A. The Specialty Emerges

The development of pain specialties resulted in part from new surgical techniques: the expanded use of medical treatments for surgical and malignant pain, the development of regional anesthetic techniques such as epidural anesthesia, and later more advanced techniques such as implanted intrathecal infusion pumps and spinal cord stimulation. Concurrently, a different group of doctors was pioneering new organizational approaches to managing and treating complex pain that focused on patients' experiences of pain. That aspect of the movement was led by Dr. John Bonica, an anesthesiologist who wrote the enormously influential Management of Pain in 1953^{Reference Fishman, Ballantyne and Rathmell22} and introduced the concept of multi-disciplinary pain management clinics;^{Reference Tompkins, Hobelmann and Compton23} and Dr. Raymond Houde, a physician and pharmacologist,^{Reference Gudas24} who around the same time, began to test and introduce the means of evaluating subjective responses to pain and pain relief provided by analgesics.^{Reference Seymour25} In the 1970s, Houde, Dr. Kathleen Foley, a neuro-oncologist, and nurses like Ada Rogers and Nessa Coyle, among others, created the pain and palliative care service at Memorial Sloan Kettering Cancer Center, using multidisciplinary pain management to treat patients suffering from cancer.²⁶

In the 1980s and early 1990s, the new techniques pioneered in interdisciplinary pain management clinics — many of which have proven successful over time — became standardized and formalized as specialties. The American Board of Medical Specialties (ABMS) approved the first board examination in pain management by the American Board of Anesthesiologists in 1991, and today there are Certificates of Special Qualification in pain management through multiple specialties, including anesthesiology, emergency medicine, neurology, and, as of 2014, family medicine.^{Reference Buffer27}

B. Bringing the Focus on Pain from the Specialty to Primary Care

The recent emergence of a pain subspecialty in family medicine — a primary care field — is just the most recent manifestation of the broader influence of the specialization of pain management on general medicine. This migration into primary care came about in large part because of efforts on the specialty side to combat what was viewed as an epidemic of under-treatment of pain. In 1991, the American Pain Society (APS), emphasizing the need to respond to patient pain, piloted quality assurance standards for the treatment of pain in the acute and cancer context. At the time, Dr. Mitchell Max, the president of the American Pain Society, noted the pervasiveness of under-treatment of pain. “[S]urveys in the past 20 years suggest that, even in the world's leading hospitals, a large proportion of these patients are not given sufficient doses of these analgesics to maintain comfort,” he wrote. “Pain relief has been nobody's job.”^{Reference Max28}

The APS continued to promote prescribing and a patient-centered approach. In 1996, Dr. James Campbell, then President of the APS, introduced the famous phrase “pain as the fifth vital sign” as a part of his presidential address to the society, noting: “If pain were assessed with the same zeal as other vital signs are, it would have a much better chance of being treated properly.”²⁹

The movement quickly gained momentum. In October 2000, the Department of Veterans Affairs mandated pain assessments in all of its facilities, publishing the “Pain as the 5^th Vital Sign Toolkit.”³⁰ That same month, Congress passed and President Clinton signed an act that established the “Decade of Pain Control and Research” as, following the “Decade of the Brain”, the second congressionally declared, medically-related decade.^{Reference Brennan31} The act aimed to incentivize research around pain and pain management by encouraging future congressional funding and to organize efforts through pain-related conferences.^{Reference Nelson32}

But the most important moment came in 2001, when The Joint Commission on Accreditation of Health Care Organizations (The Joint Commission), turned its focus to pain. The Joint Commission, a highly influential entity, is responsible for evaluating hospital performance and establishing practice standards for hospital physicians. As just one indicator of its impact, a hospital's reimbursement under Medicaid and Medicare depends in part on its compliance with The Joint Commission's standards.

In its 2001 report, The Joint Commission suggested new “standards for organizations to improve their care for patients with pain.” The standards established that “[p]atients have the right to appropriate assessment and management of pain.”³³ To meet the standard, health care organizations — of all kinds, not just specialty pain management — had to “plan[], support[], and coordinate[] activities and resources to assure the pain of all patients is recognized and addressed appropriately.”

Critically, the newly proposed standards required accredited health care organizations, including primary care offices, hospitals and ambulatory practices, to ask every patient about their pain at every point of contact. The standards mandated conducting an initial assessment into each patient's pain and conducting necessary follow ups.^{Reference Phillips34} The assessment, as a Joint Commission example suggested, could include asking a patient to evaluate the intensity of her pain — a suggestion that effectively evolved into to the question that became ubiquitous in medical encounters: “rate your pain on a scale from one to ten.”³⁵

Further reflecting the influence of the campaign, in 2006, the Centers for Medicare and Medicaid Services (CMS) included pain-related questions on the Hospital Consumer Assessment of Healthcare Providers and System (HCAHPS). One question, for example, asked patients, “During this hospital stay, how often did the hospital staff do everything they could to help you with your pain?” Those surveys allow for comparisons across institutions and can have profound effects on medical practice because of their effect on reimbursement.

These moves created strong incentives for doctors to treat pain aggressively. With patients effectively grading their doctors' pain management work for purposes of reimbursement, aggressive pain treatment became a central goal sand likely contributed to overtreatment. In a 2012 survey study of physicians, nearly half of the respondents reported feeling that the HCAHPS led to poor care and the inappropriate prescription of opioids.^{Reference Zgierska, Rabago and Miller36} In an attempt to ameliorate the problem, the CMS recently replaced questions about patient satisfaction with pain treatment with questions focusing on patient-doctor communication about pain.³⁷ Rather than asking about how effectively pain was controlled, for example, one new question asks patients, “[d]uring this hospital stay, how often did hospital staff talk with you about how much pain you had?” Whether changes of this nature reduce overtreatment remains to be seen in.

