²⁸

Taken together, these provisions offer general criteria for drawing the line between prescription and OTC products. The FDA’s Division of Nonprescription Drugs provides a helpful summary of the features that typically characterize OTC drugs, including the following:

• benefits that outweigh their risks,

• low potential for misuse and abuse,

• the ability for consumers to use them for self-diagnosed conditions,

• the ability for consumers to use them without the need for health practitioner involvement, and

• the ability to be adequately labeled for consumer use.²⁹

These criteria leave room for substantial discretion, however, and have been criticized as “vague and indeterminate.”³⁰ Indeed, at the margins, there may be some OTC drugs that pose greater risks than those available only by prescription. Consider, for example, the risks associated with hormonal contraceptives versus acetaminophen. Although oral contraceptives carry a small risk of blood clots, that risk is lower than the risk of blood clots during pregnancy and the postpartum period.³¹ Acetaminophen overdose, by contrast, is one of the most common poisonings, can lead to life-threatening liver failure, and can occur when the drug is taken at just over twice the recommended maximum daily dose.³²

In recent years, the FDA has increasingly acknowledged that it may be appropriate to critically reexamine the potential for greater use of OTC status. For example, the agency’s Nonprescription Drug Safe Use Regulatory Expansion (NSURE) initiative aims to provide additional regulatory options for innovative OTC development programs. This approach has the potential to expand which drugs can be considered nonprescription in response to public health needs, concerns about under-treatment of common diseases and conditions (for example, migraines, asthma, and hypertension), and “larger trends in healthcare toward patient autonomy and self-management.”³³ In 2012, the FDA held a public hearing on the issue, during which it considered the possibility of relying on new technologies, nonphysician providers, and other approaches to support safe use—ranging from pharmacy counseling to self-diagnostic aids to extended refills—while also considering issues related to insurance coverage, liability, innovation, and competition.³⁴ Building on these efforts, in 2018, the FDA published a draft guidance for industry describing “innovative approaches” to demonstrate safety and effectiveness for an OTC product in cases where typical consumer labeling alone would be insufficient.³⁵

These are important steps in the right direction. Efforts to better understand acceptable justifications for prescription gatekeeping—and their limits—can provide further support for the FDA’s initiative in this area. They can also help guide work to specify which categories of drug products could appropriately be reconsidered for OTC status.

Justifying Prescription Gatekeeping

In general, competent adult patients should be treated as autonomous decision-makers who have the legal and ethical right to make informed decisions about the direction of their own medical care, unencumbered by paternalistic interference. This principle does not, however, generate an absolute right to autonomy in the medical domain. To the extent that there are situations where infringing on personal autonomy is appropriate, the relevant question that arises with respect to prescription drug access is whether competent adult patients who prefer to select their own pharmaceutical interventions from FDA-approved options should instead be compelled to obtain a prescription.

Third-party harms.

The strongest justification for restricting patient autonomy through a prescription requirement is to prevent third parties from suffering serious harm. This justification is rooted in the libertarian constraint acknowledging that personal liberty extends only to the point that it begins to infringe on the liberty interests of others.³⁶

In the pharmaceutical context, third-party harms might arise through individual misuse of a drug, intentional or otherwise, that occurs at a scale large enough to threaten the community’s welfare. Antibiotic resistance exemplifies this concern. Permitting patients to self-medicate with antibiotics could lead to wide-spread overuse and misuse of these medications, which, in turn, could lead to the spread of drug-resistant infections that threaten the public’s health.³⁷ Even in the face of existing prescription requirements, antibiotic resistance from overprescribing is a serious problem, one that OTC availability of antibiotics would likely exacerbate.

Drugs can also lead to third-party harms even when they are used appropriately. For example, products that can produce an altered mental state, such as extreme drowsiness, could be subjected to a prescription requirement to increase the likelihood that patients receive guidance from a health care provider. This approach could help ensure that patients are aware of the need to take precautions to safeguard others, such as making alternative arrangements for driving or avoiding operating machinery. Finally, when pharmaceuticals have qualities that make them particularly susceptible to diversion for illicit use—as in the case of opioids—prescription requirements can help protect third parties by limiting the quantity of the products available for diversion. While prescription status is not a panacea against third-party harms, it can nevertheless help mitigate certain foreseeable consequences.

