Clinical stakeholders’ perceptions of patient engagement in outpatient medication treatment for opioid use disorder: A qualitative study

1. Introduction

Drug overdose claims the lives of more than 100,000 Americans each year (Spencer et al., 2022). Most overdose deaths involve opioids (CDC, 2021). With U.S. opioid settlement funds now available to states, a unique opportunity exists to expand access to medications for opioid use disorder (MOUD)—including methadone, buprenorphine, and naltrexone—that lower risk of opioid overdose and promote long-term recovery through MOUD treatment programs (SAMHSA, 2021b). However, to inform such investment requires understanding how best to implement MOUD programs.

Based on 2020 estimates, just 11% of the 2.5 million individuals (aged 12 years or older) with OUD in the U.S. received MOUD (SAMHSA, 2021a). Among those who initiate MOUD, treatment dropout increases risk for return to illicit opioid use, overdose, and all-cause mortality (Bentzley et al., 2015; Glanz et al., 2023; Santo et al., 2021), and limits the number of individuals who achieve long-term remission (Williams et al., 2017). Studies show wide variation in MOUD retention, with a median of 58% retention at six months (O’Connor et al., 2020). To strengthen treatment programs to keep patients engaged and promote long-term recovery requires understanding the underlying reasons for low retention.

Multiple studies have examined patient perspectives on barriers and facilitators to MOUD retention (e.g., Beharie et al., 2022; Filteau et al., 2022; Kahn et al., 2022; Poulsen, Asdell, et al., 2022; Scorsone et al., 2020). Far fewer have explored clinical and administrative stakeholders’ views of MOUD programs, though their insights are essential to improving care. Prior studies largely focused on willingness and barriers to MOUD adoption (Brown, 2022; Cioe et al., 2020; Lister et al., 2020; Mackey et al., 2020), with only a few examining perspectives on MOUD program delivery (Filteau et al., 2022; Kapadia et al., 2021; Lai et al., 2021).

With this study, we sought to understand elements of MOUD program implementation that may constrain or facilitate patient engagement in treatment from the clinical stakeholder perspective. We conducted a qualitative case study within an outpatient addiction medicine program in a geographically diverse region of Pennsylvania. The multi-clinic program, housed within an integrated health system, allowed for an in-depth and multifaceted exploration of how clinical stakeholders conceptualize and prioritize patient engagement in MOUD program implementation.

2. Methods

2.1. Study design

We applied a case study approach (Crowe et al., 2011) to explore patient engagement in MOUD from the perspective of clinical stakeholders. We use the term engagement to refer to patients’ sustained interaction with the treatment program (e.g., regularly attending appointments), which facilitates MOUD retention. The case under study was an outpatient addiction medicine program in a single integrated health system comprised of four specialty clinics that provide MOUD. We conducted semi-structured interviews with three stakeholder groups involved in delivering MOUD who represented all four clinics and the program’s administration. We collected clinical stakeholder interview data as part of a larger study on patient engagement in MOUD, with patient perspectives reported elsewhere (Poulsen, Asdell, et al., 2022). Geisinger’s Institutional Review Board approved the study.

2.2. Study setting

The outpatient addiction medicine program is embedded within Geisinger, an integrated health system that serves central and northeast Pennsylvania. Pennsylvania is among the Appalachian states with a high prevalence of drug overdose (Beatty et al., 2019), with an age-adjusted overdose mortality rate of 42.4 deaths per 100,000 total population in 2020 (CDC, 2020). The outpatient addiction medicine program began in 2017 and by 2019 had expanded to four specialty clinics. The clinics were geographically dispersed throughout Geisinger’s service area to serve areas with high demand and low supply of MOUD services—two serve metropolitan areas and two serve more rural areas (Table 1). One rural clinic is in a county designated as a mental health care professional shortage area by the Health Resources and Service Administration.

Table 1.

Setting characteristics of the outpatient addiction medicine clinics included in the study

	Metropolitan		Rural
	Scranton Clinic	South Wilkes-Barre Clinic	Bloomsburg Clinic	Williamsport Clinic
Population density of county (population per square mile)	470	366	134	93
Population size of city/town in which clinic is located	~76,000	~44,000	~13,000	~28,000
County drug overdose death rate per 100,000 population (2020)*	30-45	45-60	15-30	30-45

^*Data available only as a range from the Pennsylvania Department of Health.

Together the four clinics provide MOUD to ~2,000 patients with OUD annually. About 45% of patients are female. Most patients (roughly 95%) are non-Hispanic White, reflecting the demographics of the region; the two clinics in metropolitan areas serve a slightly higher percentage of Black and Hispanic patients. A majority of patients use Medical Assistance (Pennsylvania’s needs-based health insurance). Most patients self-refer into the program; the remainder of patients are referred from other drug and alcohol programs, a primary care provider, or other sources (e.g., emergency department, correctional institution) (Poulsen, Santoro, et al., 2022).

Most patients receive sublingual buprenorphine/naloxone, with a substantial minority receiving extended-release buprenorphine. Providers also prescribe extended-release naltrexone, but less frequently. Clinics follow a uniform model of care and day-to-day operations and have an identical staffing structure, including a physician with fellowship training in addiction medicine who leads each clinic and clinically supervises advanced practice providers, nurses, and addiction care coordinators (ACCs) (Barbour et al., 2020). A team of one physician, one advanced practice provider, one nurse, and one ACC typically manages a panel of 300 patients. The two clinics serving metropolitan areas have a slightly larger team (e.g., additional support from an advanced practice provider) and patient panel as compared to the rural clinics.

