Buprenorphine, Pain, and Opioid Use in Patients Taking High-Dose Long-Term Opioids: A Randomized Clinical Trial

Importance: Guidelines recommend dose reduction or discontinuation of long-term opioid therapy when harm outweighs benefit, but strategies to help patients do so are limited. Objective: To test optionally switching to buprenorphine as a strategy for improving pain and reducing opioids among patients prescribed high-dose, full agonist long-term opioid therapy. Design, Setting, and Participants: In this pragmatic, multisite, 12-month randomized clinical trial with masked outcome assessment, patients treated at Veterans Affairs primary care clinics were recruited from October 2017 to March 2021, with follow-up completed June 2022. Eligible patients had moderate to severe chronic pain despite high-dose opioid therapy (≥70 mg/d for at least 3 months). Patients were randomized to having the option to switch to buprenorphine or not having the option to switch. Interventions: The buprenorphine option was discussed with eligible patients as part of a larger trial of collaborative pain care interventions. Those who switched had structured follow-up to optimize dosing and address adverse effects. Main Outcomes and Measures: The primary outcome was Brief Pain Inventory total score at 12 months. The main secondary outcome was opioid dose in morphine milligram equivalents at 12 months. Results: Of 207 included participants, 185 (89.4%) were male, and the mean (SD) age was 60.9 (10.2) years. A total of 104 were randomized to the buprenorphine option and 103 to the no buprenorphine option. In the buprenorphine option arm, 27 participants (26.0%) switched. Over 12 months, the mean (SD) Brief Pain Inventory score improved from 6.8 (1.5) to 6.1 (1.9; adjusted mean difference [AMD], -0.59; 95% CI, -0.89 to -0.29) in the buprenorphine option arm and from 6.8 (1.6) to 6.3 (1.7; AMD, -0.50; 95% CI, -0.81 to 0.20) in the no option arm (between-group AMD, -0.09; 95% CI, -0.52 to 0.34). Over 12 months, mean (SD) opioid dosage decreased from 157 (75) mg/d to 94 (98) mg/d in the buprenorphine option arm (AMD, -61.0 mg/d; 95% CI, -74.1 to -47.9) and from 165 (88) mg/d to 107 (89) mg/d (AMD, -58.5 mg/d; 95% CI, -71.6 to -45.4) in the no option arm (between-group AMD, -2.5 mg/d; 95% CI, -21.1 to 16.0). Conclusions and Relevance: In this trial, outcomes did not differ between groups; both had small improvements in pain and substantial reductions in opioid dosage, but the proportion of participants who switched to buprenorphine was low.