Purpose

The ESP Coordinating Center (ESP CC) is responding to a request from VA’s Health Services Research and Development Service (HSR&D) for an evidence brief on barriers to and facilitators of use of buprenorphine and extended-release naltrexone for treatment of opioid use disorder (OUD). This review synthesizes evidence on the barriers and facilitators at the patient, provider, and health care system levels. Findings from this review will be used to inform prioritization of questions for a September 2019 State-of-the-Art (SOTA) conference.

Background

Mirroring national trends, the proportion of opioid overdoses among Veterans due to illicit opioids is increasing. Opioid overdose among Veterans increased from 14.5 per 100,000 person-years in 2010 to 21.1 per 100,000 person-years in 2016, mostly due to overdoses from heroin and synthetic opioids such as fentanyl rather than prescription opioids.¹ This finding highlights the need for opioid risk reduction efforts beyond those aimed at increasing opioid safety in the in the healthcare setting. As a component of the Veterans Health Administration’s (VHA) overall response to the crisis of opioid-related morbidity and mortality, VHA provides treatment for those with opioid use disorder (OUD), a diagnosis made based on symptoms, such as drug cravings, and behaviors, such as compulsive drug seeking and drug use despite negative consequences.² Methadone, buprenorphine, and extended-release naltrexone are medications approved by the Federal Drug Administration (FDA) for treatment of OUD. Of note, the term medication-assisted treatment (MAT) has been commonly used to describe these medications but is no longer favored given that medication is considered an option for first-line treatment for OUD (rather than an adjunctive treatment).³ Guidelines from the Veteran’s Administration/Department of Defense (VA/DoD) recommend offering patients with OUD treatment with medication. These guidelines state that while patients receiving methadone should be offered individual counseling and/or contingency management (rewards to reinforce positive behaviors), there is insufficient evidence to recommend for or against any specific psychosocial intervention in addition to office-based medical management with buprenorphine.⁴

Methadone and buprenorphine treat OUD by reducing opioid cravings and preventing withdrawal, which are both potent drivers of ongoing opioid use. Methadone is a “full opioid agonist,” meaning that it activates opioid receptors in the brain and body in the same way as other prescription opioids and illicit opioids such as heroin. Buprenorphine is referred to as a “partial opioid agonist” because it does not act on opioid receptors in the same way as full agonists like methadone. Specifically, it binds tightly to the mu opioid receptor (which is responsible for analgesic and euphoria effects but also harmful effects such as respiratory depression) but does not activate this receptor as much as full agonists like methadone. It also has a ceiling effect, meaning that the receptor is not more activated with more exposure to the drug, and it is therefore thought to have a lower risk of overdose.⁵

Buprenorphine used for OUD treatment is a higher dose than buprenorphine formulations approved to treat chronic pain and is typically co-formulated with naloxone, which helps deter misuse by causing opioid withdrawal if the medication is crushed and injected or snorted instead of being used under the tongue as prescribed. Buprenorphine is also available as an implantable medication and an extended-release injection form, although currently neither of these options are widely used or available on VA formularies.⁶

Naltrexone is an “opioid antagonist,” meaning that it blocks opioid receptor activity. Naltrexone is available in an injectable form and oral form, although treatment retention and adherence may be lower with the oral form and it is not as commonly used.^4,7 Extended-release naltrexone is the preferred option for patients who would like to avoid taking any form of opioid or for whom methadone and buprenorphine are contraindicated (such as with severe alcohol use). Naltrexone works differently than methadone and buprenorphine to treat OUD – by blocking the effects of opioids, it has been shown to reduce cravings over time and increase opioid abstinence.^2,8

Several factors determine which type of medication may be recommended for an individual patient, including his or her treatment preferences, unique history of opioid use and other substance use, co-morbidities, living situation and social history, and risk of relapse and other adverse outcomes. Similar to other chronic medical conditions, patients with OUD may need a higher or lower level of care when they are first diagnosed, and this need may fluctuate at different times during their disease course. The highest level of care for OUD is delivered in “Opioid Treatment Programs” (OTPs), which is a specific term used to denote a treatment program certified by the Substance Abuse and Mental Health Services Administration (SAMHSA).⁹ OTPs can exist in multiple settings including intensive outpatient programs, residential programs, and hospital settings. OTPs are highly regulated, with requirements that patients receive psychosocial supports, complete frequent urine drug tests, and receive a limited number of take-home medications (thereby requiring frequent visits).⁴ As of 2018, VHA operated 32 OTPs nationwide, but also contracts with an unknown number of OTPs in the community.⁶

In addition to OTPs, OUD care is delivered in “office-based” settings, a term that refers to outpatient settings other than OTPs, including primary care and general mental health clinics.¹⁰ In some health systems, these levels of care are linked. For example, in the “hub and spoke” model developed in Vermont, OTPs are connected with office-based care settings and patients are triaged to the level of care that best meets their needs.¹¹ Intensity of care delivered in office-based settings varies by practice structure and the scope of available psychosocial services.¹¹ Office-based care may be viewed as more patient-centered and accessible and preferred by some patients compared to OTPs.¹² Incorporating OUD treatment into routine chronic disease management, similar to conditions commonly treated in primary care such as asthma, diabetes, and hypertension, may help reduce the stigma associated with opioid use and addiction.¹³ Stigma may be defined as “the general public identifying and labeling differences between themselves and a specific group [and] conferring these differences with negative attributes.”¹⁴

