Background
In the last decade, opioid use disorders (OUDs) have emerged as a public health crisis, with an estimated 2.4 million American adults living with an OUD [
1]. In 2016, 42,249 people died from opioid-related overdoses in the United States [
2]. OUDs are closely intertwined with injection drug use, hepatitis C (HCV), and HIV [
3]. Expanding medication treatment for OUD and improving patient retention in treatment are critically needed to combat the epidemic and prevent adverse health outcomes, including early mortality.
Opioid agonist therapy, with either methadone or buprenorphine, has been shown to reduce illicit opioid use [
4] and downstream complications such as overdose [
5,
6], HIV [
7], and HCV [
8‐
10]. Buprenorphine, a partial µ-opioid receptor agonist, offers unique advantages over methadone, including less subjective euphoria and lower risk for side effects such as sedation [
11]. Furthermore, buprenorphine can be prescribed by any trained/waivered provider in an office setting, as opposed to methadone, which must be dispensed in federally regulated opioid treatment programs. Thus buprenorphine delivery has the advantage of being scalable and patient centered [
12]. Research on patient preferences has suggested that some patients choose buprenorphine over methadone for greater convenience and autonomy as well as less stigma [
13,
14].
Unlike licensed methadone treatment programs, office-based buprenorphine treatment programs do not provide medications through directly observed therapy (DOT); instead, they rely on patients to independently take their medications as prescribed. This delivery model risks non-adherence to buprenorphine compromising patients’ treatment outcomes.
Medication adherence has been defined as the extent to which a person’s medication-taking behavior follows the recommendations of a health care provider [
15]. Non-adherence is a common problem that undermines the potential benefits of new pharmacotherapies for a number of chronic diseases, including treatment for substance use disorders. It is estimated that adherence to long-term therapy for chronic illness in developed countries is as low as 50% [
16,
17]. For certain diseases there is a wealth of literature on adherence and interventions to improve adherence. A recent systematic review of interventions for adherence found that the most frequently targeted conditions were HIV, depression, chronic obstructive pulmonary disease, cardiovascular disease, hypertension, and diabetes [
18]. For HIV medications alone, a recent review of qualitative studies on adherence identified 127 studies [
19], and such research has been critical to inform theoretical frameworks and the development of efficacious adherence interventions [
20]. As with antiretroviral treatment for HIV (and many other chronic disease medications), the success of opioid agonist treatment for opioid use disorders is dependent on medication adherence, yet the topic has been relatively unexplored.
Buprenorphine has some unique aspects compared to medications for other chronic medical conditions. On the one hand, the potential for the development of physical tolerance and consequent withdrawal symptoms without use may be a strong deterrent for buprenorphine non-adherence among patients. On the other hand, there is a higher risk for diversion of this medication given the nature of its being a controlled substance, which may drive non-adherence. Diversion is the intentional or non-intentional transfer of medication to an unintended party through selling and/or trading, giving away, or loss of medication [
21,
22]. In a sample of injection drug users with opioid use, illicit buprenorphine use was reported by 45% [
23], and a recent study of persons entering inpatient treatment for medically supervised opioid withdrawal found that among those who had been prescribed buprenorphine, half admitted to “sharing” and a quarter reported selling medication [
24]. Clearly, for some patients the desire to help other individuals or the temptation of financial gain through diversion may be a threat to adherence with their own medication.
Research suggests that buprenorphine non-adherence may be fairly common and associated with poorer outcomes. A study of patients treated with buprenorphine in an office-based setting that used computerized caps on medication bottles to monitor adherence reported that on average patients were adherent to their medication only 71% of the time [
25]. Non-adherence to buprenorphine is associated with illicit drug use and non-retention [
25‐
28]. Buprenorphine has a high affinity for opioid receptors and, when taken regularly, will alleviate opioid cravings and block the effects of other opioids, providing little incentive for illicit opioid use. However, buprenorphine’s effectiveness is dependent on adherence. Methadone is associated with better treatment retention rates compared to buprenorphine [
29‐
31], perhaps in part due to delivery through a model of directly observed therapy.