We conducted interviews from December 2020 to March 2021, during the COVID-19 pandemic in the U.S. With the pandemic’s emergence, telemedicine became prevalent across healthcare, including in addiction medicine. Geisinger’s addiction medicine program had initiated a small pilot telemedicine MOUD program prior to the pandemic, and at the time of the interviews, all providers utilized telemedicine. Providers did not initiate MOUD remotely, but after patients stabilized on a sublingual buprenorphine/naloxone dose, providers conducted follow-up visits and prescribed MOUD to patients through various telemedicine configurations (primarily patient remote and provider remote or in clinic, although at one remote hospital, a nurse facilitated virtual visits with patients, with addiction medicine providers conducting the visit from their own clinic) (Poulsen, Santoro, et al., 2022). Drug screening occurred via remotely observed saliva toxicology tests that patients mailed to a reference lab.

2.3. Study participants

We invited all individuals involved in the day-to-day administration or supervision of the addiction medicine program (hereafter, “administrators”), physicians and advanced practice providers employed in the clinics (hereafter, “providers”), and ACCs from each clinic to participate in an interview. Recruitment occurred via email. The majority agreed to participate, including four administrators, seven providers, and four ACCs. We also interviewed an individual employed as a Certified Recovery Specialist (CRS; a trained peer with lived experience of addiction), given their familiarity with the addiction medicine program. Both interviewers agreed that the study reached sufficient data saturation with the participating individuals. Non-participants did not respond to follow-up email invitations, and so did not provide a reason for not participating.

2.4. Data collection

We developed interview guides with open-ended questions tailored to each stakeholder type. Questions focused on the participant’s role and responsibilities in the addiction medicine program; their perceptions of the program’s objectives, strengths, challenges, environmental context, and resources; and their perceptions of the program’s patient engagement efforts and barriers to patient retention (Supplement 1). We developed interview guides with input from study team members with expertise in addiction medicine (WB) and implementation science (AKR). Interviews were semi-structured and so the interviewer maintained discretion to follow leads or reframe questions (Bernard, 2006).

We conducted interviews via video conference or telephone, depending on participants’ preferences. We explained the purpose of and motivation for the study and obtained verbal consent prior to beginning each interview. Two study authors conducted interviews: MP, a PhD researcher with qualitative research expertise; and PA, an MD research assistant trained in qualitative interviewing. Interviews lasted approximately one hour and were audio-recorded and transcribed verbatim. We offered participants a $50 e-gift card for participating.

2.5. Data analysis

Two study authors conducted the analysis: MP and SR, a medical student trained in qualitative data analysis. We used the Framework Method of data analysis, a thematic case-based approach that involves five key steps: data familiarization, developing of a thematic framework, coding data, charting data, and interpretation (Gale et al., 2013). We gained familiarity with the dataset by reading all transcripts and listening to audio recordings, taking note of our impressions. Next, we developed a thematic framework using a combined deductive/inductive approach to develop codes. We pre-selected deductive codes based on the interview guides and generated inductive codes through open coding of a sample of transcripts. We used an iterative process to finalize the thematic framework, applying the codes to a second sample of transcripts and refining the framework through comparison and discussion. One researcher then coded each transcript using word processing software, with the second researcher reviewing applied codes. We resolved any disagreements through discussion. We then charted the coded data into a framework matrix—a data reduction step that is a defining feature of the Framework Method (Gale et al., 2013). We used a process of memo-writing for data interpretation, with memos focusing on theoretical concepts identified throughout the analysis process and connections between framework matrix categories. We used the COREQ checklist for reporting qualitative research to ensure explicit and comprehensive reporting of our research process and findings (Tong et al., 2007).

3. Results

We identified five themes related to patient engagement: 1) the unique opportunities for—and challenges to—facilitating access to MOUD for an addiction medicine program integrated within a health system; 2) how the program’s prioritization of patient engagement shaped its policies, practices, and clinical environment; 3) the variation in clinical practices—and potential consequences for patient engagement—that accompanies providers’ differing philosophical approaches to treatment; 4) specific services that can bolster patient engagement in MOUD; and 5) the importance of staff well-being for patient engagement. The following sections discuss these themes, accompanied by illustrative quotes attributed to the participant’s role and identifier [#]. We observed consistency in themes across metropolitan and rural clinic settings, although participants emphasized telemedicine as particularly important for increasing access to care in rural settings (discussed in section 3.4.3).

3.1. Facilitating access to addiction treatment

According to study participants, “an extraordinary disease burden” of OUD in Geisinger’s catchment area motivated the health system to initiate the outpatient addiction medicine program in 2017. System leadership reportedly provided “unflinching” support for the program’s development and expansion through provision of space, staff, and other resources, and championed critical components such as a fully integrated electronic medical record to facilitate coordinated care.

Participants described increased access to addiction treatment in the region as a priority for the program. Being housed within a health system offered opportunities for patients seeking care in other departments to be linked to addiction treatment, and the program also facilitated and accepted external referrals. As compared to inpatient treatment, participants viewed outpatient treatment as providing lower barrier access to MOUD. As one administrator described, “coming to an outpatient clinic that’s in your city is a lot easier for folks to take” than the “psychological commitment” of going to an inpatient facility. Yet the stigma around OUD and lack of awareness regarding its treatment required the program to provide ongoing education about the treatment services it offered, both within Geisinger and the communities surrounding the four clinics.