Besides health-related considerations, third-party harms stemming from OTC access to pharmaceutical products could include financial burdens on society. If patients were free to access riskier products without a prescription, at least some of the costs incurred as a result of misuse would be imposed on others through insurance premiums and tax obligations. These downstream financial consequences, however, are insufficient to justifiably constrain patient autonomy. Competent adults are permitted to ride motorcycles, base jump, hang glide, scuba dive, and engage in all sorts of risky activities that may result in injuries, the costs of which others are forced to bear as part of the price of societal freedom. The constraints on individual autonomy that would result from a standard that precluded any financial impact on third parties would be unbearably stifling. Therefore, although we endorse prescription drug requirements for products that could undermine the health and welfare of third parties, the need to protect society from the downstream costs of patients’ use of pharmaceutical products is not a sufficient reason to reject OTC status.

Paternalism.

For many pharmaceutical products, the potential harms resulting from self-medication will be primarily self-regarding, with patients themselves experiencing the most direct and tangible impacts. Intervening to protect patients from self-regarding harms through a prescription requirement may be paternalistic but nevertheless justified when the conditions for making autonomous decisions are unlikely to be satisfied for a substantial segment of the population.

No drug is universally safe or effective. Safety and effectiveness are relative and contextual concepts that require consideration of, among other things, the severity of the patient’s disease or condition, available alternatives, potential safeguards, and the intended use of the drug.³⁸ Even a product that has been approved as safe and effective can still be dangerous when not used as directed. For example, serious adverse reactions can be caused by taking too much of a drug, taking a drug for an inappropriate indication, taking certain drugs in combination, improperly administering a drug, or failing to monitor for physiologic changes that can signal concerns. Beyond adverse reactions, if a patient takes a drug meant for one condition when they actually have another, this can prevent them from getting the care they truly need.

A prescribing clinician can help patients avoid these risks, but some patients may be willing to accept them in exchange for the freedom to access approved products without a prescription. Such patients may believe that they can adequately educate themselves and take appropriate precautions without relying on a clinician’s oversight—and sometimes, they may be right. Alternatively, patients might choose simply to accept uncertainty and take the risk.³⁹ Or they might disagree with a prescriber about whether the risks of a particular drug are acceptable for them.⁴⁰ In these cases, requiring patients to secure a prescription for pharmaceutical access in order to protect them from harm would interfere with their valid, autonomous decisions—decisions not based on any mistake or other source of involuntariness.⁴¹ This sort of interference would be unjustifiably paternalistic if we considered only the interests of those individual patients who express a preference for self-medication, at least once a patient’s voluntariness has been established. However, because regulations apply to populations rather than to individuals, respecting one patient’s autonomy may create a circumstance in which other patients who cannot act autonomously without support are insufficiently protected from their own decisions. In this regard, it may be necessary to interfere with some patients’ decision-making in order to adequately protect other patients—an approach that recalls our previous argument about justified interference to protect the interests of third parties.

Forms of Paternalism

Taking a step back, paternalism refers to circumstances in which a state, institution, or individual interferes with the choices or actions of another person with the primary motivation of protecting that person from harm or making them better off. Philosophers distinguish among several general types of paternalism that vary in terms of their ethical acceptability.⁴²

Hard paternalism.

Hard paternalism is the most extreme version of paternalism; it entails interference with an individual’s decisions and behavior to prevent harm to that individual, even when they are acting with sufficient voluntariness. Thus, a defender of hard paternalism would find it acceptable to require a motorcyclist, even one who fully understands and accepts the risks of not wearing a helmet, to wear one in order to protect that individual from the foreseeable consequences of their choice.⁴³ A hard paternalist would not invoke harm to third parties or other negative externalities to justify the helmet requirement. Similarly, where access to a potentially harmful drug is concerned, a hard paternalist would consider prescription gatekeeping to be justifiable interference on the grounds that it is necessary to protect patients—including those capable of making a voluntary and knowledgeable choice about using a drug—from potential harm.⁴⁴ Opponents of hard paternalism criticize this approach for its failure to respect the liberty interests of individuals who act as autonomous agents.

Soft paternalism.

