Affecting at least 50 million individuals at an annual cost of $70 billion,1 chronic pain poses a significant health burden in the United States. While therapy for cancer pain has improved dramatically over the past decade,2 treatment of nonmalignant chronic pain remains a challenge for many practitioners and patients. Difficulties persist despite an array of medications to treat pain. Narcotics are known to be effective in malignant pain, and are a mainstay in the palliative care of cancer patients.3–9 However, narcotic use for nonmalignant chronic pain conditions commonly treated in primary care has been controversial and less well studied.
In this issue of the Journal of General Internal Medicine, Reid et al.10 characterize 98 patients receiving narcotic prescriptions for non–cancer-related pain in 2 academic primary care settings in Connecticut. The estimated prevalence of patients with chronic narcotic prescriptions in their study was low, only 0.7% in a clinic sample and 3% in a VA sample. Back pain ranked highest among diagnoses for which patients received narcotics, followed by injury-related, non–low back degenerative joint disorders, and diabetic neuropathy. A similar recent study by Adams et al.11 in 12 family practice clinics in Wisconsin also noted a low prevalence of narcotic prescriptions for chronic noncancer pain in their overall population, and found the same broad spectrum of pain-related diagnoses. As in previous studies of chronic pain patients,12–14 both studies found a considerable amount of psychiatric comorbidity, particularly depression.
Primary care providers' fear of contributing to opiate addiction is a frequently mentioned barrier to narcotic use in the management of chronic pain. While Reid et al.10 acknowledge that the criteria they used for assessing prescription drug abuse behavior have limitations, they conclude that “a significant minority (24% VA and 31% primary care clinic) of patients had documentation of prescription opioid abuse behavior.” Previously reported rates of prescription narcotic abuse, mostly from specialty pain clinics, have ranged from 3% to 20%,15,16 while Adams et al.11 documented an abuse rate of 6% in their primary care population. Reid et al.10 also attempted to identify predictors of prescription drug abuse behavior and noted that a “lifetime history of a substance use disorder” and younger age put patients at higher risk.
As the authors discuss, it is not surprising that chronic pain patients with histories of substance abuse are at higher risk for prescription drug abuse behaviors. While they acknowledge that inadequately treated pain may have been responsible for the higher rates of these behaviors, an important issue not addressed in this study is whether patients gained any pain relief or functional improvement from narcotics. Moreover, given the complex relationships we know exist between substance abuse and mental disorders,17–20 and chronic pain syndromes and mental disorders,12–14 the knowledge of how prescription drug abuse relates to other substance abuse while controlling for depression would be instructive. While the bivariate relationship between depression and prescription drug abuse did not meet selection criteria for inclusion in Reid et al.'s model of prescription drug abuse (P < .09), it does suggest that depression is related to increased inappropriate use. Information on whether depression was treated, a variable not measured in the study, would also have been informative.
The most revealing findings by Reid et al. concern the types of opioids prescribed for patients with chronic pain. Sixty percent of patients in the VA sample and 46% of the patients in the primary care clinic sample were given only short-acting opioids. Does this imply less severe pain among these patients? Unlikely, since 88 of the 98 patients in the study were prescribed daily doses of narcotics, while only 10 participants received as-needed doses. According to pain specialists, long-acting opiates should be the mainstay of treatment if narcotics are used for persistent pain.6,21 The role of short-acting opioids should be limited to initial dose titration and for breakthrough pain.
These results raise the question of how well primary care physicians are trained in treating chronic pain conditions and, specifically, how well they are informed about prescribing narcotics for these purposes. Attitudinal surveys highlight the ambivalence primary care providers feel about this issue. A survey of 161 family physicians found that while only 2% of respondents were unwilling to ever prescribe narcotics as needed for patients with noncancer chronic pain, 35% were unwilling to ever prescribe around-the-clock opioids for these patients, as would be recommended.22 A national survey of nearly 7,000 physicians in multiple specialties indicated that the negative effects of opioids and concerns over regulatory consequences inhibit many from prescribing narcotics for patients with nonmalignant chronic pain.23
To further complicate the issue, there are few widely accepted evidence-based guidelines or consensus statements readily available to help primary care providers manage chronic nonmalignant pain. Since the early 1990s, individual pain researchers and specialty organizations have produced several disease-specific guidelines for this purpose,24–29 and in 1997 the American Pain Society issued a broad consensus statement on the use of opioids in the treatment of chronic nonmalignant pain.30 However, rather than clarifying the situation, guidelines with differing views on the role of narcotics for these patients have added to the confusion.
In addition, information on the effectiveness of narcotics in chronic nonmalignant pain largely consists of conflicting observational data, as reviews have shown.1,31,32 The few existing randomized trials suggest a benefit from opioids but follow-up is often short, leaving unclear the effects of long-term treatment with these medications.33–40 In the absence of clear evidence amidst concerns about the addiction potential of opiates, management of chronic pain patients with narcotics may become susceptible to conscious or unconscious value judgements on the part of providers. Recent reports of abuse of sustained-release Oxycontin41 may complicate narcotic treatment decisions even further for primary care physicians. Thus, although the use of opiates has become more accepted in the management of chronic noncancer pain, primary care physicians are often left to rely on their own experience and guesswork, with assistance from pain and other specialties if available.
Reid et al. have added to our understanding of the current use of narcotics for chronic nonmalignant pain in 2 academic primary care settings; their work also underscores what we do not know about chronic noncancer pain in primary care. Little information exists on pain relief and functional outcomes of chronic nonmalignant pain patients treated in primary care, particularly in those receiving narcotics. We have limited knowledge about the long-term benefits and risks of narcotics in chronic pain patients, and multiple questions remain regarding selection of patients for these medications, when to initiate the therapies, and appropriate follow-up. While the potential for prescription drug abuse exists, it is unclear how factors such as comorbid mental health disorders, their treatment, and type of narcotic influence the risk of addiction in this population. Efforts should focus on consolidating management guidelines, and physician education about chronic noncancer pain and narcotics should be systematically incorporated into medical training and continuing medical education. Given the multiple facets of chronic nonmalignant pain, the effectiveness of multidisciplinary approaches to the management of these patients should be formally evaluated as well.
Primary care physicians excel at managing multiple chronic conditions when they have the appropriate tools. For our patients with chronic nonmalignant pain, we hope we will have what it takes. —Yngvild Olsen, MD,Division of General Internal Medicine, Johns Hopkins Medical Institutions, Baltimore, Md, and Gail L. Daumit, MD, MHS,Division of General Internal Medicine, Johns Hopkins Medical Institutions and Bloomberg School of Public Health, Baltimore, Md.