[W]e are constantly looking for innovation and different ways to educate not only our patients, but…our entire Geisinger family. … This is such a high stigma disease and I feel that we strive to constantly look for new ways to innovate our care and educate others on what this disease is and what services we provide. –Administrator[4]

3.1.1. Integration within the health system

Participants viewed the addiction medicine program as playing an essential role in meeting Geisinger’s aim to “take care of a person’s overall wellness,” without which, other aspects of medical care would be less successful.

[P]atients with substance use disorders interact with every point of care that Geisinger offers. They’re in the emergency department, the hospital, primary care, acute medicine clinics, behavioral health clinics. …we are going to be unsuccessful in those other programs unless we also treat substance use disorder. … It is impossible to manage someone’s hemoglobin A1c if they continue to use 20 bags of heroin a day. –Administrator[1]

Referrals from other departments had reportedly increased over time, facilitated by a referral system within the medical record. However, it had taken time for providers to understand what the outpatient addiction medicine program offered. Participants described a lack of understanding of addiction and its treatment by providers in other departments as having led to missed referral opportunities.

Participants identified system improvements to increase patients’ access to addiction treatment, including enhanced linkages within the system’s hospitals and emergency departments to overcome a “lack of knowledge of what resources are available and what to do next” when patients overdose or seek help for their OUD in these settings. CRSs had been placed in one of Geisinger’s ten hospitals and in emergency departments across the system as part of this effort. Participants saw the CRS model as successful but expressed concern that the current level of service did not “scratch the surface” in terms of the number of patients needing support.

Administrators described a vision for a system-wide hub-and-spoke model whereby primary care providers would prescribe MOUD and manage stable patients, with addiction medicine specialists available for consultation and to manage severe cases. Such a model would expand access to MOUD, particularly in remote areas where transportation challenged patients’ treatment access. According to participants, “little pockets” of successful examples had emerged, such as a telemedicine program in one rural hospital that linked remote patients to an addiction medicine specialist. Yet participants described stigma and inexperience with MOUD as having led to an unwillingness by “mainstream medicine” to treat patients with OUD. Participants noted that patients with substance use disorders were often deemed “difficult” and “brushed off’ and “lost in the system.” Opening the addiction treatment clinics had reportedly improved such patients’ care, offering a resource for treatment referrals for patients with OUD, but integration of MOUD into primary care remained limited.

The fact that so many of our [primary care] colleagues who could very easily take on prescribing, just say naltrexone, never mind buprenorphine, and they won’t. You can’t tell me that’s anything but stigma. –Administrator[2]

As we were starting to roll out the hub-and-spoke model where providers would be seeing patients within Geisinger community practices, you would hear, ‘we don’t want those people in our waiting rooms. We need security here if they are going to be here. ’ So, the stigma is alive and it is well and we are battling it. –Administrator[3]

Their big fear is, I am going to mess up, the patient is going die because I didn’t do something. So, it is really just us proving to them we are giving you somebody that has been in recovery for two years and is very stable… –Administrator[3]

3.1.2. Community partnerships

Community partnerships with drug treatment courts, Single County Authorities, local medical offices, and others in the recovery community had facilitated treatment access via referral to Geisinger’s addiction medicine program. Participants viewed stigma around MOUD as the primary barrier limiting community partners from fully embracing the clinics, arising from beliefs that abstinence is required for recovery and that MOUD is “substituting” one addictive drug for another. Participants had observed these attitudes within dominant influences in the recovery community, including 12-step programs and drug treatment courts.

Participants viewed the four clinics as having varying degrees of community acceptance, with support increasing as people came to recognize the benefits to the community. Participants described the longer established clinics as better known and accepted, with one clinic having reportedly become the “clinic of choice” for referral of “more difficult patients,” such as those who had not had success with prior treatment programs. In contrast, participants described the newest clinic’s legitimacy as “slowly” being recognized, with its familiarity increasing as clinic staff promoted the clinic at fairs and community recovery events and as patients shared positive feedback within the community.

3.2. Centrality of patient engagement in evidence-based treatment for OUD

Study participants held a shared vision of the addiction medicine program’s objective as “keeping people alive” by reducing fatal overdoses, reflecting its grounding in harm reduction. Participants viewed patient engagement as central to this objective, rather than a distinct programmatic focus, particularly among patients who struggled most with recovery.

[Retention] is not real high as a priority for me because I do feel like from the get-go, we treat people like we want them to come back and like they have a disease and not a moral weakness. And we don’t expect them to be perfect, so I don’t know what more we could do. –Provider[7]

[T]here are patients who just are not doing well no matter what… so we are basically keeping them engaged to stay alive. That’s the bottom line of what we do. –Provider[6]

Participants perceived the delivery of high-quality care using evidence-based practices as the program’s top priority and vital to meeting the objective of saving lives. Within this context, the program’s policy evolution toward a harm reduction orientation was frequently referenced.

3.2.1. Policy evolution toward harm reduction

With the first clinic opening in 2017, the addiction medicine program had evolved over a short period of time, adopting a harm reduction acuity pathway after having observed low retention with the original “functional improvement” pathway, which required patients to participate in some form of addiction counseling.

I feel like the prior functional improvement track was very rigid. We expected these patients to be substance free within 6-8 weeks. Average engagement was about 60-80 days, which is horrific for a chronic disease. –Provider[2]

The acuity level determines the cadence of visits, and when the program began, the frequency of visits was “driving people out of our clinics,” as one administrator described. Another administrator reported the program’s addiction counseling requirement was a cause of patient disengagement for some patients. In response, the program shifted to a two-tiered acuity system. Patients who wanted to regain “function” in their lives followed the original pathway. The second, harm reduction pathway supported patients who “simply don’t want opiates in their lives” but continued to use drugs. Though encouraged to attend counseling, it was no longer required. According to administrators, these changes led to a “dramatic shift in how well patients engaged,” backed up by the program’s engagement data.