Soft paternalism avoids this critique by promoting respect for individual liberty interests. Under this approach, interference in another’s choices to protect them from harm is justified only when there is reason to believe that the individual may not be acting voluntarily, perhaps as a result of a mistake, an insufficient understanding of the relevant facts, or inappropriate pressure from external influences.⁴⁵ The soft paternalist’s interference should be directed first at addressing these problems and advancing the individual’s capacity to act autonomously; if autonomous decision-making remains in question, further constraints may be acceptable. For the motorcyclist, soft paternalism could justify requiring that the rider take a safety course in motorcycle riding as a condition of securing a helmet-free motorcycle license. If the rider could demonstrate understanding of the relevant risks, however, requiring a helmet would veer toward unjustified hard paternalism, assuming no other justification exists. In the context of prescription gatekeeping, soft paternalism would support limiting access to drugs when there is reason to believe that patients’ decisions would not be sufficiently autonomous without the sort of expert guidance that a clinician can provide. However, such limitations would need to be removed if and when a patient adequately demonstrated their voluntariness.

Justified in this way, prescription requirements would initially constrain autonomy but ultimately serve to support an individual’s autonomous choices. Thus, a patient who is willing to accept substantial risks after having been properly informed, perhaps through a required clinician consultation or educational intervention, and who could demonstrate that they understand those risks, would be entitled to avoid further constraints on access intended to reduce the harm they might experience as a result of self-medication.⁴⁶ In contrast to current approaches, the prescription in this scenario would not necessarily signify that the clinician views the drug as the right, or medically appropriate, option for the patient. Instead, it would signal only confirmation of the patient’s informed, voluntary choice.

Group soft paternalism.

So far, neither hard nor soft paternalism could justify anything more than prescription consultation—not full-fledged prescription gatekeeping as currently practiced. But if one accepts the premise of soft paternalism and acknowledges that prescription versus OTC determinations about particular products must be made for patients as a group rather than on an individual basis, it becomes clear that the type of soft-paternalist assessment that would likely be necessary to evaluate patient voluntariness and informed choice will not always be practicable.⁴⁷

Individual patients will typically need different levels of intervention to support autonomous decision-making. For example, nurses, physicians, and pharmacists who are themselves in the role of patients are likely well suited to making autonomous choices about the types of pharmaceutical products they wish to use and may therefore experience pharmaceutical gatekeeping as unnecessary. However, other types of patients with poor health literacy may struggle to satisfy the minimum criteria for making autonomous decisions about which drugs to use without substantial support from a prescriber. Between these extremes, distinguishing those patients who have the capacity to autonomously self-medicate from those who do not is difficult, if not practically impossible, without a valid, feasible mechanism for assessing understanding and voluntariness before providing access to a potentially dangerous product. But is it acceptable to limit the autonomy of patients who would be capable of using a product safely on their own (or safely enough in their own estimation) in order to protect those who would not?

Interference to protect the least capable or most vulnerable within a population comprised of individuals with a wide range of capacities has been described as group soft paternalism.⁴⁸ This approach maintains that expecting individuals who are capable of making an informed and rational choice to sacrifice some of their autonomy is justifiable when there is strong reason to believe that a sizeable proportion of individuals within a group would not be able to act autonomously with respect to the choice in question and it would not be feasible to distinguish those who could do so in individual circumstances.

In the context of prescription gatekeeping, group soft paternalism would support restricting patients’ access to approved drugs that a substantial proportion of the general adult population would struggle to use safely on their own. This would likely be the case for drugs intended to treat diseases and conditions not easily diagnosed by a layperson and drugs that require complex dosing schedules, careful management of interactions with other drugs, and regular physiologic monitoring to determine whether and what type of continued use is appropriate. It may also be true for products that are particularly susceptible to off-label uses that may be unsafe or for which there is insufficient data to inform decision-making. Because these criteria describe a wide variety of drugs, group soft paternalism can likely support broad use of prescription gatekeeping, even if the FDA’s default approach moved to a presumption in favor of OTC status.

Nevertheless, given that prescription requirements represent a sacrifice of autonomy for at least some patients, it is important to make sure that these requirements are necessary in specific instances and are not overused. As noted above, sponsors seeking OTC status for their drugs are sometimes expected to complete labeling-comprehension, self-selection, and actual use studies.⁴⁹ Rather than treating these studies as exceptional, it may be appropriate to expect that they be routinely conducted for all new drug applications for which OTC status might plausibly be considered. Then, only if such studies raise significant concerns about patients’ ability to comprehend a consumer label, select the product in appropriate circumstances, and use it correctly, would prescription status be warranted. This evidence-based approach would be preferable to relying on hunches or stereotypes about patient capacity, although we acknowledge that there are trade-offs surrounding the added cost and delay that may be associated with additional studies.