The Joint Commission itself was sued at the end of 2017 by several localities in West Virginia — the state hit hardest by the opioid crisis thus far — for its role in facilitating the epidemic with those 2001 standards.^{Reference Asbury38} The Joint Commission recently announced its plan to implement new assessment and management requirements for accredited hospitals, which now include requirements like “setting realistic expectations” for patients and “[p]romoting safe opioid use by identifying high risk patients”; accredited hospitals must also now provide non-pharmacologic treatment options and improved educational materials on assessing pain and safe prescribing.³⁹

C. Pharma's Influence

Physicians, of course, were only one of the key actors. While it would be speculative to guess about the initial marketing plans made by Purdue, Pfizer, and other pharmaceutical companies during this time, there was certainly an awareness then that commonly available opioid, anticonvulsant, and antidepressant medications were being prescribed more frequently, and that there was the potential for sales growth. In 1996, the sustained release formulation of oxycodone (OxyContin) was released; it quickly became a “blockbuster” drug.

With respect to OxyContin, at least a part of the company's plan was to use the initial research coming from the pain and palliative care literature (much of which had been paid for with pharmaceutical industry support),^{Reference Zee40} as well as hired specialty physician advocates, to convince primary care practitioners that stronger time release opioids could be used safely for non-malignant pain. Some have alleged that the American Pain Society, The Joint Commission, and the Federation of State Medical Boards (FSMB) allowed themselves to be too heavily influenced by the pharmaceutical industry in making their recommendations.^{Reference Keefe41} Pain experts used to market opioids in the late 1990s also have been made defendants in several major court cases addressing the crisis.^{Reference Addy42}

D. Results and Risks of Backlash

We conclude this part with some thoughts about the arc of the specialty of pain management. The relatively recent recognition of the importance of the field and the creation of the specialty were hard fought gains, and there is a risk that backlash from the current crisis will undermine some of those advances. At the same time, there have been positive developments as a result of the current crisis. As recognition of the inherent risks of opioid therapy has increased, scrutiny of these risks has increased, too, leading to a revision of treatment; clinicians now speak more forthrightly about the risks inherent in the use of opioid pain medications and the importance of proper management, and they may refer more patients to specialists for complicated pain management. There is a feeling of responsibility on the part of the specialty to show leadership on these issues.

On the other hand, some emerging data,^{Reference Kertesz, Gordon and Satel43} plus some early anecdotal evidence from the palliative care context, indicate a risk that the pendulum may have swung too far: opioids may be becoming difficult to obtain for patients with severe pain who truly need them, creating a challenge between finding the necessary balance between safety and availability of treatment.^{Reference Atkinson, Schatman and Fudin44} Many states have passed laws restricting access to long-term opioids, and some pharmacies have endorsed similar practices. Some doctors, faced with new federal guidelines that would limit the amount of opioids that can be prescribed, may not feel comfortable seeking available exceptions for their patients; but their patients may not have insurance coverage for alternative therapies. In almost the mirror image of the controversial Joint Commission guidelines of 2001, several major groups this year — the Pharmacy Quality Alliance, the National Quality Forum, and the National Committee for Quality Assurance — have endorsed “an opioid quality metric that counts the number of patients receiving over 120 MME against the performance rating for a physician or other care provider.”⁴⁵ Physicians have come to learn that their prescription practices are being monitored by state databases as well. Fears of malpractice or even prosecution, sometimes known as “opiophobia,” may lead to a return to under-prescribing.

There is also the possibility of the rise of a new kind of stigma. Some patients may be scared off by the crisis, or now be too embarrassed to have conversations about pain and pain management that are central to the specialty practice. Whereas addiction treatment has long suffered from stigma as a barrier for some who would otherwise seek help, now some may feel similarly less comfortable talking to doctors about their pain. Specialists have a particular responsibility both to educate patients about the risks of pain and specific forms of pain management, as we discuss below, but also to be vigilant in insuring that needed access to pain is not unduly foreclosed.

A. Assessment

It is notoriously difficult for doctors to assess the underlying causes of chronic pain and its impact on a patient. Pain is inherently subjective; chronic pain, by definition, represents a syndromic manifestation of maladaptive neurologic changes in the body. It can result from a multitude of causes, and frequently persists after the initial injury or illness has already resolved. Unlike acute pain, which typically has a clearly identifiable cause and can be lifesaving and protecting (e.g. leading someone to move her hand away from a hot fire), chronic pain serves no known protective purpose and is in fact a disease in itself.

Adding further complexity, the experience of chronic pain can vary significantly between patients who may have differing pain generators: For one patient, “severe” pain can be driven largely by emotional distress, maladaptive coping behaviors, and lack of self-care; for another patient, with pain of the same self-reported severity, pain can be due more directly to neuroanatomic pathology, or brain conditions.^{Reference Gatchel48} This can be very difficult for the provider to untangle, particularly if they do not have significant experience or training in pain medicine. Furthermore, there is no clinically available blood test or other diagnostic study that can “prove” a patient is suffering from pain, so self-report is the only tool that providers have to make their assessment. Diagnostic tests may help to identify underlying triggers and can be useful to guide care, but pain assessment should focus on pain's severity, impact on quality of life, and impact on function.

