The Experience of Peer Mentors in an Intervention to Promote Smoking Cessation in Persons with Psychiatric Illness

Abstract

Peer support is an important component of services for persons with psychiatric illness but the experience of peer mentors is not well understood. This study explored the experiences of peer mentors, all former smokers and persons with psychiatric illness, who provided smoking cessation counseling as part of a 6 month professionally-led intervention. Data was obtained from 383 contact log entries and in-depth interviews with 8 peer mentors. Qualitative analysis indicated that mentor roles were unexpectedly varied beyond the focus on smoking cessation. Of the two aspects of “peer-ness,” shared smoking history was more prominent, while the shared experience of psychiatric illness was sometimes overlooked. Peer mentors experienced multiple challenges trying to help participants to change their smoking behaviors. Nonetheless, they described their experience as personally rewarding. Future interventions may be improved by anticipating peer mentor role complexity and the inherent tension between providing person-centered support and promoting behavior change.

Introduction

Peer specialists, persons with lived experience of psychiatric l illness who act as an addition to and in partnership with professionally-delivered services, are now being widely integrated into mental health services for people with psychiatric illness. Like peer interventions for other health conditions, peer support for people with psychiatric illness is based on the principle that people who have endured and overcome adversity can offer support, encouragement, hope, and mentorship to others facing similar situations (Davidson, Chinman, Sells, & Rowe, 2006). Evidence from multiple studies show that peer specialists can reduce inpatient service use, improve engagement in care, and enhance empowerment, patient activation, and hopefulness for recovery (Chinman, et al., 2014). Despite these outcomes, previous limited study of peer providers’ experiences suggests that they may face challenges regarding tensions in the work environment, difficulty in defining their role, and emotional distress in response to the work, but also benefit from their role as helpers (Ahmed, Hunter, Mabe, Tucker, & Buckley, 2015; Moran, Russinova, Gidugu, & Gagne, 2013; Moran, Russinova, Gidugu, Yim, & Sprague, 2012; Walker & Bryant, 2013).

A variation of this model is to employ peer specialists as partners with professionals (often called peer mentors) to help clients focus on specific health behaviors with agreed-upon goals and an explicit agenda for behavior change (Goldberg, et al., 2013; Jerome, et al., 2012; Swarbrick, 2013). Smoking cessation is an important target among people with psychiatric illness because the prevalence of their smoking persists at epidemic levels and is a primary cause of premature mortality (Dickerson, et al., 2014). Interventions combining pharmacologic and counseling strategies have been developed for people with psychiatric illness but quit rates are modest (Tsoi, Porwal, & Webster, 2013). The addition of peer mentors may be particularly useful because of the strong demonstrated connection in the general population between quitting smoking and interpersonal associations with non-smokers (Christakis & Fowler, 2008). People with psychiatric illness are often in environments where smoking is common and they may have few role models of successful quitting (Ziedonis, et al., 2008). Therefore, peer mentors who share both experience with psychiatric illness and previous smoking may be able to provide “been there” encouragement, strategies, and role modeling that would increase the potency of smoking cessation interventions (McKay & Dickerson, 2012).

On this basis, we developed and evaluated a well-specified peer mentor program that was added to a professionally-led group intervention for smoking cessation tailored for people with psychiatric illness.(Dickerson, et al., In press). Although other qualitative explorations of peer specialists have been conducted, to our knowledge there has not been a detailed analysis of the experiences of the peer mentors, themselves, in such programs, despite important implications for intervention refinement.

Methods

Participants

Eight peer mentors took part in this study; each was 1) a former smoker (at least 100 lifetime cigarettes but none in the past year) and willing to discuss their quitting and maintaining abstinence; 2) a past or current recipient of mental health services for psychiatric illness; 3) age 18+; 4) ≥ high school education, and ≥12 months work or volunteer experience in the last 3 years; and 5) an effective communicator ascertained via an application and interview. Applicants were hired as part-time employees of the principal investigator’s hospital. All had experience providing human services but none in smoking cessation counseling. They included 6 women, 7 Caucasians, and their mean age was 43 years (s.d. 12.5, range 24–60); 5 were college graduates. Five self-reported a diagnosis of bipolar disorder, two recurrent major depressive disorder, and one schizoaffective disorder. They had smoked an average of 17.8 years (s.d. 11.5, range 5–34) and had been abstinent from smoking an average of 8.0 years (s.d. 10.1, range 1–31).