[T]he plan was to have everybody be happy, smiling in progressive recovery. In reality that did not happen. …[W]e eventually changed the paradigm…and tried to retain them longer, hoping that things would improve, and it turned out that…worked. … [O]ur death rate was very, very low and… their addiction severity index score went down the more they were engaged with us… –Provider[1]

Providers emphasized how the current program structure helps patients achieve their own recovery goals. For some patients, successful treatment meant regaining a “normal life” by rebuilding relationships, obtaining a job, and improving physical health. For others, success lay in reducing opioid use and continuing MOUD.

What we’re trying to do is not to find the perfect solution for all of our patients, but to try to get them to where they want to go. –Provider[4]

3.2.2. Harm reduction policies and practices

According to participants, the harm reduction philosophy manifested through the prioritization of at-risk patients (e.g., with a recent overdose or pregnant) for immediate enrollment, providing “as many second chances as possible” by enrolling patients in MOUD who had been discharged from other programs, and not penalizing patients for illicit drug use.

[If]we have good evidence they are taking their buprenorphine, am I going to kick him out? I don’t think so, they are taking their medicine. Heroin or oxy is a lot more likely to kill than cocaine, so why would I even think about kicking him out? –Provider[7]

Although providers tolerated non-opioid drug use among patients, participants described the addiction medicine program as maintaining a structure of accountability through requirements related to appointment frequency and urine toxicology screening. Participants viewed such accountability as a strength and an important ethical standard that differentiated the program from profit-oriented addiction treatment programs.

We have heard from patients that have come from other programs that although we are strict and very blunt in what we expect, we hold them accountable, and that structure and accountability is everything to them. They feel safe. They feel that they can share their struggles. –Administrator[3]

[I]t is not as simple as just handing out a bunch of [buprenorphine]… [T]he handshake that you make with the patient is ‘I’m going to take care of you and I’m going to prescribe you drugs that have some abuse potential and certainly have a street value, but I’m going to prescribe them with accountability.’ … If you do a terrible job at [implementing the MOUD program], you will make the [addiction] problem worse. If you do it correctly with a lot of accountability, you may improve it. –Provider[1]

Participants believed the clinics “make it easy” for patients to remain engaged in treatment.

They described discharge as rare, with providers prioritizing keeping patients in treatment.

Because this is a place of second chances, we really go out of our way to prevent discharging folks, as long as they’re taking medicine appropriately and having some kind of accountability. –Provider[4]

Further, clinics permitted patients to return to treatment if discharged for reasons other than abusive behavior. Provider participants described having a “high tolerance” for difficult patients based on the observation that many were “struggling,” but all agreed that violence or verbal threats were never tolerated.

3.2.3. Return to illicit opioid use, re-engagement, and creating a “safe haven”

Most participants considered current retention rates in the clinics to be high, though some noted that disengagement had increased with the COVID-19 pandemic. However, several provider and ACC participants saw some degree of disengagement as unavoidable, explaining that periods of low motivation and return to opioid misuse are “a natural course of the disease,” and that lack of ‘readiness’ is one of the major barriers to patient engagement in treatment. Participants viewed it as futile to persuade patients to stay in treatment “if they are just not ready to stop using opiates” and that efforts were better spent “finding opportunities where people are ready to change…rather than trying to bludgeon people into what you think is the next right thing.”

[P]eople have moments where they want to get better and sometimes they’re fleeting. That’s just addiction. … [W]e’ve been seeing a lot of patients that dropped off before kind of cycle back in, which is…just the nature of the disease. So I think we’re always below goal for retention, but I don’t know how realistic it is to think that we’re going to get above that goal when we’re dealing with this population. –ACC[3]

This is not a ‘start and finish’program. For how many times they might fall off we take them back. –Provider[5]

Accepting return to illicit opioid use as a “medical concern” without judgment and treating patients “like a human being,” were seen as essential to facilitating engagement.

There is just this readiness that needs to happen…if we continue to be available to them, whether it’s the fifth time or the 20th time, that they always feel like they can come back is important. –ACC[3]

Especially within the healthcare system, let alone the outside world, [individuals with OUD] have been disregarded or [are] disillusioned by poor treatment, treated as junkies. And so to see their humanity and have that humanity celebrated, really I do believe is the key toward patient engagement and retention. –Provider[3]

Patient engagement was thus prioritized through the creation of a “safe haven” where patients felt respected, humanized, and comfortable returning to after disengaging.

[M]ost of our patients who disengage end up coming back…they will disengage for a month or two after they relapse. Then they will call our front desk and will say, ‘I want to reengage. I really need you guys. You guys were awesome. You made [me] feel valuable. You made [me] feel…important.’ –Provider[2]

I’ve had patients tell me a number of times they just feel like they’re being treated with dignity and respect in a clean doctor’s office environment, and I think that that helps keep them coming back for treatment. –ACC[2]

Such an environment was facilitated by a small program with deeply empathetic clinicians and staff, each of whom “hold the patients accountable in a way that creates a very safe and non-stigmatizing environment.”

[In] a lot of [national] programs, you are calling someone in Michigan to make your appointment and then doing your assessment with someone…in Rhode Island and then you‘re meeting with a whole new team when you actually get to your appointment… When they get to [our] clinics, they are meeting with the same individuals that they spoke to [to make their appointment]. –Administrator[3]

[W]e do this from the heart. … We really care about our patients, and we really show them how much we care. We have candy dishes for them. We provide food for them. The whole team makes them feel welcome. –Provider[3]

I have often been behind my hours because I spend too much time, but I think that helps retain people, if you show a genuine interest in their well-being. –Provider[7]

When patients did not show up for appointments, clinic staff followed up through phone calls and letters. As one administrator explained, these efforts let patients know that “it’s okay that you messed up, we still accept you and want you to get well.”