Unfortunately, most initial assessment methods, especially those used by busy primary care doctors who need to address a broad array of other disorders, tend to focus only on unidimensional self-reported pain severity (e.g., the “rate-your-pain-on-a-scale-ofone-to-ten” question). This numerical rating scale (or the similar visual analog scale) fails to account for the impact that chronic pain has on a patient's life, including how it affects general activity and function, and therefore may be of limited clinical efficacy, despite how widely it has been adopted in primary care.^{Reference Krebs, Carey and Weinberger49}

By contrast, specialists may have greater facility with other diagnostic methods, including more comprehensive multi-dimensional assessment tools like the McGill Pain Questionnaire or the Brief Pain Inventory. When necessary, pain specialists also may use other validated questionnaires such as the Opioid Risk Tool (ORT) to predict when patients may have an increased risk for aberrant behavior during opioid therapy.

Regrettably, these challenges are complicated by the lack of pain management and opioid safety education in medical schools.^{Reference Mezei and Murinson50} For example, in 2011 only 5 out of the nation's 133 medical schools had a mandatory course of pain management.⁵¹ Graduate medical education is not doing much better: A more recent survey of surgical residents found that fewer than 10% of surveyed trainees reported having received any formal training in pain care or opioid prescribing.^{Reference Chiu52} It is not surprising that providers often complain about lacking the confidence to treat patients complaining of pain.^{Reference Yanni53}

In the arena of pain assessment, pain specialists thus have several advantages over their primary care colleagues. By virtue of their training, pain specialists have a more robust understanding of the mechanistic etiologies of chronic pain, how to assess its severity and impact, and how to use the various diagnostic tools at their disposal to evaluate pain triggers further. And specialists may benefit from their experience when gauging the typicality of a patient's pain response, as compared to previous patients who have presented similar problems. Because of the narrower scope of a specialist's practice, pain specialists are also able to focus their assessment in a way that primary care doctors — tasked with caring for a broad array of disease states — simply cannot. They may also be more sufficiently resourced, which allows them to use the more comprehensive but cumbersome clinical assessment tools, and may have greater comfort using advanced imaging and diagnostic tools. Finally, since pain specialists are a relatively scarce resource in the United States, patients who are seen by a pain specialist may be more motivated to engage in this care and more able to afford the care that is recommended.

B. Safety Monitoring and Predicting Misuse

Compounding the problem, non-specialist providers frequently fail to follow recommended opioid safety monitoring guidelines, such as checking urine drug tests for patients on chronic opioid therapy. Highlighting the severity of this challenge, a 2011 retrospective cohort study conducted in 8 university-affiliated primary care clinics found that only 8% of patients on chronic opioid therapy received a urine drug test.^{Reference Starrels54} Although imperfect, urine drug testing can help doctors monitor for treatment adherence and screen for diversion or use of non-prescribed controlled substances; the Centers for Disease Control recommends that physicians use drug screenings as a precondition of ongoing opioid prescribing.^{Reference Dowell, Haegerich and Chou55}

It is unclear why so many non-specialist providers fail to conduct urine drug screenings, but, unfortunately, confidence in prediction alone is generally unfounded: For example, a small study found that, of patients who primary care clinicians thought were not at risk for misuse of medications, 60% had urine drug tests showing illicit drugs or the absence of prescribed drugs.^{Reference Bronstein56} Although there is a paucity of data on the utilization of recommended safety monitoring tools by pain specialists, we hypothesize that the specialist does this better for the same reasons that they are more empowered to assess pain, described in the preceding section. That said, it is noteworthy that the data described above predates the release of the 2016 CDC Guideline on the use of opioids for chronic pain, so it will be worth monitoring to see if the use of safety monitoring tools by primary care providers improves as familiarity with the guideline expands.

C. The Impact of the Structure of the Doctor-Patient Encounter

The typical patient-doctor encounter also varies dramatically across practice settings. This can result in different treatment regimes.

The amount of time a health care provider has for a patient is one important difference. In general, primary care providers report not having enough time with their patients.^{Reference Geraghty, Franks and Kravitz57} Although the average length of time for appointments may not vary much between settings — in primary care, studies find appointments tend to last between ten to twenty minutes;^{Reference Dugdale, Epstein and Pantilat58} in specialty care, it may be greater than 21 minutes^{Reference Shaw, Davis, Fleischer and Feldman59} — primary care providers typically see a significantly greater number of patients and patients come in with a broader range of issues. On average, a primary care physician addresses three patient problems per visit.^{Reference Beasley60} The primary care doctor must treat a host of other ailments with which a pain specialist does need to assess or manage (e.g., hypertension, diabetes, cardiac disease, etc.). She also must direct preventative health efforts (e.g., breast cancer screening, colon cancer screening, vaccines, etc.), and quarterback the care of any other specialists and consultants. As a result, primary care providers may not be able to spend as much time outside an appointment working on patient issues or focusing specifically on pain during the appointment. Moreover, in the specialty practice, the 21 minutes the doctor spends with the patient may be complemented by additional time with support staff on preliminary intake or screening.

Reimbursement also plays an important role. In general, primary care providers are reimbursed at a lower rate than specialists when providing the same types of services, which drives the need to see more patients. All primary care services are collapsed into a limited number of reimbursement categories, which depend on encounter length, while specialists can bill (at higher rates) for particularized forms of treatment or interventions. Primary care providers also are often not reimbursed for key and time-consuming functions they nevertheless perform, including maintaining a patient's electronic records or coordinating a patient's care. Although CMS recently introduced new billing codes for primary care, there are concerns that core primary care services like patient evaluation and counseling remain undervalued.⁶¹

Depending on their training and the focus of their practice, some pain specialists can offer entirely different (and frequently non-opioid) forms of care. For example, some specialists often provide procedural-based care (e.g., epidural steroid injections, nerve blocks, etc.) that primary care providers are not trained to provide.^{Reference Gungor62} Specialists may also be more aware of and more comfortable using non-opioid analgesics to manage chronic pain, and may also be more comfortable and familiar with a variety of other kinds of medications — including many antidepressant and neurological drugs, such as anti-seizure medications — that are commonly prescribed off-label for pain.