A total of 30 people with psychiatric illness from local psychiatric rehabilitation programs participated in the smoking cessation intervention. Eligibility criteria included 1) diagnosis of psychiatric illness; 2) current smoker (daily smoking for most days in the past month); 3) Contemplation or Preparation stage of change for quitting smoking (DiClemente, et al., 1991); 4) age 18–75.

The study was approved by the institutional review boards of Sheppard Pratt and the University of Maryland School of Medicine; all peer mentors and intervention participants provided written consent according to the approved protocol.

Intervention

The study’s smoking cessation intervention consisted of 1) a smoking cessation group of 24, one-hour meetings twice weekly over 3 months, and 2) peer mentoring relationships starting during the group and extending an additional 3 months. The intervention was delivered at three outpatient psychiatric programs between May, 2013 – June 2014; most peer mentors participated at more than one site.

The group was adapted from one for co-occurring drug abuse in persons with psychiatric illness (Bellack, Bennett, Gearon, Brown, & Yang, 2006). Led by a professional interventionist and two rotating peer mentors, it included motivational enhancement, smoking and cessation education, skills training, and goal setting. Participants received three dollars for each meeting attended. Nicotine replacement therapy was available to interested participants. Within the group meetings, the peer mentors provided modeling and encouragement regarding smoking cessation by sharing relevant personal narratives, and discussing group topics with participants. Mentors also recorded participants’ weekly goals and administered a carbon monoxide Breathalyzer which indicated the level of smoking in the previous six hours.

Additionally, each peer mentor was paired with approximately 2 participants during each group cohort. The mentors met individually with their participants in person or by phone for 1–2 hours each week and focused on reinforcing the group curriculum messages, providing encouragement, helping participants implement their goals, and problem-solving toward reducing and quitting smoking. Procedures were detailed in a manual and supported during weekly group supervision with the PI.

Prior to the above, peer mentors were trained by the PI in group sessions totaling 22 hours. Training covered the topics of peer counseling, the intervention curriculum, communication skills; cognitive behavioral cessation strategies; sharing personal experiences; boundaries; and role plays practicing these skills.

Data Collection & Preparation

Data for this study came from two sources: 1) post-intervention interviews with the peer mentors, and 2) interaction logs that peer mentors completed immediately after each individual meeting with an assigned program participant.

Peer mentors were individually interviewed one week after concluding their work with study participants by author AL, an experienced qualitative researcher not involved in the intervention. (One mentor was not interviewed until one month after ending.) Interviews followed a semi-structured guide and were audio-recorded. Questions included: What was it like for you to be a peer mentor? What has been the impact on you personally? In what ways did you feel you were helpful as a peer mentor? What were the major challenges that you faced? Mentor interviews averaged 51 minutes (s.d. 13, range 27–71). Two members of the research staff (FD and CS) then listened to each interview independently and made detailed notes in lieu of a verbatim transcript (Miles & Huberman, 1994).

During the study, peer mentors completed a written interaction log which was turned in to study staff weekly. The peer mentor recorded practical details, techniques used and topics discussed from options provided, and a narrative summary of the interaction, and was invited to answer two optional questions: “Please make any comments about your own emotional reaction and the impact of this specific interaction on your own experience as a peer mentor” and “Please use the space below to address any questions/concerns you may have about this interaction,” the responses to which were examined for this study. Of the 1078 interaction logs turned in, 512 included some response to the optional questions, but 129 mentioned nothing about the peer mentor role. Therefore, our interaction log data consisted of 383 responses; each peer mentor contributed an average of 48 logs to the data (s.d. 29, range 18–96).

