A Review of Chronic Noncancer Pain: Epidemiology, Assessment, Treatment, and Future Needs

The 2011 Institute of Medicine report (3) estimates that an astounding 100 million American adults are affected by chronic pain, exceeding the numbers of those affected by diabetes, cancer, and heart disease combined. These estimates of chronic pain do not include adults affected by acute pain, children with either acute or chronic pain, or adults living in long-term care facilities, serving in the military, or serving in prison (5). Pain affects millions of Americans throughout their lifetime, increases disability, consumes resources in the health care system, and results in a significant economic burden for the entire nation. The staggering socioeconomic burden of pain is thought to exceed one-half trillion dollars per year (3).

One of the current challenges of chronic pain is that it is difficult to obtain clear numbers on the incidence and prevalence of pain, because it is often treated as a symptom of another medical condition. Thus, research indirectly targets conditions associated with pain such as arthritis. However, it is unclear whether pain is the overriding symptom prompting a patient to seek medical care. Similarly, the reason for physician visits may be associated with a specific diagnosis, such as chronic pancreatitis, but it is unclear which visits may have been prompted primarily by pain as opposed to other symptoms. The Institute of Medicine report (3) cites a number of reasons that contribute to the paucity of data on pain prevalence in the United States, including lack of standardized research methods to obtain population-level data on pain, exclusion of key groups from national population-based health surveys (e.g., children, people serving in the military, adults living in nursing homes or long-term care facilities, and individuals serving in correctional facilities), inclusion of only a few painful diseases in existing research studies, and failure to distinguish between acute and chronic pain.

Tsang et al. (6) estimated that the age-standardized prevalence of chronic pain conditions (headache, back or neck pain, arthritis or joint pain, and other chronic pain) is 37.3% averaged across all developed countries. In the United States, the age-standardized prevalence of chronic pain conditions is 43% (6). In both developed and undeveloped countries, a higher prevalence of chronic pain was found among females and with increasing age. In addition, depression and anxiety disorders were associated with chronic pain in both developed and undeveloped countries (6). Overall, according to the Institute of Medicine report (3), musculoskeletal pain (joint pain and back pain) is the most common type of chronic pain, but most people with chronic pain have multiple areas of pain. The annual prevalence of chronic low-back pain ranges from 15% to 45% (7). According to the 2002 National Health Interview Survey (8), low-back pain was the most frequently reported pain, followed by migraines or severe headaches, neck pain, and face or jaw pain. The survey showed that 26.4% of respondents reported low-back pain in the past 3 months, corresponding to an estimate of >54 million adults in the United States experiencing back pain. The survey results also revealed that females were more likely to report back pain. In addition, American Indians and Alaska Natives reported the highest prevalence of back pain, whereas Asian Americans reported the lowest prevalence (8). From a socioeconomic standpoint, the prevalence of back pain decreased with higher levels of education, and higher levels of family income (8). Similarly, certain groups have been identified as having an increased risk for chronic pain, including those with English as a second language, particular racial/ethnic groups (African Americans, Hispanics, American Indians, and Alaska Natives), people with lower levels of income and education, females, children, military veterans, people with cognitive impairments, surgical patients, cancer patients, and those at the end of life (3).

The prevalence of chronic pain is likely to continue increasing because of a number of factors, including aging of the U.S. population and an increasing prevalence of geriatric illnesses associated with chronic pain, the rising prevalence of obesity and associated chronic diseases (diabetic peripheral neuropathy) and orthopedic issues (joint pain) with pain, a greater proportion of people surviving catastrophic injuries with subsequent chronic pain, acute and chronic pain resulting from surgery, greater public awareness of chronic pain syndromes, and increased health care access (3).

Chronic pain has a myriad of deleterious effects on both the individual and society as a whole. One notable effect is impaired performance of activities of daily living. In patients with low-back pain, increasing self-reported pain severity has been associated with increasing pain interference with function as well as impairment while working (9). Work productivity is a significant issue, as noted by the Institute of Medicine report (3). People with severe pain missed an average of 5.0–5.9 more days of work per year compared with people with no pain. The resulting cost of workdays missed ($11.6 billion to $12.7 billion), work hours lost ($95.2 billion to $96.5 billion), and wages lost ($190.6 billion to $226.3 billion) to pain is staggering (3).