The addiction medicine program’s goal to “squash” stigma furthered the creation of an environment conducive to engagement and recovery. Participants described efforts to educate the community, healthcare providers, families, and patients to help them understand OUD and the clinics’ services. Participants viewed stigma from family members who saw OUD as “a disease of weak will” and MOUD as “trading one drug for another” as affecting patients’ engagement in treatment, with families sometimes pressuring patients to request an opioid antagonist rather than an agonist, or to taper from MOUD “too fast.” Patients reportedly internalized these beliefs, leading them to develop self-stigma and view MOUD as a “crutch.” Providers encouraged patients to bring family members in to meet with the physician and learn about the patient’s addiction treatment with the goal of correcting misconceptions.

[Families] see that…we’re not just dispensing medications and having like a pill mill. We are really engaging [patients]. We are using evidence-based [medicine]. … A lot of them when do see it, say ‘okay, this person is really advocating for my family member.’ –Provider[3]

3.3. Clinic autonomy and variation in clinical practices

With a “common set of policies and procedures” established by the addiction medicine program’s leadership, the four outpatient clinics followed a uniform model of care and day-to-day operations. However, participants noted that the clinics were not “micromanaged,” and providers retained substantial autonomy in terms of clinic management and clinical decision-making, allowing care to be tailored to individual patients. Provider and ACC participants appreciated such autonomy, deeming it important because, as one administrator explained, “in addiction medicine…you could have two people with the same exact use of the same exact substance and nothing…in their recovery is similar at all.” However, with such differences in practice came concerns regarding variation in patient engagement.

3.3.1. Differing philosophical approaches to treatment

With lead providers in each clinic playing the “quarterback role,” their differing philosophical approaches to treatment appeared to influence clinical practices. Although their treatment philosophies aligned with their common fellowship training in addiction medicine and the program’s overall harm reduction approach, administrators observed variation in providers’ acceptance of the new harm reduction acuity system. Differences appeared regarding providers’ harm reduction orientation.

Each of our clinics is very, very, very different. Just because each of us has a slightly different philosophy. Some of my colleagues are a little more authoritative with their patients. –Provider[3]

Participants described the treatment philosophy at two of the clinics—one serving a metropolitan area, and one serving a rural area—as prioritizing listening, compassion, and “being there” for patients when they showed up.

I think we do better by our patients if we cut them some slack… If a diabetic patient ate a box of donuts, I would not put them out of my clinic. It’s the same way [with OUD]. It’s a disease, and if things aren’t going the way we planned, please come in and we’ll talk about it, and we can see how we can help you better. –Provider[6]

Providers from the two other clinics tended to prioritize “holding patients accountable,” viewing this as beneficial to an individual’s recovery and the well-being of the entire patient population.

We’re not sure if recovery is based on holding them accountable or being their buddy… it’s kind of both, but holding them accountable [is] more important than being their buddy. –Provider[5]

We provide [patients] structure because their life is so chaotic. We are just trying to narrow what every other human interaction is going to be like. You are supposed to be on time, you are supposed to be honest, you are supposed to be present. Those things are expected anywhere, whether it is school, whether it is court, whether it is work. [We are] only trying to reflect what the rest of the society is expecting… –Provider[4]

One administrator reflected upon the difficulty of pinpointing factors that underlay providers’ varying treatment philosophies and suggested a range of factors could influence providers’ approaches to treatment, from their personal background to professional training. Participants suggested that clinics’ patient volumes and characteristics of the local patient populations may influence providers’ perspectives on the most effective treatment approach. For example, participants described the clinic led by a provider who had firmer expectations of patients as having a patient population of primarily educated “pill popping professionals.” In contrast, a clinic led by a provider with a strong harm reduction orientation was described as serving a “majority poor, IV drug using population using street drugs” with lower education levels and higher addiction severity.

3.3.2. Variation in clinical practices

Providers’ degree of concern regarding nonadherence to treatment and medication diversion reflecting their diverging philosophies.

We have got both perspectives in our group. People saying, ‘Look, I’m ethically obligated, even though this person has broken every rule and they continue to use…they have a substance use disorder, that’s expected. We are treating them for the disease. We’re not going to kick them out because of the disease.’ But the other…argument is, ‘Look, I know that this patient is selling [buprenorphine] to buy heroin, and I have reasonable evidence to support that, and despite continued and ongoing efforts to change that, they’re not…’ And they kick them out of the clinic. –Administrator[1]

Several participants reflected on the lack of scientific evidence that would privilege one perspective over the other and noted that external guidelines regarding MOUD implementation remains vague. The addiction medicine program had established practice guidelines, but providers retained autonomy to practice within these parameters.

[O]n the one hand, there is a concern…coming through the DEA that buprenorphine not be diverted into the street. On the other hand, there is good science to say that even if a person is getting buprenorphine in the street, it protects against overdose. And so you’ve got these two competing drivers and in between there, there are some guidelines. But the guidelines are by no means as specific as we might wish …there is very, very little in the literature, for instance, around how frequently doing toxicologies helps to keep an individual patient motivated and headed in the right direction and how frequently it needs to be done for the treatment team to know how to adjust the treatment plan. … And so there’s a lot of space for people to abide by the guidelines but have variation. –Administrator[2]

Providers’ differences in treatment philosophies and varying degrees of concern about nonadherence and diversion appeared to influence clinical practices, particularly regarding bridge prescriptions and discharge practices. For example, one provider who was described as more willing to provide bridge prescriptions explained, “it is a little early in the recovery process to hold [patients with OUD] to the same standards [as patients without addiction].” In contrast, another provider did not provide bridge prescriptions for missed appointments, but would prescribe clonidine to address withdrawal symptoms, thereby holding patients accountable without forcing them to experience a painful withdrawal.

