Co-Location of DEA-Waivered Buprenorphine Prescribers and Behavioral Health Clinicians

Background and Purpose: Medication for opioid use disorder (MOUD), or medication with behavioral therapy and/or other psychosocial services, is currently the most effective, evidence-based intervention available for treating opioid use disorder (OUD). However, rates of co-location between behavioral health clinicians (BHCs) and waivered prescribers are unknown. This study assessed 1) What percent of waivered providers are co-located with BHCs; (2) If rates of co-location vary by provider type including social workers; and (3) If rates vary by state, rurality, and other geographic measures.

Methods: Behavioral health providers were identified through the DEA Drug Addiction Treatment Act of 2000 (DATA) waived provider list as of March 2022 and the Center for Medicare and Medicaid Services (CMS) National Plan and Provider Enumeration System’s (NPPES) National Provider Identifier (NPI) database. Provider addresses were geocoded to latitude and longitude coordinates using ESRI StreetMap® and ArcGIS software. Multivariable logistic regression was used to estimate differences in co-location rates by prescriber type, waiver size, expanded scope of NP practice, and county rurality.

Results: Across all prescriber types, nearly 48% were co-located with a BHC. Prescriber waiver size predicted co-location, with prescribers who had higher waiver sizes being less likely to be co-located (n=71,292 waivered prescribers). Across all provider types, those with the smallest waivers (< 30 patients) were more likely to have higher rates of co-location. Prescribers located in metro counties were more likely to be co-located than prescribers in non-metro counties (49% vs. 39%), with the differences most pronounced for physicians. Nurse practitioners were more likely to be collocated than physician assistants (OR = 1.15, CI = 1.25-1.05) and physicians (OR = 1.28, CI = 1.22-2.50). Rural counties were less likely to be co-located compared to urban counties (OR range = 0.76-0.85). Prescribers in states with expanded scope of practice laws for nurse practitioners had 1.39 times higher odds of co-location, an effect observed for all prescribers.

Conclusions and Implications: Less than half of all waivered prescribers outside of hospitals are co-located with BHCs. Given projected increases in the nurse practitioner and physician assistant workforces, expansion of prescribing-eligible workforces offers promise for OUD treatment in co-located settings. However, state-level scope of practice laws and policies impact in what setting and by whom comprehensive OUD treatment is delivered. The use of tele-health models may also be obscuring the rate of co-location between waivered prescribers and BHCs. New X-waiver policy changes will also impact how h co-located teams providing MOUD are monitored and tracked going forward. Findings imply a need for establishing additional interprofessional teams to support MOUD delivery, as more than half of all waivered prescribers and BHCs are not co-located. Future research could explore the relationship between prescribers’ co-location with a BHC, and social worker specifically, and their willingness to deliver MOUD services.