Health-related quality of life assesses the quality of life in the context of both clinical medicine and research, and it encompasses physical, emotional, and cognitive functioning as well as the ability to participate in meaningful activities within the family, workplace, and community as a whole (10). In the context of chronic pain, health-related quality of life measures patients’ perspectives on their pain experience and the effect it has on their lives (10). A number of chronic pain conditions negatively affect quality of life, including neuropathic pain (11), osteoporosis (12), chronic pelvic pain (13), and neck pain (14).

Chronic pain also has a negative effect on an individual’s psychological well-being. At an extreme, the Institute of Medicine report (3) notes that the risk for suicide is doubled among people with chronic pain compared with control groups, with a lifetime prevalence of suicide attempts ranging from 5% to 14% (3). Psychiatric comorbidities are common among patients with chronic pain. For example, patients suffering from chronic daily headaches present with comorbid depression and anxiety disorders, further increasing the risk for suicide (15). Furthermore, these psychiatric comorbidities are likely to affect the prognosis and treatment of chronic pain conditions.

Overall, chronic pain has significant effects on the individual, which are related to physical functioning, quality of life, and psychological well-being. In addition, the annual cost of pain in the United States is estimated at $560 billion to $635 billion when the cost of health care and lost productivity to pain is combined, resulting in a significant economic burden on society as a whole (3).

Although many instances of pain start after an injury or trauma, the natural history of pain has not been well characterized. Surgery represents a useful real-life context, in which patients undergo reproducible injury, resulting pain, and expected recovery. The vast majority of the 45 million Americans undergoing surgery every year experience acute pain (16), and it is postulated that 10% of these patients will not experience resolution of postoperative pain, leading to the development of chronic postsurgical pain (17). Understanding this transition is critical for a number of reasons. Specifically, future treatments could be targeted to promote pain resolution in patients undergoing surgery. Most importantly, there may be important parallels between the mechanisms promoting continued postsurgical pain and the indefinite persistence of pain leading to chronic pain.

Current research has emphasized the assessment and reduction of perioperative pain intensity and pain thresholds, rather than characterizing this temporal aspect of acute pain (pain duration). Typically, postoperative pain is assessed through examination of intermittent time points (1, 3, and 6 months postoperatively) after surgery, with assessment of pain severity at that time. This methodology does not necessarily capture waxing-and-waning pain symptoms across time, and it may introduce certain incorrect assumptions regarding postoperative pain recovery. For example, a patient may report that he or she is no longer experiencing pain at the surgical site on the day of assessment, and it may be assumed that this patient has recovered after surgery. However, the patient may undergo physical therapy the next day and may experience resumption of pain and symptoms. Capturing this scenario is only possible with more longitudinal and close follow-up. By examining intermittent time points after surgery, reported predictors of postsurgical pain include anxiety (18–20), state anxiety (21), catastrophizing pain (19, 22), preoperative pain (both remote and at the surgical site) (20, 21, 23), severity of acute postoperative pain (21, 24, 25), preoperative opioid use (21), and depression (26, 27). These predictors will be valuable targets for interventions to reduce postoperative pain intensity in future studies.

Prescription opioids are one of the primary modalities for postoperative pain management (28–32). Although the amount of opioids prescribed for any particular surgery may be fairly uniform, patients exhibit wide variability in opioid needs after similar surgeries (33, 34). Furthermore, continued opioid use after surgery has the potential to lead to opioid misuse, abuse, and even addiction. In the context of increasing death rates from prescription opioid overdose, prescription opioid sales, and substance abuse treatment admissions (35–37), patients undergoing surgery represent a particularly vulnerable population experiencing injury and obligatory prescription opioid use in the postoperative period. We previously described a novel approach to characterizing time to postoperative prescription opioid cessation and time to pain resolution through a Cox proportional hazards regression model (38, 39). In our prospective study, patient follow-up spanned 4–859 days after surgery. Prescription opioid use and pain cessation were longitudinally assessed through daily phone calls for the first 3 months, weekly phone calls up to 6 months after surgery, and monthly calls thereafter. Among this surgical cohort, pain duration, pain severity at the time of opioid cessation, preoperative baseline pain severity at the surgical site, preoperative baseline pain severity at locations other than the surgical site, preoperative illicit opioid use, and any preoperative opioid use all did not predict delayed opioid cessation (38). By contrast, elevated preoperative depressive symptoms, preoperative self-perceived susceptibility to addiction, and legitimate preoperative opioid use each predicted delayed cessation of prescription opioid use after surgery (38). In addition, a follow-up factor analysis of the preoperative depressive symptoms in this surgical cohort showed that self-loathing symptoms, rather than more motivational and emotional symptoms of depression, independently predicted a significant decrease in opioid cessation after surgery (39). Future work is necessary to replicate these findings examining other surgical populations.