As noted, specialty practices also generally see patients with more resources. Private insurance may allow specialists to ensure that patients receive more integrated care: One author, for example, only sees patients who simultaneously receive physical therapy and, if needed, psychosocial pain services. Other specialists have addiction specialists, psychiatrists and physical therapists on staff. Whereas Medicaid often either does not cover or inadequately covers many forms of non-pharmacologic, evidence-based pain management (e.g. physical therapy or behavioral modification),^{Reference Meier and Goodnough63} private insurance plans, although flawed^{Reference Thomas and Ornstein64} and recently rightly blamed for incentivizing access to opioids,⁶⁵ may have components that include the kinds of behavioral therapies that recent research has shown quite effective at treating pain without medication. Some private insurers have even been early adopters of mental health counseling in conjunction with opioid treatment, which reduces the risk of abuse.^{Reference Lee66}

Beyond resource constraints, differences in the nature of the patient-doctor relationship across practice types might also affect treatment options. Primary care providers often manage the entirety of a patient's care; they do so over extended periods of time. This may cause them to be vulnerable to the unique pressures involved with using opioids to manage pain. The decision to stop an opioid can trigger a grief-like reaction in patients that prescribers must anticipate and be trained to handle, and patients can exert pressure on doctors to prescribe or continue prescribing opioids. (That pressure may have been particularly acute when patient satisfactory surveys tied to remuneration included pain management questions, as discussed above.) Specialized training prepares specialists for the pressures of the patient-doctor relationship in the context of pain; so does the fact that specialists are typically only consulted for a particular pain related issue, and may therefore be less concerned with maintaining long-term patient relationships.

D. Clinical Support and Referrals

Finally, the primary care physician typically does not have the same levels of support as her specialty counterpart. As noted, specialty clinics often employ non-physician clinicians, including medical assistants and nurses who are specially trained to support a pain management practice (e.g., administering pain-related questionnaires, doing pill counts, collecting urine toxicology tests, etc.). In an outpatient primary care office, the clinical staff rarely has this kind of focus or training, which means that diagnosis and treatment for pain symptoms is exclusively the domain of the primary care provider.

One goal of the Affordable Care Act (ACA) was to encourage a more integrated and coordinated approach across all kinds of care, including the use of the so-called Patient Centered Medical Homes, to emphasize multi-disciplinary team treatment and chronic disease management.⁶⁷ Although small studies suggest that Patient Centered Medical Homes may enhance the quality of chronic pain and addiction in certain primary care settings,^{Reference Elder68} there is little evidence that the ACA did much to support primary care more broadly. Other ACA innovations relating to primary care also have yet to have shown significant results. For example, the ACA's promotion of accountable care organizations (ACOs) — integrated medical practices intended to improve care and reduce costs — has had an uncertain effect.⁶⁹

A. History and Context

The use of opioids to induce euphoria predates modern medicine. As early as 3400 B.C., there is historical evidence that the opium poppy was used by the Sumerians and later the Assyrians and Egyptians to induce a sensation of joy.^{Reference Brownstein70} The recreational use of morphine became broadly problematic in the 1800s, and somewhat ironically, heroin was subsequently introduced by Bayer in the late 1800's as a “non-addictive” alternative to help patients break their morphine habit.⁷¹

Unfortunately, misuse and addiction continue to complicate the use of chronic opioid therapy in pain management. In a systematic review of 38 studies, opioid misuse (defined as using an opioid differently or for different reasons than prescribed) was identified up to 29% of the time, while incident addiction was seen in 8-12% of cases.^{Reference Vowles72}

The recognition that addiction develops in approximately one out of every ten patients prescribed opioids to treat chronic pain has important medical, legal, and societal implications. The recreational use of opioids by a person who has never taken them before may be volitional, but the statistics related to prescription opioid use tell a different story. This observation has helped drive widespread acceptance of the science of addiction medicine; we now understand addiction not as a choice but as a neurochemical brain disorder characterized by drug cravings, compulsive drug use, loss of control, and continued drug use despite evidence of harm.⁷³ Criminal behaviors may accompany addiction as the disorder devolves, but the criminal system alone cannot successfully address addiction.

Instead, one type of treatment — medication-based treatment (frequently known as “Medication Assisted Treatment” or “MAT”) — has emerged as the clinical gold standard when treating opioid-use disorder. MAT, which can be accompanied by behavioral therapy, places doctors at the forefront of treating a patient's opioid addiction. Two main medications are used with medication-based treatment: methadone and buprenorphine. (A third, vivitrol, is also approved for medication-based treatment and may be equally effective for certain subsets of patients.^{Reference Armstrong74}) Developed during World War II, methadone is longer acting than other opioids. As a result, its use is not associated with the same level of opioid cravings and withdrawal symptoms.⁷⁵ Because of these properties, transitioning a patient with opioid-use disorder to methadone can put the symptoms of addiction into remission. Methadone maintenance first emerged as a viable treatment in 1960s but was almost immediately stringently regulated because of rising abuse; by the early 1970s, methadone, which is classified as a schedule II-controlled medication, could only be dispensed for substance use disorder treatment in federally-licensed clinics.^{Reference Rettig and Yarmolinsky76}

Unlike methadone, buprenorphine, which was synthesized in the mid-1960s, is classified as a schedule III-controlled medication, due to the fact that it is a partial opioid antagonist with a better safety profile and lower risk of overdose than full agonist therapy.⁷⁷ In 2000, Congress passed the Drug Addiction Treatment Act of 2000, which opened a pathway for certain physicians (see below) to prescribe buprenorphine from their offices. Over time, this has helped enable physicians—specialists and primary care physicians alike, who may now both prescribe buprenorphine as a part of medication-based treatment—to treat opioid-use disorder. Like pain, primary care providers have shouldered much of the treatment load.

