Hospitalizations related to substance use disorders (SUD) are surging.1 Despite this, hospital providers who work with patients with SUD report feeling unprepared, unsupported, and burned-out,2, 3 and commonly perceive patients with SUD to be violent, manipulative, and aggressive.4,5,6 This can lead to reliance on punitive approaches, feeding the cycle of negative attitudes towards patients with SUD.3 Furthermore, hospitalized patients with SUD are often mistrustful of providers and may feel discriminated against due to their substance use,7, 8 leading to worse outcomes, lower patient satisfaction, and higher rates of leaving the hospital against medical advice (AMA).9, 10

Emerging research in ambulatory settings suggests that the use of peer mentors (PMs) can improve substance use outcomes, including reduced relapse rates, increased treatment retention, and increased patient satisfaction.11, 12 Studies show that PMs can extend emotional, informational, and instrumental support,13 and that PMs may act as a bridge between service users and behavioral health providers.14, 15 Studies also suggest that outpatient PMs, while generally satisfied in their work, may feel unappreciated and stigmatized by their non-peer co-workers.16,17,18

To our knowledge, no study has specifically examined the role of PMs in the care of hospitalized patients with SUD. This study aims to better understand (1) how peer mentorship affects the hospital experience and care for hospitalized patients with SUD and (2) how working in a hospital affects PMs’ sense of professional identity.


Setting and Study Design

This study is part of a formative evaluation of an interprofessional addiction medicine consultation service at an urban, academic hospital in Oregon—the Improving Addiction Care Team (IMPACT).19 IMPACT includes care from addiction medicine physicians, social workers, and peers with lived experience in recovery. Patients with known or suspected SUD are referred to IMPACT from inpatient medical and surgical providers. IMPACT performs an initial comprehensive assessment; elicits patient-centered goals around the acute hospitalization and SUD; initiates SUD treatment; coordinates outpatient treatment linkages; and offers harm reduction support. Earlier work describes IMPACT and the hospital environment.3, 19, 20 Prior to IMPACT, hospitalization frequently did not address addiction and providers commonly felt distressed caring for patients with SUD. IMPACT has shifted culture and improved care.3 PMs meet patients during hospitalization and can serve a bridging role after discharge. IMPACT PMs are hired in partnership with a peer-run mentoring agency. A separate manuscript details peers’ role, supervision, and logistics of integrating PMs into hospital care.21

IMPACT started in July 2015 and the first PM joined in April 2016. We interviewed IMPACT providers, IMPACT patients, PMs, and a PM supervisor between April 2016 and June 2018. Patient interviews were collected as part of a larger IMPACT evaluation. We supplemented interview data with direct participant observation of PMs and ethnographic field notes. Study protocol was approved by the Oregon Health & Science University institutional review board.

Participants and Data Collection

We invited all IMPACT patients in the study window to participate in a baseline interview during hospitalization and a follow-up phone interview 30–90 days after discharge. Patient participants completed two survey-style interviews, each lasting approximately 15–20 min. Following a closed-ended survey, researchers asked patients open-ended questions about their experience with IMPACT (Appendix A online). Though questions did not explicitly ask about PMs, patients commonly discussed PMs. Of the 222 patients who completed interviews, 118 were eligible for this study, either because study records indicate that they worked with PMs or because they spontaneously mentioned PMs during their qualitative interview. Of those, 46 mentioned PMs in their interview and were included in this study (see Table 1 for patient participant characteristics). One author (DC) transcribed qualitative responses verbatim at the time of data collection and immediately reviewed each transcript for accuracy after completing each interview. All participants received a $25 gift card.

Table 1 Patient Characteristics (n = 46)

We conducted in-depth, semi-structured individual interviews and focus groups (FGs) with 5 IMPACT physicians, 3 IMPACT social workers (SWs), 1 IMPACT outreach liaison, 1 peer supervisor, and 2 IMPACT PMs (Table 2). Three PMs worked on the IMPACT during the study period; however, one PM transferred to a different position prior to the start of in-depth provider interviews. PMs have been de-identified as PM1, PM2, and PM3 in our findings. We chose to use a focus group format in situations where multiple participants occupied a similar role (e.g., physicians) to elicit a range of perspectives and experiences through participant interaction. We utilized individual interviews when participants had unique roles, or when scheduling barriers prevented FG participation.