One provider who worked in the clinic described (in section 3.3.1) as serving a low-income population with high addiction severity noted that while diversion “is a huge concern, if you overfocus on that, then you’re going to undertreat people too.” Reflecting a strong harm reduction orientation, the provider stated, “success is a process and a path to get our patients where they want to be. It’s based on what their goals are in care… They’re directing me where they want to go in their journey, and my only goal is to support them and to encourage them along the way.”

In contrast, the provider who worked in the clinic that served “pill popping professionals” viewed diversion as actively harming patients, stating “I can’t be a part of that.” Noting the importance of patients’ readiness to stop using opioids for treatment success, the provider cautioned against prioritizing retention of patients who “have no intention to get better” over ensuring that “vulnerable” patients who are “trying to get their life in order” do well. The provider explained that such patients “poison the well” for others because “they are in the waiting room and going to the same meetings and sitting next to the same people,” making “therapeutic discharge” a necessity “for the sake of the community.” Reflecting a more “authoritative” approach to treatment, the provider stated, “This is the only [medical field] that we can’t listen to what your patients say blindly.”

Despite these differences, all provider interviewees strove to strike a balance between holding patients accountable and keeping them engaged in treatment. As one provider explained, “you can’t be so hard that you’re inert to pain, you can’t be so soft that you’re run over by everybody, because you won’t be effective that way either.” The provider with the strongest harm reduction orientation recognized that “every human being needs some degree of accountability” to motivate them and valued approaches such as urine toxicology screening to provide this motivation. Likewise, the providers who tended to be firmer in their expectations of patients would not discharge patients for an occasional nonadherent behavior, instead observing “patterns of behavior” over time.

Without evidence as to the most effective approach, participants noted the challenge in knowing when holding patients to a higher degree of accountability may be beneficial or harmful, particularly because patients’ needs differ. For example, referencing the addiction medicine program’s “3-strikes” discharge policy, one provider explained:

[I]f patients are really serious about their recovery, they get one non-adherence and they are like, ‘oh my God,’ and then they just are good. Some… [are] so far gone in their drug use that you could [give] three of them and the patient wouldn’t even know. …we have team meetings every morning to figure out…maybe that person would need a non-adherence to wake them up or we can’t give another one over here because he has two… [T]here’s no real guidelines, so you’re trying to…come up with a plan…that you can’t look in a book for… –Provider[5]

3.4. Meeting patients’ needs

Several services emerged as key to bolstering patient engagement in MOUD, including addressing patients’ physical healthcare needs, the importance of psychosocial care, and the convenience of telemedicine.

3.4.1. Integration within the health system

Participants primarily viewed the integration of the addiction medicine program within the health system as an asset that helped set it apart from other treatment options. They described how integration facilitated coordinated care for MOUD patients, helping to ensure medical care for the “highest need, medically complex patients,” connecting patients to primary care providers, coordinating care with surgeons to facilitate appropriate pain management, and ensuring pregnant patients attended prenatal visits. They also viewed health system integration as crucial for maintaining a focus on quality-of-care. Provider participants expressed gratitude for working within a relatively well-resourced program, allowing them to “go above and beyond” for patients by having fellowship-trained providers and other qualified staff in a professional environment. However, even within this context, they described a need for more nursing support and other staff to meet demand. Additionally, some participants perceived the extent of integration as suboptimal.

[W]e’re more subspecialists who are really siloed within the Geisinger system. We make referrals out to psychiatry, make referrals to psychology, to family medicine, but they’re not super well integrated into our medical structure, to our organizational structure. –Provider[3]

3.4.2. Importance of psychosocial services

While provider participants viewed MOUD as an essential tool to facilitate patients regaining a “normal life,” they did not suggest it met all of patients’ recovery needs, with psychosocial services playing a vital role in recovery. For example, one provider described recovery as having four components—biomedical, psychiatric, social, and spiritual—explaining that the MOUD program only met the biomedical component. While this component was “vital for most individuals to help achieve…stable recovery,” the provider believed that the lack of full integration of care was a weakness of the program. The importance of psychosocial services in helping patients endure the mental challenges of recovery was illustrated by one participant, who stated, “I always tell [patients], you put down the drugs and the alcohol, you’re left with your head, and that’s why you need the counseling…”

Although the program did not require counseling for patients to continue treatment, nearly all providers and ACCs strongly encouraged patients to participate in counseling, a 12-step program, or to engage with recovery groups. ACCs connected patients with counseling within the surrounding community, though options for counseling specifically related to drugs/alcohol were noted to be limited in the more rural areas. Many participants believed patients would benefit from having psychological services available within the clinics, and some strongly emphasized this would help with patient retention. Participants attributed the lack of integrated psychosocial services to insufficient resources, a lack of space, and a payment model that made in-house counseling financially unsustainable. Similarly, provider participants described how having a CRS within each clinic would help with patient engagement by giving patients “somebody else to talk to” and “be the bridge” to help them access other recovery supports. Commonly reported barriers to recovery included patients’ lack of social support or a social network comprised of individuals who actively use drugs.