B. Addiction in the Primary Care Setting

There are a number of reasons why patients frequently get their addiction care in the primary care setting. First, the specialty is new, and there are very few board-certified addiction medicine specialists, which limits access to specialists.^{Reference Vestal78} Although some internists and psychiatrists have focused on addiction-related issues for decades, the American Medical Association did not officially recognize Addiction Psychiatry until 1993 and the American Board of Medical Specialties (ABMS) did not officially recognize Addiction Medicine as a unique subspecialty until 2015. As a result, there are far too few specialists to meet treatment needs. Making matters worse, specialists often only accept private insurance or cash. Meanwhile, primary care providers are willing and able to use standard evaluation and management billing codes to provide buprenorphine maintenance treatment.

Second, going to an addiction specialist or drug treatment center can be associated with significant stigma, which can stand as its own barrier to access. Receiving buprenorphine from a primary care doctor allows for a degree of anonymity; a person can sit in the same waiting room with patients suffering from the common cold and no one will know why he is there. Additionally, patients may also appreciate the continuity of care their primary care physician provides, as well as the ability to receive all of their medical care from a single provider.

Nevertheless, primary care has significant disadvantages in this context, as in the pain setting. Time and limited physician support are still issues on this side. In a recent, large study of rural primary care physicians, these two factors were the most commonly cited barriers against incorporating buprenorphine maintenance treatment.^{Reference Andrilla, Coulthard and Larson79}

A failure of training continues to play a role, too. One study suggests that primary care physicians may lack the confidence to use buprenorphine effectively as a part of medication-based treatment.^{Reference Suzuki80} There are also training barriers to some aspects of addiction treatment. Most significantly, as noted, although a physician may prescribe an opioid for pain with no specific training, unless they are board certified in addiction medicine, 8 hours of additional study is required before that same doctor can apply for a waiver to prescribe buprenorphine — the most commonly used drug to treat opioid use disorder in an office-based setting. Nurse practitioners and physician's assistants face an even steeper hurdle, with 24 hours of additional required training.⁸¹ Several studies show that factors like a lack of institutional support or training deter providers from becoming buprenorphine waived.^{Reference Quest82} Even after a waiver is granted, barriers to providing treatment persist. There are legal limits to the number of patients that can be treated by those providers based on how long the waiver has been in effect:^{Reference Sharma83} Initial certification only allows doctors to provide treatment to 30 patients at one time; after a year, they may treat 100 patients; and, as of 2016, after another year, they may treat up to 275 patients.⁸⁴ Nurse practitioners and physician assistants may treat up to 30 patients at once.⁸⁵

Once these initial barriers have been overcome, treatment of addiction in primary care continues to be complicated. Successful delivery models in primary care frequently utilize interdisciplinary and coordinated care services to guide a person through treatment, but the absence of this kind support in the context of addiction may make care more difficult.^{Reference Klasa86}

Moreover, complex privacy laws that target addiction, detailed elsewhere in this volume, can make it difficult to share addiction treatment information with other providers:^{Reference Hudgins, Rose, Fifield and Arnault87} An important federal regulation, 42 C.F.R. Part II, which protects information relating to patients' substance use, prevents providers from disclosing substance-use information without receiving prior patient consent. This can have important real-world consequences, such as a lack of information sharing between surgeons and primary care doctors who should collaborate to manage postoperative pain. Similarly, when a patient is enrolled in a methadone treatment program, the dispensed drug is not reported to the state prescription monitoring program database, thus raising the risk that the primary care provider may inadvertently precipitate a drug-drug interaction or prescribe other controlled substances inappropriately.

C. Care of Opioid Use Disorder Outside of Primary Care

For some patients, the anonymity of treatment for addiction in primary care can be an isolating experience. Some patients may take comfort in knowing that they are not the only person in the doctor's office struggling with addiction. Indeed, peer support and peer counseling are important benefits to participation in a drug treatment program.

Similarly, while a primary care practice may offer flexibility to working patients and those with multiple demands on their time, others may benefit from the generally more rigid schedule and set of rules that a drug treatment center requires. The need to present daily to a methadone treatment center at the beginning of treatment may be a barrier for many, but for some patients this requirement may help them to find structure and gain control over an otherwise chaotic life.

Care in the specialty setting may also be particularly helpful for those patients suffering from both mental illness as well as a substance use disorder. Known as “dual diagnosis,” this syndrome of co-occurring disorders is relatively common and adds significant complexity to the care plan; the 2014 National Survey on Drug Use and Health estimated that 7.9 million Americans suffer from this syndrome.^{Reference Hedden88} An addiction medicine practice or drug treatment center with embedded behavioral health care, enhanced social work support, and availability of an addiction psychiatrist may be best situated to manage these patients.

