There is no question that pain of all types is common and that the resulting epidemiologic, economic, and human burden is substantial. In 2011, the Institute of Medicine estimated that 100 million American adults are affected by chronic pain of any type and that associated costs (including lost productivity) are as high as $635 billion [1]. Pain is also an expected outcome of surgery. In 2010, 51.4 million inpatient surgical procedures were performed, and outpatient surgeries accounted for approximately 63.6 % of total surgeries in community hospitals [2, 3]. Cancer-related pain is also an important problem in people with current disease and in individuals with a history of cancer. In 2007, the pooled prevalence of pain across all cancer types was estimated to be greater than 50 % [4]. Furthermore, breakthrough pain in cancer patients can represent a challenging complication. While published estimates of the prevalence of breakthrough pain vary widely (due in part to use of different definitions and inclusion criteria), three estimates published in the last few years provided estimates of approximately 40 %, 44 % (incident pain), 41.5 % (spontaneous pain), and 73 %, suggesting that a significant proportion of people with cancer have relatively complex pain [5–7]. People with current cancer may have pain related to the tumor or from surgery, chemotherapy, or radiation treatment. It is somewhat ironic that as individuals with cancer live longer, the prevalence of cancer-related pain may also increase, even after there is no remaining evidence of disease [8].

While the drugs available to treat pain come from a wide variety of pharmacologic classes, mechanisms of action, and routes of administration, the opioid analgesics are among the most important of these tools. At the same time, there remain important gaps in understanding the pathophysiology of pain, opioid pharmacology and toxicology, and the clinical, economic, and human outcomes of pain and of its treatment [9, 10]. The purpose of this chapter is to review and discuss three issues that pertain to this interface: opioid-induced hyperalgesia, measurement of the clinical effectiveness of the opioids, and efforts to reduce abuse, misuse, and diversion of opioids (and other controlled substances) as part of efforts to reduce drug-related adverse outcomes and deaths.

Hyperalgesia refers to an increased response to a painful stimulus at either a normal or increased threshold [11]. In contrast, the term opioid-induced hyperalgesia (OIH) is broadly defined as patients becoming more sensitive to pain due specifically to the use of opioids [12–15]. Yet, there are significant gaps in understanding the mechanisms underlying OIH as well as the clinical relevance of this phenomenon [12, 16].

Pathophysiology. The pathophysiology of OIH is incompletely understood, and it remains important to emphasize that decreased analgesia from opioids may reflect hyperalgesia, tolerance, worsening disease, or a combination of these phenomena [17].

Having said that, the science underlying increases in pain sensitivity is growing rapidly. Specifically, hyperalgesia appears to occur in a wide variety of clinical settings and patient populations, including persons on maintenance or withdrawal of opioid use, as well as persons receiving methadone for opioid maintenance and surgical patients [18]. There also appears to be variation in the type of increased painful response to different stimulation. For example, in persons being maintained on methadone, hyperalgesia was found to occur to cold, but to a lesser degree for electrical stimuli and not to mechanical pain. In persons who underwent surgery, the development of hyperalgesia appeared to be associated with receiving higher opioid doses, although this finding has not been consistent, nor is it always easy to separate hyperalgesia from tolerance.

Risk factors. Risk factors for development of OIH are not yet well understood. In the 2010 research guideline on future research pertaining to the use of opioids for chronic non-cancer pain, the authors note that several critical topics have yet to be addressed, including the epidemiology of OIH, the likelihood of developing OIH as a function of clinical and demographic characteristics, the interaction between hyperalgesia and acute pain, and, notably, dose and duration of therapy [16]. While facets of a patient’s clinical presentation such as dose, duration of therapy, and quantitative sensory testing may be helpful to identify patients at increased risk for altered pain sensitivity, these factors are neither sensitive nor specific.

Treatment. While data on treatment for suspected opioid-induced hyperalgesia is limited, suggested treatment regimens generally include dose increase (to rule out tolerance), dose decrease, and adding (or changing to) N-methyl-d-aspartate receptor antagonists (such as ketamine and methadone), clonidine, and dexmedetomidine. Buprenorphine may be useful, and multimodal analgesic regimens may be useful due to the opioid-sparing effect of these regimens. Propranolol may also have some role in preventing OIH. Remifentanil, a potent opioid used primarily in surgical settings, is known to produce a post-infusion hyperalgesia. In a recent study of ten patients, persons who received remifentanil and propranolol did not experience mechanical hyperalgesia, while individuals who received remifentanil and placebo did have hyperalgesia [19]. In a separate publication, these authors also demonstrated in a study of adults with chronic nonradicular low back pain who used sustained-release morphine, these authors demonstrated that tolerance can occur independently from OIH [20].

Evidence-based medicine plays an important part of clinical practice in many countries, from reimbursement decisions through pay for performance incentives. Yet, the evidence base for the use of opioids is often uneven. Although there are a variety of clinical practice guidelines (CPGs) focused on the treatment of pain, there are also important gaps, such as recommendation for the treatment of pain in children and adolescents [10]. Furthermore, CPGs for conditions like hypertension and hypercholesterolemia tend to be relatively prescriptive. In contrast, CPGs for pain management tend to be very general. The result is that, in the absence of high-quality evidence, clinicians are often left to practice according to habit, which may or may not result in optimal outcomes.

