U.S. flag

An official website of the United States government

NCBI Bookshelf. A service of the National Library of Medicine, National Institutes of Health.

National Academies of Sciences, Engineering, and Medicine; Health and Medicine Division; Board on Health Sciences Policy; Committee on Pain Management and Regulatory Strategies to Address Prescription Opioid Abuse; Phillips JK, Ford MA, Bonnie RJ, editors. Pain Management and the Opioid Epidemic: Balancing Societal and Individual Benefits and Risks of Prescription Opioid Use. Washington (DC): National Academies Press (US); 2017 Jul 13.

Cover of Pain Management and the Opioid Epidemic

Pain Management and the Opioid Epidemic: Balancing Societal and Individual Benefits and Risks of Prescription Opioid Use.

Show details

6Opioid Approval and Monitoring by the U.S. Food and Drug Administration

As the federal agency responsible for protecting the public's health by assuring the safety, efficacy, and security of drugs, the U.S. Food and Drug Administration (FDA) has a central role to play in addressing the opioid epidemic. The agency is responsible for approving new drugs and reformulations, giving it an important gatekeeping function, and also, along with the U.S. Drug Enforcement Administration (DEA), helps monitor the use of available opioid products. In this chapter, the committee provides recommendations aimed at improving the FDA's regulation of opioid analgesics, including by informing the agency's development of a framework for opioid approval and monitoring that takes into account the range of benefits and harms associated with the use of opioid analgesics, incorporating both the needs of pain patients and the relevant public health considerations.

Federal regulation of opioid medications has a long history. The original Pure Food and Drug Act of 1906—the first piece of U.S. federal legislation regulating the pharmaceutical marketplace—was passed in part because of widespread use of morphine in the so-called patent medicines of the 1800s, particularly in products aimed at children, such as Mrs. Winslow's Soothing Syrup, which was promoted for treating colic. The Pure Food and Drug Act required that products containing morphine indicate the quantity of the drug on their labels. The 1938 Food, Drug, and Cosmetic Act (FDCA) built on these rules by additionally requiring manufacturers to test their products for safety in human patients prior to approval. In the 1962 Kefauver-Harris Amendments, the FDA was given the further authority to ensure that drugs showed substantial evidence of efficacy from adequate and well-controlled investigations prior to approval.

As the FDA's authorities have evolved over the past century, so have the types of opioids available to U.S. patients. After the first synthetic opioid medications were developed in the 1910s, manufacturers continued to develop new products and formulations. In the 1960s and 1970s, the FDA approved short-acting combination products such as oxycodone/acetaminophen (Percocet, 1976). In the late 1980s and early 1990s, the FDA approved long-acting formulations of older opioid products, such as morphine (MS Contin in 1985) and oxycodone extended-release (ER) (OxyContin, 1995). Most recently, starting around 2010, the FDA has approved a cohort of opioids with supposedly abuse-deterrent properties, including tapentadol ER (Nucynta ER, 2011) and hydromorphone ER (Exalgo ER, 2010), although controversy arose when the agency approved hydrocodone ER (Zohydro ER) around this time without an abuse-deterrent formulation (ADF) (see Box 6-1).

Throughout all of these approvals, as well as other regulatory actions, the FDA generally has reviewed opioids through the same lens used for other drugs. The committee believes that the preceding chapters of this report establish a scientific and epidemiological basis for special treatment of opioids by the FDA that would involve greater integration of public health considerations at the time of preapproval testing, during regulatory review and approval, and during routine post-approval oversight.

Box Icon

BOX 6-1

Approval of Zohydro Extended-Release.

In making the case for this approach, this chapter begins with an overview of the FDA's current regulatory oversight of prescription drugs. This overview is followed by a discussion of public health dimensions of FDA drug regulation, which includes examples of previous cases in which the agency has successfully incorporated public health considerations into its regulatory decision making and an examination of those public health considerations specifically relevant to the approval and monitoring of opioids. Next, the chapter lays out the key elements of an integrated framework for opioid regulation that incorporates these considerations. Finally, the chapter presents the committee's recommendations for the implementation of such a framework; these recommendations are summarized at the end of the chapter.


This section of the report briefly reviews the key principles of FDA drug regulation and their application to opioids.

FDA Review and Approval of Prescription Drugs

Drug development often begins with the identification of cellular targets and corresponding candidate compounds, with the most promising compounds moving on to preclinical studies. Preclinical in vitro and in vivo animal studies seek to establish initial pharmacologic activity and, importantly, potential for toxicity. FDA oversight at this stage is limited,1 although the agency has promulgated requirements for good laboratory practice.2

Once a compound has demonstrated sufficient preclinical activity to warrant investigation in humans, an Investigational New Drug (IND) application is filed with the FDA. Information required in an IND application includes drug chemical and manufacturing information, pharmacologic and toxicologic information from preclinical data, a summary of any prior human data, a protocol for each planned study, and a brief outline of the clinical study plan. The FDA reviews the application, which goes into effect 30 days after being submitted unless the FDA imposes a clinical hold. Once the IND has gone into effect, clinical studies may proceed, typically occurring in three phases. Phase 1 studies usually enroll a few, often healthy, volunteers to explore pharmacokinetic and pharmacodynamic parameters of the drug based on a small number of doses. Phase 2 studies begin to test the drug's optimal dosage in patients with the condition of interest, and may provide a first look at the drug's therapeutic potential.3 Phase 3 studies (if they are performed) enroll hundreds or thousands of patients and may require years to complete, although one review found that two-thirds of all new drugs are approved on the basis of trials lasting 6 months or less (Downing et al., 2014). These latter studies account for the majority of the spending on drug development, and “are intended to gather the additional information about effectiveness and safety that is needed to evaluate the overall benefit-risk relationship of the drug and to provide an adequate basis for physician labeling.”4 While the manufacturer controls the organization and execution of the trials, manufacturers can, and frequently do, consult with FDA staff at various times to receive advice on trial design and outcomes.5

At the conclusion of the clinical trials, the manufacturer submits a New Drug Application (NDA). There is a 60-day filing review period during which the FDA ensures that all the necessary information is organized within the NDA. The drug is reviewed under a standard 10-month pathway; however, drugs that appear to represent therapeutic advances may be granted a 6-month priority review schedule (FDA, 2014c). The raw data and the study reports submitted in the NDA are reviewed by teams of FDA staff with expertise in chemistry and manufacturing, pharmacology, toxicology, statistics, clinical medicine, and any other relevant fields to determine whether the data show that the drug is safe and that there is substantial evidence of its effectiveness. To meet the substantial evidence standard, the FDA traditionally interpreted its statute as requiring two adequate and well-controlled studies, each convincing on its own, because results from any single trial “may be subject to unanticipated, undetected systematic biases” (FDA, 1998). However, the Food and Drug Administration Modernization Act of 19976 amended the FDCA to allow efficacy to be demonstrated by one adequate and well-controlled trial under certain circumstances, and about one-third of all drugs are currently approved on the basis of a single pivotal trial (Downing et al., 2014).

The FDA synthesizes the efficacy and safety data that make up the NDA into a structured qualitative benefit-risk assessment (discussed in detail later in this section), leading to a determination as to whether the benefits of approval outweigh the risks for the particular clinical indication sought by the manufacturer. During the review process, the FDA may engage advisory committees of outside experts to obtain additional input and to provide a public forum for discussion of the drug. These committees also include at least one consumer representative and one nonvoting industry representative. One study examining more than 200 advisory committees between 2008 and 2012 found that approval was recommended 74 percent of the time, and approval subsequently was granted by the FDA in 79 percent of those cases (Ciociola et al., 2014).

The FDA also reviews the manufacturer's proposed labeling describing use related to the indications sought, and this labeling is finalized at the time of approval. This labeling contains, among other things, the drug's approved indications, directions for use, dosing frequency and duration, and route of administration and preparation, as well as clinically significant adverse reactions, safety hazards, or other limitations on its use. The labeling must be revised to include warnings of new, clinically significant hazards as soon as reasonable evidence of a causal association with the drug exists.7

Drug Reformulations

Many drugs approved by the FDA are reformulations of previously approved products. Reformulated drugs, which include nearly every opioid product approved in the past few decades, can be approved via an abbreviated pathway described in section 505(b)(2) of the FDCA through which the application relies on published literature or on an FDA finding of safety and/or effectiveness for an approved drug product. In these cases, the manufacturer provides data to bridge to the FDA's prior findings for the approved product, as well as data necessary to support any differences between the two formulations (FDA, 1999).8 The FDA can require studies to establish efficacy and safety, as well as additional safety studies should unforeseen safety signals arise (FDA, 2014b).

Application to Opioid Approval

The requirement that prescription drugs be subject to prospective clinical trials that provide data on their safety and efficacy is an essential component of the regulatory apparatus that protects patients, allows for collection of rigorous data that can guide clinical practice, and promotes a well-functioning prescription drug marketplace by preventing the widespread use of ineffective products. The FDA's standards for new drug approval, therefore, serve a key public health function. However, the investigational drug evaluation process also has important limitations, particularly with respect to the approval of opioids.

For example, showing that a drug has substantial evidence of efficacy does not necessarily mean that the drug is more effective than currently available therapies, or that the efficacy demonstrated is clinically meaningful. In the case of hydrocodone ER (see Box 6-1), the drug was tested against a placebo. Also, while the hydrocodone ER case showed a statistically significant improvement in pain outcomes, it is not clear whether the slight numeric difference in the pain scale is clinically meaningful for patients with pain, particularly since pain worsened overall over the course of the trial among both the subjects receiving hydrocodone ER and those receiving placebo (FDA, 2017b).

In addition, clinical trials sufficient to meet the FDA's efficacy standard can be conducted in a brief, highly protocolized setting and often exclude many patients who would be expected to get the drug following its approval. In the case of hydrocodone ER, the entire pivotal study was conducted among patients with lower back pain, and did not include patients with cancer, arthritis, or other conditions who may receive opioid medications for pain (FDA, 2017b). Clinical trials could be designed with more robust follow-up periods or be prospectively powered to ensure that well-known side effects are adequately measured. However, the FDA bases its approval decision on the data provided by the manufacturer at the time of the NDA and does not require that trials of investigational drugs be conducted with particular characteristics.

Post-Approval FDA Authorities

The FDA's regulatory authority continues following the initial marketing approval of a drug. Pre-approval prospective clinical trials cannot comprehensively assess the risks of drugs. Therefore, it is not unusual for specific questions to arise that do not preclude marketing but nevertheless warrant further investigation after approval. Additionally, risks observed in the clinical trials may require ongoing evaluation and mitigation, and post-approval monitoring may necessitate timely communication with health care providers and the public. These activities take place against a backdrop of industry activities promoting use of the drug to providers and patients.

Spontaneous Adverse Event Reporting and Active Surveillance

Traditionally, the FDA has relied on passive collection of spontaneous adverse event reports submitted by health care facilities, providers, drug manufacturers, patients, and others as its primary source of information about post-approval drug safety. Manufacturers are required to submit to the FDA within 15 days any reports of adverse events that are both serious and unexpected. Other reports manufacturers receive are to be submitted to the FDA quarterly for the first 3 years post-approval and annually thereafter. The FDA's Adverse Event Reporting System (FAERS) database stores all such reports. Physicians and patients may also submit reports to FAERS, but do so voluntarily (and rarely). FAERS data are available to the public, but often contain less information than may be needed to fully assess the relationship between the drug and the event in question (Findlay, 2015). Nevertheless, past examples of FAERS data being used to identify safety signals provide evidence that this type of passive post-approval surveillance does have some value (FDA, 2014d).

In addition to receiving and processing adverse event reports, nearly all brand-name manufacturers conduct active post-market surveillance of their products, which may include observational studies or other safety-related research. They report results of these pharmacovigilance activities to the FDA, along with their adverse event reports, in Periodic Adverse Drug Experience Reports, Periodic Safety Update Reports, or Periodic Benefit-Risk Evaluation Reports. The FDA now also has the capacity to actively monitor safety outcomes related to drugs in the post-approval setting. Based on a pilot program launched in 2008, the Sentinel System allows the agency, through an independent contractor that has established a secure distributed data network, to assess an emerging drug risk using data from a broad array of electronic health care data. While Sentinel has not yet facilitated rapid drug safety assessment and improved regulation, it holds promise for regulatory decisions to be based on big-data tools that help in organizing and evaluating evidence (FDA, 2015a).

Post-Market Commitments and Requirements

At the time a product is granted marketing approval, the FDA can impose various post-market requirements, or the agency and the manufacturer can agree on post-market commitments, intended to help address questions that arise during the review of the pre-approval data, to help assess a known serious risk or a signal of a serious risk, or to identify an unexpected serious risk when data suggest the potential for such a risk. These requirements and commitments can include clinical trials, observational studies, or the creation of patient registries, which can be used to help adjust the labeled indication or safety warnings, and even can lead to withdrawal of the approved indication (OIG, 2016). Yet despite additional authorities granted to the FDA in 2007,9 post-market requirements and commitments often are delayed or not completed (Fain et al., 2013).