Although chronic pain is often thought of as a lifelong condition, its natural history is not well characterized. Through examination of postsurgical pain, important mechanisms underlying pain persistence may be revealed. Through further research, measures to prevent continued pain after surgery and the development of chronic pain could be developed to target these underlying mechanisms. Current therapeutic modalities that may be beneficial for reducing persistent pain after surgery include the use of regional anesthetic techniques, perioperative ketamine, gabapentanoids (gabapentin, pregabalin), intravenous lidocaine, and management of patient expectations (40, 41). However, research will be necessary to also determine whether these modalities are able to promote pain resolution and decrease pain duration after surgery.

Prescription opioids currently remain one of the most commonly prescribed medications to treat various chronic pain conditions. During the past decade, there has been a dramatic increase in the prescription of opioids for the management of chronic pain conditions. This transition toward increased opioid prescribing in the 1990s was sparked by interpretations of the safety and efficacy of opioids in treating chronic noncancer pain conditions, concerns regarding the long-term safety of alternate analgesics (nonsteroidal anti-inflammatory drugs), promotion of opioids to primary care physicians, and the endorsement by national organizations to encourage liberal opioid use for chronic pain conditions (42).

The long-term effectiveness and safety of opioids in the treatment of patients with chronic noncancer pain remain to be determined. Variable conclusions have been presented by both systematic reviews and recent randomized controlled trials. Despite this context of uncertainty and the lack of a clear consensus of the utility and efficacy of opioids in the treatment of chronic noncancer pain, opioid prescribing for noncancer pain has dramatically increased during the past decade (43, 44). Systematic reviews of opioids for chronic noncancer pain examining randomized controlled trials have shown at least moderate improvements in pain, function, and disability for opioids compared with placebo (45–51). However, a major limitation of systematic reviews examining only randomized controlled trials is the shorter duration of studies over weeks, combined with a high attrition rate (46, 47). These issues preclude conclusions regarding the safety and efficacy of long-term prescription opioid therapy to treat chronic noncancer pain conditions. In comparison, Noble et al. (52) included any prospective studies, such as randomized controlled trials and case series, in their systematic review examining the efficacy and safety of long-term opioids for chronic noncancer pain. The most prevalent specified painful conditions in this systematic review were back pain, osteoarthritis, neuropathic pain, myofascial pain, osteoporotic vertebral fractures, and trigeminal neuralgia. The most common adverse effects included gastrointestinal side effects (constipation, nausea, or dyspepsia), headache, fatigue, and urinary issues (hesitancy or retention) (52). In addition, the authors estimated a 0.14%–0.27% rate of addiction among the included studies. Although the authors stated that patients taking oral opioids for 6 months demonstrated clinically meaningful pain relief, significant heterogeneity of the study results precluded any firm conclusions. Similar to the previously mentioned systematic reviews, participant dropout as a result of adverse effects and insufficient pain relief also introduced bias regarding the efficacy and safety of long-term opioid therapy for chronic noncancer pain (52). Beyond systematic reviews, an examination of individual studies provides contradictory evidence pointing to the association between long-term opioid use and minimal improvements in pain, function, and quality of life (53, 54). Baseline opioid use was associated with higher 6-month disability in a prospective cohort study of patients with low-back pain in primary care (55). Overall, insufficient evidence currently exists to determine the safety and efficacy of long-term opioid therapy for chronic pain. It should be noted that this conclusion could be made of most pharmacological treatments for chronic pain and is not restricted to just opioids. Ultimately, better data regarding outcomes beyond 1 year are necessary to provide a complete narrative on the efficacy and safety of prescription opioids (and all therapies) in the treatment of chronic pain conditions.

Prescription opioid–related overdose deaths in the United States, prescription opioid sales, and substance abuse treatment admissions have climbed in parallel during the past decade (35–37). Medical and nonmedical use of prescription opioids has created an epidemic of prescription opioid overdose, emergency department admissions, addiction, diversion, and death. The economic cost of nonmedical prescription opioid use in the United States is >$50 billion annually (37). In fact, nonmedical prescription opioid use costs insurance companies $72.5 billion annually in health care costs (36).