Analysis

We used a thematic analysis approach (Braun & Clarke, 2006) starting with the detailed interview notes. Two team members, CS and FD, re-read all interviews to derive broad categories of the peer mentors’ experiences and then discussed and merged their outlines, forming the preliminary coding guide. CS and FD then independently coded the interview notes with this guide, discussing and merging their results and modifying the outline via consensus and in discussion with AL. CS and FD then applied the revised coding outline to the interaction logs’ data, making several additional changes to yield the final code set for the study. CS and FD re-coded all data with the final code set, resolving minor discrepancies by consensus. Lastly, FD reviewed all final coded data and summarized themes and variations within each category. The most illustrative quotations or comments were chosen as examples.

Results

The combined mentor interview and interaction log data yielded five themes describing the peer mentors’ experiences.

Peer Mentor Functions

A major theme was how varied peer mentors’ roles became, encompassing five different and often interwoven role functions.

The first role, as specified by the intervention, involved tasks directly related to the smoking cessation agenda. In meetings with participants, peer mentors completed worksheets focused on smoking cessation topics such as high risk situations for smoking and the cost of smoking. They also set and reinforced weekly smoking-related goals. In the second role function, and in service of the first, the peer mentors noted spending time establishing a supportive relationship: “I just mostly reassured her that she can call me with questions and concerns and asked about her weekend and holiday, I want her to feel comfortable with calling me.” Peer mentors noted the link between relationship building and the goal of quitting smoking but in some cases questioned this link: “I think playing cards helps develop an enjoyable relationship for the two of us. I am not sure if it will further the goal of quitting smoking.”

Peer mentors frequently commented on a third and related role function: engaging in activities whose main purpose was to provide companionship. The aim of these activities often overlapped with rapport building, but was distinct when pursued for their own sake. Peer mentors noted that participants were interested in activities such as playing cards, doing crafts, eating at local restaurants. In fact, mentors often said/wrote that many participants were more amenable to spending time together doing things that were not demanding or related to smoking cessation. Thus “companionship” sometimes became disconnected from the study’s smoking cessation agenda.

A fourth role function involved engaging in activities, usually initiated by the mentors, that were focused on improving participants’ lives and health other than smoking. Different from rapport-building or companionship (although one activity could serve multiple functions) such activities included taking long walks, working on personal appearance, going to the local library. “The things we are going over here - wasn’t part of the study but part of a holistic view of working with a person... The unmet needs were with their diet, with their fitness, and with their social skills…”

Fifth, peer mentors described sometimes becoming advocates for participants. One mentor made calls with her participant to a hotline to help find him housing after he abruptly became homeless. In other cases, the advocacy was to promote smoking cessation (overlapping with the intended primary mentor role function). For example, several peer mentors emphasized the importance of smoking cessation to a participant’s family members or mental health staff and sought their support. In one instance, a mentor interceded between two mental health providers to help arrange an alternate daytime activity that would better support the participant’s smoking abstinence: “It’s difficult… to help him stay stopped [smoking] without intervening on his behalf when it appears that he may be in jeopardy of picking up smoking because he’s lonely, depressed and has no friends.”

Challenges of Promoting Smoking Cessation

A second major theme was the challenge of promoting smoking cessation. Although all participants agreed voluntarily to join a smoking cessation intervention, many seemed more interested in general social support than in help with quitting smoking. “He had a lot of other things going on, like family and girlfriends…and it was hard for me to bring it back to smoking all the time… “

The peer mentors observed their own reluctance to push the topic of smoking in the face of participants’ preference not to address it. “I’m so glad that she finally called back. We did not discuss a lot of smoking-related issues because I wanted her to be more comfortable with calling me just to chat, because then I can work smoking into the conversation. I’m afraid she won’t call if she’s worried it will always be questions about smoking.” Yet at other times mentors felt they should have more actively promoted quitting: “Getting participants to set quit dates – I felt I wasn’t firm enough; I was worried that I would push the participant away from the study.”

Conversely, in a minority of cases, peer mentors felt that their interactions became too narrowly focused on smoking, limiting their relationship and, paradoxically, impeding the opportunity to leverage rapport for quitting. “I tried to always bring it back to smoking and stick to that. In some ways I wish I would have been more like a friend figure.”

