Effect of Different Interventions to Help Primary Care Clinicians Avoid Unsafe Opioid Prescribing in Opioid-Naive Patients With Acute Noncancer Pain: A Cluster Randomized Clinical Trial

Abstract

Importance: Prescription opioids can treat acute pain in primary care but have potential for unsafe use and progression to prolonged opioid prescribing.

Objective: To compare clinician-facing interventions to prevent unsafe opioid prescribing in opioid-naive primary care patients with acute noncancer pain.

Design setting and participants: We conducted a multisite, cluster-randomized, 2 × 2 factorial, clinical trial in 3 health care systems that comprised 48 primary care practices and 525 participating clinicians from September 2018 through January 2021. Patient participants were opioid-naive outpatients, 18 years or older, who presented for a qualifying clinic visit with acute noncancer musculoskeletal pain or nonmigraine headache.

Interventions: Practices randomized to: (1) control; (2) opioid justification; (3) monthly clinician comparison emails; or (4) opioid justification and clinician comparison. All groups received opioid prescribing guidelines via the electronic health record at the time of a new opioid prescription.

Main outcomes and measures: Primary outcome measures were receipt of an initial opioid prescription at the qualifying clinic visit. Other outcomes were opioid prescribing for more than 3 months and a concurrent opioid/benzodiazepine prescription over 12-month follow-up.

Results: Among 22 616 enrolled patient participants (9740 women [43.1%]; 64 American Indian/Alaska Native [0.3%]; 590 Asian [2.6%], 1120 Black/African American [5.0%], 1777 Hispanic [7.9%], 225 Native Hawaiian/Pacific Islander [1.0%], and 18 981 White [83.9%] individuals), the initial opioid prescribing rates at the qualifying clinic visit were 3.1% in the total sample, 4.2% in control, 3.6% in opioid justification, 2.6% in clinician comparison, and 1.9% in opioid justification and clinician comparison. Compared with control, the adjusted odds ratio (aOR) for a new opioid prescription was 0.74 (95% CI, 0.46-1.18; P = .20) for opioid justification and 0.60 (95% CI, 0.38-0.96; P = .03) for clinician comparison. Compared with control, clinician comparison was associated with decreased odds of opioid therapy of more than 3 months (aOR, 0.79; 95% CI, 0.69-0.91; P = .001) and concurrent opioid/benzodiazepine prescription (aOR, 0.85; 95% CI, 0.72-1.00; P = .04), whereas opioid justification did not have a significant effect.

Conclusions and relevance: In this cluster randomized clinical trial, comparison emails decreased the proportion of opioid-naive patients with acute noncancer pain who received an opioid prescription, progressed to treatment with long-term opioid therapy, or were exposed to concurrent opioid and benzodiazepine therapy. Health care systems can consider adding clinician-targeted nudges to other initiatives as an efficient, scalable approach to further decrease potentially unsafe opioid prescribing.

Trial registration: ClinicalTrials.gov Identifier: NCT03537573.

Conflict of interest statement

Conflict of Interest Disclosures: Dr Gordon reported grants from the National Institutes of Health, Veterans Affairs, and US Department of Health and Health Services during the conduct of the study as well as personal fees from UpToDate and board service with the American Society of Addiction Medicine, Association for Multidisciplinary Education and Research in Substance use and Addiction, and International Society of Addiction Journal Editors outside the submitted work. Dr Wright reported grants from the Patient-Centered Outcomes Research Institute (PCORI) during the conduct of the study and grants from Pfizer and grants from the National Institute on Drug Abuse outside the submitted work. Drs Cochran, Fischer, Gellad, and Hamm reported grants from PCORI during the conduct of the study. Dr Wasan reported consulting fees from Greenwich Biosciences and grants from Parallel outside the submitted work. No other disclosures were reported.

Copyright 2022 Kraemer KL et al. JAMA Health Forum.

Figures

Figure 1.. CONSORT Diagram of Primary Care Practice Participation

CDM indicates the National Patient-Centered Clinical Research Network (PCORnet) Common Data Model.

Figure 2.. Rate of Opioid Prescribing at the Qualifying Clinic Visit

A, Opioid justification vs control. B, Clinician comparison vs control. The figure shows opioid prescribing rates by month during the 15-month intervention phase of the trial. Month 0 is when interventions were initiated at the participating practices. The vertical lines indicate 95% CIs. Opioid justification includes participants from the justification and justification/comparison groups. Clinician comparison includes participants from the comparison and justification/comparison groups.