Finally, an addiction medicine specialist will, like a pain specialist, have more focused and expanded training on the management of opioid use disorder than the typical primary care provider and may therefore be more up to date with the evolving science of addiction care. To become a board-certified addiction medicine specialist, a provider must either complete a dedicated fellowship training program in addiction medicine (after residency training) or must submit documentation of a minimum of 1,920 hours in which they were engaged in the practice of Addiction Medicine at the subspecialty level.⁸⁹ Additionally, addiction specialists' time with patients will be dedicated to the addiction, while, as with pain, a primary care physician must address a variety of patients' medical needs. Although primary care physicians are often able to treat patients adequately,⁹⁰ patients with severe disease, multiple prior failed attempts at sobriety, and those with multiple co-existing addiction disorders may particularly benefit from the expertise of a specialist, who is better positioned to manage co-existing diseases and disorders.

A. Altering Expectations About Pain

Patients may misunderstand the nature of chronic pain and may have unrealistic expectations about their treatment goals or what the health care system can offer them. For example, patients may have the incorrect belief that their care has “failed” and that they need an escalation of their opioid analgesia if they are not fully pain-free. As discussed above, The Joint Commission may have fed this belief by overemphasizing rudimentary and subjective pain-rating scales. But the very nature of chronic pain as a syndrome of complex maladaptive neurologic processes makes it very unlikely that a patient suffering from a chronic pain syndrome will be fully pain-free. Indeed, due to the risks of opioid overdose and analgesic tolerance, maintaining a pain-free state is rarely achievable or safely sustainable with chronic opioid therapy.

Communication is essential. One author, for example, attempts to reset patient expectations by explaining patients' natural capacity to regulate pain, and the potential for opioids to disrupt this process. Indeed, opioids may induce hyperalgesia, a phenomenon in which opioid users experience heightened sensitivity to pain.^{Reference Lee91} Building from informed consent guidelines that encourage providers to discuss the risk of substance use disorder, providers should also explain to patients the extraordinary high risk that chronic opioid users will develop substance use disorder. Another author proposes that providers use the acronym “SMART” — specific, measurable, action-oriented, realistic, and time-bound — to set treatment goals with patients that are optimized around reasonable pain management.

Of course, providers must also commit to taking their patient's pain seriously. As previously described, doctors have historically underdiagnosed and under-treated pain syndromes. As with all of medicine, doctors must conduct an assessment to determine the underlying etiology and personalize an evidence-based care plan that seeks to optimize efficacy while preserving safety. Sometimes opioid analgesia is appropriate; sometimes, only counseling, reassurance, and nonpharmacologic interventions are necessary.

B. Addressing Stigma Around Addiction

The treatment of opioid use disorder often suffers from a similar misalignment of expectations and care delivery, largely due to issues related to stigma. For years, the media and the criminal justice system have demonized substance users, leading to many misconceptions about the illness. For fear of this bias, patients may fail to seek care in the first place,^{Reference Cunningham92} and stigma may affect patients' willingness to stay in treatment.^{Reference Louma93} Similarly, providers may avoid becoming waivered to prescribe buprenorphine for fear that their practice will rapidly devolve into care for an unruly collection of “addicts.” Beyond the walls of the medical office, patients may continue to suffer from the effects of stigma manifested as unemployment and social isolation. In short, stigma has destructive effects on nearly all the pathways through which providers can and must interact with patients.

To a large extent, this stigma has been propagated by the language of addiction which has not kept pace with our medical understanding. Many commonly used terms (e.g., “junkie”, “dirty urine”, “staying clean”, “drug abuser”, etc.) suggest that patients with opioid use disorder are morally culpable for the brain disease that leads to the disorder. The White House and prominent physicians and medical journals have only recently issued recommendations on non-stigmatizing language to describe substance use.^{Reference Botticelli and Koh94}

Similarly, pharmacotherapy for opioid use disorder also suffers from stigmatizing language that negatively impacts care. For example, the use of methadone or buprenorphine to treat opioid addiction is commonly referred to as opioid “replacement” or “substitution” therapy and so reinforces the misconception that such medications are merely substituting one form of addiction for another. Even the more modern term of “medication assisted treatment” (MAT) misrepresents the nature of addiction care in that it suggests pharmacotherapy is simply an adjunct to non-prescription treatment, when the science says otherwise.^{Reference Schwartz95} Considering diabetes as an analog, it would be preposterous to call the use of insulin “medication assisted treatment.”

An awareness of these challenges — and paying attention to words chosen — must be incorporated into professional health education. Beyond that, research shows that various interventions can be helpful to reduce stigma when treating addiction, including group-based acceptance and commitment therapy, motivational interviewing, and communicating positive stories of patients who have successfully managed their substance use disorder.^{Reference Livingston96}

C. Relieving the Burden on — and Empowering—Primary Care

The first step toward relieving the burden on primary care doctors is to ensure that they have sufficient training to provide competent care to their patients. With modest additional education, the treatment of chronic pain syndromes and opioid use disorder can be substantially improved within the primary care setting.^{Reference Tobin97} The treatment of chronic pain, the safe use of opioids, and the management of opioid use disorder all should be required coursework in medical school. Proficiency in these areas also should be considered a core competency of graduate medical education. Further, states should require either a competency assessment or nominal additional training before granting or renewing a practicing provider's controlled substance registration; alternatively, a similar requirement could be adopted by the Drug Enforcement Administration (DEA). Although this creates an additional requirement, in the long run doing so will help ensure that physicians are comfortable prescribing appropriately.