Risk Evaluation and Mitigation Strategies

One particularly important type of post-market requirement is a Risk Evaluation and Mitigation Strategy (REMS). The FDA can require that a manufacturer develop a REMS to provide safeguards for the use of high-risk medications when the FDA determines that such safeguards are necessary to ensure that benefits of a drug outweigh its risks (Sarpatwari et al., 2014). A REMS may simply involve disseminating means of educating prescribers and patients about the drug, but it may also require manufacturers to implement “elements to assure safe use,” such as mandatory training or certification for prescribers and pharmacies, restrictions on dispensing, and targeted patient follow-up and testing that can rely on the establishment of registries (Sarpatwari et al., 2014). Although elements to assure safe use often target prescribing and dispensing practices, it is drug manufacturers, not health care providers, that are responsible for ensuring that REMS requirements are met (Zettler, 2015). Brand-name manufacturers also are required to periodically monitor and assess the success of their REMS.10

Evidence about whether REMS can substantially affect prescribing and dispensing practices is conflicting. In a 2013 report, the U.S. Department of Health and Human Services (HHS) Office of the Inspector General raised concerns about the effectiveness of REMS in improving safe use of drugs (HHS OIG, 2013). An evaluation of post-FDA approval use of bosentan (Tracleer), a treatment for pulmonary hypertension, uncovered a high level of nonadherence to liver function tests required among the elements to assure safe use in the REMS for the drug (Blanchette et al., 2015). On the other hand, some REMS with elements to assure safe use may be effective in reducing non-evidence-based off-label drug prescribing. One rigorous study found that the REMS for the thrombopoietin agonist eltrombopag (Promacta), which before the FDA eliminated the REMS required such elements to assure safe use as a signed acknowledgment of drug risks and semiannual patient monitoring, decreased off-label use of the drug (for an indication later approved by the FDA) (Sarpatwari et al., 2015). As mentioned previously in this report, the FDA has required a REMS with elements to assure safe use for ER/long-acting (LA) opioids, which currently requires manufacturers of these drugs to provide education to prescribers based on an FDA prescriber education “blueprint” (FDA, 2017e).

Individual professional schools have produced their own online REMS teaching modules based on the FDA REMS blueprint for ER/LA opioids. Boston University's SCOPE (Safe and Competent Opioid Prescribing Education) of Pain program was funded by an independent education grant awarded by the manufacturers of ER/LA opioid analgesics, known collectively as the REMS Program Companies or RPC.11 Boston University School of Medicine partnered with the Federation of State Medical Boards and the Council of Medical Specialty Societies in the development, execution, and promotion of the SCOPE of Pain program (Alford et al., 2015). The committee notes that education through REMS represents one source of education on safe opioid prescribing, but is not a substitute for fundamental knowledge of multidisciplinary pain care that utilizes nonopioid and nonpharmacologic strategies for managing acute pain and especially chronic painful conditions.

Communicating Drug Safety Information

The combination of data from passive adverse event reporting, the Sentinel System, and other surveillance activities conducted by the FDA and manufacturers, together with post-market commitments and requirements, can point to the need to update a drug's labeling. While the FDA can, under certain conditions, require the manufacturer to update the label with new safety information, primary responsibility for keeping labeling up to date for brand-name drugs lies with the manufacturer.12

A boxed warning (also called a black-box warning)—the most prominent safety warning on a drug's label—is appropriate when an identified hazard poses a risk of death or serious injury.13 A boxed warning can be required at the time of drug approval or after a drug is already on the market and, in tandem with the media coverage it inevitably generates, can reduce prescribing (Dorsey et al., 2010). However, some boxed warnings fail to change practice as substantially as expected, and physicians commonly prescribe drugs without regard to information in these warnings (Lasser et al., 2006).

When a label change is made after a drug's approval, it is often accompanied by a Drug Safety Communication. Between 2010 and 2016, 233 Drug Safety Communications were issued (39 in 2010, 66 in 2011, 29 in 2012, 32 in 2013, 16 in 2014, 30 in 2015, and 21 in 2016) (FDA, 2017c). One review found little impact of FDA drug risk communications on prescribing behaviors (Dusetzina et al., 2012).

Regulating Industry Promotion

After a drug has been approved, its manufacturer promotes it to prescribers and patients. Promotion to prescribers includes detailing (face-to-face interactions between a sales representative and a prescriber); educational programing; provision of drug samples; and direct financial incentives, such as meals, travel expenses, grants, and consulting fees (Pew Charitable Trusts, 2013). Research shows that pharmaceutical marketing to physicians has a strong, consistent, and specific effect on driving prescribing practices toward the product being promoted, particularly when it is not necessarily the first-line or most cost-effective therapeutic option available (Avorn et al., 1982; Manchanda and Honka, 2005). Similarly, direct-to-consumer (DTC) promotion affects prescribing by changing how patients interact with their health care providers—for example, by prompting patients to ask for a particular drug and increasing the likelihood that patients will be prescribed both appropriate and inappropriate medications (Kravitz et al., 2005; Skeldon et al., 2015; Spence et al., 2005). In the opioid context, McKinlay and colleagues (2014) conducted a study that involved showing primary care physicians two different video-based scenarios in which actors played patients with sciatica-like symptoms. In one of the scenarios, the “patients” requested oxycodone; in the other, they requested no specific pain medication. After viewing each scenario, physicians were interviewed about how they would manage the case: after viewing the scenario in which the “patients” specifically requested oxycodone, 19.8 percent of physicians prescribed that drug, compared with 1 percent following viewing of the scenario in which no specific pain medication was requested (McKinlay et al., 2014).

The FDCA prohibits false or misleading prescription drug labeling and advertising,14 and the FDA regulates the promotion of prescription medications and certain medical devices to both prescribers and patients by encouraging companies to portray products in a way that is truthful, balanced, and accurate (FDA, 2010b). Advertisements must provide fair and balanced information with respect to the risks and benefits of a drug, reveal material facts related to the representations in the advertisement, give comparable prominence to risk and benefit information, and not overstate efficacy or safety.15 If the FDA becomes aware of promotional material that it believes violates the law (e.g., states or implies that a drug can treat a condition when the FDA has not approved it for such use, overstates a drug's benefits, omits or downplays information about a drug's risks), it sends the company a letter asking that the promotional material be removed and/or corrected (FDA, 2015b). Improper prescription drug marketing also can violate other laws, including the federal antikickback statute; state consumer fraud statutes; and federal and state false claims acts, which permit the government to recover payments made for prescriptions (such as through Medicare or Medicaid) as a result of fraudulent advertising.

After a drug has been approved, prescribers ordinarily may use it in ways that the FDA has not approved (known as “off-label” use), a practice that is common in the field of pain medicine (Radley et al., 2006). When an off-label use is particularly risky and non-evidence-based, the FDA can factor this consideration into its post-approval regulatory decisions. For example, when data emerged showing that antipsychotics used off-label in elderly patients with dementia increased the risk of mortality, the FDA added a boxed warning that helped reduce such dangerous prescribing. Off-label use for opioids contributes to misuse and opioid use disorder (OUD), and the inevitability of such off-label use of opioids is another justification for the development of an opioid-specific FDA review framework (discussed later in this chapter).

While off-label use is common, industry promotion of off-label uses violates the FDCA by causing the drug to be misbranded or to be an unapproved new drug (Cortez, 2016). In recent years, constitutional questions have been raised about the FDA's ability to limit manufacturers' off-label marketing.16 In test cases, the drug industry and libertarian advocacy organizations have had some success in persuading courts that the FDA violates industry's First Amendment rights when enforcing its policies against off-label promotion.17 The agency is “currently engaged in a comprehensive review” of its regulatory framework for medical product promotion.18

Drug Quality and Supply Chain Security

Another key area of oversight for the FDA is drug quality and security. The primary means through which the FDA regulates drug quality is its Current Good Manufacturing Practice (CGMP) requirements. The agency's CGMP regulations cover the methods, facilities, and controls used for the manufacture, processing, packing, holding, or preparation of a drug.19 These requirements include standards for the qualifications of the personnel involved in drug manufacturing, for the design of facilities and equipment, and for sanitation and cleaning. The purpose of these regulations is to help ensure that a drug is safe and has the identity, strength, quality, and purity that it is represented as possessing. Before approving a drug, the FDA reviews compliance with CGMP requirements,20 and it continues to monitor compliance after approval.21

The FDA also oversees the security of the drug supply chain. The 2013 Drug Quality and Security Act amended the FDCA to create an electronic, interoperable system to “track and trace” many prescription drugs throughout the supply chain.22 Once fully in effect in 2023, the system will include product identifiers for certain prescription drug packages; information on who handles a drug each time it is sold in the United States; requirements that industry stakeholders investigate products suspected to be counterfeit, substandard, or otherwise illegitimate; and processes for notifying the FDA and others when illegitimate drugs are found.23 Various requirements will apply to drug manufacturers, wholesale drug distributors, repackagers, third-party logistics providers (entities that help coordinate distribution of a drug but never take ownership of it), and dispensers. The intent of this expansive system is to enable the FDA and industry to verify the legitimacy of drug products; enhance detection of counterfeit, substandard, or otherwise illegitimate products; and more easily conduct drug recalls.24 Although the track and trace system is designed to prevent illegal drugs from entering the pharmaceutical supply chain, it may help identify some instances of opioid diversion by providing more information about drug distribution.

Application to Safety Monitoring of Opioids

Opioids have been the subject of numerous post-approval strategies to address the serious safety concerns associated with these products, although thus far these approaches have had little effect in terms of stemming harms. For example, as reports of misuse and diversion of oxycodone controlled-release mounted, Purdue Pharma and the FDA fashioned a risk management plan in 2001 encouraging improved surveillance and the education of prescribers about the risks of the drug. In addition, the label was updated to include a boxed warning calling attention to the potential for misuse and diversion. But neither of these interventions appeared to have much effect on diminishing the rate of opioid overdose, which crested over the next decade.

Recent actions by the FDA have included requiring manufacturers of immediate-release (IR) opioid analgesic products to update the safety information in their product labeling (FDA, 2016b) and requiring additional warnings about interactions between opioids and benzodiazepines (FDA, 2016c). In 2012, the FDA imposed a REMS for all ER/LA opioid analgesics. As discussed in Chapter 5, the REMS requires manufacturers to provide unrestricted education grants to accredited continuing education providers to develop and provide voluntary prescriber education programs. To date there has been little evidence that the REMS has had much effect on prescribing practice or on curbing opioid-related harms. The current opioid REMS also has been criticized for providing inadequate checks on unsafe opioid prescribing practices (FDA, 2016d). Propelled by the unrelenting increase in opioid-related deaths in the United States, one element of the FDA's Opioid Action Plan, launched in 2016, is to expand the REMS for opioids to incorporate pain management, include a broader range of health care professionals involved in the management of patients with pain, include IR opioid analgesic manufacturers, and evaluate approaches for implementing mandatory pain management education for prescribers (FDA, 2017e).

Similarly, passive adverse event surveillance and active use of such systems as Sentinel have proven insufficient with respect to opioids or medications to treat substance use disorder (SUD), because of delay in reporting and detecting problems. The International Classification of Diseases (ICD) has multiple ICD codes for chronic pain, and there are known challenges with diagnosis and documentation in medical records and billing for stigmatized conditions. The most recent post-marketing requirements for ER/LA opioids include studies to validate better mechanisms for extracting these data from medical records (FDA, 2014e).

Recent efforts to augment the post-market surveillance of opioid medications include, but are not limited to, the development of the Researched Abuse, Diversion and Addiction-Related Surveillance (RADARS) system for active, real-time surveillance, with the aim of using this information to guide risk reduction interventions. Developed by Purdue Pharma after the FDA provided suggestions and comments, it collects this information through regular surveys of individuals entering or being assessed for SUD treatment, experts in SUD, and law enforcement agencies, as well as analysis of exposure calls to poison control centers pertaining to misuse and diversion of licit and illicit drugs, including prescription opioid analgesics (Cicero et al., 2007). Around the same time, the National Addictions Vigilance Intervention and Prevention Program (NAVIPPRO) was developed to provide post-marketing surveillance, signal detection, signal verification, and prevention and intervention programs for scheduled therapeutics. The surveillance component of NAVIPPRO integrates multiple data streams to monitor drug use both temporally and spatially at a product-specific level, in part by collecting data from a national network of SUD treatment centers on substances used by adult individuals entering treatment (Butler et al., 2008).

Another recent step taken by the FDA was to update the shared list of post-marketing requirements for all ER/LA opioid analgesics in 2014 (see Annex Table 6-1 at the end of the chapter), such that the holders of the NDAs for the entire class would be responsible for performing 10 observational studies to assess the known serious risks of misuse, OUD, overdose, and death associated with these products, as well as one clinical trial to assess the risk of hyperalgesia associated with long-term, high-dose opioid therapy (FDA, 2014e). These required observational studies focus on the development and validation of algorithms or measures to identify patients exhibiting signs of SUD, including through electronic health records and other data, and the use of these algorithms to support studies of patients prescribed these products long-term to determine the risk and risk factors for the known serious adverse events. Beyond these class-wide studies, manufacturers of individual opioid analgesics can be subject to additional requirements related to safety signals and other issues that arose during NDA review or have arisen in the post-market context.

