Our mixed methods study compared physician prescriber perceptions of efficacy and barriers to OUD treatment across three dominant MOUDs. In terms of barriers, we focused on office-based treatment settings (i.e., naltrexone and buprenorphine prescribing), but also asked about referral to methadone clinics. We also compared responses from those physicians with and without a waiver to prescribe buprenorphine. The survey data complemented by qualitative responses provides new and timely information on MOUD treatment beliefs and challenges.
Our study found that insurance barriers, specifically prior authorization requirements, were the most commonly cited barrier to buprenorphine and extended-release naltrexone prescribing. While few other studies have explored barriers to extended-release naltrexone prescribing (26, 30), partly owing to its relatively recent FDA-approval for OUD, other studies have likewise found that insurance requirements are a strong barrier to buprenorphine prescribing(21, 34, 35). By confirming results from these other studies, our study lends further support to the need for federal and state governments to intervene in decreasing insurance barriers to MOUD. For example, federal and state authorities should strengthen enforcement of parity laws and sanction violations related to inequitable non-quantitative treatment limitations applied to OUD treatment as compared to other chronic health conditions. These barriers may be quantitative (e.g., the number of days of treatment coverage) or non-quantitative barriers (e.g., types of treatment covered; fail first requirements or prior authorization requirements). Furthermore, given Medicaid’s important role in ensuring OUD treatment (36), states should eliminate prior authorization requirements for buprenorphine and extended-release naltrexone covered by Medicaid programs.
Interestingly, we found that regulatory barriers were ranked lower than other barriers to buprenorphine prescribing, despite the existence of relatively unique regulations for buprenorphine prescribing such as patient limits and special education requirements. Possibly this is due to our sampling strategy, which oversampled physicians with a preexisting waiver to prescribe buprenorphine (approximately 40% of our sample) even though only approximately 2% of U.S. physicians have a waiver (20). Individuals who do not view buprenorphine prescribing regulations as a salient barrier may have self-selected into the group that has already obtained a waiver. Future studies should further examine the perception of regulatory barriers among a representative sample of physicians who have not yet obtained a waiver. Some previous studies may have oversampled physicians without a waiver; and physicians without a waiver may overestimate the difficulty of adhering to patient limits, completing special education requirements, and applying to SAMHSA for a waiver. Alternatively, over time, physicians may find it easier to meet regulatory requirements, especially as the availability of online education courses has increased. Also, the institutions in which physicians work may be increasing their support of buprenorphine prescribing over time, thereby giving physicians time and funds to complete the waiver process. Future studies should examine the impact of educational availability and institutional support on perceptions of regulatory barriers.
Our study found higher perceptions of efficacy in treating OUD for methadone and buprenorphine than for extended-release naltrexone. This discrepancy may be explained by greater awareness of methadone and buprenorphine (which were FDA-approved prior to extended-release naltrexone) and fewer published studies about extended-release naltrexone—a point noted by focus group participants. Recently, some studies have found similar efficacy between buprenorphine and extended-release naltrexone for OUD (29, 37), while another more recent study found lower efficacy of extended-release naltrexone in terms of overdose protection (9); but these studies were unavailable or very recent when we implemented our survey. Additionally, since many physicians in our sample reported that no one in their practice was prescribing extended-release naltrexone, they may also have limited experiential knowledge of the medication. In other words, even if prescribers know that extended-release naltrexone has demonstrated efficacy in controlled trials, they may be unsure about its effectiveness in real world settings. However, our interpretation is limited by the fact that we did not ask survey participants whether they are currently prescribing extended-release naltrexone. Finally, our participants may feel that extended-release naltrexone is less effective for patients who are not yet opioid-abstinent (as required for extended-release naltrexone) and are unwilling or unable to do so, even though the medication may be effective for patients in other practices who have already completed the detoxification process.
Participants believed that buprenorphine is slightly more effective than methadone at preventing opioid overdose, although the scholarly literature suggests that their efficacy is comparable, with methadone being slightly more effective at retaining patients in treatment than lower doses of buprenorphine (38). Respondents may simply have been less familiar with the literature about methadone and with real-world effectiveness of methadone, since they cannot prescribe it in office-based settings. Additionally, patients who seek OUD treatment in office-based settings may have stronger pre-existing preferences for buprenorphine than for methadone(39, 40), making providers in such settings less likely to seek out education about methadone or to refer patients to methadone treatment.