The biggest challenge, noted by all peer mentors, was dealing with participants’ difficulties in reducing and quitting smoking, especially fluctuating motivation and diverse barriers. “I thought perhaps one of my people was going to quit smoking and he didn’t…it was kind of surprising that his interest just kind of petered out.” Another remembered, “I had someone at the program who specifically said, ‘I’m not quitting.’ I realized how hard it is to quit smoking.”

Many mentors were distressed by participants’ not quitting or staying abstinent. “I was really down and a little resentful that he hadn’t followed our plan to call me if he was about to use [smoke]….I guess he was never actually ready for an uninterrupted period of cessation.” Even more strongly, one lamented, “This woman is killing herself [by smoking] and I feel helpless! If she were threatening suicide, we could help, but since she is doing it quietly, we can’t.” Such feelings appeared to test the resolve of the mentors: “I was actually disappointed [that he resumed smoking] but I tried really hard to just focus on optimism and encouragement. I just keep hoping he didn’t see through me.” One mentor was unsure about continuing with the participant: “It reached a point towards the end that I really felt like I didn’t want to go on [working with him because he was not interested in quitting]. I was wasting my time.” Sometimes the peer mentor felt personally responsible: “I feel like I am not doing my job well because neither of my participants seems as motivated as some other participants. “

Some peer mentors found that reflecting on their own histories provided a useful perspective. “I need to consciously accept that she is no longer motivated to quit smoking without getting discouraged or disappointed as it is natural to the stages of change. I am glad that she is comfortable with being honest about her lack of motivation, and based on my own history. [I] can definitely relate.”

Interpersonal Issues

Peer mentors also described interpersonal challenges. One prominent one was posed by participants’ psychotic symptoms. “I am concerned that I need to decipher what is real and what is not since she is sometimes delusional when she speaks.” In some instances, mentors could use their own history empathically, “Even though he didn’t come out and say it, I could tell by his expressions on his face and the way he would look away from me because I’ve had issues with that myself so I knew what the signs were.” But at other times certain symptoms were beyond their own illness experience: “The paranoid symptoms was something new…. There was one time where she [was] crying and said there were people outside the door that were making fun of her. I was shocked, that was one I didn’t know how to handle, that I wasn’t used to.”

A second interpersonal challenge was figuring out how to respond to aspects of a participant’s self presentation that were off-putting. One mentor wrestled with whether a participant was lying, saying, “He is quite a story teller. I’m beginning to realize most of the things he tells me are completely false.” Some such behaviors or views made peer mentors feel uncomfortable or offended: “He is racist and at times he uses the ‘N’ word… I told him he would make most people uncomfortable, angry, or hurt… I found it really hard to tolerate today.”

Other times, participants’ behavior led peer mentors to feel rejected. “Especially in the beginning, he never wanted to meet…that felt bad, ‘What was wrong with me and my skills this guy didn’t even want to meet with me?’” And another recalled, “We had an incident where I had to cancel an appointment and she got very upset and was like ‘I never want to talk to you again.’ That was like, ‘Oh my God, I must have screwed up so bad.’ “

And, different yet again, occasionally mentors were challenged by participants’ inadvertently triggering the mentors’ symptoms or history. “It brought up again my social anxiety, my difficulty in getting to know someone and feeling comfortable with small talk and chit chat.” And, “He brings up traumatic events a lot…it causes me to recall when I have been handcuffed and forced to the hospital.”

“Peer-ness” in the Peer Mentor Experience

The peer mentor role in this project was unusual in that had two aspects of “peer-ness”: history of smoking and psychiatric illness. Peer mentors’ reported talking about the former much more than the latter: “They never really asked me personal questions about my mental illness or hospitalizations or anything like that. I think they just took my word for it that I had that experience… They did ask me about how I quit smoking.”

At times a shared psychiatric illness history did foster connection: “There was some more commonality there…it made it a little more comfortable to approach and work with him, even though we have different diagnoses.” However, more often it was not a very active part of the relationships. In part this was due to uncertainty: “I never brought it [my mental illness] up much because I didn’t know how they felt [about their mental illness.].” However, more often mentors attributed it to the considerable differences between their and participants’ lives: “I think that [mental health topics] floated away pretty quickly…. [When] they see the peer mentors, they are people who have jobs, they come in, they have their name badges…. They don’t seem like, upon first impression that these are people who are mental illness, too… because we are not actively struggling with as many symptoms, or we are, but [we’re] keeping it under wraps in a different way.” And, “They were kind of boxed in to a given situation…Whereas I felt that the reason that I have had successes in my own life that have transcended the mental health dysfunction is because I am able to go in a lot of different directions.”