Finally, the FDA's rules concerning marketing and promotion did not stop manufacturers from engaging in illegal off-label marketing, as well as dissemination of advertisements that overstated the benefits of opioids and downplayed the risks of addiction.25 As discussed in Chapter 1, one well-publicized example involved Purdue Pharma's marketing of oxycodone ER for chronic noncancer pain during the years after its approval. During that marketing campaign, Purdue Pharma promoted oxycodone ER to prescribers and also engaged in DTC promotion through brochures, videotapes, and a “Partners Against Pain” website (VanZee, 2009). That marketing effort drove oxycodone ER sales from $48 million to more than $1 billion as the drug became the most prescribed brand name opioid for moderate to severe pain. Therefore these promotional practices were a strong contributor to the subsequent and ongoing increase in oxycodone misuse and oxycodone-related deaths (Dhalla et al., 2011; GAO, 2003). State and federal prosecutors have sued opioid manufacturers for allegedly fraudulent marketing in violation of the law.26 However, the penalties imposed in these cases invariably fall well short of the billions of dollars in revenues earned by opioid manufacturers as a result of these marketing campaigns.

Scheduling of Opioids Under the Controlled Substances Act

As discussed in Chapter 5, the five schedules for drugs covered by the Controlled Substances Act (CSA) (see Table 6-1) were designed to provide a structure for balancing the nuanced requirements of perceived safety, medical utility, and “abuse potential” (Spillane, 2004). Scheduling status affects prescribing authority (e.g., manner of prescribing and limits on refills), triggers requirements for supply chain record keeping, and determines the degree of criminal punishment for illicit trafficking. The most restrictive controls on use cover Schedule I and II substances.

Decision Making About Scheduling

The CSA allows the DEA to place a drug temporarily in Schedule I when it believes the drug may pose “imminent hazards to public safety.” The substance may be retained in Schedule I for up to 3 years, after which it must be removed or permanently scheduled.27 The DEA has used this temporary scheduling authority for more than 35 synthetic drugs since 2002. Most recently, the DEA has used it to place several synthetic opioids temporarily in Schedule I.28

TABLE 6-1. Schedules Under the Controlled Substances Act.


Schedules Under the Controlled Substances Act.

The CSA's somewhat ambiguous designation of authority to make permanent scheduling decisions is the result of a compromise that was reached at the time of its passage. The American Medical Association (AMA) resisted providing broad regulatory authority to the regulatory agencies, preferring that particularized decisions be made for each drug, similar to the approach of drug-by-drug approval used under the FDCA. Physicians distrusted the ability of federal regulatory agencies to accurately assess the therapeutic and research value of any given drug (Spillane, 2004), and pharmaceutical manufacturers feared that strict controls could have a serious impact on profitability. The FDA was uncomfortable with wielding enforcement power and ceded that power to the U.S. Department of Justice (DOJ). DOJ wanted to have the authority to control a drug quickly to address incipient issues of abuse. The resulting shared authority reflects an attempt to address all of those concerns.

Under the CSA, “If, at the time a NDA is submitted to the Secretary for any drug having stimulant, depressant or hallucinogenic effect on the central nervous system, it appears that such drug has an abuse potential, such information shall by forwarded by the Secretary to the Attorney General.”29 That determination by the FDA triggers a coordinated response by the FDA and the DEA designed to limit the potential for such abuse by assigning the drug to an appropriate “schedule.”30 The CSA requires HHS and the attorney general (usually acting through the FDA and the DEA) to consider eight factors in determining whether and at what level to schedule a drug: “(1) the drug's actual or potential for abuse, (2) scientific evidence of the drug's pharmacologic effect, (3) the state of current scientific knowledge regarding the drug, (4) the drug's history and current pattern of abuse, (5) the scope, duration and significance of abuse, (6) risk to public health, (7) the drug's psychic or physiologic dependence liability and (8) whether the substance is an immediate precursor of a substance already controlled under the CSA.”31

The FDA begins the process by making a recommendation as to whether the drug should be “controlled or removed as a controlled substance” and if so, “the appropriate schedule, if any under which such drug . . . should be listed.”32 In so doing, the FDA is directed to consider factors (2), (3), (6), (7), and (8) above, as well as any other relevant scientific or medical considerations.33 The FDA may require the drug's manufacturer to provide relevant data pertaining to its abuse potential as part of the NDA requirements.34 The FDA's recommendation is binding on the DEA, although any specific scheduling recommendation is not.35 If the FDA has recommended that the drug be controlled as a scheduled drug, the baton passes to the DEA administrator for consideration of the above eight factors in the context of appropriate scheduling.36

The FDA and the DEA thus play different roles in their evaluation pursuant to the CSA. The FDA's role is to perform a risk assessment of the drug's abuse potential. If the FDA determines that such a potential exists, it may require appropriate labeling for both physicians and patients. As discussed above, under the REMS authority, the FDA may also require various measures intended to ensure safe use of the drug, including the provision of patient medication guides, prescriber and/or patient agreements, and prospective registries. It can recommend extensive education for prescribers and counseling for patients as well. The DEA may also have input into the drug's labeling, and is responsible for licensing manufacturers of scheduled drugs and prescribers and for setting quotas for Schedule I and II drug production. The DEA has enforcement authority for violations under the CSA.

Once a drug has been placed in a schedule, that placement is unlikely to be changed (Henningfield and Schuster, 2009). “Down-scheduling,” or moving a drug to a tier with fewer controls, is very rare; however, “up-scheduling” has occurred in recent years in the context of increased prescribing and misuse of opioids. Schedule changes may be initiated by the FDA, the DEA, Congress, or any other interested party.37 In such cases, as in the original scheduling, the DEA seeks scientific and medical advice from the FDA and then acts through formal rulemaking. In the case of hydrocodone combination products, for example, a physician specializing in treatment of SUD petitioned the DEA to reschedule those products from Schedule III to Schedule II in 1999 (DEA, 2014). Five years later, following its review of the abuse potential of the drug, the DEA forwarded relevant data on the petition to the FDA for a scientific and medical evaluation. When the FDA undertook a review based on the eight factors listed earlier in 2008, it paid special attention to how rescheduling might affect prescribing practices. It found, for example, that rescheduling might result in the need for additional physician visits, that prescribers might then opt for oxycodone rather than hydrocodone combination products since they were in the same schedule, and that patients might receive inadequate pain relief (FDA, 2012). The FDA then recommended that hydrocodone combination products be maintained as Schedule III drugs. In 2009, after receiving another petition for rescheduling, the DEA sent additional data to the FDA providing further information about misuse in 2009, and the FDA undertook another review. In 2013, after an advisory committee voted 19-10 to recommend a scheduling change, the FDA forwarded a letter to the DEA recommending the rescheduling of hydrocodone combination products to Schedule II. The DEA issued a final rule to that effect in 2014.

Effects of Scheduling on Medical Practice

The design of the CSA reflects the inherent tension between optimizing the medical benefits of the controlled drugs and minimizing the dangers associated with their misuse. This tension is reflected in the CSA's tiered classification scheme, which anticipates that the responsible agencies will balance these considerations in making scheduling decisions. The tension is also evident at the level of the individual prescriber, given that placing a drug in the higher schedules can have a chilling effect on medically appropriate prescribing. As discussed in Chapters 2 and 5, prudent clinical judgment is required in deciding whether, when, and how to taper or terminate prescribing of opioids for patients reporting chronic pain. Well-meaning providers may be concerned about whether continued prescribing over the long term might be regarded by law enforcement or licensing agencies as being without “legitimate medical purpose” on the part of a practitioner “acting in the usual course of his professional practice,” and therefore in violation of the federal or state CSA.

Despite a DEA guidance document that attempts to clarify those terms,38 they may create enough concern that providers may choose not to prescribe controlled substances at all. For providers who do prescribe controlled substances, the CSA's tiered scheduling has had a more nuanced effect. Schedule tiers impose different prescribing requirements; CSA scheduling also affects how state law may impose additional requirements on prescribing of drugs assigned to the various tiers. Schedules III–V do not impose stringent prescribing limitations, but for Schedule II substances, prescriptions may not be refilled, the amount of drug or duration of use that may be prescribed on a single prescription is limited, and the prescription is required to be in written form. Many states require triplicate forms for Schedule II drugs and limit prescriptions to a short duration.

Schedule II requirements may increase providers' reluctance to prescribe substances that are so classified. Scheduling requirements do not provide incentives for providers to find other avenues for treatment, and they are not coupled with education. Making prescribing difficult for all providers, regardless of patient population, may result in denying access to individuals who need these drugs (Noah, 2003). The reclassification of hydrocodone combination products in 2014 has provided a natural experiment with which to study the effect of moving a drug to Schedule II. As noted in Chapter 5, early evidence shows that the reclassification substantially reduced the prescribing of these drugs (Chumpitazi et al., 2016; Jones et al., 2016), but whether health outcomes have improved as a result remains to be seen. Indeed, concern has been raised that rescheduling opioids to Schedule II is an unduly blunt instrument with which to limit overprescribing, and that it may have serious offsetting effects for individuals who need adequate pain treatment (Dineen, 2016). If rescheduling were simply to deter prescribing, the objection raised by the AMA when the CSA was adopted would be validated. More research is needed to study the effect of scheduling to Schedule II on pain treatment.

Three points emerge from the committee's review of CSA scheduling. First, the CSA requires explicit trade-offs between the effects of regulatory decisions on legitimate medical use and the harms associated with misuse and OUD. Second, the rescheduling of hydrocodone combination products reveals the diverging perspectives of the FDA and the DEA in exercising regulatory judgment on these issues and the inefficiency thus produced. Finally, because the FDA has many tools available under the FDCA for balancing these interests, the experience with hydrocodone combination products highlights the virtues of harmonizing the regulatory analysis undertaken under the two statutes, especially in relation to opioids.

Current FDA Benefit-Risk Framework

Currently, after the FDA reviews an NDA, it lays out the key details to help guide its decision making. The benefit-risk table shown in Figure 6-1 had its origins in an FDA initiative of 2009 “to develop a structured approach for drug benefit-risk assessments that could serve as a template for product reviews, as well as a vehicle for explaining the basis for the FDA's regulatory decisions in drug approvals” (FDA, 2013, p. 1). In 2012, section 905 of the FDA Safety and Innovation Act formalized this commitment by requiring the agency to implement a structured benefit-risk framework in its new drug approval process.39 The FDA states that while “quantitative assessments certainly underpin” any regulatory decisions, this approach is “designed to support the identification and communication of the key considerations in FDA's benefit-risk assessment and how that information led to the regulatory decision” (FDA, 2013, p. 4).

The framework displayed in Figure 6-1 has been applied explicitly in a number of cases since 2012, although not yet for an opioid product. For one product, pimavanserin (Nuplazid), a treatment for the hallucinations and delusions of Parkinson's disease, the framework revealed that the FDA considered the unmet clinical need for a treatment in the “Analysis of Condition” row, the on- and off-label use of other available drugs for this purpose (and their outcomes) in the “Current Treatment Options” row, summaries of the pivotal efficacy trial (“Benefit” row) and serious adverse event profile (“Risk” row), and any key post-market surveillance activities in the “Risk Management” row (FDA, 2017a).

FIGURE 6-1. Current FDA benefit-risk framework.


Current FDA benefit-risk framework. SOURCE: FDA, 2013.

The committee believes that this framework was developed thoughtfully and that it achieves its patient-centered goals by clearly organizing the components of the benefit-risk decision leading to a drug's approval. The committee also believes that this framework can be adapted to specifically integrate public health considerations, and can be incorporated into a much more comprehensive approach to gathering and reviewing the available information and making decisions accordingly to guide the FDA's regulation of opioids. As described below, the FDA routinely considers broader public health goals in its assessment of drugs, and the committee believes there is a growing public health mandate to apply this flexibility in certain ways to the approval and oversight of opioids.


