Three trajectories of implementation of medications for opioid use disorder in primary care.

BACKGROUND

The opioid epidemic has prompted nationwide efforts to expand access to medications for opioid use disorder (MOUD). Primary care settings have been identified as a critical access point for patients who may benefit from MOUD treatments. Despite implementation efforts, there is limited understanding of how MOUD practice capability in primary care settings evolves over time or what factors influence clinic-level implementation trajectories.

METHODS

We conducted a longitudinal study of 95 primary care clinics in California from 2019 to 2024. MOUD practice capability was measured using the Integrating Medications for Addiction Treatment in Primary Care (IMAT-PC) index across three timepoints. Using latent class growth analysis, we analyzed implementation growth trajectories and examined their associations with clinic characteristics and MOUD implementation outcomes (e.g., patient reach and provider adoption).

RESULTS

Three distinct implementation trajectory classes emerged: elevated improving (41.0%), moderate improving (47.4%), and low improving (11.6%). All clinics demonstrated improvements in MOUD practice capability over time. Elevated improving clinics primarily consisted of smaller clinics (< 15,000 patients) and achieved significantly higher number of patients receiving MOUD compared to moderate ( = 0.03) and low improving clinics ( = 0.04). Clinics serving medically underserved populations disproportionately represented the low improving class ( < 0.01). Increase in the number of providers prescribing MOUD did not differ significantly across trajectory classes.

CONCLUSIONS

Although all clinics increased MOUD capability, we found significant associations between implementation trajectory classes and changes in patients receiving MOUD over time in primary care-based MOUD programs. Implementation supports may be more effective and efficient if selected and delivered based upon clinic contextual factors, particularly in resource-constrained and underserved settings.