Perceived Barriers and Facilitators to Implementation of Peer Support in Veterans Health Administration Primary Care-Mental Health Integration Settings

Abstract

Peer support is increasingly recognized as consistent with the goals of integrated primary care and is being implemented in primary care settings as a patient-centered approach that increases patient activation and access to care. Within the Veterans Health Administration (VHA), peer support specialists (PSSs) have traditionally worked in specialty mental health settings and only recently started working in Primary Care-Mental Health Integration (PC-MHI) settings. Prior research has identified implementation challenges, such as role confusion, when integrating peer support into new settings. In this qualitative descriptive study, we conducted semi-structured interviews on perceived barriers and facilitators to implementing peer support in PC-MHI with 25 key stakeholders (7 PSSs, 6 PSS supervisors, 6 PC-MHI providers, and 6 primary care providers). We used conventional content analysis to code responses within four a priori implementation categories: barriers, initial facilitators, long-term facilitators, and leadership support. Perceived barriers included poor program functioning, inadequate administrative support, role confusion, and negative stakeholder attitudes. Key perceived facilitators of initializing and maintaining peer support were similar; administrative support was emphasized followed by program functioning and team cohesion. Stakeholder buy-in and access/visibility were perceived to facilitate initial implementation, whereas evidence of success was believed to facilitate maintenance. Stakeholder buy-in and administrative support were considered key elements of leadership support. Results were consistent with prior research from specialty mental health settings, but identified unique considerations for PC-MHI settings, particularly clarifying the PSS role based on local PC-MHI needs, obtaining buy-in, and facilitating integration of PSSs into the primary care team.

Peer support is a recovery-oriented approach in which “a peer with a history of mental illness who, having experienced significant improvement in their condition, offers services and support to a peer considered to be not as far along in their own recovery process” (Jain, McLean, & Rosen, 2012, p. 481). The shared experience is a unique feature that allows the patient to not only feel understood and receive practical and social support from someone who has been in similar circumstances, but also have a positive role model for recovery (Gidugu et al., 2015).

The goals of peer support include promoting coping skills and problem-solving for self-management of illness, sharing lived experiences and recovery stories to instill hope, and encouraging engagement with treatment and community resources (Chinman et al., 2014). The specific roles and duties of those who deliver peer support, commonly termed peer support specialists (PSSs), vary widely by setting (Cronise, Teixeria, Rogers, & Harrington, 2016) but include a combination of direct patient work and administrative tasks (Jacobsen, Trojanowski, & Dewa, 2012). Peer support in specialty mental health settings improves clinical and recovery outcomes among patients (Repper & Carter, 2011) and positively impacts other providers on the team and the overall system (Chinman, Young, Hassell, & Davidson, 2006; Solomon, 2004).

Peer Support in VHA

Peer support is believed to be especially powerful within military and Veteran populations given the emphasis on unit cohesion and teamwork within military culture, and increased credibility and trust from having a shared experience of military service (Money et al., 2011). The Veterans Healthcare Administration (VHA), the nation’s largest integrated healthcare network, adopted peer support as part of a shift toward a more patient-centered, strengths-based approach to mental health care (Goldberg & Resnick, 2010). VHA began hiring PSSs in 2005 (Chinman et al., 2008) and is now the largest single employer of PSSs (Rogers & Swarbrick, 2016). VHA PSS positions must be filled by Veterans who self-identify as “recovered or recovering from a mental health condition” and are trained and certified by a VHA-approved training organization (Chinman, Henze, Sweeney, & McCarthy, 2013, p. 14). Peer support in VHA was initially focused on Veterans with serious mental illness (SMI; Goldberg & Resnick, 2010) but has expanded to a variety of patient populations, such as substance use (Tracy, Burton, Nich, & Rounsaville, 2011) and homelessness (Tsai & Rosenheck, 2012).