To approve a drug, the FDA must determine that the drug is safe and efficacious “under the conditions prescribed, recommended, or suggested in the proposed labeling.”40 The FDA has long interpreted this approval standard as meaning that a drug's benefits must outweigh its risks. Since at least the early 1990s, the agency also has acknowledged that it has “flexibility” in applying the approval standard, and in “determin[ing] the kind and quantity of data and information an applicant is required to provide” to demonstrate that a drug meets the standard.41

One of the ways in which the FDA exercises this flexibility is by integrating public health considerations into its benefit-risk determinations. Public health considerations may include how the availability or use of the product will affect an unintended population or the broad public health impact resulting from the aggregated effects on patients taking the drug. For drugs with the potential for misuse, for example, NDAs must include “studies or information related to abuse of the drug,”42 which, of course, is not information about the use of the drug as directed in the proposed labeling. The FDA's authority to consider the broad impact of its pre- and post-approval decisions on the health and well-being of American patients and consumers is an extension of the FDA's primary role as a public health agency.43

Indeed, various provisions of the FDCA and FDA regulations make clear that the FDA has considerable discretion in determining what information is relevant to its regulatory decisions. Consistent with this flexibility, the FDA considers the public health consequences of its approval decisions in many aspects of its oversight of prescription drugs. For example, it may require a REMS, safety labeling changes, or post-market studies or trials to address risks of misuse, SUD, and overdose associated with a drug.44 When requiring a REMS, the agency also must consider the broad context within which the drug will be used, including the burden on patient access and the health care delivery system.45 As another example, as noted earlier, the FDCA requires holders of approved NDAs to report to the agency any adverse drug experiences, regardless of whether the drug was used as directed.46 Likewise, the FDA's Sentinel initiative is intended to identify and analyze a broad range of drug risks, not limited to those associated with the intended patient population using the drug as directed.47

The following examples of FDA decision making with respect to testosterone products, transmucosal IR fentanyl (TIRF) products, antibiotics, and prescription acetaminophen products further illustrate that the FDA is able to integrate, and has integrated, public health considerations into its drug approval and withdrawal decisions pursuant to its existing authority under the FDCA. In addition, this integration of public health considerations into regulatory decisions has encompassed decisions regarding the content that must be in drug labeling and REMS requirements. The following examples are not exhaustive. The FDA has incorporated public health considerations into numerous other decisions not described in depth in this report, including its approval of vaccines, requirements for misuse warnings on all opioid labeling, and certain requirements for labeling of over-the-counter (OTC) drug products, among others. Since the agency already incorporates these issues into its decision making in various contexts, integrating public health considerations into its regulation of opioids—including its approval decisions on new opioids—would be consistent with both its past practice and a generally accepted understanding of its statutory authority.

Examples of the FDA's Taking a Public Health Approach to Regulation

Example 1: Testosterone Products

In 2009, the FDA received a series of adverse event reports of children who had not been prescribed testosterone gel suffering serious side effects after inadvertent exposure to the products. After reviewing these cases, the FDA determined that the labeling for the products failed to adequately protect children from unintended side effects because some patients for whom they were being prescribed did not follow instructions, and as a result, children were coming into direct contact with the patients' treated skin (FDA, 2009b). In response, the FDA required manufacturers of certain formulations to include a boxed warning on the products' labels and implement a REMS that included a medication guide providing more thorough instructions for the user. At the time, this regulatory action drew some attention because it was the first instance of the FDA's requiring a REMS designed exclusively to protect a third party rather than the patient. (The manufacturers did not contest the label changes.)

Example 2: Transmucosal Immediate-Release Fentanyl Products

TIRF products are intended to manage breakthrough pain in adults with cancer who are already taking, and are tolerant to, other opioids for their consistent pain. TIRF products, however, pose significant public health risks, including diversion, misuse, and overdose.48 These risks are particularly acute for off-label use among nonopioid-tolerant patients and for accidental exposure and toxicity in children, because TIRF products come in a variety of easy-to-ingest forms, including sublingual and buccal tablets, lozenges, nasal sprays, and buccal soluble films.49

Because of these risks, FDA regulation of TIRF products provides an example of how the agency has dealt with concerns about the use of a drug in unintended populations. The FDA reviews published at the time of the TIRF product approvals address the risks of use in nonopioid-tolerant populations and accidental exposure and overdose in children, suggesting that the agency considered those risks as part of its approval decisions.50 That the agency considered these risks in its approval decisions is further apparent from the TIRF products' approved labeling. All TIRF product labeling contains a contraindication for nonopioid-tolerant patients and a warning explaining that TIRF products contain fentanyl in a dose that can be fatal to a child, and advising that patients ensure the products' proper storage and disposal.51

Beyond these measures, the TIRF REMS is designed to mitigate the risk of exposure to nonopioid-tolerant patients and children. Express goals of the TIRF REMS include “prescribing and dispensing TIRF medicines only to . . . opioid-tolerant patients” and “preventing accidental exposure to children and others for whom [the TIRF product] was not prescribed” (FDA, 2015c). To accomplish the first of these goals, the REMS requires prescribers, dispensers, and patients to confirm that they are aware of the risk of TIRF products for nonopioid-tolerant patients and the contraindication for that population. To accomplish the second goal, the REMS requires a prescriber–patient agreement form in which the prescriber documents that she or he has counseled the patient on the risk that TIRF products pose to children and on proper storage, and in which the patient documents that she or he understands this information. In sum, the FDA has considered the risks of the use of TIRF products by unintended patient populations in its approval and labeling decisions, as well as in the design of the REMS, for these products.

Example 3: Antibiotics and Resistance

Antibiotic resistance has been recognized as a problem since the late 1960s (Swann et al., 1969), and the FDA has struggled with how best to regulate antibiotic use in humans and animals in light of this problem, which poses risks not only for the patient or animal being treated but also for the population broadly. Use of antibiotics in animals has been widespread, not only for treatment or prevention of illness but also because such drugs promote weight gain and feed efficiency. The FDA has been legitimately concerned that such use of antibiotics in animals leads to greater antibiotic resistance in humans.

The FDA first announced its intent to withdraw approval for penicillin and tetracycline for livestock production uses in 1977, but for decades, the agency struggled to provide conclusive evidence that such use posed risks to humans. Finally, in 2003 the FDA determined that while it did not have full proof of the resistance risks posed by livestock production use of antibiotics, it could not conclude that such use was safe. Accordingly, it issued a guidance document describing a risk-based assessment process for new antimicrobial animal drug applications (FDA, 2003). This document explained that the FDA expected new animal antimicrobial drug applications to demonstrate not only safety and efficacy for the intended animal use but also “reasonable certainty of no harm to human health” from that use. The document was followed by additional guidance further explaining the agency's thinking on mitigating the risks of resistance associated with use of antibiotics in animals.

With respect to human use, in 2003 the FDA published a final rule52 requiring specific language on labels for human antibiotics encouraging providers to limit prescription of the drugs. This language advises providers to prescribe antibiotics only when bacterial infection is strongly suspected and warns against the potential for antibiotic resistance. These admonitions appear at least four times on each label: at the beginning of the label, in the section on indications and usage, and twice in the precautions section. The precautions section also provides specific guidance for physicians in counseling their patients about the proper use of antibiotics. Additionally, the 21st Century Cures Act, enacted in 2016, amended the FDCA to create an approval pathway for antibiotic drugs intended for patients with unmet medical needs that would require the drugs' labeling to caution prescribers that the drug is intended only for a limited population.”53

Example 4: Prescription Acetaminophen Products

Acetaminophen is an active ingredient in many prescription combination drug products for pain, such as hydrocodone/acetaminophen (Vicodin), as well as in OTC pain relievers, such as Tylenol. It has been a persistent cause of liver injury, and acetaminophen overdoses—both intentional and unintentional—are a leading cause of acute liver failure in the United States. The FDA has taken numerous steps to address the problem, including working with the National Association of State Boards of Pharmacy to label prescription medications containing acetaminophen more clearly, organizing a 2002 advisory committee meeting regarding OTC acetaminophen products, launching a patient education campaign in 2004, initiating an internal agency working group on acetaminophen in 2007, requiring changes to OTC drug labeling in 2009, and holding another advisory committee meeting in 2009 focused on both OTC and prescription products (FDA, 2009a).

In January 2011, the FDA published a Federal Register notice announcing that it was taking two additional steps;54 requiring a warning about hepatoxicity on the labeling of prescription drugs containing acetaminophen55 and asking prescription drug manufacturers to limit the maximum amount of acetaminophen per dosage unit to 325 mg (previously, some products had dosage units with as much as 750 mg). The agency explained that if manufacturers did not comply voluntarily within 3 years, it would use its authority under section 505(e) of the FDCA to withdraw approval of any prescription acetaminophen products that exceeded the new maximum dosage unit strength. The agency ultimately was successful in removing all high-dose acetaminophen products from the market by March 2014 (FDA, 2014a).

These actions all involved consideration of the broader public health implications of acetaminophen use. In particular, in explaining its rationale for planning to withdraw approval of prescription acetaminophen products that did not comply with the new maximum dosage unit strength, the agency pointed to some conventional individual health considerations, including the lack of evidence suggesting that the benefits of the higher-strength products outweigh their risks and the need to establish a larger margin of safety because of uncertainty about the precise toxicity threshold for different patient populations. But the agency also discussed various public health considerations. One basis for its decision was the high risk of unintentional overdose—in other words, the risks associated with the drugs when patients do not use them as directed. The FDA also discussed some of the societal impacts of acetaminophen-associated overdoses, including the estimated 56,000 emergency room visits, 26,000 hospitalizations, and 456 deaths per year caused by such overdoses (numbers far lower than those for opioid overdoses) (Nourjah et al., 2006). Additionally, the agency pointed to the contribution of prescription acetaminophen products to the high incidence of acetaminophen-related liver injury as another reason for withdrawing approval of the higher-dose products.

Public Health Considerations Relevant to Opioid Regulation

Some reasons why opioid analgesics warrant a unique regulatory approach are summarized in Table 6-2. As discussed in previous chapters, in addition to pain relief, opioids can produce feelings of pleasure, relaxation, and contentment (NIDA, 2017). Misuse and diversion associated with seeking these effects, facilitated to some degree by variability in prescribing practices and suboptimal management of pain, have fueled the development of black markets for opioids and counterfeiters.

Accordingly, a public health orientation assumes particular importance in the case of opioids and opioid derivatives, which are associated with nonmedical use and OUD and often are diverted from the lawful system of medical distribution. Related problems arise for opioid agonists and partial agonists (e.g., medication-assisted treatment for OUD), which share many of the same chemical properties and, like opioids, are diverted from lawful medical distribution or used by others beyond the patients for whom they are prescribed. Even opioid antagonists (e.g., the overdose reversal drug naloxone) are used legally and regularly to great benefit beyond the individuals to whom they are prescribed, and sometimes must be administered to the individuals to whom they were prescribed by other persons.

As discussed in Chapter 4, an approach to opioid regulation that actively takes public health considerations into account also requires the recognition that actions taken with respect to one opioid will affect the use and misuse of other opioids, opioid derivatives, and forms of pain management and consideration of the social system shaping use of those drugs.

TABLE 6-2. Special Biological and Social Characteristics of Opioids and Opioid Derivatives.


Special Biological and Social Characteristics of Opioids and Opioid Derivatives.

The interrelations among regulatory decisions concerning different drugs are a prominent feature of the opioid marketplace. The FDA recognizes that continued decisions about drug A require updating the information about the benefits and risks of drug A and its alternatives. For example, in the wake of placebo-controlled randomized trials linking the nonsteroidal anti-inflammatory drugs (NSAIDs) rofecoxib (Vioxx) and celecoxib (Celebrex) to increased cardiovascular risk, the FDA required that new boxed warnings about cardiovascular risk also be added to the labels for older, nonspecific NSAIDs, even though the strength of the evidence implicating those drugs was based on observational data. In the case of opioids, the various drugs in the class interact in the legal and illegal markets and often substitute for one another, so regulatory decisions about opioid A should not be based solely on predicted outcomes among users of opioid A. As an extreme example, it might be excellent policy to approve a new opioid formulation that was expected to cause 500 overdose deaths per year if that new opioid reduced overdose deaths from all other opioids by 5,000 per year. Likewise, randomized clinical trials might show that an ADF of opioid B was safer than an ADF of opioid C, but for various reasons, opioid B would achieve little market penetration and would be used primarily by people who would not develop OUD in any case, while opioid C was positioned to displace use of the current dangerous non-ADF more successfully. Opioid B might win a head-to-head competition in a traditional clinical trial, but a more circumspect decision to instead approve opioid C would save more lives.

Finally, the social system surrounding opioids is a key driver of the committee's recommendations in this report. Integrating public health considerations into regulatory decision making helps in considering distributional effects of those decisions over time. Important state-to-state and regional differences in opioid prescribing and problems have been observed since oxycodone ER was introduced in 1995 (Cicero, 2005). Thus, an optimal regulatory process will consider not only what is best for the country as a whole but also the possibility that what is best for the country as a whole might create unacceptable problems in certain states or regions that are more vulnerable because of established opioid trafficking routes, migration patterns, poverty and unemployment, and other social determinants of opioid misuse and OUD. Similar logic applies to key subpopulations, such as pregnant women (see the discussion in Chapter 4) and persons with mental health conditions that historically have been heavily impacted by SUD (Edlund et al., 2010). The effects of a policy action on the health and welfare of these subpopulations could also serve as a “warning signal” for the population at large.