While methadone can only be used in the OTP setting due to federal regulations, federal legislation over the past 20 years or so has made it possible for buprenorphine to be prescribed in the office-based setting. VHA guidelines and program changes have also helped promote buprenorphine and naltrexone use. See Figure 1 for a timeline of federal legislation and VHA changes related to OUD medication prescribing, including the addition of extended-release naltrexone to VA formularies. The first legislative milestone to expand access to OUD medication was the Drug Addiction Treatment Act of 2000 (DATA 2000), which allowed providers to prescribe buprenorphine in the office-based setting after completing an 8-hour training, obtaining a SAMHSA waiver, and obtaining a specially designated Drug Enforcement Agency (DEA) registration number. Initially, DATA 2000 set a cap on prescriptions of 30 patients per physician, with the option to increase to 100 patients after a year. However, the 2016 Comprehensive Addiction and Recovery (CARA) Act increased the cap for experienced providers to 275 patients and expanded the waiver option to nurse practitioners (NPs) and physician assistants (PAs), although many states do not allow NPs and PAs to prescribe buprenorphine without oversight by a physician who also has a waiver.¹⁵ Because naltrexone is not an opioid and has no abuse potential, it is not subject to additional federal regulations and may be prescribed in any setting just as other non-controlled medications.

Figure 2

Timeline of federal and Veterans Affairs events affecting the availability and use of medications for opioid use disorder.

Despite evidence of the effectiveness of medications as first-line treatment for OUD, even without psychosocial interventions, and guidelines from the Veterans Affairs/Department of Defense (DoD), Centers for Disease Control & Prevention (CDC), and the American Society of Addiction Medicine recommending their use, these medications are underutilized within VHA populations and the general community.^4,16,17 Within VHA, the percentage of Veterans with diagnosed OUD on medication was approximately 39% at the end of the second quarter of 2019.⁶ While improved compared to treatment rates of approximately 34% in 2016, this rate of OUD medication use is still far short of what is needed given the scope of the opioid crisis.

Reasons why OUD medications are still underutilized even several years into the opioid crisis are not fully understood, but are likely due to multiple interconnected factors at the patient, provider, and systems levels. Patients and providers can both experience barriers at several places along the pathway towards starting OUD medication. Patients must first recognize and accept a diagnosis of OUD (either self-recognition or recognizing and accepting a diagnosis made by a provider), decide to seek treatment, identify a provider with prescribing capacity who accepts the patient’s insurance and/or has affordable out-of-pocket fees (in non-VA settings), navigate transportation or child care needs, and start treatment, which involves ceasing illicit opioid use. From the provider perspective, use of buprenorphine presents several potential barriers. Providers must be interested and willing to prescribe buprenorphine, work in a practice setting that supports buprenorphine prescribing, complete training and obtain a SAMHSA waiver and updated DEA registration, possibly update their hospital/clinic credentials, educate staff regarding buprenorphine inductions (process for starting the medication), ensure adequate clinic space, and navigate insurance prior authorization requirements (in non-VA settings) before writing a prescription. Low reimbursement for OUD treatment (in non-VA settings) may also be a barrier.

A qualitative study based on interviews with a national sample of administrative and clinical staff at VHA sites in 2006–2007 found that provider barriers to prescribing buprenorphine included lack of education regarding buprenorphine treatment, negative perceptions of patients with OUD, perceived lack of resources, and the thought that OUD care was best delivered outside the VA.¹⁸ While it is possible that many of these same barriers still exist, the opioid crisis itself has changed since 2006–2007, with a higher prevalence of heroin and synthetic opioid use.¹⁹ Societal attitudes towards opioids and people with OUD may have also shifted in the past decade with greater public awareness of the problem. Understanding current barriers and facilitators to OUD medication use, and which barrier and facilitator types are most relevant to the VHA patients and providers, is necessary to design and evaluate interventions to increase prescribing capacity and patient uptake. The aim of this review is to synthesize the evidence on provider, patient, and system-level barriers and facilitators to use of OUD medications to help inform VHA stakeholders engaged in policy development, program planning, and OUD research. Because buprenorphine and extended-release naltrexone can be prescribed in office-based settings and have the most potential for widespread use, this review is limited to those 2 medications rather than methadone. We also focus our discussion on barriers that VHA can address, rather than federal regulatory barriers outside of VHA’s control.

Scope

This rapid evidence review addresses the following key questions and eligibility criteria:

Eligibility Criteria

We included studies of all types (quantitative and qualitative) of adults eligible for OUD treatment (excluding pregnant women) in the US. We included studies of initial OUD medication use (not treatment retention). Included studies had to have a specific aim of identifying barriers and facilitators or factors associated with OUD medication use or prescribing (such as patient demographics or provider characteristics). Full inclusion and exclusion criteria are available in Appendix A.

Searches and Study Selection

To identify articles relevant to the key questions, our research librarian searched MEDLINE, PsycINFO, CENTRAL, Cochrane Database of Systematic Reviews, and EMBASE from database inception to March 2019 using terms for opioid use disorder, buprenorphine, naltrexone, barriers, and facilitators (see Supplemental Materials Appendix B for complete search strategies). Additional citations were identified from hand-searching reference lists and consultation with content experts. We limited the search to published and indexed articles involving human subjects available in the English language. Study selection was based on the eligibility criteria described above. Due to the large volume of studies we identified in our search, the need to accommodate a compressed rapid review timeline, and the goal of providing VHA clinicians, researchers, and policy-makers with the most current and relevant synthesis of evidence on OUD medication barriers and facilitators, we limited selection of studies to those published since 2014. Limiting our study selection in this way allowed us to prioritize evidence in the context of recent VHA and federal policy changes including the addition of extended-release naltrexone to the VHA formulary in 2014, updated VHA substance use disorder guidelines in 2015, and the 2016 CARA Act aimed at expanding OUD medication prescribing. Titles and abstracts and full-text articles were reviewed by one investigator and checked by another. All disagreements were resolved by consensus or review from a third investigator.

