Association for Public Policy Analysis & Management

Background: Hospital visits for opioid-related causes have risen sharply over the last decade. Hospitals are being prompted to initiate treatment with buprenorphine. Induction in hospital settings can facilitate continued treatment in the community with an outpatient buprenorphine prescriber, but continuity of care after discharge is recognized to be a problem. However, limited supply of providers able to prescribe buprenorphine may be one impediment to post-hospital continuity of care. Additionally, it is unknown whether physicians who prescribe buprenorphine are affiliated with hospitals through integrated health systems. Our objective was to characterize the waiver rates among community physicians based on their health system affiliation status, and to examine variation in waivered rates across geographic areas.

Data and Methods: We obtained the full sample of PCPs and psychiatrists from SK&A extracted in September 2018. The sample was limited to California, Florida, Georgia, Maryland, Ohio, Rhode Island, Wisconsin, and West Virginia. SK&A includes office locations and information on the health system affiliation of a physician office. We merged the data to the list of waivered buprenorphine prescribers from the U.S. DEA.We stratified PCPs and psychiatrists and we compared waivered rates based on health system affiliation. Among waivered physicians in each group, we considered percentages with different patient limits based on health system affiliation. We calculated area-level differences by creating scatterplots of waivered status by hospital affiliation status within hospital referral regions (HRRs).

Results: About 40% of all PCPs in the nine states were practicing in health systems. Only 3.6% of PCPs in health systems had a buprenorphine waiver compared to 8.2% of PCPs practicing outside of health systems. Among waivered PCPs, those in health systems were much more likely to have a 30 patient limit than those outside health systems. About 26% of all psychiatrists in the nine states were practicing in health systems. Psychiatrists in health systems were more likely to have a buprenorphine waiver than those outside of health systems. However, waivered psychiatrists in health systems were more likely to only have 30 patient limit than those outside health systems. Consistent with our aggregated analysis, we find that for PCPs, the percentage with a waiver is much lower in health systems than outside health systems. Indeed, in 20% of the HRRs there were no PCPs affiliated with health systems with a waiver, compared to 2 of the HRRs where there were no PCPs unaffiliated with health systems that have a waiver. By contrast, for psychiatrists, the percentage with a waiver was much higher in health systems than outside health systems.

Discussion: Many buprenorphine-waivered physicians are practicing outside of integrated health systems. Exclusion from health systems has several practical implications. It can limit the continuity of community care after a patient is hospitalized for an opioid overdose. In general, referral patterns after hospital discharge are known to be highly concentrated within networks. If buprenorphine-prescribing physicians are disproportionately excluded from health system networks they may also have more limited ability to refer their patients out to other providers.