Table 2 IMPACT Provider, PM, and Supervisor Characteristics (n = 12)

We asked IMPACT providers and the PM supervisor about their experiences working with PMs and how they felt PMs influenced care (Appendix B online). We asked PMs about their experiences working in the hospital and on IMPACT (Appendix C online). Although interviewing team members about one another creates a potential for bias, we have tried to reduce this possibility in several ways. We designed the interview guides to assess challenges and negative experiences from both IMPACT providers and PMs. Additionally, interviews and focus groups were conducted by an author who is not a part of the IMPACT clinical team (DC), and they were confidential and immediately de-identified for analysis. All interviews and FGs lasted approximately 1 h, were audio-recorded and transcribed verbatim.

Finally, we utilized participant observation and field notes to supplement interview data. One author (DC) observed daily IMPACT “huddles” and weekly quality improvement meetings and took ethnographic field notes during the study period. Two authors (DC and JA) shadowed PMs twice and documented patient-PM interactions using ethnographic field notes.

Data Analysis

Transcripts and field notes were transferred to Atlas.ti. We conducted a thematic analysis using an inductive approach.22 In other words, we identified themes that emerged directly from the data (as opposed to using a deductive approach, where we would look for themes based on the literature or our own theories). We analyzed data at a semantic or explicit level, as opposed to a latent level (i.e., we focused on what participants said, not the underlying ideas that are theorized to shape the semantic content of the data).22

In keeping with principles of thematic analysis, we used a subjective heuristic to identify themes. Included themes needed to (1) be expressed by multiple participants; (2) be expressed as a central concern; (3) relate to the research topics explored; and (4) have clinical or policy relevance.22 The multidisciplinary analysis team included the IMPACT director (HE), an IMPACT clinician (JG), a general internist (CN), a peer program director/supervisor (DJG), a medical student (JA), and a sociologist (DC). Using an iterative process, two authors (DC and JA) developed a codebook and independently coded all data, meeting regularly to discuss and reconcile coding patterns until reaching consensus. Coders identified preliminary themes. They then reviewed and finalized themes iteratively, during multiple discussions with the full team. We used the same codebook across all qualitative interviews and field notes. Study participants’ reflections were based on their experience working with three PMs. However, after the study window, IMPACT employed six more PMs and utilized member checking to ensure the durability of our findings.


We organize our findings into five sections. First, we describe PMs unique role within the hospital. Second, we show how PMs contextualize patient experiences to IMPACT providers. Third, we explain that PMs “translate” provider care plans to patients. Fourth, we describe how PMs can transfer trust towards hospital providers and systems. We represent these findings in Figure 1. Finally, we explore challenges and growth for PMs’ professional identity.

Figure 1
figure 1

PMs’ effects on hospitalization for patients with SUD.

PMs Have a Unique Role Within the Hospital Ecosystem

Patient respondents frequently described PMs as trustworthy and relatable, highlighting shared lived experiences. Patients explicitly identified PMs sharing “the same background” or “similar paths,” which they felt supported mutual understanding and open communication. Patients often described PMs as uniquely honest, caring, authentic, emotionally engaged, and supportive.

Patients framed small, ordinary acts from PMs as extremely meaningful. Patients described that PMs “just stopping by and chatting” and making daily visits was important, particularly for patients who were otherwise socially isolated. As one patient described (referring to a PM), “I got somebody I can talk to here… I don’t have any resources or friends. I don’t have anybody who’s coming to see me.” Participants noted that humanizing gestures are particularly meaningful in a hospital, where staff only provide brief, clinical interactions. An IMPACT physician recalled:

[Humanization] can be hard in hospitals in particular because of the busy aspect and the sick patients. I remember my first interaction with [PM1]- I walked into the room and [PM1] was doing [the patient’s] hair – I’d never seen that in a hospital before.

Patients described that PMs did not judge their behaviors or readiness for treatment. As one patient recalled, “[PM1] basically said, if you wanna quit, great. If you don’t wanna quit, great. Maybe we can get a plan figured out. She put the ball in my court and she didn’t judge me.” Many patients described a restored sense of agency that helped them “take the reins” and feel empowered by working with PMs. This contrasted with patient descriptions of non-peer hospital staff as pushy, intrusive, or controlling, particularly around pain medications, behavioral contracts, and room searches. As one patient explained,

I wish [IMPACT doctors and social workers] would have left me alone in the beginning… I don’t like to be told what to do… Sometimes that can be detrimental to my health… But because of PM1, she played a big role, I was able to make the decision on my own to become clean and sober.