There are also mechanisms that can be used to improve the potential for safe use without a prescription. The FDA proposed, in its 2018 draft guidance, additional channels for delivering nonprescription drug information, including interactive displays, that would go beyond the back-of-package consumer “Drug Facts” label. The FDA also suggested that, prior to purchase, patients could be required to “view and affirm that they viewed text or images in a video that describes how to appropriately use the nonprescription drug product” or to answer a set of questions about the drug via a “self-selection test” in a mobile application or pharmacy kiosk, which would “affirmatively indicate … that the consumer is an appropriate candidate to use the nonprescription drug product.”⁵⁰ These approaches need empirical testing. However, they have the potential to protect patients who need additional support to satisfy the conditions of autonomous decision-making while minimizing infringement on the autonomy of those capable of making informed and rational choices.

In the context of the NSURE initiative, the FDA also proposed making greater use of pharmacists and pharmacies to assist patients in self-selecting appropriate nonprescription products and to provide patients with opportunities for follow-up monitoring. For example, the FDA suggested that pharmacists or consumers might be able to confirm diagnoses or engage in routine monitoring using diagnostic tests that could be made available in a pharmacy.⁵¹ Pharmacists might also take the role of “assessing whether the consumer has any conditions or other risk factors that would indicate that the drug should not be used, or assisting the consumer in choosing between various drug products.”⁵² Note that this is different from the behind-the-counter approach that has been used for some products, an approach in which pharmacists are simply tasked with checking consumer identification to avoid product diversion for illicit drug production (of pseudoephedrine, for example) or to satisfy age restrictions (as previously applied to emergency contraceptives).⁵³ It is also different from an approach in which pharmacists would be expected to confirm that a patient is making an informed and voluntary choice. Instead, the point of the NSURE model is to facilitate the conditions necessary for a patient to act autonomously without a prescription. Another approach the FDA has considered to advance this goal involves requiring a prescription for initial use but allowing refills to be provided without one.⁵⁴

Overall, the principles of group soft paternalism can provide normative justification for prescription gatekeeping in many of the circumstances in which prescriptions are currently required. However, this reasoning must not be applied uncritically. Where less restrictive alternatives could be used to achieve the same goal of promoting patient autonomy across a population—including nonprescription status with innovative approaches to support understanding and safe use—they should be given serious and routine consideration.⁵⁵

Inadequate Justifications for Prescription Gatekeeping

Despite constraining patient autonomy, both arguments aimed at safeguarding third parties and those intended to protect individuals who would have difficulty making autonomous decisions about using certain drugs can support prescription gatekeeping, so long as it is done in the least restrictive manner. Other justifications for prescription gatekeeping fail.

One function currently served by prescriptions is to signal to third-party payers that a product is clinically indicated for a particular patient based on a clinician’s expert judgment. In other words, it is “reasonable and necessary” for the payer to use limited and shared resources to pay for the medication (although there remain many disputes between clinicians and insurers about coverage). At present, prescribers function not only as gatekeepers to drugs; they also function as gatekeepers to the collective funds most patients need to pay for drugs. Prescription gatekeeping is not, however, the only way to steward resources. Instead, patients who are willing to pay out of pocket could be permitted to access a product without a prescription, whereas patients who intend to seek third-party payment could be required to secure a clinician’s (or payer’s) authorization for use of the OTC product. This is not an infringement on the latter group’s autonomy, per se, but rather a reasonable prerequisite for access to a shared resource.

Insurers already pay for several categories of OTC drugs without patient-specific authorization, such as some contraceptives, aspirin for individuals at high risk of cardiovascular disease, smoking-cessation aids, and folic acid supplements for pregnant women.⁵⁶ This approach could be extended to additional product categories, and insurers also could be encouraged (or potentially mandated) to extend coverage for OTC products that a qualified medical professional has deemed “necessary” for a specific patient.⁵⁷ If insurers refused to do so, any improved patient access from reduced prescription gatekeeping as proposed herein could be negated by inability to pay.⁵⁸ Because this is not a matter of a product’s safety and effectiveness, it is legally outside the realm of the FDA’s consideration, although we nonetheless acknowledge it as a serious ethical and practical concern. However, the appropriate response to coverage and affordability constraints is not to artificially impose prescription gatekeeping where it is not truly justified, holding all patients to a prescription requirement regardless of their individual willingness or ability to pay out of pocket. Instead, the appropriate lever is to adjust insurance coverage—not to cover all OTC products, since some truly are discretionary, but rather to cover those that are appropriately validated as being reasonable and necessary for the patient’s condition, whether they are prescription or nonprescription.⁵⁹ Attention to drug pricing and the affordability of drugs that become newly available OTC would also be critical, although it is possible that pharmaceutical companies would adjust their prices depending on whether their products were available by prescription or OTC. In our view, it is best to identify prescription drug policies that appropriately respect patient autonomy and then focus on ways to avoid detrimental financial impacts.

