To the Editors:

Pain represents a major reason why patients present for rheumatic and musculoskeletal care. Even though pain is such a common part of the patient's presentation, how we treat pain is by no means standard. Specifically, the role of opioid analgesics is a point of great controversy. Three senior academic rheumatologists informally polled about their use of opioids for patients with rheumatic or musculoskeletal conditions gave the following responses: “I never need to use those drugs since other treatments are relatively effective.” “Opioids are an important part of many patients' analgesic regimens.” “I just don't use opioids.” This wide range of opinions epitomizes the lack of consensus about opioid analgesics and motivated us to begin a series of studies examining how they are used in patients with arthritis and other musculoskeletal conditions.

In our article, we described the frequency of opioid use and how this frequency has changed over time. In our first set of analyses we found that in 2001, 15% of adults with rheumatoid arthritis (RA; mean age 80 years), received at least 1 prescription in a given calendar year; and 5% of the patients with RA received opioids long-term. For patients with low back pain (also with a mean age of 80 years), 8% received at least 1 prescription for opioids, and 2% received them long-term. These patterns were steady over the 6-year study period.

Vogt et al suggest that the frequency of opioid prescribing we describe is “relatively low” compared with a study they conducted among patients with low back pain enrolled in a commercial insurance plan (1). In their study, 38% of the patients had received a prescription for an opioid in 2001. The difference in estimates between the 2 studies may be explained by important variations in age and comorbid conditions. Patients in our study had an average of 2 comorbid conditions and because of their advanced age, clinicians may have opted for nonopioid analgesics. Therefore, the patient population in the study by Vogt et al may have been better candidates for opioid treatment than the patient population in our study.

We also compared our results with prior literature. One study using data from the National Ambulatory Medical Care Survey found that in the year 2000, 11% of patients with acute musculoskeletal pain and 15% with chronic pain received an opioid analgesic (2). Data from the Third National Health and Nutrition Examination Survey found that 3.4% of adults ages ≥17 years reported use of opioid analgesics in the prior month (3). A study including 25,479 patients attending a spine center found that 3.4% had opioids included in their plan of care (4). Although these data are from different populations and time periods, they suggest that our result (8–15% of adults with a musculoskeletal symptom received at least 1 prescription for an opioid in a given year) is within the bounds of other studies. As Vogt and colleagues note in their original study, the prevalence of opioid use in their study population “is much higher than previously reported” (1).

Understanding the current frequency and predictors of opioid prescribing is an initial step at developing a clear picture of the role for these drugs. Future studies must be conducted to determine the relative benefits versus risks of these agents and better understand the prejudices of physicians and patients about their use. Although the Hippocratic oath of “first do no harm” should guide much of our therapeutic approach, as Charles Mayo noted, “Of all the symptoms for which physicians are consulted, pain in one form or another is the most common and often the most urgent.”

Daniel H. Solomon MD, MPH*, Jerry Avorn MD*, Philip S. Wang MD, DrPH*, George Vaillant MD*, Danielle Cabral BA*, Helen Mogun MSc*, Til Stürmer MD, MPH*, * Brigham and Women's Hospital, Harvard Medical School Boston, MA.