Peer Support in Integrated Primary Care

Researchers have recently begun to examine the role and outcomes of PSSs in primary care settings (Mayer et al., 2016). Peer support specifically for diabetes management has been studied extensively by Peers for Progress (see Acheson & Fisher, 2015; Fisher et al., 2015). Otherwise, little research has been conducted on peer support in primary care (Daaleman & Fisher, 2015), or in VHA integrated primary care settings in particular. The lack of research is due to the newness of peer support in VHA primary care; of approximately 1,100 PSSs employed by VHA, “only a few” have worked in primary care (Chinman et al., 2017a, p. 2). At the same time, the potential contributions of peer support to the patient-centered medical home model are being increasingly recognized (Daaleman & Fisher, 2015; Fisher et al., 2015). Peer support has been deemed an “untapped resource” (Tellez & Kidd, 2015, p. 84) given the potential for PSSs to facilitate wellness- and recovery-oriented whole health care to help address the mental and behavioral health needs of primary care patients (Swarbrick, 2013; Swarbrick, Tunner, Miller, Werner, & Tiegreen, 2016). For example, PSS may serve as wellness coaches and deliver interventions such as Whole Health Action Management (see Swarbrick, 2013; Swarbrick, Murphy, Zechner, Spagnolo, & Gill, 2011; Swarbrick et al., 2016).

VHA began its Primary Care-Mental Health Integration (PC-MHI) initiative in 2007 to integrate mental and behavioral health care into the primary care setting (Kearney, Post, Zeiss, Goldstein, & Dundon, 2011). PC-MHI facilitates early detection, prevention, and easy to access intervention for common mental and behavioral health conditions, while decreasing stigma (Post, Metzger, Dumas, & Lehmann, 2010; Zeiss & Karlin, 2008). Peer support has great potential to supplement the work of primary care and PC-MHI providers to help increase access to care and enhance patient-centeredness. Recognizing the good fit between PC-MHI and peer support, an Executive Order was issued in 2014 assigning PSSs at 25 VHA sites to primary care to improve mental and behavioral health care among primary care patients. Sites were recruited and initial implementation occurred throughout 2016 and 2017 (see Chinman et al., 2017a).

Implementation Challenges with Peer Support

As with any new model of care, introducing peer support services into new clinical settings often results in significant challenges. For example, early implementation of peer support in VHA specialty mental health settings was hampered by role confusion (i.e., confusion among PSSs and staff regarding PSS job duties), staff resistance (e.g., negative attitudes toward PSSs), and unequal treatment of PSSs (e.g., poor or no compensation; Chinman et al., 2006, 2008). Research from community mental health settings has identified challenges related to lack of role clarity (Crane, Lepicki, & Knudsen, 2016) as well as inadequate training for PSSs, staff, and supervisors (Cabral, Strother, Muhr, Sefton, & Savageau, 2014). A review by Vandewalle et al. (2016) identified numerous perceived barriers to implementation of peer support in mental health settings from the perspective of PSSs, including role confusion, stigma and negative attitudes from staff, confusing boundaries between PSSs and patients, poor team functioning, limited opportunities for PSS career advancement, and inadequate training, supervision, and logistical support for PSSs. However, this review focused specifically on mental health settings, assessed only barriers to implementation, and included studies sampling only PSSs.

Two recent qualitative studies examined implementation of peer support in community PC-MHI settings. Mayer et al. (2016) interviewed 18 staff from four clinics that varied in size, extent of integrated care, and type of peer support program. Facilitators of successful integration included co-locating PSSs in primary care, using technology to coordinate care (e.g., electronic medical record), PSSs and staff communicating frequently (e.g., spontaneous huddles), and having a clear plan for implementation with regular team meetings. Barriers included limited clinic space, PSS role confusion, low trust of PSS by staff, and unclear financial sustainability. Siantz, Henwood, and Gilmer (2016) reported findings from interviews and observations of 24 clinics with various integrated care models, 15 of which had peer support programs. Challenges in peer support implementation included stigma among racial/ethnic minority patients toward PSSs with mental illness, lack of perceived need for PSSs among clinic leaders, limited training and supervision resources for PSSs, and lack of guidelines on PSS hiring and roles. These studies begin to shed light on implementation challenges specific to PC-MHI, but research is needed within VHA PC-MHI given the unique features of that healthcare system and patient population.