A public health perspective is necessary but not sufficient for rational opioid regulation. Rather, as reflected in the committee's recommendations in this chapter, public health considerations need to be embedded in a regulatory framework that is flexible enough to capture and weigh an array of diverse outcomes occurring at multiple levels, from individual to societal. This integrated framework needs to facilitate informed regulatory decisions throughout a drug's life cycle, and include built-in periodic monitoring of each decision's consequences instead of decisions being treated primarily as self-contained events. If correctly formulated, this integrated framework will minimize mistakes and allow the community to recover expeditiously from any that are made. This is the ideal scenario for decision making in the face of uncertainty, which is the hallmark of regulating new drugs. This approach may also be applicable to other drug classes with similar concerns related to risk of misuse, SUD, diversion, and illicit market–based substitutes, likely with some alterations for the specific issues those drug classes present, although in the context of this report the discussion is focused on opioids.

The starting point for the development of an integrated decision-making framework for the FDA's regulation of opioids is recognition that attempting to introduce considerations beyond the clinical trial and other scientific data presented in the NDA will substantially increase the complexity of the agency's decision-making process. To promote rational, data-driven, and transparent decision making under such conditions, the FDA would need to (1) identify all relevant outcomes; (2) quantify those outcomes to the extent feasible; and (3) integrate those outcomes into an evaluative framework, including a common metric that would facilitate comparison and balancing.

Step 1: Identifying All Relevant Outcomes

An integrated framework for opioid regulation would include all relevant outcomes with an impact on public health. For most drugs, these outcomes are adequately summarized by the potential benefits and risks for individuals for whom the drug is indicated in the labeling. For example, a regulatory analysis of a cholesterol-lowering statin drug would need to consider its impact on users' cardiovascular risk reduction (the potential benefits) and on the development of diabetes and muscle and liver injury (the potential risks), as well as any other outcomes that are anticipated to affect the health of the drug's users. Another important consideration in the regulation of many drugs is the quantification of “risk compensation.” For example, if statins make users feel less concerned about their diets, they may revert to less healthy eating habits, therefore partially offsetting the hoped-for benefits of the drug.

As discussed in more depth in the next section, application of this step to an opioid would involve consideration of its impact on such outcomes as users' short- and long-term pain relief and functional improvements (the potential benefits); hyperalgesia, misuse, OUD, overdose, and death (the potential risks); and the possibility of risk compensation. It could also include outcomes that are not directly experienced by the drug's users but affect families, communities, and society as a whole.

Other outcomes that might need to be considered include illegal markets for diverted prescription opioids, illegal opioids such as heroin and illicitly manufactured fentanyl, and counterfeit pills that look like prescription opioids. Approval or withdrawal of a prescription opioid on the legal market can affect levels of use of black market opioids such that the total net effect on mortality may be very different from the apparent effect if one considers only outcomes directly related to the approved or withdrawn opioid. Indeed, one survey found that roughly three-quarters of people who used heroin in the past year misused prescription opioids first, and seven of ten people who used heroin in the past year also misused prescription opioids over the same period (Jones, 2013). The challenge of monitoring indirect effects also is illustrated by the introduction of ADFs, which may prevent misuse through specific modes of administration (e.g., injection or insufflation) but may have unintended impacts (see discussions in Chapters 4 and 5).

Step 2: Quantifying the Outcomes

Traditionally, the FDA appropriately relies on randomized trials, observational studies, and other patient experiences to quantify drugs' benefit-risk profiles. However, because supplying opioids for long-term use outside of medical facilities creates additional risks from misuse and diversion, many of the relevant outcomes cannot be identified or quantified using the FDA's usual research tools. In these situations, regulation would need to be informed by data on behaviors of intended users or others designed to evade or neutralize desired outcomes (including intentional efforts to defeat the system, such as by physician shopping or operating “pill mills” in the case of opioids). To accomplish this important task, the FDA might need to monitor nontraditional data sources (e.g., prescription drug monitoring programs, relevant online message forums, and special populations such as people in treatment for OUD) to quantify the extent of these behaviors in support of its regulatory decisions.

Evaluation of this full spectrum of outcomes is an inherently interdisciplinary task that requires alternative data sources and inputs from experts in epidemiology, economics, and other social and behavioral sciences. Although improvements in measurement and surveillance are under way, the precision and completeness of the tools available to measure the many relevant outcomes are not ideal (Secora et al., 2014), and may never be given the illicit nature of most opioid misuse. However, sound regulatory decisions need not overlook important benefits and risks just because they are difficult to quantify. In addition, incorporation of the full range of considerations need not be postponed until all pertinent data sources have been developed, but may proceed tactically and strategically, incorporating available outcomes and data sources as they are developed and improved.

The outcomes would ideally be measurable in at least one extant surveillance system. Risks could generally reflect mortality (e.g., risk of fatal overdose) or substantial morbidity (e.g., measures of OUD among women of childbearing age). A denominator reflecting the drug's availability or its potential for misuse or diversion at the local level could be applied to aid in comparing across the components (Butler et al., 2008; Secora et al., 2014). Because geographic trends could be especially useful in risk-benefit considerations, preservation of the lowest possible geographic unit for numerator and denominator might be optimal. Another important choice would be what to consider as the denominator. For example, morphine milligram equivalent (MME) availability could be used because data on dispensed medications are readily available at high levels of specificity (zip code, county), and because availability of illicit drugs can be captured at this geographic unit level and equated with prescription opioid–generated MME to provide a more accurate measure of the relevant health outcomes. Diversion and corruption of the drug's access mechanisms could be anticipated based on information on comparable products captured by government and private datasets. These “secondary” outcomes of the opioid under consideration could be estimated at the patient, provider, manufacturer, and distribution levels.

One of the FDA's major challenges would be to evaluate the currently available data sources addressing these outcomes and to work with the sponsoring agencies or institutions to improve these sources, such as by identifying gaps in the data and collaborating with partners to close those gaps or generate new datasets. Appendix C of this report provides a tabular summary of current data sources, as well as their strengths and limitations.

Step 3: Integrating Outcomes into an Evaluative Framework

Beyond creating a comprehensive list of outcomes and quantifying those outcomes, a key conceptual challenge is determining how to integrate many outcomes into a single framework that permits a transparent comparison of policies with differential effects on each outcome.

For most drugs, the procedure for weighing benefits and risks typically involves a mix of quantitative estimates (e.g., findings from clinical studies) and qualitative judgments (e.g., opinions of advisory committees). A 2012 Institute of Medicine report proposes a framework for assessing a drug's benefit-risk profile (IOM, 2012). For opioids, however, the weighing of benefits and risks is more complex than is the case for other drugs because the relevant consequences affect intended and unintended users as well as third parties, operate at multiple levels (individual, household, community), and encompass a wide array of fatal and nonfatal outcomes. Weighing benefits and risks in this context requires a decision-analysis framework that can adequately capture the dynamic interrelations among the many variables involved. One possibility, discussed briefly in Chapter 5, is building a mathematical model of the opioid system that simulates the expected outcomes. However, developing and testing such a model is likely to take several years, and the committee believes the need to expand the FDA's regulatory framework, including by incorporating unquantified elements and “best estimates,” warrants action to meet that need in the meantime.

Another challenge is to weigh the risks avoided by tighter regulation of an opioid against the pain, functional limitations, and other adverse effects experienced by patients who would benefit from that drug if its access were not restricted. The FDA's current approach informally weighs the available measures of pain utilized in clinical trials against estimated increases in misuse and OUD and the derivative risks. Although this approach will remain necessary for the immediate future, the committee also encourages and expects the FDA to explore use of a common yardstick (e.g., quality-adjusted life years) to incorporate all the outcomes of interest within a single metric.

The committee recognizes that no single quantitative exercise, even an integrated one using a common metric, can replace the agency's regulatory judgment for every decision. However, the FDA could quantify the outcomes as fully as possible given the available data and integrate these outcomes into a transparent framework that utilizes a common metric for measurement to the extent feasible. In the next section, the committee provides its recommendations for how this transparent framework might look and how it might be implemented.


In the committee's judgment, the FDA should take steps toward the implementation of an integrated, transparent framework for opioid regulation at three different stages of its decision-making process: clinical development, drug approval, and post-approval monitoring. Box 6-2 contains the committee's overarching recommendation framing this discussion.

Stage 1: The Clinical Development Stage

The FDA can first intervene to implement a new approach to opioid regulation after the submission of the IND application. During the investigational clinical trial period that follows submission of an IND application, crucial data currently are collected on the drug's pharmacodynamics, safety, and efficacy for intended users, but data also could be collected on its potential public health consequences. To date, evidence generation for opioids, as for many drugs, often has involved short-term trials involving narrowly defined patient populations (e.g., patients with back pain). A more comprehensive approach to organizing pre-approval trials could encompass

  • testing the drug in subpopulations at high risk of harmful outcomes, including those in locations of the country with high rates of misuse, OUD, or diversion;
  • including patients with mental health disorders and OUD and other populations in which opioid drugs are known to be widely used to ensure a representative sample of patients in the pivotal clinical trials;
  • measuring outcomes reported by household members or other third parties expected to be affected by the product (to partially overcome underreporting of misuse and OUD);
  • conducting continued testing of ADFs to understand the mechanisms of manipulation that might be used to defeat them; and
  • understanding interactions with other drugs (both prescription and illicit) commonly used with opioids or by people who use opioids illicitly, including how the drug interacts with antiretrovirals or anti–hepatitis C virus (HCV) medications.

While the committee understands that not all of these outcomes could be collected for every opioid being tested, this also may not be a comprehensive list—the particular public health outcomes would need to be specific to the opioid and its predicted effects. To that end, the FDA could issue a guidance document delineating the specific public health data that are likely to be most relevant to different types of opioids and that would need to be collected during pre-market clinical trials. This guidance document would explain the agency's current thinking on the overall development program and clinical trial design for opioids intended to treat acute and chronic pain. In addition to commenting on public health outcomes, the guidance could address the current state of the evidence on the essential features of trials for new opioids or opioid formulations, such as the duration necessary to collect appropriate outcomes. Such a document could also specifically address how the agency will handle new applications through the 505(b)(2) pathway frequently used for opioid reformulations or dosing changes. While reformulations may undergo less drug efficacy and safety testing, they may entail important public health considerations based on ongoing experience with the formulations that are currently being marketed. Similarly, studies have documented a positive opioid dose–harm relationship (with respect to OUD and death in particular). FDA approval of opioids through the 505(b)(2) process would need to involve the same rigorous data evaluation process as that used for approvals made under the traditional pathway.

Box Icon

BOX 6-2

Overarching Recommendation for Development of an Integrated Framework for Regulation of Opioids.

Communication of the types of public health outcomes sought by the FDA for a particular opioid could be communicated during the meetings that manufacturers are permitted to have with the FDA after each stage of testing, and at other times with sufficient notice. At these meetings, manufacturers may discuss plans for the design and outcomes of their trials, as well as the early evidence on the drug that has emerged. The FDA can impart useful advice during these meetings on optimal trial designs that can meet the considerations outlined in this chapter; indeed, according to one review, manufacturers that had an end-of-phase-2 meeting were far more likely to have their drugs approved than those that did not (Booz Allen Hamilton, 2006). Some manufacturers are diligent about having these meetings, while others are not. These meetings could serve as a useful mechanism for encouraging a new paradigm for opioid testing. The FDA guidance could suggest that manufacturers developing new opioids or new opioid formulations request a certain number of pre-approval meetings before submitting an NDA. These meetings could also help build a paper trail to inform FDA post-approval surveillance and help regulators understand why any recommendations about measurement of public health outcomes are not being implemented.

While the committee did not wish to make specific recommendations on what the FDA should do if manufacturers' development plans were to diverge substantially from the above guidance or if signals of potential problematic public health outcomes were to arise (such as evidence of diversion or misuse even in the highly structured environment of a clinical trial), issuance of a clinical hold is a strategy the FDA can use to delay additional proposed clinical studies or suspend an ongoing study. Reasons why a clinical hold may be issued under the current regulations include an unreasonable risk for subjects participating in the clinical research or a protocol for a phase 2 or phase 3 trial that is clearly deficient in design to meet its stated goals.56 Twenty-nine clinical holds were issued between 2008 and 2014 (Boudes, 2015), a remarkably low number given the number of investigational drugs being tested during those years. A clinical hold, if needed, could be issued as soon as possible after the IND was submitted or after the FDA received new information about ongoing opioid development trials, thereby reducing disruption for manufacturers and clinical trial enrollees. For example, if a manufacturer sought to bring a new LA formulation of an opioid to market without a tamper-resistant formulation, the FDA could decide to act at this point to hold the clinical trial until the company's rationale could be assessed. In this case, the proposed formulation could present an unreasonable risk of contributing to harmful outcomes among the subjects of the trial, and the trial would clearly be deficient in design, assuming that one of its stated goals would be to obtain FDA approval of the product.

As another example, if the FDA observed that a proposed pivotal trial for a new opioid or opioid formulation had not been designed to be of sufficient duration to enable collection of the necessary public health outcomes, this could be the basis for issuing a clinical hold until the trial had been redesigned. In this case, the agency might conclude that the trial was clearly deficient in design, assuming that one goal of the trial was to support FDA approval. The FDA could create an internal system to prioritize review of opioid INDs to facilitate the issuance of clinical holds, when warranted, and develop a similar system for integrating new information it received about opioids later in the development process to help in deciding whether clinical holds would be needed at any point.