Quality Assessment & Data Extraction

We took a categorical approach to assessing study quality. This approach involved evaluating whether each study met minimum methodological quality standards to distinguish the subset of studies that adequately performed the most critical methods from those that did not. Using predefined criteria from existing tools,^20,21 we focused on assessing adequacy of the following features that are common across study types (qualitative and quantitative) and most critical for interpreting findings: 1) sampling methods, 2) adequacy of survey or interview questions in capturing the desired information, 3) population descriptions, 4) setting descriptions, and 5) barrier/facilitator detection methods. For quantitative studies, we also assessed 6) whether appropriate statistical analyses were conducted, and for qualitative studies we assessed 7) whether the study used a formal process for recording, transcribing, and coding themes from interviews or open-ended responses. We then categorized studies as either “met minimum criteria” or “did not meet minimum criteria.” Additionally, we informally commented on the status of additional study features that can help in identifying the highest-quality studies, including whether survey studies compared responders to non-responders, or whether qualitative studies had independent analysis of data by more than one researcher or examined the potential effect of the author’s background, education, or perspective on the results.

We abstracted data from all studies, including study characteristics, setting, population, type of medication discussed, and barriers/facilitators identified. All data abstraction and quality assessment was first completed by one reviewer and then checked by another. All disagreements were resolved by consensus. Full data abstraction and quality assessment tables are available in Appendix D.

Strength of Evidence Assessment

We informally graded the strength of the evidence as a whole based on the AHRQ Methods Guide for Comparative Effectiveness Reviews.²² This approach incorporates 4 key domains: risk of bias (includes study design and aggregate quality), consistency, directness, and precision of the evidence. It also considers other optional domains that may be relevant for some scenarios, such as strength of association (magnitude of effect), and publication bias. While AHRQ’s strength of evidence tool is designed to evaluate intervention studies, we felt the principles could be applied to this evidence base of non-intervention studies. Strength of evidence ratings generally range from high to insufficient, reflecting our confidence that the evidence reflects the true effect. For this review, we applied the following general algorithm: evidence comprised of qualitative studies and uncontrolled observational studies with consistent and precise findings received a rating of “low”; whereas the same type of evidence with few studies and/or indirectness and inconsistency would be downgraded to “insufficient.”

Synthesis of Data

We synthesized the evidence qualitatively but, due to heterogeneity, did not perform a quantitative synthesis. We prioritized synthesis of studies of Veterans populations and/or those conducted in VHA settings. When VHA studies were not available, we focused on studies with the most applicability to VHA by prioritizing 1) studies of patients with OUD (reasoning that Veteran and non-Veteran patient populations with OUD have much in common) and 2) studies of providers in primary care and other office-based settings (given the VHA’s aim to expand OUD medication use widely and not just in OTPs). As much of the data was qualitative (open-ended surveys and interviews), we conducted content analysis of barriers and facilitators in prioritized studies using coding based on an iterative process without pre-defined categories. We considered a barrier/facilitator to be any factor endorsed by at least 1 participant in the study that either inhibited or helped them adopt medication treatment or that theoretically would inhibit or help them adopt medication treatment (as opposed to assigning a cut-off point for the percentage of participants that would need to endorse it for it to be considered a barrier). As we reviewed studies, we grouped content into existing categories or created a new category. We later identified sub-categories within each of the major categories. All codes were first conducted by one reviewer and checked by a second.

A draft version of this report was reviewed by peer reviewers as well as clinical leadership (see supplemental materials for disposition of peer review comments). The complete description of our methods can be found on the PROSPERO international prospective register of systematic reviews (http://www.crd.york.ac.uk/PROSPERO/; registration number CRD42019133394).

Literature Flow

The literature flow diagram (Figure 2) summarizes the results of the search and study selection processes. Among 2,753 potentially relevant citations, we included 26 studies published since 2014.^7,23–46 Of these, we prioritized evidence synthesis of 16 studies with the most applicability to VHA, as discussed in more detail below.^{7,28,32,34–47} See supplemental materials for list of excluded studies and full data tables.

Figure 3

Literature Flowchart.

KEY QUESTION 1. What are the patient, provider, and systems-level barriers and facilitators to use of buprenorphine and extended-release naltrexone for OUD?

Patient-identified Barriers to OUD Medication Use

Patient-identified barriers to buprenorphine use (none examined naltrexone use) may be categorized into 4 main categories – stigma, logistics, treatment experiences (current and prior) and beliefs, and knowledge of OUD medications. These barrier types are likely overlapping and mutually reinforcing. Of these, social stigma associated with OUD and high out-of-pocket costs were the most commonly described barriers. See Table 1 for a summary of patient-identified barriers and Table 2 for an overview of barrier categories and subcategories.