We really want our own certified recovery specialist… They are there to offer support both during clinic hours and after hours and meet the patient for a cup of coffee, maybe take them to a meeting… [They have] a much better picture of what is really going on in the person’s life. –Provider[7]

Participants also described an overwhelming need for greater access to psychiatric care for patients. Many patients suffered from psychiatric comorbidities, which, as one participant stated, “portends a poor prognosis for their recovery.” A shortage of psychiatrists both within Geisinger and externally made referrals difficult.

It’s very difficult to get a psych evaluation completed unless you are actively suicidal or homicidal. – I’ve had patients [who have] been told they can be seen in three months. That’s just not sufficient for someone who is suffering. –ACC[2]

A pilot project using a collaborative psychiatric model had been introduced in one clinic, in which a psychiatrist worked with the addiction medicine physician to prescribe for patients with psychiatric disorders, but participants identified a much greater need for mental healthcare.

3.4.3. Telemedicine

With the emergence of the COVID-19 pandemic, telemedicine, which previously comprised a small pilot program in one rural hospital, rapidly scaled up throughout the clinics. As one administrator explained, patient engagement became the priority for MOUD programs nationally, and “put a different emphasis on how we want patients to engage with us in terms of how frequently they need to show up, what they need to show us they’ve been able to do between one appointment and the next.” Participants viewed telemedicine as having greatly improved engagement during the pandemic and hoped to see it expand, both in terms of the number of patients using it and the breadth of locations served.

Participants saw telemedicine as particularly beneficial to patients in rural areas, where community resource were lacking, helping patients overcome commonly reported barriers to attending MOUD appointments, including a lack of transportation and childcare and difficulties taking time off work to attend appointments.

[With telemedicine] you don–t have to get childcare figured out, transportation. If you have a DUI, you don’t have a car, transportation could be a big challenge, driver’s license, etc., or you’re on parole, house arrest, those kinds of things are very real to a lot of people. –Administrator[l]

Telemedicine also allowed clinics to shift efforts from “supervision” of “stable” patients to keeping “acute cases afloat,” with those entrusted to use telemedicine reacting positively to such trust.

[I]t was amazing how the patients responded… ‘Wow, they trust me to not come in and do a urine.’ … That’s a big part of [what makes telemedicine work]. It gives them credit. –Provider[5]

3.5. Prioritizing team well-being

On the frontlines of providing care for MOUD patients, clinic providers and staff kept patients engaged in treatment; thus, participants viewed team well-being as an important consideration in patient engagement. Addiction medicine was seen as a uniquely challenging field in which to work.

The main challenge is the difficult behaviors of our patients…in any medical venue, 10-15% will be difficult patients, maybe in the emergency room 25% are difficult patients. 100% of our patients are difficult. –Provider[4]

Administrators were acutely aware of these challenges and described the importance of retaining providers and staff, facilitating their wellness, and preventing burnout. One administrator described their role as “talking to the bureaucratic side of Geisinger” to advocate for the clinic teams, as the addiction treatment model did not always fit well into the system’s expectations regarding appointment scheduling.

The [health system’s] goal is 90% slot utilization, which is basically 90% of the day having a scheduled patient. We are currently in the high 70’s and I think we are approaching issues of provider burnout and staff burnout just because there is an emotional toll taking care of this population… –Administrator[1]

Likewise, provider participants described prioritizing their team’s wellness by building their team’s resilience, emphasizing the meaningfulness of the work, and discharging patients for threatening or violent behavior in part to maintain staff wellness.

I can’t tolerate [violent and threatening behaviors from patients] because it wears on our staff’s spirituality… It has to be a sustainable, spiritually fit place because we are taking care of the sickest class of society. So, my staff’s wellness matters more than the individual patient. –Provider[4]

Many participants described a strong sense of working as a team within their clinics. They valued daily morning huddles at each clinic as helping to facilitate a “team-based structure” and inspire the team to provide quality care. As one administrator explained, these huddles provide a chance for the full team to not just “run the list” of patients, but also to see—through data—the overall success of the clinic’s efforts. Communication across clinics was less common, though monthly provider meetings offered an opportunity to review challenging patient cases and some provider participants described reaching out to other providers when they faced challenging decisions.

Participants described staff turnover as low, except for one clinic where nursing and care management staff had frequently turned over, due in part to interpersonal conflicts, potentially exacerbated by a small physical space. Some participants expressed dismay in front desk staff who “show frustration” to patients, but also gave examples of other staff going out of their way to make patients feel welcome, bringing food and clothes to those in need. ACC participants suggested that staff interactions with patients could be improved by training clinic staff on addiction, such as an “addictions 101 course for all staff, even those working at the front desk.”

4. DISCUSSION

Improving MOUD implementation to bolster patient engagement is essential to overcoming the overdose epidemic—and to do so requires understanding the perspectives of individuals involved in MOUD delivery. Through this qualitative inquiry into clinical stakeholders’ conceptualization and prioritization of patient engagement in MOUD, we identified clinical practices and services that study participants viewed as facilitating or constraining patient engagement. Together with prior findings from the patient perspective, these findings provide insight into potential implementation strategies to improve patient engagement in MOUD programs broadly.

Integrated within a large health system, the MOUD program at the center of this study was distinctive in terms of its delivery context. Our findings indicate the strength of such integration is the potential for bidirectional referrals, linking patients seen in other clinical departments to addiction treatment and holistically addressing current MOUD patients’ health needs through connections to other clinical departments, potentially improving MOUD retention as patients’ physical and mental health needs are addressed. However, participants described widespread stigma within the medical community around addiction as hampering MOUD access. These findings parallel previously-reported patient perspectives (Poulsen, Asdell, et al., 2022) as well as numerous studies documenting barriers to MOUD adoption in primary care settings (Cioe et al., 2020; Mackey et al., 2020) and highlight the harm caused by stigma surrounding OUD and MOUD.