The FDA could also specially consider the public health implications of opioid approval when making use of the multiple pathways leading to approval of investigational drugs. In addition to the 6-month priority review option, drugs can receive four other special designations to expedite their development or approval (see Table 6-3). While the expedited access provided by these pathways can be highly useful in cases of transformative new products or drugs intended to serve an unmet medical need, shortened development and review times have also been associated with negative public health outcomes. Drugs approved shortly before their regulatory deadlines have been found to be more likely to have post-marketing safety problems—including safety-related withdrawals and the need for added boxed warnings—relative to drugs approved at any other time (Carpenter et al., 2008, 2012). Drugs receiving faster reviews also have more spontaneous reports of drug-related adverse events (Lexchin, 2012; Olson, 2008; Reaves, 2009).

In the case of opioids, it would be inadvisable to truncate the development time in the absence of extraordinary circumstances. Instead, opioids and their secondary effects need to be fully investigated and the normal amount of time allotted to reanalyze the results of that investigation (currently 10 months for standard-review drugs). Because it is highly unlikely that a new opioid would satisfy the criteria for an expedited review or development pathway (e.g., fills an unmet medical need or offers a substantial improvement over available treatments for a serious condition), guidance might be issued defining how these pathways apply to opioids and other drugs with addiction potential. Recently, the 21st Century Cures Act of 2016 permitted supplemental approvals—for newly approved indications for drugs already on the market—to be granted on the basis of summaries of the data, rather than full FDA review of the underlying data. Again, the committee believes this truncated pathway is inappropriate for opioids, and instead review of the underlying data for supplemental NDAs for these drugs is necessary in all cases. Box 6-3 contains the committee's recommendations to the FDA for the clinical development stage.

TABLE 6-3. The U.S. Food and Drug Administration's Expedited Drug Development and Approval Pathways.


The U.S. Food and Drug Administration's Expedited Drug Development and Approval Pathways.

Stage 2: Drug Approval

The next major intervention point for the FDA in its regulation of opioids is the time of market authorization, when it is considering an NDA for a new opioid molecule or formulation. As indicated above, a decision usually is made at this stage based on the efficacy and safety data related to the specific drug for the intended clinical use. In making this decision, the FDA conducts a formal, qualitative benefit-risk assessment and ultimately arrives at a decision as to whether a drug's benefits to patients for whom it is prescribed outweigh its risks. The committee believes, given the evidence presented thus far in the report, that formal incorporation of public health considerations into the existing assessment process is warranted since the risks of opioids are so profound, and their diversion is so prevalent. To this end, using its existing legal authority to take into account the public health considerations outlined in Recommendation 6-1, the FDA would consider use by the individual patient (including, for example, the possibility that the drug would not be used as intended) or by unintended persons (such as household members), as well as the broader societal consequences of likely use, such as the scale of diversion and the overall impact of addiction on the health and well-being of patients who develop OUD. One would expect a thorough regulatory analysis of a new opioid within this framework to consider the drug's broad impact on untreated pain, the risk of diversion/OUD, the risk of overdose/death, and an assessment of the expected number of persons who would experience each of these outcomes. Relevant considerations for each of these factors could include the following:

  • Impact on untreated pain
    • expected prevalence of patients who would be served by the drug in question (versus with other opioids or with nonopioid treatment regimens);
    • pain relief observed in clinical trials (number of people benefiting and average improvement on pain and/or functioning scales);
    • differences between short-term effectiveness in highly protocolized clinical trials with selected patients and long-term effectiveness in health care settings (i.e., avoiding assuming that real-world impact on pain relief will correspond directly with outcomes of randomized controlled trials); and
    • prevalence of untreated pain if the drug were not approved and prescribed for the desired indication.
  • Impact on diversion/OUD
    • expected prevalence and frequency of nonmedical use, which could be extrapolated from data on people currently using a related compound nonmedically;
    • expected diversion and impact on existing black markets, again extrapolated from data on people diverting (e.g., giving, selling, exchanging, buying, or otherwise receiving from someone other than one doctor/one pharmacy) related drugs that have already been approved; and
    • expected prevalence and frequency of SUD involving the drug in question if approved and involving use of substitute opioids (e.g., other prescription opioids or illicit opioids).
  • Impact on overdose/death—estimated rates of fatal and nonfatal overdoses associated with or involving (1) the drug in question, (2) the compound in question, (3) other prescription opioids not of the same compound, and (4) illicit opioids.
  • Other public health outcomes—if the drug is injectable, the risk of transmission of infectious diseases (e.g., HIV and HCV) caused by such use.
Box Icon

BOX 6-3

Recommendations for the Clinical Development Stage.

Potential effects of the drug on individuals for whom it is indicated and prescribed—as well as those whose use of the drug is unintended and not as prescribed—can be anticipated during the pre-approval stage and, if the drug is approved, can then be monitored post-approval. The factors outlined above could fit into an opioid-specific expansion of the FDA's current benefit-risk framework presented earlier in Figure 6-1 (see example Table 6-4), used when making approval decisions on applications for new opioids, new opioid formulations, or new indications for approved opioids.

The proposed expanded framework includes measurable, opioid-specific considerations relevant to public health, including patient and public safety. Should the information thus amassed suggest to the FDA that an opioid product should not be granted marketing approval, the committee believes the current FDA practice of providing a response letter complete with the rationale for the decision and suggestions for positioning the application for subsequent approval would remain appropriate. Complete response letters traditionally are not made public, but recent research has shown that manufacturers' press releases often misstate the reasons for disapproval. Because of the significant public health concerns associated with opioids and the need to be able to evaluate the FDA's new regulatory processes accurately, the FDA may want to reexamine its policies relating to publication of complete response letters and consider what steps it needs to take to ensure that all complete response letters related to opioids are publicly released at the time of issuance. Notably, an FDA Transparency Task Force in 2010 proposed that releasing certain relevant documents currently kept confidential, including the agency's letters to drug, biologic, and device manufacturers when their products are not approved, would be consistent with existing agency rules related to safeguarding commercial information (FDA, 2010a).

TABLE 6-4. Example of an Adapted Benefit-Risk Framework for Approval of Opioid Products.


Example of an Adapted Benefit-Risk Framework for Approval of Opioid Products.

Box Icon

BOX 6-4

Recommendation for the Approval Stage.

The final rows in the opioid-specific framework in Table 6-4 relate to post-approval mitigation strategies, which are discussed in the following section. Box 6-4 contains the committee's formal recommendation to the FDA for the drug approval stage.

Stage 3: Post-Approval Monitoring

When the FDA makes an approval decision or after a drug is on the market, the agency can establish post-approval commitments and requirements, including whether the opioid requires a REMS. As detailed in prior chapters, prescribing of opioids for long-term use for chronic pain has led to numerous safety concerns that cannot be adequately addressed or anticipated in limited, prospective pre-approval trials. For this reason, the committee believes that rigorous, active post-approval monitoring of the ongoing safety and effectiveness of opioids is essential.

The key question is through what mechanisms optimal monitoring can occur such that the benefits of opioids are maximized and their risks minimized. As the FDA considers how to optimize its current post-approval monitoring authority for opioids, a useful way to integrate the review of the collected pre-approval data with the prospect of post-approval monitoring can be found in a three-step decision-making process previously proposed by an Institute of Medicine committee (IOM, 2012). In that process, each step corresponds to one of the three fundamental requirements for rational decision analysis under uncertainty reviewed in the previous section:

  • In the first step of the analysis, the FDA would define the public health question that prompted the need for a regulatory decision under the applicable statute. This step would include identifying the specific characteristics of the drug and the health problem at issue, available information about the drug, alternative treatments that are available, and plausible regulatory actions and their potential consequences. This stage would be aimed at identifying the information needed for a regulatory decision.
  • In the second step of the analysis, the FDA would evaluate the quality of evidence on both the benefits and the risks associated with the drug, including any new information that has triggered the need to consider regulatory action. The output of this step would include estimates of the likelihood and magnitude of a drug's benefits and risks and a characterization of the scientific evidence on which the estimates are based.
  • The third step of the analysis would involve synthesizing and integrating the estimates of benefits and risks and the quality of the evidence on which these estimates are based (Step 2) with the public health question (as specified in Step 1); deciding on the appropriate regulatory actions, including whether further study should be required; communicating the decision; implementing the regulatory actions; evaluating the effects of the regulatory actions; and, particularly in the case of complex or difficult decisions, evaluating the decision-making process and the impact of the actions taken on the public's health. Note that this step would involve deciding whether immediate regulatory action is warranted, or holding a decision in abeyance in anticipation of better information from additional study would justify the costs and consequences of further delay.

With this model in mind, the committee suggests a number of specific actions (reflected in Recommendation 6-5 in Box 6-5 at the end of this section) relating to the FDA's use of post-approval monitoring for opioids. The first set of actions relates to the use of currently available authorities, such as REMS, safety labeling changes, and risk communications. Currently, ER/LA opioids must be incorporated in a shared REMS, and the FDA has said that it intends to update the opioid REMS requirements to include IR opioids as well (FDA, 2017e) (the committee supports such a step). The current REMS for opioids is intended to reduce the serious risks associated with these formulations while maintaining access to the drugs for patients in need by educating providers about the limitations, benefits, and continued abuse potential of these formulations. However, the REMS may instead provide a false promise of risk mitigation (Nelson and Perrone, 2012). As discussed previously in this report, evidence is conflicting as to whether REMS can substantially affect prescribing and dispensing practices and is lacking on the effectiveness of the REMS for opioids. As part of efforts to improve its post-approval oversight of opioids, the FDA could make better use of REMS components that have been shown to improve prescribing practices (see Recommendation 6-5).

Meanwhile, the FDA could evaluate the data on the performance of the existing REMS, collecting additional data if needed, and change the features of the REMS so it would more optimally ensure the evidence-based use of opioids while reducing unsafe prescribing. For example, the FDA could consider additional supplemental education strategies when strengthening the opioid REMS, similar to the SCOPE of Pain program discussed in the first section of this chapter. Related considerations include how heightened prescribing restrictions might affect the supply of prescribers willing or able to prescribe opioids to patients with legitimate pain needs. Thus, it would be important to actively solicit the perspectives of prescribers and patients who are independent from the pharmaceutical industry in developing an optimal REMS. Development of an optimal opioid REMS also could be facilitated through collaborations between the FDA and other relevant government agency stakeholders, such as the Substance Abuse and Mental Health Services Administration, National Institute on Drug Abuse, U.S. Centers for Disease Control and Prevention, Health Resources and Services Administration, and U.S. Department of Veterans Affairs, among others.

Similarly, the boxed warning on opioids was strengthened in August 2016 to indicate that opioids carry “serious risks, including misuse and abuse.” It may be instructive for the FDA to study whether this change and the publicity surrounding it helped more prescribers and patients better balance the benefits and risks of opioid prescribing. For example, FDA efforts to communicate risk and safety information to providers and the public through advisories and warning labels appear to have variable impact (Dusetzina et al., 2012). If no clear effects are observed, the FDA could further modify opioid labels to include more specific statements about particular clinical situations, such as the management of chronic noncancer pain, in which there is clear evidence that the risks of opioids outweigh their benefits. These enhanced warnings could be included in the boxed warning, or disseminated through Drug Safety Communications and other media intended for a broad audience of prescribers and patients.

When new opioids are approved, requirements for boxed warnings or post-approval monitoring strategies such as REMS could be used as a way of justifying approval of a drug with an important safety risk, recognizing that the heightened post-approval surveillance or data from additional tests could inform changes to the label or even the marketing status of the product that might prove necessary. Before the FDA relied on such a strategy, however, the committee believes it would be best to study such post-approval actions as applied to opioids rigorously to ensure that they offer the real prospect of safety protections or timely acquisition of necessary data.

Another component of the committee's recommendation concerning post-approval monitoring pertains to the gathering of emerging information about the use of prescription opioids and how they are being used in both safe and unsafe ways. The collection of such information is part of the FDA's oversight of the safety and effectiveness of drugs in widespread use. Following approval of a new opioid or opioid formulation, initial estimates of the drug's risks and benefits would be informed and updated by data on the cumulative impacts of the drug as used in the community. As indicated above, the FDA might seek to impose post-marketing commitments or requirements to conduct ongoing studies. As the committee proposes in Recommendation 6-5, the FDA should engage in active surveillance of data on the use and misuse of approved opioids. This surveillance might include monitoring of new data that emerge from post-market commitments or requirements or the REMS program, which could be acted on efficiently and integrated with spontaneous adverse event reports and other observational data conducted through the Sentinel System. Other mechanisms for monitoring and generating new data might include periodic literature searches for independent reports of potential concern and the organization of prospective studies to respond to safety or other signals that might arise. Given the unique considerations related to opioids outlined in this report, the FDA might consider establishing a special center for opioid oversight to coordinate these activities in the Office of Surveillance and Epidemiology and work with the secretary of HHS to ensure adequate funding for its work. Newly emerging information might require changes to an opioid's labeling, although decisions to change the label wording or add safety warnings would ideally be guided by knowledge of whether past changes to opioid labeling have positively affected prescribing practices. Should such changes be deemed necessary, clear information dissemination plans that go beyond the Drug Safety Communication mechanism currently in place, for which there is insufficient evidence of effectiveness, would be essential.