Stigma

Studies discussed several forms of stigma including social stigma, manifesting as not wanting to be seen or associated with OUD treatment, and self or internalized stigma, manifesting as shame related to having an addiction and needing treatment.³⁸ Another form of stigma is related to treatment specifically. For example, stigma associated with buprenorphine can manifest as the view that using buprenorphine is “a crutch” and that people on buprenorphine are not really sober.^32,38,47 These forms of stigma may affect patients with OUD along the entire care continuum, including their ability to accept an OUD diagnosis and willingness to engage in treatment, and were discussed in all studies of patient-identified barriers to OUD treatment.

Logistics

Logistical barriers were also common and include high out-of-pocket costs including insurance copays and costs associated with “cash-only” providers who do not accept insurance, finding a buprenorphine provider, provider waiting lists, the need to “first fail” abstinence-based treatment prior to receiving buprenorphine, and having access to transportation and childcare to attend treatment visits. Of these logistical barriers, cost was identified in most (4/5) studies of patient-identified barriers. The exception was a study of former inmates with OUD in New York City, 90% of whom had Medicaid coverage of OUD medication.⁴⁷

Prior treatment experiences, beliefs, and knowledge

Beliefs about OUD treatment generally and prior experience with buprenorphine/naloxone and/or other OUD treatment were also identified by patients as barriers to use of OUD medication. In a qualitative study based on interviews with previously incarcerated individuals at addiction treatment centers, some participants stated that a reason why they had not sought OUD treatment in the past was a belief that personal characteristics such as willpower and readiness to change are more important to opioid abstinence than medical treatment.⁴⁷ Two studies based on interviews with patients with a history of OUD treatment (one of patients in New York City at a syringe exchange program and the other in Alaska) discussed negative past experiences with treatment, including being treated poorly by treatment center staff, not feeling supported by staff, or not trusting staff.^32,38 Lack of knowledge about where to find treatment was a less commonly cited barrier, discussed in only 1 study in which 50% of patients surveyed at a syringe exchange program did not know where to access buprenorphine.³² While use of illicit buprenorphine was more often discussed as a facilitator, as described below, some participants in a study based on interviews with former inmates with OUD described negative experiences with illicit use of buprenorphine, which study authors speculate may have may been due to use of illicit buprenorphine concomitantly with other opioids and therefore precipitated withdrawal, underscoring the need for greater access to prescribed buprenorphine and proper instruction on how to use it.⁴⁷

Table 1

Patient-identified barriers to OUD medication use.

Patient-identified Facilitators to Use of OUD Medications

The most commonly cited facilitator of receiving buprenorphine was receiving personal support from others, including peers who had used OUD medication, family, and treatment providers. A study based on interviews of patients with OUD in Alaska highlighted participation in a structured program as a facilitator (although participants also noted that restrictive programs structures may also be perceived as a barrier later in the treatment process).³⁸

Illicit use of buprenorphine as a treatment facilitator

Three studies discussed use of illicit buprenorphine as a facilitator to seeking treatment with prescribed buprenorphine.^32,36,47 In a survey of syringe exchange program participants in New York City overall interest in buprenorphine was greater among those with a history of illicit buprenorphine use versus those without (54% vs 36%), although this difference was not statistically significant.³² The perceived likelihood of starting prescribed buprenorphine if it was offered as part of harm reduction agency services was also higher among those who had used illicit buprenorphine versus those who had not (82% vs 50%). These findings are consistent with another survey of patients in OUD treatment (N=303) indicating that among those who had used illicit buprenorphine, this use was intended for prevention of withdrawal symptoms (79%) or to maintain abstinence from other drugs (67%).³⁶

These findings suggest that providers who cite concerns about buprenorphine diversion (selling or giving away prescribed medication) as a barrier to prescribing it could benefit from knowing how diverted buprenorphine is used, in case they would be more likely to prescribe it in the context of harm reduction. Interest in illicit buprenorphine to manage withdrawal or maintain abstinence is likely due to insufficient access to prescription treatment. It is possible that with increased access to prescribed buprenorphine, illicit use may decrease.³²

Table 2

Categories of patient-identified barriers and facilitators to OUD medication use.

Provider-Identified Barriers to Prescribing OUD Medications

Similar to patient-identified barriers to using OUD medications, provider-identified barriers to prescribing OUD medications be categorized into 4 main types – stigma related to OUD, logistical barriers, treatment experiences and beliefs, and knowledge gaps. Just as with patient-identified barriers, the barriers affecting providers’ willingness and ability to prescribe buprenorphine are likely overlapping and mutually reinforcing. See Table 3 for a summary of provider-identified barriers and Table 4 for an overview of barrier categories and subcategories.

Stigma

Four studies of providers identified barriers to OUD medication prescribing related to stigma.^6,16,34,37 Similar to patient-identified barriers related to stigma, providers described social stigma related to OUD manifesting as lack of desire to be known as an OUD treatment provider⁷ and concerns about attracting more patients with drug use to their practice.³⁴ In a study by Huhn et al based on a survey of US physicians through the American Society for Addiction Medicine and American Medical Association Listervs, most providers (87%) had a buprenorphine waiver, but among those who did not have a waiver, 29% cited concerns about being “inundated” with requests for buprenorphine as a reason for not prescribing it.³⁹ It is possible that stigma related to patients with OUD, which can also manifest as the perception that patients with OUD are “difficult” or not trust-worthy,^7,37 contributes to concerns about requests for treatment from a large number of patients. However, concerns about time may also underlie the fear of being “inundated” with buprenorphine requests. The Huhn et al study does not provide enough detail to know which factors drive a lack of desire to treat high volumes of patients with OUD.