Patient engagement emerged as central to the addiction medicine program’s core objective, evidenced by its policy evolution to incorporate a harm reduction acuity pathway and clinical practices that centered on offering “second chances.” As noted by past research (Lai et al., 2021) and reflected by the patients served by the addiction medicine program (Poulsen, Asdell, et al., 2022), all patients’ needs cannot be met with a uniform approach. Through its two-tiered approach, the program had built in the flexibility needed to allow patient to choose different recovery goals. Mirroring findings from a low-threshold MOUD program in Philadelphia (Lai et al., 2021), acceptance of return to illicit opioid use as a natural course of disease and the intentional creation of a non-judgmental, humanizing “safe haven” within the clinics were fundamental to encouraging patient engagement. Studies of patient perspectives bear out the significance of a program’s philosophical orientation and receipt of nonjudgmental therapeutic support in encouraging continued engagement (Beharie et al., 2022; Filteau et al., 2022; Kahn et al., 2022; Poulsen, Asdell, et al., 2022; Scorsone et al., 2020). Study participants also identified key services to bolster engagement in MOUD, including psychosocial services and recovery supports such as CRSs to help patients endure the mental challenges of recovery; psychiatry to address comorbid mental health issues; and telemedicine to reduce barriers to attending MOUD appointments, particularly for rural populations.

Notably, despite a common set of policies across clinics, provider autonomy in clinical management and care decision-making appeared to have led to variation in practices across the four clinics that some participants believed differentially impacted patient engagement. Clinicians have long been identified as a key influence on substance use disorder treatment retention, though pinpointing the characteristics that account for such variance is challenging (Najavits et al., 2000). Consistent with prior research (Kankanam Gamage et al., 2023), our findings suggest providers varied in their degree of concern regarding nonadherence to treatment and medication diversion, leading to dissimilarities in the provision of bridge prescriptions and discharge practices. This divergence in treatment philosophy appeared to reflect differences in providers’ espousal of the harm reduction approach and dismissal of the “traditional” substance use disorder treatment model that prioritized abstinence, as has been observed in other programs (Kapadia et al., 2021). It may also reflect a lack of specific evidence-based guidance regarding optimal MOUD implementation, and how treatment practices should be tailored to benefit individual patients, a key challenge identified by study participants. Some participants believed these differences had impacted patient engagement but highlighted the lack of evidence to privilege one perspective over another. Research on patient perspectives reinforces such ambiguity, with some patients expressing a desire for greater accountability, while others report a detrimental impact of more punitive approaches on the therapeutic alliance between patients and providers (Lai et al., 2021; Poulsen, Asdell, et al., 2022).

Finally, given the challenges of working in addiction medicine, study participants viewed staff well-being as critical to MOUD engagement. A team that is burned out or with high turnover was not viewed as providing the compassionate care required to effectively treat addiction. A study conducted among rural MOUD program staff demonstrated how high patient burden led to staff stress and burnout (Filteau et al., 2022). Administrators strove to avoid burnout by advocating for providers at the system level and granting providers autonomy in their practice, while providers prioritized staff wellness through team building and clinic policies such as discharging abusive patients.

Strengths of this study include its novel exploration of clinical stakeholders’ perspectives on patient engagement in MOUD; the data triangulation achieved by interviewing participants with varied roles across four clinics, increasing credibility; and study of an addiction medicine program that serves a geographically diverse population, including rural populations that face significant challenges to MOUD engagement (Lister et al., 2020). This study is limited, however, in its focus on a single MOUD program embedded within a health system. Interpretation of the results may not transfer to other MOUD programs, particularly those that are not part of a health system. Additionally, findings may have been strengthened with additional data sources; to gain a thorough understanding of cases, the case study approach commonly includes multiple sources of evidence (Crowe et al., 2011). Although we referenced findings from patient interviews, we relied primarily on results from qualitative interviews with clinical stakeholders for this analysis. Finally, the cross-sectional design did not allow us to capture changes in perspectives that may have occurred as the program evolved.

Understanding the perceptions of those who administer and deliver care is critical for identifying barriers and facilitators to patient engagement in MOUD and ultimately informing optimal delivery of MOUD programs. Although our study findings highlight the (well-known) detrimental impact of stigma around MOUD, which continues to hamper patient access to and engagement in MOUD, they also suggest potential opportunities for ensuring a patient centered and robust MOUD program. These include integration with other services to meet patients’ comprehensive healthcare needs; adoption of a harm reduction approach; creation of nonstigmatizing and welcoming clinical environments; investment in key services that bolster MOUD engagement such as psychosocial services, psychiatric care, and telemedicine; and prioritization of staff wellness. Finally, findings indicate a need for scientific evidence to inform practice guidelines to better tailor MOUD delivery to meet individual patient needs, highlighting a potential area for future research.

Highlights:

• We explored stakeholder views on retention in medication for opioid use disorder

• Harm reduction and welcoming clinical environments supported patient engagement

• Varied treatment philosophies and clinical practices ambiguously affected engagement

• Psychosocial and telemedicine services and staff wellness were key to engagement

• We identify opportunities to improve patient engagement in treatment

Abbreviations:

ACC

Addiction Care Coordinator

CRS

Certified Recovery Specialist

MOUD

Medication for opioid use disorder

OUD

Opioid use disorder