Beyond financial considerations, a potential unintended consequence of making more products available OTC is increased off-label use without prescription gatekeepers monitoring the purpose for which a patient is seeking a drug. It is not clear that any such increase would be likely, especially since off-label prescribing is already common—and because patients may nonetheless need clinicians to authorize product use to justify insurance coverage, as just discussed. But if off-label use were to become more prevalent as patients self-medicate, would it leave sponsors without a strong incentive to study the uses of a drug beyond those for which it is labeled? We do not foresee such an outcome. Because restrictions on off-label promotion apply to both prescription and OTC drugs, sponsors should have the same incentive to study novel indications in either scenario—namely, to permissibly market them.⁶⁰ The desire to secure third-party payment would also encourage sponsors to generate needed evidence. Accordingly, concerns about inhibiting the production of information cannot justify prescription requirements, in contrast to marketing approval requirements themselves.

In theory, prescription status would facilitate better monitoring of adverse drug events, which is especially important for new drugs that do not yet have a track record of use in patients. Health care providers should presumably have greater expertise to identify safety concerns, as well as the wherewithal to accurately and consistently report them, compared to the patients using a drug OTC. However, current evidence suggests that professionals and patients report adverse events at similar rates.⁶¹ Moreover, health care professionals notoriously underreport adverse drug events to the FDA for a variety of reasons, including difficulty identifying causation, lack of time, and uncertainty about reporting procedures.⁶² Therefore, this reasoning also does not support a preference for prescription status in practice.

Lastly, nonmedical considerations related to partisanship or professional territorialism should have no bearing on prescription gatekeeping. For example, debates as to whether emergency contraception should be available OTC were largely influenced by political and moral considerations unrelated to the broad justifications for prescription status advanced above.⁶³ Similarly, the fact that patients’ access to OTC products could bypass clinicians in ways that might have significant financial implications for the medical profession or otherwise threaten the profession’s prerogatives cannot support constraints on patient autonomy.

Policy Implications

The value of respecting patient autonomy—as well as the potential public health benefits of increasing access to a range of pharmaceutical products for which patients may otherwise face significant barriers, as discussed below—suggests that drugs should be made available OTC absent a compelling reason to require a prescription. Changing the FDA’s default approach to new drug approval in this way would appropriately prioritize patient autonomy while recognizing critical limits that serve to protect both third parties and individuals who may struggle to make autonomous choices. This balance places our proposal in contrast to more extreme attempts to increase access to pharmaceutical products, such as “right-to-try” legislation and broader recommendations that would allow companies to market products after only the earliest-phase trials and on the basis of weak evidence.⁶⁴

Besides offering a more robust normative account than has been presented to date of when and why prescription gatekeeping is justified—and when it is not—our approach has important practical implications. Although we anticipate that many products would remain appropriately under the prescription umbrella, both the FDA and sponsors would be expected to critically assess whether prescription status is truly necessary, an approach more respectful of patient autonomy than simply assuming that OTC status would be problematic. In addition, changing the default in favor of OTC status would help ensure that products that could reasonably go either way would not be made prescription-only due merely to inertia on the part of sponsors or regulators.

One practical question about changing the default is whether sponsors should be permitted to opt out. In other words, should they be allowed to seek prescription-only—rather than OTC status—at their own discretion, as is currently the case, or should the FDA have the exclusive authority to determine a product’s path? In our view, the best way to make the default rule meaningful and to advance the aim of promoting patient autonomy is for sponsors to be expected to submit their application with the data necessary to support OTC use, unless they can convince the FDA that such use would be inappropriate on either of the grounds discussed above—because it would lead to unacceptable harms to third parties or would leave a significant proportion of the general adult population unable to safely use a drug without clinician support. Although this conclusion may sometimes be obvious, addressing whether a particular new drug application is appropriate for the OTC pathway will likely demand more labeling, self-selection, and actual use studies than have traditionally been carried out, or the use of other methods to demonstrate that a product truly is not suitable for self-medication. In at least some cases, we anticipate that these studies will support OTC status.