Prescription opioids were involved in 16,917 overdose deaths in 2011 (56), and 31% of these deaths also involved benzodiazepine sedatives (56). Overall, the death rate from prescription opioids (4.2 per 100,000 population) was four times the rate for heroin and cocaine combined in 2010 (57). The trend among women is particularly concerning. Deaths from prescription painkiller overdoses have increased 415% in women since 1999 compared with 265% in men, perhaps because of increased opioid prescribing to women (57). However, men are still more likely to die of prescription painkiller overdoses (57).

Despite the intense scrutiny of prescription opioids, it is important to recognize that many unintentional or accidental polysubstance deaths in the United States involve a combination of opioids and benzodiazepines (58–60). In West Virginia, 41% of unintentional pharmaceutical overdose fatalities involving opioid analgesics also included benzodiazepines (61). Similarly, in a study of unintentional overdose deaths in New York City from 2005 to 2010, 53% of deaths involving an opioid analgesic also involved a benzodiazepine (62). Research continues to pinpoint the markedly increased risk for fatal prescription opioid intoxications in decedents concurrently using one or more benzodiazepines (63). These findings are mirrored outside of the United States. In Norway, a 1-year observational study reported that 76% of fatal poisonings involved opioids, and 76% of fatal poisonings involved benzodiazepines (64). In the midst of mounting evidence of the combined role of prescription opioids and benzodiazepines in unintentional overdose deaths, many professional societies have released guidelines to encourage safe opioid prescribing in the context of preexisting or concurrent benzodiazepine use (65–68).

The treatment of comorbid chronic pain and substance use disorders is particularly challenging. The incidence of dependence ranges from 0% to 24%, and the prevalence of dependence ranges from 0% to 31% in patients receiving prescription opioids for pain relief (69). In a randomized trial of a more conservative versus a more liberal opioid-prescribing strategy of 135 patients in a Veterans Affairs hospital pain clinic, 27% of patients were discharged because of opioid misuse or noncompliance issues, including concurrent alcohol/illicit substance abuse, medication misuse, or noncompliance with clinic procedures (e.g., refusing to complete urine toxicology testing). There was no difference in the rate of opioid misuse or noncompliance between groups (70). Similarly, in a study of 705 patients receiving four or more opioid prescriptions in the preceding 12 months for noncancer pain, rates of lifetime and current opioid dependence concordant with DSM-IV criteria were reported as 35.5% and 26%, respectively (71). In addition, current opioid dependence was associated with age <65 years, opioid abuse history, higher severity of opioid dependence, major depression, and use of psychotropic medications (71). Clinicians need to be aware of both preexisting substance use disorders and the development of substance use disorders while treating chronic pain patients with prescription opioids.

In the context of treating chronic pain, potential risks associated with persistent opioid use are numerous and include increased risks for medication overdose, opioid abuse, and overall morbidity (72). The adverse effects of prescription opioids are well described. The presence of tolerance and physical dependence can occur even at prescribed doses (73–78). Opioids are associated with immunosuppression and opioid-induced endocrinopathy (sexual dysfunction, depression, or decreased energy) (79–85). Opioids are implicated in opioid-induced hyperalgesia or increased pain sensitivity despite increasing doses of opioids. Despite opioid-induced hyperalgesia being demonstrated in animal studies and acute perioperative studies, its existence and meaning in chronic pain populations remain unclear (86–88). For instance, in the largest clinical trial to date, testing for the presence of opioid-induced hyperalgesia was only able to demonstrate tolerance without demonstrable opioid-induced hyperalgesia (89).

More concerning is the association between long-term prescription opioid use, affective disorders, and substance abuse. Increased rates of substance abuse and depression exist in long-term prescription opioid users compared with nonusers with chronic pain, and pain intensity does not predict treatment with opioids versus nonopioid analgesics (90). Other research shows that depression and anxiety contribute to substance use disorders among long-term opioid users (91). In addition, opioid addicts >45 years old have higher rates of severe pain and psychiatric illness (92). These results point to the importance of acknowledging psychological vulnerability or comorbidity in the development of substance abuse disorders in patients with chronic pain treated with long-term prescription opioids.