Also similar to patient-identified barriers to OUD medication use, 2 studies of providers identified stigma specific to buprenorphine, specifically the view that it is “substituting one addiction for another”^7,37 and the Huhn et al study discussed above found that reasons why waivered providers were not prescribing at capacity included lack of belief in buprenorphine therapy. Further explanation is not provided, so it is not known whether a lack of belief in buprenorphine is based on experience prescribing it or view that it is “substituting one addiction for another” or another unspecified reason.³⁹

Logistics

Overall, studies of provider-identified barriers to OUD medication prescribing found that logistical barriers are most common. Time was identified as a barrier in 7/7 studies and was the first or second most commonly cited barrier in 5 of these studies. Although no studies were conducted in VHA settings, identification of time as a barrier to OUD medication prescribing is still applicable to VHA settings given that concerns about time and workload are common among frontline providers.⁴⁹ As discussed above in the section on stigma, these concerns may be exacerbated by negative views of patients with OUD among some providers. For example, the perception that patients with OUD are “difficult,” which was cited in 2 studies^7,37 (one based on physician interviews and the other a survey with open-ended questions), may make providers even more concerned about the time commitment required to take on buprenorphine prescribing for these patients and make them fear being “inundated” with buprenorphine requests.³⁹

Other logistical barriers include concerns about insurance coverage, prior authorizations, and reimbursement. While not directly applicable to VHA settings, these barriers apply to VHA contracted care providers, and may impact VHA’s ability to find community providers to partner with, particularly in rural areas. A study evaluating how often providers with buprenorphine waivers in Ohio accept insurance payments found that among active prescribers, 52.7% accepted insurance for buprenorphine, 20.8% accepted insurance for non-buprenorphine services but not for buprenorphine, and 26.5% did not accept insurance for any services (ie, accepted cash only).²⁸ The findings of this study highlight the potential that access may be much more limited than numbers of waivered providers might suggest, and also help explain why patients without means to pay out-of-pocket for treatment may encounter even steeper barriers to treatment than others.

Knowledge gaps and beliefs about OUD medication

Knowledge gaps regarding OUD medications and OUD generally were also cited as barriers and are likely to impact VHA providers in the same way as providers in other practice settings. In a qualitative study by Andraka-Christou et al of 20 US-licensed physicians (mostly psychiatrists and primary care providers) in 4 states, 75% of participants cited lack of addiction education in medical school and residency as a barrier.⁷ Lack of knowledge also translates into low confidence in buprenorphine use, cited as a barrier in a large survey of 1,124 rurally located providers with buprenorphine waivers.³⁴ Lack of training regarding buprenorphine may also manifest as a perception that buprenorphine is not effective or that patients have a low need for it.^7,37 In a survey of clinicians obtaining an initial buprenorphine waiver or an increase in authorized patient limit, lack of patient demand for treatment was cited as the primary barrier to prescribing. It is unclear whether this lack of demand is based on provider perception, which could indicate a knowledge gap, beliefs that patients with OUD do not need (and are therefore not requesting) buprenorphine, or other subjective or objective factors.⁴¹

The qualitative study by Andraka-Christou et al discussed above was the only one that examined barriers specific to buprenorphine and extended-release naltrexone (and the only study overall to discuss barriers and facilitators to naltrexone).⁷ In this study, 20 US-licensed physicians identified 3 main barriers to buprenorphine prescribing – regulatory requirements, fear of liability, and restrictions imposed by the criminal justice system (such as not allowing or discouraging buprenorphine use). Barriers related to extended-release naltrexone were more specific to its pharmacologic properties. Because naltrexone is an opioid antagonist, patients must completely cease opioid use before starting it (to avoid precipitating acute opioids withdrawal) and physicians in this study expressed concerns about patient willingness to take this step without access to medically-supervised detoxification programs. Another barrier specific to naltrexone in this study was lack of awareness of its use in OUD treatment.

Concerns about buprenorphine diversion

In 5 of 7 studies, providers described buprenorphine misuse and diversion as barriers. It is unclear whether these concerns are primarily driven by negative views of patients with OUD, concerns about patient safety and the potential for overdose and other adverse outcomes, or concerns about liability and DEA intrusion (cited as barriers in 2 studies^7,34).

Barriers to prescribing at capacity

Three studies found that providers with buprenorphine waivers were not prescribing at capacity.^39,41,42 Only one of these studies, the Huhn et al study based on a survey of providers on the American Society for Addiction Medicine and American Medical Association Listervs, examined barriers specific to waivered providers who were not prescribing at capacity. In this study, 56.2% of waivered providers (N = 272) reported not prescribing to capacity (providers with patient caps of 30 and 100) and the most commonly cited reason was not having time for more patients.³⁹ More than half (54.8%) of survey respondents with a buprenorphine waiver who were not prescribing to capacity indicated that nothing would increase their willingness to prescribe at that level. Reasons for this lack of willingness based on statistical analysis by study authors included lack of belief in agonist treatment (also discussed above in the section on stigma), lack of time for additional patients, and beliefs that reimbursement rates were insufficient.

Table 3

Provider-identified barriers to buprenorphine use.