We also recommend that CMS continue to improve reimbursement for the cognitive work of primary care providers to afford them the time needed to address problems like chronic pain and opioid use disorder. One recent large study compared the complexity of patient visits in primary care disciplines with other specialties and found that patient visits in family medicine and internal medicine involved more diagnoses and were more complicated than 12 other specialties.^{Reference Katerndahl, Wood and Jaen98} Primary care providers also often perform several essential patient-related functions for which they are either not reimbursed or inadequately reimbursed, such as coordinating care with other physicians.^{Reference Berenson and Goodson99} For complicated issues like pain and opioid use, primary care providers need more time; the time available to many providers depends on how much and for what services they are reimbursed.

Reimbursement also ideally would be more closely aligned with evidence-based treatments to help incentivize the right type of primary care service. For example, although a wide range of potentially dangerous and addictive opioids are typically covered by insurers when treating chronic pain, it can be challenging to find coverage for other less dangerous alternative medications such as various topical agents (e.g., topical lidocaine) and opioids with a lower risk of overdose (e.g., buprenorphine patch (Butrans)).¹⁰⁰ Similarly, various adjunctive non-pharmacologic interventions may not be covered or may be in short supply, including cognitive behavioral therapy, chiropractic care, acupuncture, etc. Not surprisingly, in a system that does not allot sufficient time or resources, patients and providers often follow the path of least resistance which may result in a less efficacious and more dangerous care plan.

More support for behavioral interventions deserves special mention. Even though medication-based treatment without counseling has been shown to be effective, primary care providers often cite the lack of available or integrated psychosocial support as an obstacle to prescribing buprenorphine.^{Reference Johnson101} Furthermore, mental illness frequently co-exists with a substance use disorder,¹⁰² so coverage for these interventions is particularly important. Readily available behavioral treatment has been shown to improve provider comfort with prescribing buprenorphine, and likely contributes to improved outcomes.¹⁰³ Although fully integrated behavioral therapy as part of a primary care practice that treats addiction may be ideal, other less robust services may also be helpful and should be encouraged. For example, contingency management programs that reward patients for staying abstinent and adhering to therapy can be effective and may be more feasible to implement in a primary care setting where they could be administered by clinic or office staff at limited cost.

We can also lessen the burden on primary care providers through the implementation of opioid safety regulations that offer clarity about physician responsibilities and can easily be incorporated into practice. Prescription drug monitoring programs (PDMPs) — which allow providers (and pharmacists and law enforcement) to monitor other prescriptions patients are receiving — are an important example. There is evidence that requiring prescribers to check the PDMP before prescribing opioids may improve patient outcomes.^{Reference Haffajee, Jena and Weiner104} But the PDMP interface can be cumbersome and difficult to use, which may discourage providers from working with the tool. To address this, beyond simply streamlining the user interface, PDMPs should be implemented with “single sign-on” (SSO) functionality. This approach allows integrated access to the state PDMP database from within a medical practice's electronic medical record system and will improve usability and interoperability. SSO functionality has been recommended to Congress by the Department of Health and Human Services as a proposed best practice standard,¹⁰⁵ but has not been widely implemented.

Other existing regulations should be re-evaluated for feasibility. For example, some states (such as Rhode Island) require primary care providers to consult with a pain medicine specialist or justify not doing so when opioids above a certain dose are prescribed.¹⁰⁶ Although this approach may seem to be a logical safety standard, there is a national shortage of board certified pain specialists to account for the number of patients with chronic pain syndromes.^{Reference Breuer, Pappagallo, Tai and Portenoy107} Additionally, 42 C.F.R. Part II, as discussed in another article in this volume and mentioned above, bears revisiting; its stringent privacy protections, while intended to minimize patient concerns, can hinder communication and prevent providers from fully understanding the needs and risks of their patients. Reflecting some of these concerns, Substance Abuse and Mental Health Services Administration (SAMSHA) recently updated the rule to allow patients to release their information to a broader set of providers.¹⁰⁸ SAMSHA continues to study how the regulation impedes coordinated care in this context.¹⁰⁹

Providers treating opioid use disorder face regulatory barriers as well. As previously noted, to become waivered to prescribe buprenorphine for opioid use disorder, physicians must undergo a mandatory eight-hour training program and register with the Drug Enforcement Administration. Although proper training in the treatment of opioid use disorders is important, such a lengthy training requirement may be a significant disincentive and deterrent for primary care providers interested in treating opioid addiction.^{Reference Walley110} This leads us to question whether practical training on the use of buprenorphine could be accomplished with less of a time-burden, and we recommend that this be studied and piloted in practice. Moreover, as previously described, the training is even more onerous for nurse practitioners and physician assistants. The decision to allow nurse practitioners and physician assistants to prescribe buprenorphine as a part of the 2016 Comprehensive Addiction and Recovery Act marked significant progress in making addiction care more accessible,¹¹¹ but we believe that the non-physician training requirement is so burdensome that it may be impractical for many and similarly recommend that it be abbreviated.

We also question the need for stringent DEA oversight of DATA-waivered providers that impose limits on the number of buprenorphine prescriptions a provider may issue and require periodic and sometimes in-person agency reviews. There are no similar prohibitions on the use of full opioid agonist therapy to treat chronic pain, and treatment for opioid use disorder should be as accessible as possible and not limited in this way. Since partial opioid agonist therapy with a medication like buprenorphine is associated with a lower risk of overdose than full opioid agonist therapy (e.g., methadone, which can be prescribed by primary care doctors for pain without a waiver), the enhanced monitoring required for buprenorphine is somewhat counterintuitive. To be clear, buprenorphine itself has abuse potential and its use must be controlled for safety, but stringent DEA regulations out of proportion to other opioids may make primary care providers feel as if they are under constant surveillance and at risk for prosecution. Indeed, one of the authors of this manuscript endured a “routine” visit from a DEA agent just a month after receiving his waiver and before ever issuing a buprenorphine-naloxone (Suboxone) prescription; experiences like these only serve to reinforce provider reluctance to treat addiction.