Clearly, long-term prescription opioid therapy for chronic pain is not a panacea, because one-third of patients with chronic pain do not tolerate prescription opioids or find long-term opioid therapy to be ineffective (93). Before therapy is initiated, it is essential for the clinician to weigh the risks and benefits of treatment. During ongoing treatment of chronic pain with prescription opioids, and in the context of tolerance, difficulties in assessment will arise as patients continue to dose escalate either in the context of increased pain or opioid misuse (93). With continued treatment, providers will need to monitor for iatrogenic drug dependence, opioid misuse, and opioid diversion. Patients with a history of addiction are at risk for relapse when opioids are prescribed, and patients with active substance use disorders are likely to be noncompliant and difficult to engage in nonopioid strategies for pain control. Given the chronology of chronic pain conditions and the relatively small number of patients who experience complete resolution of their painful conditions, prescription opioids need to be safe and sustainable for long-term use.

The biopsychosocial model of pain and disability is currently the most widely accepted approach to understanding and treating chronic pain conditions (94). In this model, pain and disability have complex interactions among physiological, psychological, and social factors, which perpetuate and worsen a patient’s clinical condition and result in chronic, complex pain syndromes (95). The biopsychosocial model encompasses both disease and illness. Disease refers to an objective biological event, which disrupts body structures or organ systems. By contrast, illness describes the subjective experience of the disease through interaction with biological, psychological, and social factors (94). Although emotion is the immediate reaction to nociception, cognition attaches meaning to the emotional experience of pain; these pain-related cognitions can augment the experience of pain by triggering additional emotional and behavioral reactions (95). Thus, pain is an individualized and unique experience, and psychosocial factors often modulate symptoms and disability (94). The biopsychosocial model acknowledges how pain and disability trigger psychosocial responses and cause emotional distress (anxiety, depression, or fear) (96). This model contrasts greatly with a more traditional biomedical approach, which assumes that pain has a specific physical pathology. Treatment through this model often centers on cutting or blocking pain through pharmacological or surgical methods (96). Unfortunately, many chronic pain syndromes do not have clearly established mechanisms, and pathophysiology has not been clearly explained. More importantly, no “cure” has been established for many chronic pain diagnoses. Using the traditional biomedical approach to treating patients with chronic pain is likely to increase health care costs and lead to inappropriate expectations and frustrations for patients expecting complete recovery.

The biopsychosocial model of pain has diversified the treatment options for patients with chronic pain conditions and has allowed treatment to move forward beyond solely targeted biological approaches. Individualized treatment allows clinicians to acknowledge the complex physiological, psychological, and social dynamics involved in the perpetuation of pain. Clinicians should be mindful of a patient’s emotional concerns, centering on the implications regarding prognosis, function, and ability to return to work throughout treatment. Comprehensive interdisciplinary pain programs address the dimensions of the biopsychosocial model of pain, and they have demonstrated both short-term and long-term efficacy for a number of important pain and psychosocial outcomes.

Pain may initially be viewed as a symptom of an underlying disease. However, with persistent pain, an unclear etiology, or lack of improvement with initial therapy, pain becomes a disease itself that necessitates a comprehensive assessment. A multidimensional assessment is particularly important for patients with chronic pain because multiple aspects of life are affected; various stakeholders have different interests in patient outcomes (family versus third-party payers), and relevant outcomes (pain, quality of life, and health care utilization) do not necessarily improve simultaneously (96).

In the context of recalcitrant chronic pain, it is especially important to proceed with an accurate, detailed history and physical examination before establishing an interdisciplinary, comprehensive treatment plan. The description of pain should span onset, location, radiation, intensity, and character (sharp, stabbing, dull, or aching). Although patients may present with established diagnoses, the pain specialist should review all history, physical examinations, and diagnostic tests for additional pathology. Through this process, referrals to other specialists may be necessary (e.g., referral to a gastroenterologist for ongoing abdominal pain). Functional limitations related to activities of daily living, work, socializing, sex, and self-care should be quantified (97). Responses to previous treatments should be documented, including medication trials (e.g., opioids, nonsteroidal anti-inflammatory drugs, antidepressants, or anticonvulsants), surgical procedures, interventions, physical/occupational therapy, psychological interventions, and complementary and alternative medicine modalities.

Substance abuse history should span nicotine, alcohol, marijuana, prescription medications (e.g., opioids or benzodiazepines), and other illicit drugs. Patients receiving prescription opioids should describe analgesia, activity level, aberrant behaviors, and adverse effects associated with opioid therapy. Patients should quantify the degree of analgesia that they receive from their opioids as well as the amount of increased function that they associate with opioids. Patients should also be assessed for self-reported aberrant behaviors such as running out of medications early and opioid-associated adverse effects (e.g., constipation, nausea, or vomiting).