The Present Study

VHA is expanding peer support services into primary care and PC-MHI to reach more patients (Chinman et al., 2017a) and to capitalize on the contributions PSSs could make with whole health care and wellness coaching (Myrick & del Vecchio, 2016; Swarbrick, 2013). Research has demonstrated the importance of obtaining feedback from a variety of key stakeholders when implementing peer support in new clinical settings to minimize difficulties and maximize likelihood of success (Chinman et al., 2006, 2017b; Vandewalle et al., 2016). Thus, the present study was designed to explore perceived barriers and facilitators to implementation of peer support in VHA PC-MHI settings. We address a gap in the literature by examining implementation of peer support specifically in VHA PC-MHI settings, sampling a variety of key stakeholders (PSSs, PSS supervisors, PC-MHI providers, and primary care providers [PCPs]), and assessing perceived barriers as well as facilitators of implementation. Findings will benefit PSSs, staff, clinic/hospital leaders, system administrators, and policy makers (Davidson, 2015) by offering guidance regarding potential pitfalls to avoid and strategies to facilitate more efficient and successful implementation of peer support in PC-MHI settings.

Method

Study Design

We conducted an exploratory descriptive qualitative study (Sandelowski, 2000) using semi-structured interviews to elicit key stakeholders’ perspectives on perceived barriers and facilitators to integrating peer support into the PC-MHI setting. The local institutional review board approved all study procedures.

Participants and Setting

Participants included four types of VHA employees: 7 PSSs, 6 PSS supervisors, 6 PC-MHI providers, and 6 PCPs. Implementation questions were not administered for one PC-MHI provider, so the final sample consisted of 24 participants. All participants were recruited from VHA facilities in central New York, with the majority from a single medical center, except for four PSSs recruited from VHA facilities in the Northwest and Midwest (regions chosen randomly) to ensure greater diversity with respect to gender and urban/rural location. Inclusion criteria for PSSs were (a) completed state or VHA PSS training and/or (b) currently serve as a VHA PSS. The inclusion criterion for PSS supervisors was being a current or former (within past two years) clinical supervisor of a PSS. The inclusion criterion for PCPs and PC-MHI providers was being a current VHA employee working in primary care.

At the time of data collection, very few VHA sites had peer support in PC-MHI, as the national pilot project had just started (Chinman et al., 2017a); thus, no PSS participants worked in PC-MHI. The study site began piloting peer support in PC-MHI on a part-time basis during the current study (after PSS interviews were done). Therefore, some of the participants (all 6 PCPs and 4 of 6 PC-MHI providers) may have had some familiarity with the new program (approximately 1–3 months exposure), but were generally still learning about peer support.

Recruitment occurred via email invitations from the first author that described the study and invited interested individuals to contact her to schedule. The response rate was good, as 7 of 8 PSSs, 6 of 7 PSS supervisors, 6 of 12 PC-MHI providers, and 6 of 7 PCPs invited to participate enrolled in the study. We found that saturation was reached once we interviewed six participants per provider type, which is consistent with prior research (Guest, Bunce, & Johnson, 2006). We interviewed a seventh PSS who was recruited prior to confirmation of saturation.

Of the 24 participants, 13 (54.2%) were female; no other demographics were collected to protect privacy given the low sample size and the majority of participants being from a single site. Most PSSs worked at VHA medical centers in urban settings, although one was in a rural setting and one was at a community-based outpatient clinic. PSSs reported an average of 3.1 (1.9) years in the PSS role and 5.3 (2.8) years at their current facility. All PSS supervisors were team leads in outpatient specialty mental health services. Four were current PSS supervisors, and two were former PSS supervisors. PC-MHI providers reported an average of 3.1 (1.7) years in PC-MHI and 6.2 (6.3) years at their current facility. PCPs reported an average of 16.7 (3.6) years as a PCP and 10.3 (5.5) years at their current facility.

Procedure

We used a rapid assessment approach to interviewing (e.g., Sobo et al., 2002, 2003) using semi-structured interviews with open-ended questions with specific follow-up probes (see Table 1). This team-based approach is an iterative process of data collection and analysis, in which preliminary findings are used to guide future adjustments to the protocol, such as interviewing additional participants (Beebe, 2001). In this case, based on the initial interview data from PSSs regarding the importance of team acceptance, we decided to sample an additional stakeholder group, PSS supervisors, who have experience with integrating PSSs into teams.

Table 1.