Video directly observed therapy (VDOT) is a novel method of confirming medication ingestion using an asynchronous video platform via mobile phone application which is modeled after in-person directly observed therapy. With VDOT, patients are able to be observed in their own environment without the constraints necessary to attend a clinic for in-person observation or have clinical staff travel to patients. VDOT has been successfully utilized for the treatment of tuberculosis (TB) [
32,
33] and was implemented post-disaster relief to provide medical monitoring for patients affected by Hurricane Katrina [
34]. VDOT has been found to be acceptable among patients and providers for treatment of tuberculosis; however, it is unknown whether patients and providers would be receptive for its selective use with buprenorphine. While standard of care for treatment of OUD with methadone utilizes face-to-face DOT at federally approved opioid treatment programs, there is no such standard for monitoring adherence among patients who receive office-based buprenorphine. Yet, the delivery of office-based treatment with buprenorphine places a large burden for monitoring on providers and clinic staff [
35,
36]. Currently there is a buprenorphine provider shortage, and known barriers to prescribing include fears of diversion [
37]. Use of VDOT could potentially ameliorate providers’ concerns about diversion and decrease the need for frequent face-to-face visits and urine drug testing. Potentially, VDOT could preserve patient autonomy while allowing for sufficient monitoring to ensure safety and treatment success.
This study among patients and providers within a primary care office-based program aimed to examine perspectives on barriers to buprenorphine adherence, with specific focus on circumstances and reasons why patients might not take their medication regularly. A secondary intent of the study was to explore the acceptability of a mobile phone application for VDOT buprenorphine treatment for opioid use disorder.
Discussion
Through qualitative interviews with providers and current patients at an office-based buprenorphine program based in an adult primary care clinic in the northwest U.S., we identified reasons for buprenorphine non-adherence. Among both groups, perceived reasons for buprenorphine non-adherence generally fell into two categories: factors that prevented patients from being able to consistently access their medications (e.g. homelessness, transportation difficulties, and chaotic lifestyles) and those that resulted in patients not taking their medications as instructed even when medications were available. Reasons for the latter category included intentionally withholding medication in order to use illicit drugs, diversion of medication and forgetting to take medication. Generally, providers and patients alike were receptive to the idea of using VDOT for buprenorphine treatment. Perceived benefits were enhanced patient accountability, leading to more trust from providers who are concerned about diversion. On the other hand, patients and providers expressed some concerns that it would place undue burden on their patients, and some questioned the need for such an adherence monitoring for buprenorphine.
This study sheds new light on the phenomenon of non-adherence to buprenorphine treatment for patients with opioid use disorders. Buprenorphine has proven efficacy; however, in the real world, competing factors may interfere with patients’ ability to be fully adherent to medication, which can jeopardize important treatment outcomes such as retention. Although there are a number of studies that have examined predictors of buprenorphine treatment initiation and retention [
40‐
46], we are unaware of other studies that have focused on medication adherence. A prior qualitative study also explored facilitators to retention in buprenorphine treatment and found positive reaction to medication, personal commitment, and support from clinic staff to be facilitators, whereas transportation issues and competing priorities were barriers [
47]. Our study also identified homelessness and transportation to be barriers to adherence, pointing to a critical need for social services, as well a possible role for mobile health technologies which might cut back on the need for frequent face-to-face visits for patients with transportation difficulties. Patients did mention occasional forgetting to take medication, although in general this was mitigated by withdrawal symptoms. However, interviews revealed that the daily decision whether to take their medication or to use drugs was a challenge. Such challenges to adherence might be addressed by switching to use of extended release formulations of buprenorphine [
48], as well as use of mobile technology for VDOT, text messaging and other supportive features [
49‐
51].
