Future Directions in Opioid Prescribing Patterns for Patients with Chronic Pain

Context: Despite a growth in opioid-related deaths over the past 10 years, efforts to reduce opioid prescribing are complicated by lack of experience and skills in tapering opioids, disagreement with guidelines, lack of access to qualified pain specialists, the medical complexity of patients with chronic pain, and a lack of prescriber awareness about their own prescribing patterns relative to their peers.

Objective: 1) Demonstrate the feasibility of collecting data from prescription drug monitoring programs (PDMPs) for population health reporting, 2) analyze trends in PDMP and electronic health record patient data relative to CDC guidelines and 3) identify use of patient-centered measures for pain management in primary care.

Study Design: Secondary data analysis of a retrospective cohort from 2017-2023.

Setting or Dataset: 4 primary care clinics (PCCs) across Maine, New Hampshire, and Vermont.

Population: 569 adults with chronic pain treated with opioids for 5 or more prescriptions or 1000+ morphine mg equivalents (MME) per year.

Intervention/Instrument: Video training program for office staff to use PDMP; chart review for patient-centered measures.

Outcome Measures: Change in total clinic-prescribed MME per year, change in mean MME per patient per year over the most recent 2 years of observation, proportion of patients with a 30% reduction in MME over 2 years, change in concurrent benzodiazepine use, and various chart review measures.

Results: Trend lines of opioid MME prescribed by providers at each participating clinic were shared at provider meetings. At the patient level, opioid MMEs decreased from 8041 MME/yr to 6454 MME/yr [delta -1586; 95%CI 911,2262], and 50% of patients decreased daily MME by 30% or more. Concurrent use increased from 6.3 to 7.7 (delta +1.4 [95%CI -.2,+3.1]. Use of mandatory precautions such as patient consent/treatment agreement and PDMP lookup was highly utilized, but use of patient-centered measures (e.g., pain, function, and risk stratification) varied within and across practices.

Conclusions: PDMPs are a feasible source for population health reporting and for PCCs to evaluate their opioid prescribing patterns. Trends across 4 unrelated practices varied greatly, likely due to type of practice, history of prescribing opioids, and differences in patients served. Use of a common measure for pain management is needed for future studies.