Provider-identified Facilitators to OUD Medication Prescribing

Two studies described provider-identified facilitators to prescribing buprenorphine, including a survey by Jones et al of more than 4,200 providers obtaining an initial buprenorphine waiver or applying for an increased prescribing limit. In this survey, the most commonly cited facilitator was increased patient demand for buprenorphine (22%), followed by institutional support (12.5%), increased reimbursement (12.2%), having an integrated system with specialty linkages (9.3%) and an easy referral process (8.3%), and having an addiction medicine mentor (7.7%). In this same study, lack of patient demand was identified as the primary barrier to buprenorphine prescribing. How these 2 findings – lack of demand is a barrier and increased demand is a facilitator – relate to each other and manifest in clinical practice is unclear. For example, it is not known whether providers perceive increased patient demand for buprenorphine because patients are directly asking for it, or whether providers become more aware of population-level OUD treatment needs and therefore perceive that patient demand is higher.

Although it was the least commonly cited facilitator in the study by Jones et al, the concept of having a mentor was similar to findings of another provider survey identifying “being paired with an experienced provider” as a facilitator.³⁹

Table 4

Categories and subcategories of provider-identified barriers and facilitators to use of OUD medications.

KEY QUESTION 2. Do these barriers and facilitators vary by patient characteristics, provider characteristics, or setting?

Patient Characteristics

The 5 patient studies based on surveys and/or interviews did not perform subgroup analysis. Therefore, we are not able to draw conclusions about how patient-identified barriers and facilitators may vary among different patient groups. We identified 3 retrospective studies that evaluated the likelihood of patients receiving buprenorphine – one based on data from the National Ambulatory Medical Care Survey and the National Hospital Ambulatory Medical Care Survey (which provide nationally representative estimates of ambulatory care provided in the US by non–federally employed physicians) and 2 based on chart review.^44,45,48 These studies of the predictors of buprenorphine use provide indirect evidence on how barriers and facilitators to OUD medication use may vary among patient groups. See Table 5 for study characteristics and findings.

Studies consistently found that adults in the age range 30–50, white patients, and those who self-pay or are employed are more likely to be prescribed buprenorphine that those who are on the extremes of age and non-white. In addition to these factors, predictors of receiving buprenorphine included living in an area with an increase in buprenorphine prescribers,⁴⁴ living in a metropolitan area,⁴⁴ having a point-of-service insurance plan allowing a choice of providers,⁴⁴ having chronic pain,⁴⁴ and history of incarceration.⁴⁵

In a retrospective chart review of patients in a large non-profit integrated health system serving a substantial portion of the population in Washington and Northern Idaho (Group Health), older age and co-occurring alcohol and non-opioid drug dependency were associated with a higher likelihood of not receiving buprenorphine.⁴⁴ This finding may reflect a clinical practice of not using buprenorphine in more complex patients, although the study does not describe enough detail regarding treatment settings (non-specialized outpatient settings vs OTPs) to know whether patients were more likely to receive buprenorphine treatment if they were referred to higher levels of care (OTPs) or whether some patients were offered treatment but did not start it. Another retrospective study of patients receiving care at a primary care clinic in Washington found that recent polysubstance use and prior experience with OUD medications were predictors of not receiving buprenorphine.⁴⁵ This study specifically evaluated patients during the time between being offered buprenorphine treatment and starting treatment and suggests that more complex patients (indicated by polysubstance use) with a longer history of OUD (indicated by prior buprenorphine use) may face unique barriers to starting buprenorphine.

Our confidence in these findings is low given the small number of studies, observational study designs with potential for unmeasured confounders including the influence of provider attitudes towards buprenorphine treatment, and indirect measures of barriers (characteristics associated with receiving buprenorphine rather than direct measurement of a barrier). Additional studies are needed before concluding that the patient characteristic associated with receiving buprenorphine are representative of similar patterns in VHA settings. Nonetheless, these studies provide insight into patient characteristics that may make certain populations more or less likely to receive OUD medications that would be difficult to evaluate by asking patients directly, and the consistency of the findings suggests an area to focus future research.

Table 5

Predictors of receiving or not receiving buprenorphine.

Provider Characteristics and Setting

Similar to studies of patients, studies of providers did not perform subgroup analyses so we are not able to draw conclusions about how provider-identified barriers and facilitators may vary among different groups. However, findings from 3 studies of provider characteristics associated with buprenorphine prescribing provide indirect evidence on how barriers and facilitators to OUD medication use may vary among providers.

Three general findings emerge from these studies (Table 6). The first is that prescribing behavior is associated with provider-identified barriers and facilitators. For example, 1 large survey of providers found that those reporting resistance from practice partners or staff or lack of institutional support were less likely to prescribe buprenorphine (aOR = 0.14, 95% CI = 0.10–0.21).⁴¹

The second finding is that buprenorphine prescribing and presumed underlying barriers and facilitators may vary by region. For example, one survey of providers applying for a buprenorphine waiver or requesting an increase in patient caps found that practicing in the Midwest compared to the South (aOR = 1.35, 95% CI 1.03–1.78) predicted buprenorphine prescribing.⁴¹ Regional differences were also highlighted in a survey of rurally located physicians on the DEA list, which identified a greater lack of confidence among physicians in the Pacific Census Division versus other regions (22.0% vs 7.7%) and higher rates of concern about attracting drug users to their practice among physicians in the East and West South Census Divisions versus other regions (42.7% vs 27.8%). These regionally specific barriers may inform local intervention efforts.

The third finding, based on a study by Jones et al of providers applying for a buprenorphine wavier or requesting an increase in patient caps, suggests that providers who engaged in buprenorphine training and/or using other best practices in OUD treatment (such as co-prescribing naloxone) are more likely to prescribe buprenorphine.⁴¹ These findings suggest that as providers learn about and implement OUD treatment best practices, rates of buprenorphine may increase.