Given the high likelihood of comorbid psychiatric diagnoses, patients should be screened for a number of mood disorders, including anxiety, depression, and posttraumatic stress disorder. Inquiry regarding past suicidal ideations or attempts and previous psychiatric hospitalizations is also important. With any comorbid psychiatric diagnoses or symptoms, referral to a pain psychologist is further beneficial to aid in the comprehensive interdisciplinary evaluation of the patient with chronic pain. Assessment of social history relating to marital status, employment, income, and living arrangements reflects a patient’s quality of life and support system. Litigation, especially related to a chronic pain condition, can prolong disability. In addition, trauma throughout life starting in childhood is important to assess, given its connection to increased somatic symptoms such as pain (97).

A comprehensive physical examination is a necessary complement to the detailed history in the assessment of a patient with chronic pain. Goals of examination center on developing patient trust, quantifying physical functioning in the context of pain, and accurately identifying potential pain generators (2). A comprehensive neurologic examination is essential to characterize sensation, motor function, reflexes, and coordination and is combined with a focal regional examination in order to accurately diagnose or rule out common and rare pain conditions (2). A thorough physical examination can often prevent the pursuit of costly imaging studies or more invasive testing.

Choosing Wisely is a campaign specifically encouraging physicians and patients to have a discussion regarding unnecessary treatments, tests, or procedures. The American Board of Internal Medicine Foundation launched the campaign in 2012, and medical specialty societies have created lists of tests, treatments, or procedures that have demonstrated strong scientific evidence of overuse as well as the potential for harm or increased cost (98). All of these lists are available at www.choosingwisely.org. As part of the campaign, the American Society of Anesthesiologists Committee on Pain Medicine set forth a list of five items that physicians and patients should question. Specifically, the recommendations state “Avoid imaging studies (MRI, CT, or X-rays) for acute low-back pain without specific indications” (98). The committee recommends avoiding imaging for low-back pain in the first 6 weeks in the absence of specific red flags (e.g., cancer with potential metastasis, known aortic aneurysm, or a progressive neurological deficit). The committee points out that imaging may reveal incidental findings that increase the patient’s risk of undergoing an unnecessary surgical procedure. In addition, the recommendations state “Don’t prescribe opioid analgesics as long-term therapy to treat chronic non-cancer pain until the risks are considered and discussed with the patient” (98). The committee emphasizes discussing the risks associated with opioid therapy with patients (including the risk for addiction), reviewing and signing a written agreement between the physician and the patient regarding responsibilities and consequences of noncompliance, and using caution in the coprescription of opioids and benzodiazepines.

Proper patient selection is paramount, with ample consideration of the risks and benefits associated with initiating opioid therapy for chronic noncancer pain. Universal risk assessment and stratification are necessary, given that the estimates of aberrant drug-related behaviors, drug misuse, and abuse in patients with chronic noncancer pain are as high as 50% (99). As outlined above, an accurate and detailed history is essential to identify risk factors of opioid misuse, abuse, and addiction. A personal or family history of drug or alcohol abuse strongly predicts aberrant opioid-related behaviors, opioid misuse, and abuse after initiation of therapy (99). Additional risk factors for opioid use should be assessed, including euphoria associated with use, other substance abuse diagnoses, mental health disorders, male sex, younger age, preexisting substance use disorders (42, 92, 100), and specific characteristics of opioid prescriptions (e.g., longer supply, higher average doses, and use of schedule II opioids) (100). Patients with chronic pain and active substance use disorders present a particular challenge. Pain medicine specialists should work closely with an addictionologist, who can directly assess and treat the comorbid substance use disorder, during ongoing interdisciplinary care. These patients will require optimization of nonopioid and behavioral therapy as well as close follow-up for management of their chronic pain condition and substance use disorder (92).