Although little has been written about the appropriateness of prescribing methadone, buprenorphine, or extended-release naltrexone for individuals with co-occurring mental health disorders, our participants believed that methadone and buprenorphine are more appropriate than extended-release naltrexone for dual diagnosis patients. Possibly participants are aware that depression is an adverse event associated with extended-release naltrexone in about 10% of patients (41). The literature on extended-release naltrexone’s efficacy in dual diagnosis patients may also be less developed because of its novelty in treating OUD and/or because participants are more risk averse to prescribing it. Furthermore, since individuals beginning extended-release naltrexone must be opioid abstinent for at least seven days, health practitioners may feel that this hurdle is too difficult for individuals with dual diagnosis to overcome. Given the correlation between OUD and mental health disorders (42), significantly more research is needed regarding the effectiveness of MOUD for individuals with dual diagnosis and barriers to prescribing MOUD for this population.
Non-waivered participants believed methadone is highly effective for pregnant women, but waivered participants had more negative beliefs about methadone’s effectiveness in this population (mean score 4.47 vs. 3.48, p < .0005). Likewise, non-waivered participants believed buprenorphine is highly effective for pregnant women, but waivered participants had more negative beliefs about buprenorphine’s effectiveness in pregnant women (mean score 4.42 vs. 3.58, p < .0005). Possibly the waivered survey participants do not routinely treat pregnant women for OUD and are thus more risk averse to using MOUD; however, we did not ask survey participants what percentage of their patient panels consists of pregnant women. Both methadone and buprenorphine are effective for pregnant women with OUD (43, 44). Therefore, education about methadone’s and buprenorphine’s efficacy in pregnant women should be part of courses for obtaining a SAMHSA waiver, especially in light of increasing rates of OUD in pregnant women and of neonatal abstinence syndrome (45).
Participants were significantly less likely to identify diversion and licensing board oversight as barriers to extended-release naltrexone prescribing than to buprenorphine prescribing. This result is not surprising, since misuse or diversion of extended-release naltrexone is unlikely, given its office-based administration and lack of a psychoactive ingredient. Nevertheless, even for buprenorphine, participants did not, on average, believe diversion and licensing board oversight were strong barriers to prescribing (mean score 2.2 and 1.7 out of 5). However, our study oversampled waivered physicians; and physicians who have sought and obtained a waiver may as a group be less likely to have diversion or oversight concerns than physicians who have not sought and obtained a waiver. Nevertheless, real-world experiences of those actually waivered to prescribe buprenorphine are important to the extent they reflect that diversion is not a high concern with this medication, to refute longstanding stigma.
No participants were implanting Probuphine®, likely reflecting the novelty of the medication. Even though we included questions about Probuphine® in our survey, due to sample size limitations, not enough data was gathered to assess specific barriers to its utilization. Future studies should explore the extent to which the REMS certification serves as a barrier to prescribing Probuphine®, as well as barriers associated with the need to stabilize patients on oral buprenorphine prior to Probuphine® administration. Additionally, future studies should examine barriers to Sublocade® prescribing. To date the vast majority of studies of buprenorphine accessibility have been conducted on oral and sublingual formulations only.
Our study has several limitations. Our survey response rate was small relative to the population sampled, likely because the incentives offered were small and because this population may be experiencing survey fatigue. Complementing our survey results with focus group data that support our findings and offer additional insights helps to contextualize the survey responses. Our final sampled population over-represented physicians with a buprenorphine waiver, so our results may represent a bias in favor of MOUD treatment and more moderate perceptions of barriers related thereto. However, this may suggest that once prescribers become waivered and prescribe MOUDs, that the actual barriers to this treatment for OUD may be less substantial than previously perceived. Also, regulatory barriers may be less prominent to MOUD prescribing than they were in earlier years, and provider beliefs about efficacy largely track the evolving evidence base. Finally, we did not ascertain whether respondents were currently prescribing extended-release naltrexone, a potential variable related to perceptions of efficacy and barriers.
Our findings suggest that there is room for improvement in OUD treatment education. For example, less is known about newer medications—especially implantable and injectable buprenorphine—and these are areas for further training. Also, MOUD treatment in pregnant women is not well understood and warrants additional training. Finally, persistent insurance barriers to MOUD prescribing, including prior authorization, continue to merit attention and parity enforcement from regulators. Public payers can act as market leaders in generously covering MOUD and complementary treatment such that prescribers and patients do not perceive these as significant obstacles to effective care.