Notably, all of these surveys were of physicians and no studies evaluated practice patterns among NPs or PAs, who represent an important component of the workforce able to treat patients with OUD.

Table 6

Provider characteristics associated with buprenorphine prescribing and predictors of prescribing.

Barriers to Use of OUD Medications

Studies using mostly qualitative methods suggest that patients and providers experience 4 main types of barriers related to using or prescribing OUD medication – stigma, logistical barriers, treatment experience and beliefs, and knowledge gaps. Of these, stigma and logistical barriers are most common. Among patients, stigma was described as a barrier in all studies and primarily manifests as not wanting to be associated with OUD treatment, which may lead to reluctance to attend addiction-specific programs and clinics (compared to primary care or general mental health clinics). As Hewell et al point out, stigma may also be internalized so that patients are not comfortable admitting that opioid use is a problem or that they have sought OUD treatment.³⁸ Among providers, stigma manifests as not wanting to be perceived as having a practice with a high prevalence of patients with OUD and negative beliefs about patients with OUD, including the perception that patients are “difficult” or not trustworthy. Although not captured in our included studies, it is possible that stigma accounts for systems-levels barriers as well as patient- and provider-level barriers. For example, the reluctance of a practice to be known as offering treatment for OUD may underlie lack of leadership or institutional support for providers who are interested in prescribing OUD medications.

In addition to stigma, barriers related to logistics were common. Among patient studies, high out-of-pocket costs were discussed 4/5 studies. The exception was a study of former inmates with OUD in New York City, where OUD medications are covered by Medicaid. Challenges with finding a provider who is accepting new patients and does not have a long wait list were other commonly identified barriers. For providers, concerns about time were the most commonly cited logistical barrier and discussed in all 7 studies. The perception that patients with OUD are more challenging to care for and concerns about time are likely mutually reinforcing barriers. Interestingly, these findings are similar to results of a qualitative study based on interviews with a national sample of administrative and clinical staff at VHA sites in 2006–2007 which found that ill perceptions of opioid-dependent patients and perceived lack of resources were common barriers.¹⁸

Studies based on physician surveys and chart review suggest that certain patient populations are more or less likely to be prescribed buprenorphine and indirectly suggest that barriers vary by patient groups. Specifically, adults in the age range 30–50, white patients, and those who self-pay or are employed are more likely to be prescribed buprenorphine that those who are on the extremes of age and non-white. Patients with non-opioid substance use, including alcohol use, were less likely to receive buprenorphine or less likely to start it once it was offered. While some of these differences likely reflect a clinical assessment that use of buprenorphine would be potentially unsafe in patients using other substances, it is also likely that practice trends and norms for who is offered buprenorphine vary by provider, practice setting, and possibly geographic region. As no studies were conducted in VHA populations, additional research is needed to evaluate patient characteristics related to buprenorphine use within VHA.

Facilitators to Use of OUD Medications

Support from peers, family, and treatment providers was the most common facilitator for patients. This support may help to reduce the social stigma or internalized stigma associated with OUD and seeking treatment. Among providers, limited information regarding facilitators of OUD prescribing are available and one factor did not stand out as being most important. Provider-identified facilitators include increased patient demand for OUD medication, mentoring and linkages to addiction medication specialists, receipt of OUD and buprenorphine training, ability to refer patients to psychosocial support services, and institutional support.

VHA Strengths and Opportunities for Intervention

The barriers discussed above can overlap and mutually reinforce each other, but given that, it is also plausible that efforts to reduce one barrier may lead to improvements in others. For example, addressing a knowledge gap by providing buprenorphine waiver training also addresses a logistical gap. Similarly, reinforcing with providers that office-based treatment with buprenorphine is an option for first-line treatment and does not require psychosocial supports other than those provided in the context of medication management may reduce providers’ concerns about lack of access to mental health services. A 2019 consensus report from the National Academies of Sciences, Engineering, and Medicine states, “Lack of availability or utilization of behavioral interventions is not a sufficient justification to withhold medications to treat opioid use disorder.”³ As another example, if providers are aware that many patients with OUD are using illicit buprenorphine to maintain abstinence from other opioids or treat withdrawal and that prior experience with illicit buprenorphine may increase patients’ interest in seeking treatment, providers may have fewer concerns related to diversion from a harm-reduction perspective.

VHA may be well-structured to intervene to reduce knowledge gaps, as multiple educational resources are available to VHA providers including team-based trainings, online and video-based training, and support from clinical pharmacy and academic detailing. Likewise, because VHA is an integrated care system, providers working in VHA settings may not experience the same barriers related to lack of access to mental health and addiction medicine specialists as providers working in other healthcare systems or clinic settings. It is possible that connections with those who have more experience with OUD treatment may reduce concerns about time and effort to prescribe OUD medications. A recent perspective piece on “myths” associated with buprenorphine use addressed this idea with a statement about the “reality” of prescribing: “Treating patients with buprenorphine can be uniquely rewarding. In-office inductions and intensive behavioral therapy are not required for effective treatment.”⁵⁰

Regulatory barriers, specifically the requirement for providers to complete buprenorphine-specific training and obtain a SAMHSA waiver to prescribe it, are common to all US providers, and discussing efforts to change federal regulations is beyond the scope of this review. Nonetheless, VHA facilities may be able to reduce internal barriers to obtaining a waiver, such as making training available and free to providers during the workday and reducing VHA-specific administrative barriers such as the need to update hospital and clinic credentials prior to prescribing buprenorphine. While the waiver requirement is a commonly cited barrier, the finding that many providers with waivers are not prescribing at capacity reflects the need to reduce barriers to not just obtaining a buprenorphine waiver, but also to starting to prescribe buprenorphine and take on a higher patient load.