Although screening tools may be helpful for risk stratification, more data are necessary to understand how routine use of these tools predicts aberrant behaviors and ultimately affects clinical outcomes. The Diagnosis, Intractability, Risk, and Efficacy instrument is clinician administered and assesses potential efficacy and harm associated with opioid therapy. This tool identifies appropriate candidates for long-term opioid therapy based on scoring of seven factors, with risk divided into psychological, chemical health, reliability, and social support categories (99). In addition to the Diagnosis, Intractability, Risk, and Efficacy instrument, a number of self-report screening tools may be useful. The Opioid Risk Tool is a 10-item scale that is used to stratify patients at high, moderate, and low risk for development of aberrant behaviors when receiving long-term opioid therapy (42). Drawbacks of the Opioid Risk Tool include limited face validity. In addition, this tool has not yet been cross-validated in a separate population (42). The Screener and Opioid Assessment for Patients With Pain–Revised is a self-reported measure of the risk potential for aberrant medication behavior among people with chronic pain (101). This tool has been validated as a screening tool in this population and is intended for administration to patients being considered for long-term opioid therapy (101). Among a cross-validation sample for this assessment, 33.5% of patients had a score of ≥18, which was reported as a cutoff score with 80% sensitivity and 52% specificity in identifying at-risk patients. In other words, at this cutoff score, the Screener and Opioid Assessment for Patients with Pain–Revised tool has a higher chance of identifying false positive results because of this lower specificity. Nonetheless, this tool has demonstrated superiority over other self-report measures, including the Pain Medication Questionnaire and Opioid Risk Tool, to predict discharge from opioid treatment and the presence of aberrant drug-related behavior (102). Unlike these screening tools, which identify the risk potential for aberrant medication behaviors, the Current Opioid Misuse Measure assesses ongoing aberrant medication-related behaviors in patients prescribed opioids for chronic pain conditions (103). Use of the Current Opioid Misuse Measure may assist providers in determining the level of continued monitoring appropriate for any given patient receiving opioids to treat chronic pain (104). In addition, the Current Opioid Misuse Measure shows promise as a reliable self-report tool to identify prescription drug use disorder among primary care patients with chronic pain who have been prescribed opioids (105). Despite the potential limitations of these and other screening tools for predicting opioid misuse, abuse, and addiction before initiating patients on long-term opioid therapy or identifying ongoing opioid misuse and abuse, their use has been advocated by a number of clinical guidelines and will continue to play a role in the initial assessment of patients with chronic noncancer pain (106).

Evidence-based medicine is the standard for supporting clinical decision making. However, given the paucity of prospective, placebo-controlled randomized trials in pain medicine owing to the large amount of resources required to conduct such research, there is an urgent need to accurately and consistently measure relevant patient outcomes in order to characterize the most efficacious treatments. There is a need to standardize the assessment and reporting of outcomes to allow for comparison across studies and different patient populations as well as to facilitate consolidation of results in systematic reviews and meta-analyses.

In addition to prospective, placebo-controlled randomized trials, which can be difficult to generalize because of participant homogeneity and requirement of a large amount of resources (because of sample size), systematic practice-based evidence may provide more useful data in the form of prospective, observational, cohort studies (93). Standardized data capture can be included as part of ongoing routine clinical pain management from both patients and providers. In addition, combining computerized systems for patient self-reported outcomes, such as the Patient-Reported Outcomes Measurement Information System (PROMIS) from the National Institutes of Health (NIH), has the potential with electronic medical record data to rapidly identify subsets of patients who will respond favorably to specific treatments (93).

NIH developed PROMIS to create a standardized measurement system to improve interpretation and comparison of outcomes across populations and studies. The item banks of PROMIS assess multiple domains across physical, social, mental, and global health. The Pain Registry includes item banks for pain intensity, pain interference, pain behavior, anxiety, depression, anger, fatigue, sleep disturbance, and sleep interference.

The Stanford-NIH Collaborative Health Outcomes Information Registry (CHOIR; CHOIR.stanford.edu) was developed in response to the Institute of Medicine report Relieving Pain in America (3), which emphasized the need to improve the collection and reporting of data on pain. This open-source, open-standard, free, secure, electronic learning health care system is designed to capture detailed, longitudinal patient-reported outcomes data on physical, psychological, and social health. PROMIS measures have been integrated into CHOIR to maximize efficiency in data capture through cloud-based computation, with adaptive algorithms for items with wide dynamic range and computerized adaptive testing. This testing has been shown to reduce patient response burden by as much as 70% and will facilitate continued patient participation. Ultimately, CHOIR will allow for low-cost, large, prospective, observational studies on thousands of patients in a “real-world” clinic setting. CHOIR has the potential to address many fundamental questions regarding pain treatment and efficacy, and CHOIR will allow for further characterization of optimal patients for specific therapies. Other applications of CHOIR include reverse translation to identify new analgesic drug targets and to facilitate enriched patient selection for prospective, randomized controlled trials (93).