PMs’ professional networks and first-hand experiences using community resources further enabled them to establish trust and credibility when offering support, including housing assistance, bus passes, clothing and hygiene products, and legal advocacy. As one PM described:

I have a reputation for quashing warrants. So I use that as my ace-in-the-hole. I say “hey, you said you got this warrant, but guess what, I will go to the judge, I will call the P.O., I will call Trump himself, whoever we gotta call, we gonna do what we gotta do to [try and] get rid of this warrant.”

Further, unlike other providers, PM relationships are unique in that they extend past hospital discharge. One patient explained, “PM1 and PM3 still come to see me and are willing to go to meetings with me or walk around and talk to me… Having that familiar face was not as scary, having care that follows me, it makes a huge difference.”

Participants felt PMs could mitigate common power struggles between staff and patients. They often accomplished this by emphasizing their independence from traditional medical roles. As one PM explained,

I try to feel out how suspicious [a patient] is of me, and how uncomfortable. If they’re keeping me at arm’s length, then I double-down [and explain] I don’t have to report anything to the rest of the team besides mandatory reporting stuff.

While patients felt unique trust with PMs, hospital providers often did not. Particularly when the role was new, PMs described sometimes feeling that their role was misunderstood or devalued by non-IMPACT staff. As one PM recalled, “Walking through the halls here in the beginning it was really hard. I dressed differently; I talked differently, so there was some people who obviously felt like I didn’t belong.”

PMs Contextualize Patient Experiences for Providers and Influence System-Level Change

IMPACT social workers and physicians felt that PMs, who attend daily team huddles, reframed their understanding of patient experiences and informed their approaches to patient care, helping them “learn more about what’s really happening.” One IMPACT physician recalled an instance of a PM offering advice about how to best engage a patient:

It was [PM2], who said, you know, I wouldn’t use that language again, or I would re-frame that differently… I thought I was being all street-savvy. And you know, after we had left [the patient room], he quietly, very kindly said I was totally off the mark.

In another example, a PM was able to re-contextualize patient behaviors to allow for increased empathy and understanding from IMPACT:

I was able to go back to the providers and say he’s scared… What he’s going through has nothing to do with you… he doesn’t know if he can escape this lifestyle and he’s afraid to die. And so being able to tell the doctors these things that they can’t hear from the patient, they were able to humanize him and not give up on him.

Because of their liminal position between hospital providers and patients with SUD, PMs may act as a “bridge” that “anchors” providers in patients’ experiences. As one provider described, the contrast, “helps us not go down this wonky medical path.”

Respondents felt that PMs played a powerful role in identifying previously unseen concerns within hospital structures and that this in turn led physician leaders in their approach to program development. As one IMPACT physician explained,

This is an institution, and so often I feel like the peers will show us the ways in which institutions can either harm patients or not hear patients… those are the conflicts that our patients also experience. We just don’t have to see it when we’re the ones with the power.

Physician and social work participants noted that the role of PMs as “cultural brokers” had tremendous strengths, but also presented challenges. They recalled times when PMs’ emotional identification with patients or lack of indoctrination in an often rigid hospital system could feel like a PM was “fighting against the system” in ways that may not have been productive. These kinds of conflicts supported the IMPACT growth and PMs’ professional development.

PMs “Translate” Provider Recommendations to Patients, Making Hospitalization More Tolerable

Respondents described that PMs can “translate” hospital providers’ recommendations, helping patients better understand and tolerate prescribed care plans. Participants felt that PMs communicate frankly and directly, in ways that “the patient can hear it.” As one social worker recalled,

I think about a patient that [PM1] was working with where he was just kind of like, screw these doctors, nobody cares, you all just want me to fail and I’m just gonna leave… [PM1] got to be firm with him in a way that was legitimate to their relationship. If I had done that it would not have been… successful. But [PM1] was able to say… you still need to follow the rules of the hospital.

Providers felt that PMs are able to de-escalate crises, discouraging patients from leaving AMA and alleviating perceived discrimination from hospital staff. One physician recalled an instance after a patient overdosed and was resuscitated with naloxone:

He became anxious and agitated, and resentful that we had given him naloxone when he felt he didn’t need it … it was the peer who came in and said, you know, you look upset, let’s go on a walk. Let’s talk through this, and [that] de-escalated his physiologic response to withdrawal. It was really remarkable.

IMPACT providers felt that PMs “run interference,” and relieve tensions between floor staff and patients with SUD. As one physician elaborated,

Peers play a huge role in the day-to-day behavioral stuff that happens… that’s helped to de-escalate some of the angst from… front-line staff around patients with substance use disorders… Overall I think it’s moved the culture to a place where it feels more manageable.

