Adherence to chronic opioid therapy prescribing guidelines in a primary care clinic.

Objective: Characterize primary care patients prescribed opioids for chronic noncancer pain (CNCP), explore guideline-recommended opioid-monitoring practices, and investigate predictors of pain agreements.

Design: Retrospective chart review. Setting: Primary care clinic at a tertiary academic medical center. Patients: Adults prescribed chronic opioids (three or more monthly prescriptions within a year) for CNCP between April 1, 2014 and April 1, 2015. Patients without CNCP served as controls. Main outcome measure: Patient demographics, medical diagnoses, tobacco status, provider status, documentation of guideline-recommended opioid-monitoring practices, pain agreement status, and opioid prescription. Univariate statistics were used to explore differences in patient demographics, comorbidities, and guideline-recommended opioid-monitoring practices by chronic pain and pain agreement status. Logistic regression was used to investigate predictors of agreement status.

Results: The clinic had 834 (9 percent) patients on chronic opioids, with 335 on a pain agreement. Documentation of opioid-monitoring practices was lacking. Logistic regression indicated that patients were significantly more likely to be on an agreement if they were Caucasian (adjusted odds ratio [OR] 2.17 [95% CI 1.41, 3.39]), had a baseline urine drug screen (adjusted OR 10.72 [95% CI 6.16, 19.41]), were prescribed a schedule II controlled medication (adjusted OR 11.92 [95% CI 6.93, 21.62]), and had risk assessed to some degree (adjusted OR 3.06 [95% CI 1.90, 4.96]).

Conclusions: Aside from race, most patient characteristics were not predictive of pain agreement implementation. However, controlled medication of higher schedules and the use of certain guideline-recommended practices were associated with an agreement. Studies are needed to examine whether pain agreement or guideline-adherence influence clinical outcomes.