There are several categories of products that are currently prescription-only that we anticipate will fail to satisfy either of our compelling justifications for requiring a prescription and therefore should be promptly considered for an OTC switch. This is especially true if we are able to incorporate some of the technological safeguards considered by the FDA’s NSURE initiative, including mobile apps and pharmacy kiosks with comprehension tests to assess patients’ understanding of the risks, benefits, and indications for a drug, as well as symptom and history questionnaires that could help flag contraindications and the need for follow-up care. It is important to recognize that all drugs carry associated risks, including some that may be serious. However, complete absence of risks is not—and should not—be required for OTC status.

The first and most obvious category of drugs that should be made available without a prescription consists of those that have been demonstrated to be safe for use without professional oversight.⁶⁵ For example, there is strong empirical evidence that adult women can use oral contraceptives safely on their own.⁶⁶ Infringing on women’s autonomy in this case is likely due to paternalistic or moralistic political considerations that should not be permitted to guide medical decisions.

A second category of products that should be considered for OTC status are those used almost exclusively by experienced patients, such as those who have been diagnosed with a chronic condition. In this context, the process of diagnosis and chronic disease management can provide an opportunity for clinicians to relay information without necessarily requiring prescription gatekeeping. For example, bronchodilators for asthma, epinephrine autoinjectors for severe allergies, insulin for diabetes, and statins for elevated cholesterol are typically used by patients who have become highly knowledgeable about the dosing and administration of these drugs, as well as when to seek help. Moreover, these types of drugs are unlikely to be attractive to those who have not already been diagnosed with a relevant condition. Each of these products is also well characterized, making adverse events from using them as directed an unlikely outcome.

A third category would include products for conditions that affect patients who may have difficulty securing a prescription due to various barriers, including cost, inconvenience, or having a stigmatized condition. In these contexts, we might be more worried about risks from an untreated medical problem than about risks from using a drug without prescription gatekeeping. For example, products in this category might include prescription drugs for smoking cessation, some of which have already been made OTC, and HIV-prevention drugs like those under the California model described above. These harm-reduction approaches could entail temporary OTC status, requiring a prescription for long-term use once the patient was able to initiate therapy.

Broadly, potential targets for OTC status will likely have the following characteristics, many of which are already considered by the FDA when evaluating whether a drug is a candidate for an Rx-to-OTC switch, although not as part of a consistent default assessment of new drug applications. First, they will have wide therapeutic margins (in other words, a large quantity of the drug must be consumed to induce toxicity and, if taken by someone without a relevant clinical need, the drug is unlikely to be harmful). Second, they will not require complex clinical tests—either to determine whether the drug is indicated or to monitor physiologic changes across time that are not readily accessible to the patient. That said, as more clinical tests become available to patients outside the clinical setting (as with self-ordered blood tests), additional drugs might become candidates for OTC status. Third, the drugs will treat conditions and symptoms that are easily self-diagnosed or for which the consequences of misdiagnosis would not be severe or would prompt patients to seek further care. And, finally, they will be administrate by the patient without involvement of a medical professional.

The regulatory assumption that new drugs should be made available only by prescription uncritically encroaches on patient autonomy and therefore requires reassessment. The FDA should be applauded for its efforts to identify innovative methods by which drugs may be made safe for OTC use without relying on labeling alone. But it can go even further by adopting a default for new-drug approval that will consider products for OTC status unless prescription status is justified by concern for third-party harms or a need for group soft paternalism. Both sponsors and the FDA have important roles to play in encouraging evidence-based approaches to assessing patient capacity to use products safely without a prescription gatekeeper.

Our proposal is not meant to encourage patients to forgo a clinician’s care simply because they can access certain drugs on their own. It would, however, challenge twenty-first-century clinicians to help patients understand the value of their professional expertise in achieving patients’ goals, without passively—and perhaps paternalistically—relying on prescription gatekeeping to bring patients into the office for wellness visits and preventative health care. Ideally, patients will continue to seek their clinicians’ counsel because they independently find it useful, regarding their clinician as a trusted advisor and provider of expert care rather than as an unnecessary gatekeeper to a desired prescription.