Peer mentors raised questions about the boundaries delineating their relationships. Mentor training and supervision defined their role as including substantial mutuality and bi-directional sharing. At the same time, the primary focus was to serve the participant’s needs, and a professional demeanor was expected. One mentor described the difficulty combining these two aspects, “We finally came up with a word, ‘peer-fessionalism.’ One day I guess I did something that people said was not professional and I said, ‘Well, I thought I was supposed to be a peer’…so you know, defining what it means to be a peer has been interesting.” This challenge came up in diverse ways. “I fear that I over-shared and don’t think I should do that again; he seemed concerned about how I was doing when I saw him which is not a dynamic I want.” Almost the opposite, another said, “Boundaries! I want to be ‘friends’ and she does, too. After the supervisory meeting, I understand better why we can’t be… It is disappointing… I am learning.”

These complexities of “peer-ness” were also reflected as the mentor-participant relationships ended. Mentors sometimes felt conflicted: “I felt a bit uncomfortable when he asked what happens when the study is over. He asked if we will still talk, and I said maybe from time to time.” Another noted, “One part of me is glad it is over. I feel relieved to no longer have the responsibility that comes with trying to help someone change.”

Rewards of the peer mentor experience

Despite the various challenges, peer mentors also described numerous rewarding aspects of their roles. First, they felt gratified that the participants appreciated them, and fortunate to get to know the participants as individuals. “Being with him reminded me that peers are affected by mental illness in varying ways… [I] didn’t expect to be so deeply affected by interacting with peers in a quit smoking program.”

Second, mentors found it rewarding to work on an important problem, smoking, and to see participants make genuine attempts to reduce or quit, and they often admired participants for their efforts. “He is teaching me the different phases a mentee can experience in their recovery journey…He wants to and is beginning to prepare himself, which is a huge inspiration to me ….He really is walking his own journey and I’m simply there to cheerlead, educate, and support. It’s really a great experience.”

Third, they found it especially rewarding to be able to use their own lived experience. “It was rewarding and challenging…. to get paid in just being myself in working with another individual just by sharing my experience… in a way it is like, how good can it get? “ One added, “[It] felt useful to be able to turn a negative in my life (smoking) into a positive by helping someone quit and likewise use the experience of having mental illness to help someone else.”

Fourth, the peer mentors noted their own personal growth. Some observed that the role had helped them become more compassionate, kind, or patient. They appreciated learning specific counseling skills. Additionally, most said that the peer mentor job was beneficial for their careers. Furthermore, working as a peer mentor role reinforced their quitting and increased their resolve not to smoke again.

Discussion

Peer specialists of varying types are being widely integrated into mental health services. Peer-delivered interventions promoting specific health behavior change, such as this one, involve different roles than support-only peer specialist roles. Therefore, the documentation and reflections of the mentors in our intervention may have a number of implications for the future use of similar peer mentors, regarding smoking or other health behavior change.

First, the roles that mentors occupied and therefore the complexity of their interactions with participants were more numerous and varied than we anticipated. We underestimated how much the peer mentors would end up engaging in non-smoking-related companionship. This role expansion was primarily driven by participants’ mixed feelings about focusing on smoking, their ambivalence about quitting, and their substantial unmet social needs. When using peer mentors in the future, it may be helpful to anticipate and plan for similar patterns, especially when the intervention focus is a health behavior that is difficult to change. Issues of role clarity and definition have been also noted in previous reports about the experience of peer specialists in interventions not focused on health behavior change, suggesting that this is a common issue in peer-delivered services (Moran, et al., 2013).