Interview Questions by Implementation Category

Category	Interview Questions
Barriers	What would be the greatest barriers to incorporating peer support services into PC-MHI?  *Any barriers specifically for [Peer Support Specialists/ PC-MHI Providers/ Primary Care Providers and other PACT team members]? (tailored for the provider type with their own specialty) For PSS supervisors: What were the greatest barriers or challenges to incorporating peer support services into your team?  *Any barriers specifically for the Peer? Barriers for other staff?
Initial supports	What supports, tools, and/or resources would be needed to ensure successful integration of peer support services into PC-MHI and primary care?  *Any supports needed specifically for [Peer Support Specialists/ PC-MHI Providers/ Primary Care Providers and other PACT team members]? Additional question for PSSs: What supports would need to be in place for clinics to incorporate peer support services into PC-MHI? For PSS supervisors: What supports, tools, and/or resources helped facilitate successful integration of peer support services into your team?  *Any supports needed specifically for the Peer? Supports for other staff? What were the most effective strategies for helping to facilitate acceptance of the Peer by the team? Do you have any other ideas or suggestions, maybe things you didn’t do but you wish you would have, for how to facilitate the Peer being accepted by other staff as a valuable team member?
Maintenance supports	What supports, tools, and/or resources would be needed to maintain successful peer support services in PC-MHI and primary care in the long run? For PSS supervisors: What supports, tools, and/or resources are helping to maintain successful peer support services on your team in the long run?
Leadership support	What type of leadership support would facilitate integration of peer support services into PC-MHI and primary care? For PSS supervisors: What type of leadership support would facilitate integration of peer support services into a new team, such as primary care?

Note. Questions marked with an asterisk were follow-up probes asked if needed. PACT = patient-aligned care team; PC-MHI = primary care-mental health integration; PSS = peer support specialist. Questions were nearly identical for PSSs, PC-MHI providers, and primary care providers, with the exception of tailoring the follow-up probes to refer to providers within their own specialty area. Questions were slightly different for PSS supervisors as indicated.

All interviews were conducted by the first author along with the second author and/or a research assistant. Interviews lasted on average 45–60 minutes and were conducted in person for those at the study site (71%) and by telephone for those located elsewhere. Providers were not compensated for participation. Following verbal consent, a brief 1–2 minute verbal overview of peer support or PC-MHI was provided. Participants completed a self-report background questionnaire followed by an interview. Semi-structured interview guides (see Table 1) were adapted from an implementation-focused process evaluation (Hagedorn et al., 2014) and assessed factors relevant to the Adoption, Implementation, and Maintenance (AIM) dimensions of the RE-AIM framework (Glasgow, Vogt, & Boyles, 1999). Given our focus on informing future implementation efforts, rather than evaluating an existing program, we did not assess Reach and Effectiveness, and instead targeted four main areas of interest for AIM: barriers to adoption/implementation, initial facilitators to begin a program, long-term facilitators to maintain a program, and leadership support. In the interest of staff privacy, we did not audio record the interviews. Instead, research staff took field notes, including direct quotations and paraphrased material (Sobo et al., 2002, 2003).

Analysis

Field notes were consolidated to produce one final set of interview data for each participant. Data were entered into ATLAS.ti (ATLAS.ti Scientific Software Development GmbH, 2014) to facilitate qualitative coding. Consistent with conventional content analysis, codes were developed inductively based on themes that emerged directly from the data (Hsieh & Shannon, 2005). This approach was modified by initiating coding with the four broad categories selected due to their role as key implementation domains and used to structure the interview: perceived barriers, initial facilitators, long-term facilitators, and leadership support. To develop the codebook, the first author read through all interview data multiple times and identified codes that emerged from the data as well as exemplars of each code, then the first and second authors reviewed the draft codebook together and made revisions. The first and second authors independently coded two interviews and compared codes to ensure adequate calibration and discuss any discrepancies. The first and second author then independently coded the remaining interviews. Kappa, used to assess inter-rater reliability, was 0.62 for barriers, 0.76 for initial facilitators, 0.71 for long-term facilitators, and 0.74 for leadership support. Thus, coders showed substantial agreement in all categories based on the Landis and Koch (1977) guidelines. Discrepancies were resolved after discussion, and coding was finalized.

Results

Table 2 provides a summary of codes within each of the four categories as well as descriptions of each code and an exemplar quotation illustrating each code.

Table 2.