Diversion was acknowledged as a potential reason for medication non-adherence among patient and provider participants interviewed in this study. Providers, but not patients, identified VODT as a potential tool to monitor for diversion. This suggests differing perspectives on the usefulness of VDOT as a tool for adherence monitoring versus diversion control, and the distinctions therein. While diversion requires non-adherence, adherence monitoring is not the same as diversion control. Reports of misuse and diversion have emerged in the U.S. over time and have prompted recommendations for the development of diversion policies for office-based buprenorphine programs [
11,
52‐
54]. A study of buprenorphine providers found that at least half felt that diversion was at least a moderately significant problem in their communities, and another observed that among rural providers, diversion was cited as the leading barrier (48%) to prescribing buprenorphine [
35,
37]. Research shows that providers do employ various strategies to reduce diversion [
35,
36]. However, many of those strategies are not convenient to patients, such as requiring more frequent than monthly visits and regular urine drug testing, and some may even be counter-productive to engaging and retaining patients, such as limiting doses, selectively accepting patients and immediately terminating patients with suspected diversion. Though VDOT would not measure diversion directly, measuring adherence may serve to alleviate some of the unknown felt by providers. There is a need for additional diversion control strategies such as VDOT which might both mitigate provider fears and allow more flexibility for patients. The use of VDOT would not be a substitution for but rather an addition to existing clinic approaches for reducing diversion. Shortages of buprenorphine prescribers exist in many areas of the country in part because of providers’ concerns about diversion [
37,
55‐
57]; having additional tools to safeguard against diversion might increase providers’ willingness to prescribe.
Based on our qualitative interviews, it appeared VDOT for buprenorphine was acceptable for most providers and patients. This is the first study of which we are aware to solicit patient and provider views on the acceptability of VDOT for buprenorphine treatment. A previous qualitative study examined patient and provider perspectives on use of VDOT for monitoring TB treatment in the U.S.-Mexico border region and found broad support for use of the technology in both groups, as did a pilot study of its use in TB public health clinics in Maryland [
58,
59]. As with our study, providers also expressed concern for patient literacy and technical support to allow them to successfully use this technology. In contrast with those study, our study did reveal some opinions that VDOT might be unnecessary for this treatment as there are other compelling reasons to adhere (such as avoiding withdrawal symptoms), and DOT is not currently the standard of care for buprenorphine treatment. Yet some patients found the structure and greater accountability from VDOT appealing, which prior studies have suggested patients desire from treatment [
14]. Indeed, the requirement for daily observed dispensing may be one factor that contributes to better retention rates among methadone treated patients compared to buprenorphine [
60]. It is possible that retention to buprenorphine treatment could improve with the addition of VDOT among primary care office-based programs; however more research is needed to prove such benefits. While forgetting to take medication was mentioned as one reason for non-adherence, it is unclear if patients would be less likely to “forget” with prompts for VDOT. Furthermore non-adherence to VDOT (due to “forgetting” or inconvenience) could exceed that for medication non-adherence, limiting its utility. Finally, it should be noted that among the patients interviewed, there was one individual who was of the strong opinion that he would never utilize such technology due to the fear that it would put data security and privacy at risk. As such, it is clear that not all patients will be willing to partake in such an adjunct to buprenorphine treatment.
This study had a number of limitations. Participants were recruited from a single site in an urban hospital-based clinic that adopted a collaborative care model utilizing nurse care managers based on prior experience in Massachusetts [
12]. This sample may differ from other office-based buprenorphine settings. Our study sample size was relatively modest, and for patients was limited to a convenience sample. We did not collect information on length of treatment from patient participants. It is possible that perspectives on VDOT might differ among patients who are more newly engaged. We used relatively few prompts to solicit information on barriers to buprenorphine from patients and providers in order not to impose a pre-existing framework given the exploratory nature of this work, but this may have limited the number of themes that emerged. There is the possibility that social desirability bias may have influenced participants’ responses; we attempted to mitigate this by having a researcher (MG) uninvolved with clinical care conduct the bulk of the interviews.
Authors’ contributions
JIT, JMS, AJS, JHS and MEG contributed to initial conceptualization and study design. JIT and MEG contributed to data collection, analysis, and drafting the manuscript. All co-authors reviewed and contributed to the revising of the manuscript. All authors reviewed and approved of the submitted version of the manuscript.