PMs Transfer Their Credibility to Hospital Providers and Systems

Respondents noted that PMs often “transferred” their credibility and trust to IMPACT medical providers and social workers. One PM described one such instance of a PM vouching for the IMPACT team:

I tell patients… if the IMPACT team did not care about you they wouldn’t have hired me… And I feel like if a peer isn’t here to have that conversation with those patients who are AMA threats, who really need to stay, I don’t think that they would receive the care.

IMPACT providers described explicitly associating themselves with PMs in front of patients as a strategy to build their own credibility:

The fact that there is a peer that we can align ourselves with and that’s part of the team, to me, creates a level of trust. The fact that we defer to them, that’s really powerful… [There was] a patient that was just pissed off, and looked at us… and asked, “has anybody ever used heroin?” For [PM2] to be able to raise his hand, and that surprise [helped him trust us]…

Resultantly, participants described that PMs made patients more amenable to seeing IMPACT. As one patient explained, “[PM2] was a good listener, he didn’t make judgments… if it wasn’t for [PM2], I would have said I don’t want [IMPACT]. If it wasn’t for him, I wouldn’t have talked to them.” Participants felt that by encouraging patient autonomy, PMs increased patients’ confidence in their decisions. As one patient explained, “[PM2] made me feel like it wasn’t a bad choice to get on methadone. That helped a lot.” Finally, by providing “a point of positive engagement experience with the healthcare system,” participants felt that PMs may “open the door for further engagement outside the hospital.”

Although rare, patients sometimes perceived PMs as manipulative or “secret agents” of IMPACT because of their association with a hospital team. As one patient said: “[IMPACT] was persistent, and they were sneaky too. Because they sent [PM1] up there and I wound up talking to her. So their manipulation tactic worked.”

Working in a Hospital Enriched PM Professional Identity and Introduced Unique Challenges

PMs felt that their experiences working on IMPACT allowed them to advance their professional development and career goals. As one PM, who was working to apply to medical school, explained, “This was another opportunity to push my boundaries and my comfort.” Another PM noted that the role provided “a foundation of how to behave professionally.”

Hospitals may present unique challenges for PMs and their sense of professional identity, especially when PMs are new in the role and as the peer program is introduced. Respondents felt that isolation in a rigid, hierarchical environment puts PMs at risk for feelings of stress and imposter syndrome. One PM described the exhaustion of “fitting in” in the hospital setting: “I gotta let them know I’m not a thief. I gotta know I’m a hard worker. I gotta let em know that I earn this money… and it never stops.”

Furthermore, PMs described how opportunities for professional growth conflicted with fears of assimilating into professional culture and losing one’s “peer-ness.” As one PM put it:

It’s almost as if you have to be able to talk two languages - you have to talk professionally and when you’re alone with a peer you have to be able to turn it off and be down to earth with them.

At the same time, PMs expressed the need for self-acceptance and self-confidence in the face of competing role expectations:

I don’t have to try to copy the way other people dress in here, I can be myself, I don’t have to assimilate. I went through that for a minute, like I had to flat iron my hair every day… wear these little clunky shoes and these slacks like them to be accepted and it just didn’t work for me. It felt really uncomfortable. And so I realized that, you know, I can be myself.

PMs also identified the intense emotional work they go through daily with patients as a key challenge to their role. As one PM recalled, “I wasn’t prepared professionally to handle death and dying and palliative care.” They further elaborated:

You just want to grab [patients] and hold them and just walk them step-by-step through everything and just take care of them and protect them and follow them everywhere and make sure that they don’t die, because some of them are really young and beautiful… You’re just like, how can I get you to understand you’re gonna die?…it doesn’t matter what their consequences are, we are there to support [them]. It’s not my plan for them, it’s their plan. And I have to… radically accept that…. For me that’s the hardest thing….