Second, part of the mentors’ role complexity was an inherent tension between the “peer” and “mentor” aspects. Our peer mentors were given an agenda; it was their job to promote reducing and quitting smoking, but they were to do so through the role of supportive peer. Thus they were not neutral about participants’ smoking although they assumed a supportive and non-judgmental stance. Peer mentors sometimes found this duality difficult, especially when participants lost interest in quitting. Further, they felt participants’ wishes for companionship and social support sometimes usurped the relationship’s anti-smoking purpose. Thus their experience contrasts with that of peers in programs with a primary goal of providing general social support or mutual assistance (Sledge, et al., 2011). Yet, peers are now being deployed more often in structured programs like ours, where they are expected to use their lived experience to be a supportive, empowering ally, but are also expected to promote lifestyle change (Goldberg, et al., 2013; Jerome, et al., 2012; Swarbrick, 2013). Additionally, at least in our case, the peer aspect was sometimes de-emphasized due to the agenda-peer paradox.

Third, it is important to underline that one cannot assume shared identity or rapport based solely on both persons having psychiatric illness as such experiences and outcomes are highly varied. Further, working as part of a professionally-delivered intervention, peer mentors are usually seen as extensions of the professional staff which may contribute to the perceived differences between peer mentors and participants. As a result, the peer mentors in our study experienced a gap between their own lives and those of participants despite sharing a history of psychiatric illness and smoking. However, this gap may not be necessarily counterproductive. Peer specialists with psychiatric illness employed in counseling roles are typically selected for modeling “recovery” and are therefore likely to be more functional and less symptomatic than the persons they serve. Role model social comparison, defined as “an increased sense of hope and motivation as a result of comparisons with another who is effectively managing their illness,” is known to contribute to the effectiveness of peer interventions (Proudfoot, et al., 2012). Therefore, it is important to consider how social comparison is operating in order to capitalize on the benefits it may afford. The question of how, when, and what peer providers disclose to their service recipients about their own psychiatric illness history has been noted in the literature as one with which peer providers struggle (Moran, et al., 2013). In our study, self-disclosure about smoking and quitting was easier and more relevant that that than around psychiatric illness so this aspect of the “peer -ness” became dominant.

Fourth, our peer mentors found the role rewarding despite the challenges, consistent with the experience of peer providers in other studies. We observed a temporal sequence in our data regarding peer mentors’ attitudes over the course of the six month intervention: initially they were optimistic about their work with participants; this was followed by disappointment in some cases as barriers to quitting became apparent. By the end of the intervention period they tended to adopt a more measured view and expressed appreciation of having serving in the role. The “Helper therapy” principle—the idea that a helper receives benefit from their helping–is known to occur in peer support for people with psychiatric illness and in other human service roles (Proudfoot, et al., 2012; Reisman, 1965). We did not directly measure this construct but found peer mentors’ comments consistent with it.

Fifth, the peer mentors in our study did not experience some of the frustrations that have been reported by peer providers in other studies such as tensions with co-workers who are not peers or the lack of a recovery orientation in the workplace (Ahmed, et al., 2015; Moran, et al., 2013; Walker & Bryant, 2013). Given that a lack of role clarity is a challenge consistently mentioned in other reports, it is likely that the highly structured nature of the intervention minimized these potential problems. Also, our peer mentors were not employed at the agencies where the services were delivered and reported directly to research staff who supported and directed their work.

Limitations of this study include the relatively small sample of peer mentors. In addition, the persons working as peer mentors here may not be representative of persons who assume this role. Also, the experience of the peer mentors may have been shaped by the professionally-led smoking cessation intervention to which the peer component was added. An intervention which offers a different approach to smoking cessation (e.g. focused on smoking cessation medication) or which addresses a different health behavior (e.g. wellness more broadly) may lead to different peer mentor experiences. Strengths of the study include the two sources of data, interviews conducted shortly after the clinical work was finished and interaction logs completed during the intervention, which allowed us to deepen and triangulate our understanding of the peer mentor experience. In addition, the intervention was delivered in 3 successive cohorts enabling us to gather data over the course of several implementations.

Conclusions

There may be inherent tension in the peer mentor role between providing person-centered support and promoting behavior change in interventions of this kind. Anticipation of role complexity can help peer mentors navigate their roles effectively.