Common facilitators of OUD medication use for both patients and providers include support from peers. This theme highlights the potential for community – peer supports for patients and communities of practice for providers – to overcome some of the perceived barriers to OUD medication use. Notably, in a study of 108 family physicians, only 10% of whom prescribed buprenorphine, nearly three-quarters (73%) reported that they felt a personal responsibility to treat opioid addiction.³⁷ Capitalizing on a sense of responsibility to treat OUD among providers on the frontlines of the opioid crisis, paired with peer and institutional supports, has the potential to expand the number of providers who prescribing OUD medications and make it a routine part of their practice.

Limitations

Gaps and Future Research

This review highlights several important gaps in evidence on barriers/facilitators to increase use of OUD medications. Some of these gaps are specific to the VHA setting, and others are general evidence gaps.

While we do not expect that future studies in the VHA setting would uncover additional barriers and facilitators that do not relate to one of the themes highlighted in this report – stigma, logistics, knowledge gaps, and beliefs and attitudes related to OUD – studies in VHA populations would increase certainty regarding which barrier and facilitator subthemes are most significant for VHA populations and relevant to VA systems of care. For example, future VHA research may find that patients with OUD have access to psychosocial treatments but decline medication due to reasons related to stigma, or are not being referred to providers who can prescribe OUD medication due to reasons related to provider access. Another example would be finding that ED and inpatient providers who could start patients on OUD medications are not doing so due to lack of available outpatient providers to continue treatment. Comparing practices at VA facilities with high rates of OUD medication use to those with low rates of use could also help VHA replicate successful models.

Also, we did not identify studies published since 2014 that have examined how OUD medication utilization varies by patient co-morbidities. A recent VA study based on chart review found that Veterans engaged in post-traumatic stress disorder (PTSD) treatment had higher retention in buprenorphine treatment.⁵² Understanding how co-morbidities that are prevalent among Veterans, such as such as PTSD, are barriers and facilitators to starting OUD medications would also fill an important evidence gap and help direct future VHA interventions to increase OUD medication use.

Because VHA is not subject to barriers imposed by private health insurance and patients receiving VA care have low or no copays, VHA is also an ideal setting for further research into population-level disparities in OUD medication utilization, including those related to age and race/ethnicity. Whether or not those same disparities exist in VHA populations could impact the direction of future research and interventions within VA and non-VA settings. Additionally, comparisons of OUD medication use among different VHA facilities could help VA clinical leadership refine criteria for offering OUD medications. For example, finding that Veterans in one VA facility or region are receiving buprenorphine despite perceived barriers such as older age or homelessness could help broaden the population of patients who are offered buprenorphine in other facilitates or regions. VHA also has case management resources and capacity for telehealth visits. Studies of OUD medication uptake in the VHA setting if transportation barriers are reduced or removed would also be informative.

To summarize, findings of this review highlight evidence gaps specific to VHA settings in the following areas:

• Specific rates of OUD medication use among VHA populations and how utilization varies by region, age, race/ethnicity, unstable housing or current homelessness, rural versus urban location, and co-morbidities including PTSD
• Barriers and facilitators specific to VHA systems of care, including transitions of care between the ED and inpatient settings and outpatient settings
• VA practice settings in which OUD medication use is highest and lowest (eg, regional differences or provider disciplines/settings such as primary care compared to general mental health compared to outpatient substance use treatment programs)
• Which commonly cited OUD medication barriers including stigma and logistics (eg, provider time) are having the most impact on use in VHA settings, and which facilitators are having the most impact

Additionally, we identified non-VHA-specific gaps in the following areas:

• Barriers and facilitators to use of extended-release naltrexone, which was only discussed in 1 provider study and no patient studies
• Which barriers and facilitators have the most impact on OUD medication use relative to other barriers and facilitators

Operational Partners

Operational partners are system-level stakeholders who have requested the report to inform decision-making. They recommend Technical Expert Panel (TEP) participants; assure VA relevance; help develop and approve final project scope and timeframe for completion; provide feedback on draft report; and provide consultation on strategies for dissemination of the report to field and relevant groups.

• Karen Drexler, MD
  Director
  National Mental Health Program – Substance Use Disorders
• Ellen Edens, MD, MPH, MA
  Director
  VA Interprofessional Advanced Fellowship in Addiction Treatment

Peer Reviewers

The Coordinating Center sought input from external peer reviewers to review the draft report and provide feedback on the objectives, scope, methods used, perception of bias, and omitted evidence. Peer reviewers must disclose any relevant financial or non-financial conflicts of interest. Because of their unique clinical or content expertise, individuals with potential conflicts may be retained. The Coordinating Center and the ESP Center work to balance, manage, or mitigate any potential nonfinancial conflicts of interest identified.

APPENDIX A. Inclusion/Exclusion Criteria

APPENDIX B. Search Strategies

APPENDIX C. List of Excluded Studies

Exclude reasons: 1=Ineligible population, 2=Ineligible intervention, 3=Ineligible comparator, 4=Ineligible outcome, 5=Ineligible timing, 6=Ineligible study design, 7=Ineligible publication type 8=Outdated or ineligible systematic review

APPENDIX D. Evidence Tables

APPENDIX E. Peer Review Disposition