Opioids in OA: A Matter of Geography

Long-term opioid use is common among elderly patients with severe osteoarthritis (OA), and a large observational study documented substantial geographic variation in use.

Among patients with advanced knee or hip OA enrolled in Medicare, one in six used prescription opioids for longer than 3 months in the year leading up to joint replacement, according to Rishi J. Desai, PhD, of Harvard Medical School in Boston, and colleagues.

And the percentage of patients who had used opioids for at least 3 months ranged from 8.9% in Minnesota to 26.4% in Alabama, the researchers reported online in Arthritis & Rheumatology.

Osteoarthritis is one of the most common causes of chronic pain in the U.S., and typically is treated with nonsteroidal anti-inflammatory drugs, opioid analgesics, and steroids. However, a recent randomized trial found a lack of superiority among patients with chronic back pain or OA for opioid and non-opioid treatments at 1 year, and the CDC recommends that clinicians assess risk-benefit ratios every 3 months for patients using opioids.

Yet little is known about real-world long-term use of opioids in older patients with OA. “Therefore, we conducted an observational cohort study in a nationwide sample of Medicare enrollees with severe OA to describe long-term opioid use and to evaluate the role of geography and healthcare access in determining long-term opioid use,” Desai and colleagues wrote.

For the analysis, they included Medicare claims from 2010 to 2014 for diagnoses, procedures, and medications prescribed. Information about access to primary care providers was obtained from the Primary Care Service Area (PCSA) database, which contains information about patients, providers, and healthcare centers at postal-code level throughout the country.

One million randomly chosen Medicare recipients who underwent total joint replacement were included, and opioid use in the year preceding the procedure was examined.

Multiple covariates were included in the analysis, including comorbidities, socioeconomic factors, and differences in state requirements for prescribers’ checking prescription drug monitoring programs when they write opioid prescriptions.

Long-term users were those whose prescriptions were for 90 days or more. Average daily dose was calculated in morphine milligram equivalents (MME), with 50 to 90 MME being the range where the CDC advises caution because of the likelihood of harm, and 90 MME or higher representing a dose range that should be avoided.

The study included 358,121 patients with OA whose mean age was 74. Most were white women.

Of the entire cohort, 16.5% were long-term users and 42.3% were short-term (less than 90 days) users.

The median length of use in the long-term group was 7 months, compared with 15 days in the short-term group. Doses of 50 MME or higher were used by 19% and 15.9%, in the two groups, respectively. The specific agents most commonly used in the long-term group were tramadol, oxycodone, and fentanyl, at 45.8%, 32.2%, and 6.2%, respectively, compared with the short-term group, with rates of 36.8%, 21.7%, and 0.5%.

The analysis included 4,080 of the 7,144 PCSAs in the U.S., which reflected a range of socioeconomic groups and varying access to healthcare. The average number of long-term users was 17.2% in these PCSAs, but ranged from 0% to 60%.

Overall, the PCSAs with the highest long-term opioid use were located in the South and the lowest in the Midwest and Northeast.

The reference state was New York, which was chosen because of its large population and previously reported low opioid use. Significantly higher rates of long-term users compared with New York were found in 31 states, and two states, West Virginia and Alabama, had a mean difference in long-term users of more than 10 percentage points compared with New York.

For PCSAs with the highest concentration of primary care providers (>8.6 per 1,000 patients) the adjusted mean difference in long-term opioid users was not significantly different, at 1.4% (95% CI 0.8-2) than for PCSAs with the lowest concentration (<3.6 per 1,000). Similarly, no difference in opioid use was seen in PCSAs with the highest concentration of rheumatologists (>0.29 per 1,000) and the lowest (0), with an adjusted mean difference of 0.6% (95% CI -0.1 to 1.3).

Despite the substantial variation in long-term use by state and PCSAs, the association with access to primary care providers was “modestly negative,” the researchers noted.

Moreover, in their analysis they adjusted for differences in access to care, patient populations, and socioeconomic factors and found that state of residence remained independently associated with long-term opioid use, and that “regional prescribing practices play a key role in determining rates” of use.

“This finding suggests that geographically targeted interventions to ensure widespread dissemination and implementation of safe opioid prescribing guidelines are necessary to make a meaningful impact on prescribing practices,” they concluded.

Limitations of the study included the lack of information on pain severity or function or other potentially important services such as physical therapy.