D. Leveraging Access to Specialists

Access to pain management and addiction medicine specialists is a scarce resource in some parts of the country. One recent study determined that for every 100,000 patients with a diagnosis of chronic pain, there were only 4 board-certified pain specialists.¹¹² As a result, beyond advocating for training more specialists, it is important also to rationally assess how and when specialists are engaged to drive the principal care of a chronic pain or addiction disorder.

First, accessing specialty care is difficult and often confusing. Even primary care providers can struggle to find and successfully refer patients to specialist practices. For both pain management and addiction care, it is worth reemphasizing that the specialties are not homogenous. Some pain specialists employ multimodal and interdisciplinary treatment models that use a combination of behaviorally-based, pharmacologic, and procedurally-based interventions, while others only offer a single treatment modality.¹¹³ For a primary care provider to effectively collaborate with a pain specialist in her area, she needs to understand what services are available from that provider and what the expectations and limitations of that specialty practice are. Similarly, some addiction treatment programs offer only non-pharmacologic supportive counseling, while others limit their services to treatment of specific substance use disorders or use only a limited number of drug treatment options. Patients suffering from a dual-diagnosis of a psychiatric disorder and a substance use disorder may be particularly limited in what treatment options are available. Furthermore, as discussed above, both pain management and addiction specialty practices may limit the insurers they work with, which can be particularly problematic for underinsured or uninsured patients.

Although there is no easy solution to these challenges, we recommend that there be an organized approach to catalog available pain management and addiction services in each community. Such an effort could be spearheaded by a regional or state Department of Public Health and or medical board could lead to a more effective and efficient use of resources simply by promoting a better understanding of what is available and where to turn for help.

Given the public health emergency, we also urge government-developed incentives to increase the number of trainees in both addiction and pain management fellowships and get them in to the affected communities. There are precedents for such programs, such as the National Health Service Corps, which provides incentives for providers to practice in a designated healthcare shortage areas for a limited period of time. Lawsuits related to the crisis dominate the headlines and are detailed elsewhere in this volume;^{Reference Gluck, Hall and Curfman114} rewards from those lawsuits likewise might be devoted to increasing the supply of professionals needed to treat this crisis.

Secondly, it is important to reflect upon the difference between requesting a specialist consult or placing a referral to a specialist. In the consult model, the primary care provider requests guidance from a specialist to help clarify a diagnosis or implement a treatment plan, while a referral asks the specialist to assume principal care for the issue at hand.¹¹⁵ As primary care providers become more informed about these diagnoses and management of these problems through changes in medical education and regulation, the need for specialty referrals may lessen. A more consultative approach to specialty care would both empower the primary care provider to deliver evidence-based treatment with sufficient support while simultaneously unburdening the specialist so that they are more available to offer principal care to truly complex patients.

Creating a network of consultative support for primary care providers that does not ask a specialist to simply “take-over” the care of the patient can be helpful in the management of pain and addiction. For example, establishing such linkages has proven essential in the context of prescribing buprenorphine.¹¹⁶ Certain health care models, like Vermont's “hub and spoke,”¹¹⁷ have responded to this need by creating a network of primary care providers who operate in connection with trained specialists. Rather than asking primary care providers to find available addiction specialists whom they can refer to, Vermont's model establishes a formal network in which specialists are linked to and available to assist primary care providers via consultation or referral, depending on need. Spokes are also supported by treatment teams — a registered nurse and a behavioral health provider — that travel to assist with patient treatment, coordinate referrals, and perform tasks like arranging for urine drug testing or reimbursement.^{Reference Brooklyn and Sigmon118} Vermont pays for hub and spoke services through Medicaid, with funding being distributed from the hub to the spokes. So far, the Hub and Spoke model has reduced the costs of claims,¹¹⁹ and early evidence suggests that these treatment models redouble capacity and ease the strain on primary care providers.¹²⁰ Similar support networks can be effective when managing pain as well; implemented at the state level, a hub and spoke model for pain treatment would provide primary care providers with enhanced support, exposure to best practices, and make specialty referrals more accessible when needed. Evidence suggests that this can also be accomplished remotely; studies show that remote consultative support via a tele-medicine specialist network can effectively assist primary care providers managing difficult cases of chronic pain.^{Reference Doorenbos121}

Less formal support networks should be leveraged as well. For example, the Providers' Clinical Support System for Opioid Therapies (PCSS-O) and the Providers' Clinic Support System for Medication Assisted Therapy (PCSS-MAT) are federally-supported and free online peer-to-peer mentoring programs that offer enhanced education and mentoring to providers struggling to treat complex cases of pain and addiction.¹²² These resources are supported by a large consortium of member organizations and stakeholders representing both specialty and primary care, and they provide continuing educational opportunities as well as clinical support.

Finally, it is also important to recognize that each patient will have a unique experience with pain and addiction, and that specialty care needs will vary among patients and for the same patient over time. For these reasons, we argue against mandating referrals to a specialist simply because an opioid is prescribed for a certain length of time or above a certain dose, as is done in the aforementioned Rhode Island model. Instead, we call upon professional organizations and regulatory bodies to promote collaboration between primary care providers and specialists in a way that is based on evidence-based best practices and the specific needs of an individual patient and provider.