Recommendation 6-5 includes establishing a new post-approval monitoring structure for opioids to promote adequate post-approval oversight that would include periodic follow-up. During these formal re-reviews, the totality of the pre- and post-approval data available at the time could be collected and an advisory committee convened to help the FDA review the drug's real-world use and outcomes. The FDA could develop guidance on the types of data that would lead to withdrawal of the drug, the requirements to revise the label, initiation of other REMS or monitoring pathways, or other outcomes of this review process. The progress of such post-market commitments or requirements could be reviewed, giving the FDA an opportunity to examine preliminary data. Conversely, if warranted, the FDA could take an enforcement action, including imposing civil monetary penalties authorized under the FDA Amendments Act, if a manufacturer failed to comply with post-market requirements (including by failing to comply with the timetable for a study or trial).

In extreme cases, after the formal re-review, the FDA might conclude that withdrawal of an opioid was necessary because its benefits no longer outweighed its risks. Notably, the FDA cannot require a mandatory recall of an approved prescription drug. If an approved prescription opioid were later found to be unsafe because it was contributing excessively to misuse and OUD, a recall would have to be initiated voluntarily by the manufacturer in response to an FDA request. The FDA could request a voluntary recall in such extreme circumstances, or other federal or state enforcement authorities could evaluate their potential enforcement roles.

A third component of the committee's recommendation on post-approval oversight of opioids is more effective regulation of industry promotional activities. The committee bases this part of the recommendation on the fact that, as discussed earlier in this chapter, decades of research have shown that industry promotion of prescription drugs to physicians and consumers influences prescribing practices (Robertson et al., 2012). The FDA could issue new guidance outlining what it views as responsible advertising and promotion of opioids to prescribers. Just as the FDA should move to incorporate public health considerations in its approval-related decisions for opioid drugs, it should incorporate such considerations into its review of industry promotional strategies for these products. Requiring that advertising of a drug explicitly mention these public health considerations might be necessary for the advertising of approved opioids to be considered accurate, truthful, and not misleading. For example, the FDA might require that advertising mention the risk that someone in the patient's household might misuse or sell the drug if it is not safely stored, or that it include specific statements about the risks of developing tolerance and OUD after unduly prolonged use for alleviating pain. More significantly, the FDA could find that there is no way to incorporate such broader considerations fairly into broadcast media advertising, ending the practice of DTC advertising of opioids via these media. The committee urges the FDA to issue guidance on responsible practices of DTC advertising and promotion as expeditiously as possible. Violations of promotional rules related to opioids should be pursued to the fullest extent of the government's current powers; in particular, off-label marketing of opioids should be carefully scrutinized.

Box Icon

BOX 6-5

Recommendation for the Post-Approval Monitoring Stage.

Box 6-5 contains the committee's formal recommendation for post-approval monitoring.

Implications for Other Regulatory Decisions

The framework outlined in this section was designed for new opioid products and formulations, but can be applied with equal force to opioids already on the market. Thus, in Recommendation 6-6 (presented in Box 6-6 at the end of this section), the committee recommends that the FDA conduct a full review of currently marketed/approved opioids. Such a review could be carried out by an expert panel that would systematically examine the current range of approved brand-name and generic opioids to determine which of these drugs remained effective and safe; which might need revised labels, formulations, or post-market requirements; and which should be withdrawn from the market entirely. Such a model could be modeled on the Drug Efficacy Study Implementation (DESI) of the 1960s and 1970s, in which the FDA worked in concert with the National Academy of Sciences/National Research Council to classify the risk-benefit ratios of the purported indications for drugs approved between 1938 and 1962 (NAS, 2017), ultimately finding that more than 300 products were ineffective for all indications and had to be withdrawn from the market, and more than 2,400 products had labels for indications for which they were ineffective. Although modeled on DESI, the Opioid Study Implementation (OSI) process envisioned by the committee could be carried out in a much shorter time frame and with far fewer resources than DESI because it would be limited to a single drug class for which the medical literature already provides substantial evidence to help answer the questions about opioids that the expert panel might want to address.

Although the OSI process would not be prohibitively expensive—and should be overall cost-saving to the U.S. health care system given its potential to reduce the substantial costs due to opioid-related harms—it would require sufficient funding sustained until the full range of available opioid products could be reviewed. In addition, several of the ideas offered for how the FDA might implement the committee's recommendations (e.g., the creation of a special center for opioid oversight within the Office of Surveillance and Epidemiology, routine post-approval reviews of new opioid approvals) would require additional regulatory resources. The cost of such interventions could be accounted for without additional legislation as part of the FDA's discretionary budget until the next reauthorization of the Prescription Drug User Fee Act, at which time the user fees applied to NDAs could be adjusted to account for the additional costs of adequate oversight of the prescription opioid market. Funding for this work might also be donated voluntarily by opioid manufacturers interested in helping to ensure a safer opioid marketplace. Another approach, which would require congressional action, would be to add a very small surcharge to each opioid prescription, in the same way that the National Childhood Vaccine Injury Act established a trust fund to compensate those suffering vaccine-related injuries through a $0.75 excise tax on each vaccine dose. All of these approaches warrant study to ensure that the FDA has the funding it needs to modernize its approach to exercising its vital role in oversight of the opioid market.

The committee recognizes that the OSI process might lead to the removal of some of the opioid formulations or doses currently on the market because it is highly unlikely that all of these products would be judged safe and effective under the new drug approval framework proposed in this chapter should they just now be entering the market. However, the committee does not believe that this process would unduly restrict the availability of opioids for appropriate use in treating pain syndromes overall, since one of the advantages of the proposed OSI process would be its public health scope and the ability to take into account the advantages and disadvantages of removing a product in the context of the current marketplace of pain treatment modalities. Additionally, the FDA could establish reasonable time periods within which manufacturers would have to come into compliance with decisions resulting from the OSI process to minimize any disruption to treatment resulting from changes to marketed opioids (and reduce burdens on industry). Patients also would not need to be concerned that the OSI process would affect the cost of opioids as long as sufficient numbers of generic manufacturers were producing the opioid formulations remaining on the market at the conclusion of the OSI review.

The committee also believes that its recommendations may be relevant to some of the next-generation pain medications outlined in Chapter 3. Many of these products are designed to be nonaddictive, in which case they could be reviewed under the FDA's normal paradigm. But the agency might have lingering doubts about how some products will perform in long-term or widespread use, in which case it might want to apply relevant recommendations detailed in this chapter. When considering the various guidance documents suggested in this report, the FDA could indicate which recommendations it believed would also apply to novel nonopioid pain medications that nonetheless posed a potential risk for misuse, OUD, or illicit use.

Similarly, it is possible that some of the recommendations offered in this chapter could be applied to other controlled substances, such as benzodiazepines, neurostimulants, or other performance-enhancing drugs. This possibility warrants additional study, and the committee expresses no opinion on whether other drug categories should be added to the special focus it proposes for opioids.

The FDA has approved several ADFs of opioids that have physical or chemical properties to prevent misuse, as noted earlier in this chapter. A component of the FDA's Opioid Action Plan is to expand access to ADFs to discourage misuse (FDA, 2016a). While ADFs may have a role in preventing escalation of opioid misuse, as discussed in Chapters 4 and 5, multiple factors will determine the impact of a given ADF on public health. These include such factors as whether shifts in use behaviors that occur as a result of attempting to defeat the abuse-deterring properties introduce risks and whether substitutions are made for comparably harmful prescription or illicit opioids. Indeed, in June 2017 the FDA requested that the manufacturer of the ADF Opana ER remove the drug from the market because of concern that the drug's ADF properties had led to increased injection of the drug and outbreaks of HIV and HCV, as well as cases of thrombotic microangiopathy (a serious blood disorder) (FDA, 2017d). The evidence on the specific role of ADFs in efforts to curb opioid-related harms is still developing. In light of continuing uncertainty about the benefits and risks of various types of ADFs, the FDA's cautious case-by-case approach appears warranted.

While the committee's recommendations for revised regulatory treatment pertain to brand-name and generic opioid products, many other products relevant to the opioid crisis, particularly opioid reversal agents (such as naloxone) and treatments for OUD, have been discussed in this report. To the extent that these products are intended to alleviate the opioid crisis and themselves present no risk of addiction, the committee favors rigorously testing them for efficacy and safety and making them widely available to patients as expeditiously as possible. In the case of these agents, REMS and other restrictive post-approval prescribing systems might do more harm than good by making them less available to patients and providers. The public health considerations relevant to approval of these drugs are therefore quite different from those outlined in this chapter and would not fit well under the proposed approach for opioid regulation. Thus, a different set of considerations may need to be enumerated in FDA guidance for products intended primarily to treat OUD or manage the opioid crisis rather than to treat pain.

The committee believes further that the process for initial DEA scheduling—and subsequent rescheduling—of drugs also could benefit from implementation of the approach discussed in this chapter. The FDA and the DEA are already required to take “risk to public health” into account in making scheduling decisions, but the considerations included under this heading have not been enumerated in detail. For example, there may be differences in the value placed by the FDA and the DEA on different public health risks, how heavily the two agencies weight these risks, and how they balance these risks against the potential health benefits of opioids. Thus, the committee favors taking the same public health considerations incorporated in the opioid benefit-risk framework into account when the FDA and the DEA evaluate the “risk to public health” criterion in making scheduling—and rescheduling—recommendations and decisions.

Finally, predictions about the various risks of initial scheduling and rescheduling decisions to various public health parameters need to be made based on solid data, and gathering such data will require development of proper methods and data sources. While recognizing that decisions about scheduling of opioids will have to continue based on the best available data while more data are generated, the committee supports a sustained commitment among funders and policy makers in the field to better understanding the outcomes of scheduling decisions.

Box Icon

BOX 6-6

Recommendations for Other Regulatory Decisions.


Traditionally, the FDA takes a product-specific approach to drug approval decisions by focusing on the data generated and submitted by the manufacturer on the drug at hand, and balancing the benefits of the drug revealed by those data against the risks known (and unknown) at the time of the review. While this process works well in most cases, the committee believes that the regulatory oversight of opioids needs to be viewed differently. The recommendations offered to the FDA in this chapter are intended to balance manufacturers' ability to introduce new opioid products that hold promise for pain management with the agency's obligation to manage the risks posed by opioids, which extend beyond risks to individual patients. In line with the FDA's public health authorities, mission, and practice, these recommendations focus on incorporating public health considerations into the entire life cycle of drug development to create a safer prescription opioid marketplace. If implemented, these recommendations will enable both the drug companies and the FDA to evaluate the full range of benefits and risks that need to be reviewed and considered before pre-market approval as well as during post-approval surveillance.

Given the well-described individual-, household-, and society-level outcomes that have emerged from decades of experience with opioids, special considerations are necessary in the opioid development, approval, and post-approval stages that incorporate the principles discussed in this report.

Recommendation 6-1. Incorporate public health considerations into opioid-related regulatory decisions. The U.S. Food and Drug Administration (FDA) should utilize a comprehensive, systems approach for incorporating public health considerations into its current framework for making regulatory decisions regarding opioids. The agency should use this approach, in conjunction with advisory committee input, to evaluate every aspect of its oversight of prescription opioid products in order to ensure that opioids are safely prescribed to patients with legitimate pain needs and that, as actually used, the drugs provide benefits that clearly outweigh their harms. When recommending plans for opioids under investigation; making approval decisions on applications for new opioids, new opioid formulations, or new indications for approved opioids; and monitoring opioids on the U.S. market, the FDA should explicitly consider

  • benefits and risks to individual patients, including pain relief, functional improvement, the impact of off-label use, incident opioid use disorder (OUD), respiratory depression, and death;
  • benefits and risks to members of a patient's household, as well as community health and welfare, such as effects on family well-being, crime, and unemployment;
  • effects on the overall market for legal opioids and, to the extent possible, impacts on illicit opioid markets;
  • risks associated with existing and potential levels of diversion of all prescription opioids;
  • risks associated with the transition to illicit opioids (e.g., heroin), including unsafe routes of administration, injection-related harms (e.g., HIV and hepatitis C virus), and OUD; and
  • specific subpopulations or geographic areas that may present distinct benefit-risk profiles.

The committee acknowledges that the quality of data for some of these considerations (e.g., data from nontraditional sources, such as rates of transition from prescription to illicit opioids) is currently suboptimal, but nevertheless stresses the need to include these considerations in a comprehensive public health framework to inform regulatory decision making for opioids.