In addition to the broad research utility of CHOIR, the system provides computer-assisted documentation, which has proven indispensable and invaluable in delivering comprehensive, targeted interdisciplinary pain treatment. The platform is designed to be customizable to different settings (inpatient and ambulatory), providers, and disease conditions. Use of CHOIR will facilitate real-time comparison of patient data to clinic, national, and disease-specific norms. CHOIR has demonstrated efficiency, low cost, and minimal burden to staff, providers, and patients, while augmenting clinical care. CHOIR provides rapid real-time, longitudinal feedback to clinicians regarding standardized quantitative outcomes to guide decision making regarding various treatments (93).

Beyond learning health care systems, future efforts must focus on improved data collection in population-level databases. The National Pain Strategy Task Force acknowledges the need to improve both state- and national-level data to monitor changes in the incidence and prevalence of chronic pain, document treatment rates of pain and barriers to pain treatment, assess the health and societal consequences of pain, and evaluate the effect of public policy, payment, and treatment changes. Short-term priorities, as defined by the Population Research Working Group of the National Pain Strategy Task Force, include delineating standardized questions and indicators of chronic pain and identifying essential electronic medical record components. These standardized questions will be vital to assess the prevalence and effect of chronic pain in the general population and to evaluate the efficacy of public health interventions. Additional standardized questions will be imperative for clinical care to screen patients for pain, to assess patients with pain, and to guide clinical treatment decisions. Components of the electronic medical record need to be streamlined to identify patients with chronic pain and to assess the quality, safety, and efficacy of treatments.

Personalized strategies need to be researched and developed to predict who will experience chronic pain after injury and who will develop a substance use disorder during long-term opioid therapy. Moreover, strategies are needed to designate responders and nonresponders to various pain treatments, including prescription opioid therapy. Research has demonstrated the contribution of genetic factors in regard to opioid dependence, pain perception, analgesic response to opioids, and tolerance and physical dependence resulting from opioid analgesic therapy (93). To move forward with personalized opioid prescribing, genome-wide association studies will need to target individuals with carefully selected phenotypes regarding opioid analgesic response, hyperalgesia, tolerance, and physical dependence. Furthermore, research will need to account for pharmacogenomic influences on opioid pharmacokinetics and pharmacodynamics. Additional factors to consider in identifying responders versus nonresponders include accounting for clinical characteristics of chronic pain mechanisms, brain and neurotransmitter changes associated with chronic pain, and other biomarkers (93).

As noted earlier, improved tools are needed to monitor compliance with prescription opioid therapy. Beyond ongoing individual assessment, public policy has the potential to influence and develop national strategies for monitoring compliance with prescription opioid therapy. For example, an integrated national PDMP across states would provide more comprehensive monitoring of dispenser, prescriber, and patient behaviors. PDMPs have currently been implemented on a statewide level, but limitations in interpretation apply when treating patients who have recently moved from other states or who travel to other states to seek pain treatment. Beyond clinical guidelines, risk assessment tools need to be validated and tested in different populations to assess generalizability and clinical utility.

The Institute of Medicine report (3) highlights the existence of knowledge deficits on opioid-related topics among providers, especially in the context of prescription opioid diversion to nonmedical use. The overall increase in opioid use and diversion has further created reactionary backlash against the medical use of opioids, causing restrictive public policies, reduced access to opioid medications, and individual and family concerns regarding the medical use of opioids. Overall, there is a need to create a curriculum for health care professionals to improve pain education and to provide guidance on the balance of prescribing opioids while minimizing associated harms. Furthermore, public education efforts are needed to promote the proper use and disposal of prescription opioids to maximize their safety.

Given the millions of Americans suffering from pain and the enormous socioeconomic burden resulting from pain, it is necessary for many different stakeholders, including providers, patients, third-party payers, and policy makers, to understand the necessities of appropriate, precise assessment and treatment of patients with chronic noncancer pain. Given the complexities of the biopsychosocial model of pain, an interdisciplinary approach to pain management is essential. Furthermore, despite the ongoing controversy surrounding the use of prescription opioids to treat chronic noncancer pain, accurate initial and ongoing assessment of opioid misuse, abuse, and addiction is essential to curtail the epidemic of prescription opioid overdoses. Future directions in research and public policy are needed to address knowledge gaps and to ultimately develop personalized multimodal treatment strategies for all patients with pain.

The authors report no competing interests.