To cope with these challenges, PMs emphasized importance of self-care, contact with the recovery community, and strong supervision:

Stay within your hours and take care of yourself. Go to your NA meetings… at the end of the day, I’m still an addict… my ability to pick up and use at any given time is still an opportunity for me if I don’t take care of my own recovery, if I don’t get into therapy, if I don’t process with my supervisor things that are overwhelming me… [mentorship] does not equate your recovery …


This study describes the benefits and challenges of peer mentorship in a hospital setting using qualitative interviews, field notes, and participant observation. We found that PMs form meaningful relationships with patients based on trust and shared lived experiences; this contrasts starkly with other hospital providers whose value stems from professional training and expertise. PMs facilitate patient-centered care by amplifying patient experiences. Reciprocally, PMs “translate” provider thoughts and recommendations to patients in ways that patients can hear. Respondents described PMs as “cultural brokers” who have the potential “transfer” trust that they have earned with patients to providers who may otherwise be viewed as not trustworthy. Effects of hospital PMs included increasing patients’ ability to tolerate hospitalization, accepting recommended care, de-escalating crises, and reducing staff stigma towards people with SUD. We found that PMs open the door for patient engagement after discharge by bridging to aftercare and acting as a positive point of contact within the healthcare system. Finally, while PMs felt that their role led to personal and professional development, the intensity of the role put PMs at risk for emotional drain and burnout.

The mechanisms of peer support that we describe are consistent with literature describing PMs in other settings,15, 18 including the potential to encourage patient autonomy23 and improve patient satisfaction with their care.14 Our study shows that such benefits are possible in hospitals, where relationships between patients and providers may be particularly strained,7 and where completing recommended treatment may be particularly important given potentially life-threatening illnesses. Our finding that PMs may feel isolated from non-PM staff is also consistent with reports from PMs in other settings.16,17,18 This may be even more exaggerated in a hospital, which is a highly specialized clinical setting with a rigid professional hierarchy.

Our study has several limitations. As a single-site study, our goal was not generalizability, but transferability. As such, we aimed to obtain rich in-depth information that can inform PM integration in other hospitals. Because patient data was collected as part of a larger IMPACT evaluation, patients were not explicitly asked about their experiences with PMs. Despite this, nearly 40% of patients who worked with PMs offered relevant comments, and patients revealed both positive and critical perspectives on PMs. Finally, there was low racial and ethnic diversity among patients and providers, findings may not be transferable to more diverse populations.

Despite these limitations, our study has several important implications, and PMs offer great potential value for systems struggling to engage hospitalized patients with SUD. Contrary to our expectations, the notion of PM as a patient role model or inspiration was not a salient finding. Instead, patients repeatedly stressed the tremendous importance of small, humanizing gestures. This finding reinforces how alienating and traumatizing a hospital course can be for patients with SUD, many of whom are facing potentially life-threatening illnesses, are socially isolated, and perceive discrimination by their care team. By focusing on relationships, PMs can avoid common power struggles among hospital providers and patients with SUD.7 PMs, who have no agenda other than “meeting patients where they are at” and advocating for them, bring to light the pervasiveness of control, mistrust, and paternalism within hospitals, even among providers who strive to be patient-centered.

Many healthcare providers may themselves be in recovery. And while their lived experience may enrich the care they provide and allow for deeper understanding and empathy, lived experience alone does not define the full value of peer mentorship. PMs are, above all else, nonclinical. Retaining flexibility, independence, and autonomy are crucial elements of peer mentorship.

PMs’ lived experience of addiction, combined with their distance from medical culture and hierarchy, is at the core of their value. This also creates inherent challenges in integrating PMs into hospital settings. PM success in hospitals will likely depend on key implementation factors. Our findings support the importance of peer supervision, which may mitigate some of the increased stresses of hospital PMs. Supervision can help PMs navigate their “brokering” role, provide support when difficult situations arise that could potentially compromise patient trust (e.g., mandatory reporting), and help PMs balance “fitting in” in a hierarchical and highly professionalized environment while maintaining “peer-ness.” Furthermore, supervision is crucial in a clinical environment where patients are sick and dying. The high potential for re-traumatization, stress, and burnout among hospital PMs highlights the need for systems that support self-care. Beyond supervision, PMs need to feel integrated and valued in hospital settings. This important process may be facilitated through a hospital-based champion or a team setting. In a separate paper, we describe implementation factors to facilitate PMs’ integration in hospitals.21

While this study focuses on the dynamics between IMPACT clinicians and social workers, IMPACT PMs, and IMPACT patients seen by IMPACT, further research is needed to explore PM integration from various perspectives in diverse hospital settings, including academic and community hospitals, hospitals with and without dedicated addiction medicine providers, and urban and rural settings.

Engaging hospitalized adults will continue to be a key priority in addressing the SUD epidemic. As our study shows, integrating PMs is not without substantial challenges. However, PMs can be a “secret weapon” to engage patients, elevate and contextualize patient experience, and improve hospital systems.