Implementing this approach successfully will require significant changes in collection and analysis of data. One important implication is that the evidence necessary to demonstrate safety and efficacy for opioid products will necessarily broaden, and this will affect the traditional FDA review and approval process at multiple points. Specific considerations to meet these needs may extend beyond the protocolized setting of traditional clinical trials to encompass use of data from less traditional sources, such as online forums. The agency should include reports from family members or other third parties affected by the drug, as well as data on outcomes in subpopulations at high risk of OUD or with mental health comorbidities common in patients with pain. Outcomes of interest include impact on function and long-term efficacy for pain reduction.

Other data that could inform the agency's decisions include the drug's estimated impact on the demand for and availability of all other prescription and illicit opioids, as well as interactions with other drugs (both prescription and illicit) commonly used with opioids or by people who use opioids illicitly (e.g., considering how the drug interacts with antiretrovirals or anti-HCV medications). Nontraditional data sources will be needed to inform regulatory decisions for opioids. The FDA should also apply these nontraditional study design considerations in the setting of post-marketing requirements imposed as conditions of approval.

As discussed in Chapter 5, another important implication of the need to take a systems approach is that the agency, perhaps in collaboration with the U.S. Centers for Disease Control and Prevention (CDC) or other agencies, will eventually need to develop and implement a quantitative model of the opioid ecosystem and establish the data infrastructure needed to support and apply that model. An explicit model can better integrate information from different sources, articulate assumptions, incorporate dynamic processes, and assess the public health consequences of different decisions and value judgments. However, the committee recognizes that developing such a model will be a challenging task given the complexity of the opioid markets and consumption patterns and the weaknesses of the data currently available to measure several of the outcomes outlined in Recommendation 6-1. To begin the process, the agency could periodically convene experts in policy modeling to review available data and needs pertaining to opioid distribution, use, and consequences—with the eventual objective of formulating a conceptual map and a formal quantitative model of the opioid ecosystem. Doing so would enable the agency to better predict the effects of changes in policy or other changes in the opioid ecosystem.

Recommendation 6-2. Require additional studies and the collection and analysis of data needed for a thorough assessment of broad public health considerations. To utilize a systems approach that adequately assesses the public health benefits and risks described in Recommendation 6-1, the U.S. Food and Drug Administration (FDA) should continue to require safety and efficacy evidence from well-designed clinical trials while also seeking data from less traditional data sources, including nonhealth data, that pertain to real-world impacts of the availability and use of the approved drug on all relevant outcomes. The FDA should develop guidelines for the collection of these less traditional data sources and their integration in a systems approach.

Recommendation 6-3. Ensure that public health considerations are adequately incorporated into clinical development. The U.S. Food and Drug Administration (FDA) should create an internal system to scrutinize all Investigational New Drug (IND) applications for opioids. This review should examine whether public health considerations are adequately incorporated into clinical development (e.g., satisfactory trial design; see Recommendation 6-2). In implementing this recommendation, the FDA should rarely, if ever, use expedited development or review pathways or designations for opioid drugs and should review each application in its entirety.

The committee believes a commitment to transparency is critical to maintain balance between preserving access to opioids when needed by patients experiencing pain and mitigating opioid-related harms. Increased transparency would optimize the clinical development and use of opioids considering the proposed comprehensive systems approach.

Recommendation 6-4. Increase the transparency of regulatory decisions for opioids in light of the committee's proposed systems approach (Recommendation 6-1). The U.S. Food and Drug Administration should commit to increasing the transparency of its regulatory decisions for opioids to better inform manufacturers and the public about optimal incorporation of public health considerations into the clinical development and use of opioid products.

Steps the FDA could take to implement Recommendation 6-4 might include issuing a guidance document that outlines opioid-specific clinical development considerations, or releasing summary versions of complete response letters for opioid products to inform the public about the public health considerations that the FDA has determined would preclude marketing approval.

The committee believes that use of REMS that have been demonstrated to improve prescribing practice, surveillance activities, formal reevaluation of opioid approval decisions, and regulation of advertising and promotion are critical to supporting the safe use of opioids.

Recommendation 6-5. Strengthen the post-approval oversight of opioids. The U.S. Food and Drug Administration should take steps to improve post-approval monitoring of opioids and ensure the drugs' favorable benefit-risk ratio on an ongoing basis. Steps to this end should include use of Risk Evaluation and Mitigation Strategies that have been demonstrated to improve prescribing practices, close active surveillance of the use and misuse of approved opioids, periodic formal reevaluation of opioid approval decisions, and aggressive regulation of advertising and promotion to curtail their harmful public health effects.

Evidence on the effectiveness of the current REMS for opioids is conflicting and limited, and the REMS may provide a false sense of risk mitigation. To improve the data on the existing opioid REMS, the FDA could continue to evaluate the data on its performance, collecting additional data if needed and changing the features of the REMS so it more optimally ensures the evidence-based use of opioids while reducing unsafe prescribing. Maximizing the use of REMS and other post-approval oversight mechanisms for opioids may be facilitated through collaborations among the FDA and other relevant government agency stakeholders, such as the Substance Abuse and Mental Health Services Administration, National Institute on Drug Abuse, Health Resources and Services Administration, and U.S. Department of Veterans Affairs, among others.

The consistent regulatory oversight of opioid products under the committee's proposed systems framework will necessarily raise concerns about the safety and efficacy of products currently approved for market. The committee believes the FDA possesses the authority and responsibility to reexamine the opioid class of drugs, consistent with previous agency actions motivated by public health concerns with a drug class, to ensure that they remain safe and effective. Options for such a large-scale review include a process similar to that used for DESI or a process for reviewing individual applications that would give manufacturers a time frame within which to submit supplemental data necessary for the FDA's review.

Recommendation 6-6. Conduct a full review of currently marketed/approved opioids. To consistently carry out its public health mission with respect to opioid approval and monitoring, the U.S. Food and Drug Administration should develop a process for reviewing, and complete a review of, the safety and effectiveness of all approved opioids, utilizing the systems approach described in Recommendation 6-1.

Finally, the process for initial DEA scheduling of drugs could benefit from the explicit incorporation of the public health considerations discussed in this report. The FDA and the DEA are already required to take “risk to public health” into account in making drug scheduling decisions, but the considerations included under this heading have not been enumerated in detail, and the two agencies may differ in prioritizing certain benefits or risks. Moreover, the ultimate impact on health outcomes related to these decisions remains largely unknown.

Recommendation 6-7. Apply public health considerations to opioid scheduling decisions. To ensure appropriate management of approved opioids, the U.S. Food and Drug Administration and the U.S. Drug Enforcement Administration should apply the same public health considerations outlined in Recommendation 6-1 for approval decisions to scheduling and rescheduling decisions, and study empirically the outcomes of scheduling determinations at the patient and population health levels.

ANNEX TABLE 6-1. Extended-Release (ER)/Long-Acting (LA) Opioid Post-Marketing Study Requirements.


Extended-Release (ER)/Long-Acting (LA) Opioid Post-Marketing Study Requirements.




7 U.S.C. § 2131.


21 C.F.R. § 312.21(b).


21 C.F.R. § 312.21(c).


Note that “manufacturer” refers to an entity engaged in manufacturing, preparing, propagating, compounding, processing, packaging, or labeling of a product (e.g., a drug), while “sponsor” is defined in the regulations for Investigational New Drug applications as the pharmaceutical company, government agency, academic institution, private organization, or other entity that takes responsibility for and initiates a clinical investigation (21 C.F.R. 312.3). Although a drug's manufacturer may not be its sponsor, for simplicity, and because in the context of unapproved investigational opioids the committee expects that the sponsor will frequently be the manufacturer, the committee uses the term “manufacturer” throughout this chapter.


Public Law 105-115.


See FDCA 502(f)(2) for specific statutory language and 21 C.F.R. § 201.57(c)(6) for relevant regulations.


21 C.F.R. § 201.57.


Public Law 110-85.


21 U.S.C. § 355-1(d).


For more information, see http://www​.er-la-opioidrems​.com/IwgUI/rems/home.action (accessed June 27, 2017).


21 C.F.R. § 201.57(c)(6).


21 C.F.R. § 201.57(c)(1).


21 U.S.C. §§ 321(n), 352.


21 C.F.R. § 202.1.


Virginia State Board of Pharmacy v. Virginia Consumer Council, 425 U.S. 748, 748 (1976); Thompson v. Western States Medical Center, 535 U.S. 357 (2002); Sorrell v. IMS Health Inc., 131 S. Ct. 2653, 2670 (2011).


United States v. Caronia, 703 F.3d 149 (2d Cir. 2012); Washington Legal Found. v. Henney, 202 F.3d 331 (D.C. Cir. 2000); Amarin Pharma, Inc. v. FDA, 119 F. Supp. 3d 196, 224 (S.D.N.Y. 2015).


21 C.F.R. § 210.1–211.1.


21 U.S.C. § 355(d), 355(j)(4)(A).


21 U.S.C. § 351(a)(2)(B).


Public Law 113-54 127 Stat. 587 (2013).


Public Law 113-54 127 Stat. 587 (2013).


Federal Register, Vol 81, No. 181, September 19, 2016: 64175–64177, https://www​.gpo.gov/fdsys​/pkg/FR-2016-09-19/pdf/2016-22441​.pdf (accessed June 27, 2017).


U.S. House of Representatives Committee on Governmental Reform—Minority Staff Special Investigations Division, FDA Enforcement Actions Against False and Misleading Prescription Drug Advertisements Declined in 2003 (Washington, DC: Government Printing Office, January 2004).


21 U.S.C. § 811 (h) CSA and amendments Synthetic Drug Abuse Prevention Act of 2012, Subtitle D of Title XI FDASIA (P.L. 112-144).


An example is synthetic opioid 3,4-dichloro-N-[2-(dimethylamino)cyclohexyl]-N-methylbenzamide (also known as U-47700), placed in Schedule I in 2016. The DEA announced and then later withdrew temporarily the placement of kratom in Schedule I through this authority.


21 U.S.C. § 811(f). Since 1973, the attorney general has subdelegated authority for drug scheduling to the administrator of the DEA. See Exec. Order No. 11,727, 38 Fed. Reg. 18,357 (July 10, 1973); 28 C.F.R. § 0.100 (2003). Under the CSA, a scheduling proceeding “may be initiated by the Attorney General (1) on her own motion, (2) at the request of the Secretary [of Health and Human Services (HHS)], or (3) on the petition of any interested party.” 21 U.S.C. § 811(a).


As noted previously, there are five scheduled classifications under the CSA based on potential for abuse. Schedule I drugs are those that have high abuse potential and are not approved in the United States. Schedule II–V drugs are allowed to be marketed under restrictions depending on their potential (high to limited) for physical or psychological dependence (see Table 6-1 for further information).


21 U.S.C. § 811(c).


21 U.S.C. § 811(b). The FDA has a manual that outlines these procedures: MAPP 4200.3, Consulting the Controlled Substance Staff on Abuse Liability, Drug Dependence, Risk Management and Drug Scheduling.


21 U.S.C. § 811(c).


21 C.F.R. § 314.50(d)(5)(vii). This includes a proposal for scheduling under the CSA.


21 U.S.C. § 811(c).


21 U.S.C. § 811(c).


21 U.S.C. § 811(a). Congress can and does insert itself into this process. In 2000, Congress legislatively required emergency scheduling of GHB (liquid ecstasy), and the 2012 Synthetic Drug Abuse Prevention Act required permanent scheduling of a number of synthetic stimulants and opiates.


DEA Docket No. DEA-286P, Dispensing Controlled Substances for the Treatment of Pain (2006).


Public Law 112-144.


21 U.S.C. § 355(d).


21 C.F.R. § 314.105(c).


21 C.F.R. § 314.50(c)(5)(vii).


21 U.S.C. § 393 (1997).


FDCA Sections 505(o)(3) and (4); 505-1(2)(b).


FDCA Section 505-1(f)(2).


21 C.F.R. § 314.80.


FDCA Section 505(k)(3).


Abstral, Actiq, Fentora, Lazanda, Onsolis, and Subsys reviews, http://www​.accessdata​.fda.gov/scripts/cder/rems/index​.cfm?event=RemsDetails​.page&REMS=60 (accessed June 27, 2017).


See, e.g., Abstral Labeling at http://www​.accessdata​.fda.gov/drugsatfda_docs​/label/2011/022510s000lbl.pdf (accessed June 27, 2017).


21 C.F.R. 201.24.


Sec. 3042 of the 21st Century Cures Act.


76 Fed. Reg. 2691 (Jan. 14, 2011).


Section 505(o)(4) of the FDCA (added by the Food and Drug Administration Amendments Act of 2007 [FDAAA]).


21 C.F.R. § 312.42.

Copyright 2017 by the National Academy of Sciences. All rights reserved.
Bookshelf ID: NBK458654


  • PubReader
  • Print View
  • Cite this Page
  • PDF version of this title (5.4M)

Related information

  • PMC
    PubMed Central citations
  • PubMed
    Links to PubMed

Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...