Additionally, clinicians may not know which outcomes are best to assess or recommended methods (surveys, questionnaires, etc.). For example, it is generally easy to assess a patient’s pain (and, indirectly, the quality of care) using a numeric analog 0–10 or verbal rating scale, but pain intensity is just one of many facets of a complex multidimensional experience [21]. Similarly, the expected duration of pain and subsequent treatment play an important role. For example, acute pain studies typically last just a few days, with single-dose regimens commonly used, and chronic pain studies typically last no more than a few months at most.

Other dimensions of the pain experience, such as the ability to function in ways that are important to the patient, family, and caregivers, are also likely to be important. In a consensus statement regarding the core competencies for pain management across all healthcare disciplines, four domains were identified as being of primary importance: the multidimensional nature of pain, pain assessment and measurement, management of pain, and context of pain management [22]. Similarly, the domains identified in the revised American Pain Society Patient Outcome Questionnaire as reflecting the quality of care were pain severity and relief, impact of pain on activity, sleep, and negative emotions, side effects of treatment, helpfulness of information about pain treatment, ability to take part in pain treatment decisions, and use of nondrug treatments [23].

In designing clinical trials, a set of domains recommended for consideration of inclusion in clinical chronic pain studies has been published by the Initiative on Methods, Measurement, and Pain Assessment in Clinical Trials (IMMPACT) group. These recommendations emphasize the importance of these core domains: pain, physical and emotional functioning, participant ratings of improvement and satisfaction with treatment, and participant disposition, including adherence to treatment and withdrawal from the study [24]. More specifically, patients reported that the most important ways that pain affected them in daily life were pain reduction, enjoyment of life, emotional well-being, fatigue, weakness, and sleep [25] In studies of acute, chronic, and recurrent pain in children and adolescents, the domains considered to be of importance were generally similar to those suggested for studies of adults [26]. Specifically, the domains recommended for studies in children were pain intensity, satisfaction with treatment, symptoms and adverse effects, physical recovery and emotional response, and economic factors, with physical and emotional functioning, role functioning, and sleep added for chronic pain studies.

There is significant concern about the potential for opioids to be misused, abused, or diverted, resulting in adverse events and deaths. In 2011, the US Centers for Disease Control and Prevention estimated that the rate of deaths related to prescription analgesic overdose had increased by more than threefold since 1990, that the number of these events was greater than that of motor vehicle deaths and larger than deaths attributed to cocaine and heroin overdoses [27, 28]. While these data do not tell us important information such as where the drugs were obtained, whether the overdose was intentional or accidental, and if the event was due to use of one drug or multiple medications, it is clear that there is a serious problem. It is also clear that solving this problem will requires interdisciplinary efforts on a wide variety of facets to help ensure that safe and effective pain management is available for people who need it, and that access to potent medications is minimized for people who do not.

As of October, 2012, 41 US states have an operational Prescription Monitoring Programs (PMP) in place and an additional 10 (including the District of Columbia and Guam) have enacted legislation, but the program has not yet begun operating [29]. These programs generally collect information about the patient, the drug and amount being prescribed, the date the prescription was written and the date it was filled, and the pharmacy that filled the prescription. States differ in terms of what data are collected, what time lag between filling the prescription and submitting the information is allowed, who is responsible for maintaining the dataset (In Texas, it is the Department of Public Safety, in other states, the Board of Pharmacy or other groups are responsible), who is allowed to see or use the data, and whether prescribers are alerted to potential problems passively or actively. Yet, despite the prevalence of PMPs, it remains unclear how to best use this information to achieve optimally safe and effective pain management. Here are a few challenges and opportunities that remain for healthcare providers and health policy decision makers before the promise of these data can be fully realized.

Much of the PMP-related literature and discussions focus on drug abuse, often using prescriber- or pharmacy-shopping as markers of inappropriate use of controlled substances. Despite this emphasis, aside from extremely high levels of use in a short period of time, we don’t know the right number of prescribers or pharmacies, the right time period to consider, or whether there is even a right number. For example, in a literature search of “prescription monitoring program” AND “opioid,” time intervals used to define either doctor- or pharmacy-shopping included zero (concurrent use of opioids or controlled substances from different prescribers), 6 months, 30 days (prescription for the same medication from at least two practitioners filled by at least two pharmacies), and 1 year, although one consensus statement published in 2004 defined use of multiple prescribers as at least six prescribers over a 1-year period [30–37]. In a recent analysis, Wilsey and colleagues used California PMP data to assess differences between people who used only one prescriber compared to individuals who used 2–5 prescribers over a 1-year period [32]. These investigators found that people who used 2–5 prescribers were not different in terms of male sex, younger age, or location in larger geographic areas—factors thought to be associated with drug abuse.