Background: Patients with chronic noncancer pain treated with higher doses of opioids or concurrent substance use are at increased risk of adverse events. Although several national guidelines recommend maximum dosing thresholds and urine drug testing, adherence to these guidelines is inconsistent.
Methods: To identify predictors of higher-risk opioid prescriptions in 2 academic primary care clinics, the authors developed a retrospective cohort of 842 patients who were prescribed ≥5 opioid prescriptions for noncancer pain between March 2012 and March 2013. The authors evaluated odds of higher-dose opioid prescriptions and urine drug testing using multivariate logistic models.
Results: Among study subjects, 47% received prescriptions for the equivalent of ≥50 mg morphine per day. After adjustment for confounders, patients with a resident primary care provider were less likely to receive higher-dose prescriptions compared with faculty providers (odds ratio = 0.66, 95% confidence interval [CI]: 0.46-0.94), whereas patients with a nonlocal home address were more likely to be prescribed higher doses (odds ratio = 2.1, 95% CI: 1.5-2.9). Hispanic, Asian, and older patients were also less likely to be prescribed higher doses. Urine drug testing was not regularly completed (35% over 2 years), but odds of testing were higher for patients who self-identified as black, had resident primary care providers, lived locally, or were prescribed higher opioid doses.
Conclusions: In this academic clinical setting, patients with a resident primary care provider are less likely to receive higher-risk opioid prescriptions, as are Hispanic, Asian, and older patients. Black patients complete urine drug tests more frequently independent of other patient and provider characteristics. Additional studies are needed to assess why patients who travel larger distances to their primary care clinic are prescribed higher doses of opioids for chronic noncancer pain.
Keywords: Chronic pain; opioid analgesics; prescription drug misuse; substance abuse detection.