Description and Exemplar Quotations for Each Code by Implementation Category

Category	Code	Description	Exemplar quotation
Barriers
	Poor program functioning	Confusion or unfamiliarity with peer support program goals, scope, and workflows	Being too broad with what the service is leads to off target referrals or to being overwhelmed with referrals
	Inadequate administrative support	National mandates, unqualified clinical supervisors, and non-user friendly resources for supervisors	[The national office] has taken a lot of autonomy out of the process, for example with the structured interview process and defining the peer role and duties
	Role confusion	Confusion regarding PSS role and inappropriate scope of practice for PSSs	The nebulous description of what peer support is and the undefined role
	Negative stakeholder attitudes	Misconceptions or negative attitudes toward PSS from providers, staff, and patients	Peers have to prove themselves to some team members. Some staff are skeptical, some scrutinize everything about the peer at first
	Peer characteristics	PSS who have skill deficits, are too independent, or are not a fit for the role or setting	Not understanding the issues of splitting and not consulting enough with providers
	Poor team cohesion	Mistrust or disrespect of PSS by team or poor fit with team	With any provider role, some of it is fit and personality. Fit with the team is a key variable. Teams have different energies.
	None/no barriers	No perceived barriers	Personally, I see no barriers
Facilitators (Initial and Maintenance)
	Administrative support	Funding, logistical resources (e.g., office space), training and resources for PSSs and PSS supervisors, professional development and Peer networks for PSSs, appropriate clinical supervision, and autonomy for local programs	One I know that would be like pulling teeth at our hospital, I don’t know about other hospitals, is having a room for the peer support…to meet with people or make phone calls. Here space is at a premium.
	Team cohesion	Trust, respect, and acceptance of PSS by team; reciprocal communication; and inclusion of PSS in team meetings, projects, and program improvement	Having Peers feel like they are an equal part of the team and belong on the team
	Program functioning	Key aspects of peer support program including goals, scope, workflows, and feedback loops	Developing a structure for referrals and a list of concrete things [the PSS] could do
	Stakeholder buy-in	Strategies to increase buy-in for peer support among providers and staff	They’re got to experience it themselves, get the Peer Support Specialist in the room with them and the patient
	Access and visibility*	PSSs being easily accessible, reminding staff to refer, and providing handouts for providers and patients	Maybe a pager or cell phone the Peer carries so we can easily contact them
	Peer characteristics	PSS having the appropriate knowledge, skills, and personality, and being a good fit for the role and the setting	Someone in primary care needs to be flexible, approachable, personable, professional but not too professional, friendly but not too friendly
	Role clarity	Ensuring staff and PSSs understand PSS’ role and scope of practice	Setting up the right expectations for everyone so all team members know what peer support does and has clear expectations of their role and expertise
	Evidence of success**	Patient and provider satisfaction and sharing encouraging data (local or from literature) with staff and leadership	Have Peers meet with heads of departments, for example, of primary care, giving them examples of success
Leadership support
	Stakeholder buy-in	Types of key stakeholders whose support is essential and expressions of support from local and national leadership (e.g., valuing peer support, public commitment)	Leadership saying peer support is here to stay, we need it, it’s valuable
	Administrative support	Provision of funding, clinical supervision, and autonomy for clinics to tailor programs to suit local needs	A problem we have here in working as peer supports is that everybody likes the idea of having them, but nobody likes the idea of doing the extra supervision and signing the notes
	Evidence of success	Demonstrating success of the peer support program through outcome data	There has to be some visible outcome and you have to be able to demonstrate effectiveness to leadership to support the positions

Note. PSS = peer support specialist.

^*Identified as a facilitator of initial integration only.

^**Identified as a facilitator of long-term maintenance only.

Perceived Barriers

Participants identified six types of potential barriers to adoption/implementation of peer support in PC-MHI (see Table 2): poor program functioning, inadequate administrative support, role confusion, negative stakeholder attitudes, peer characteristics, and poor team cohesion.

Aspects of program functioning identified as potential barriers included confusion or unfamiliarity with the concept and goals of peer support as well as the parameters of the peer support program (e.g., appropriate referrals, workflows). Participants cautioned against a poorly articulated program goal/focus and recommended that the program scope not be so broad that PSSs are inundated with referrals, but also not be so narrow that it inadvertently limits referrals.

For administrative support, four of six PSS supervisors perceived national mandates regarding peer support as constricting. One said, “It’s not one size fits all…It’s better to have the local team decide what is most helpful.” PSS supervisors also had difficulty utilizing national resources, as they perceived many materials to be time-consuming and either “too prescribed” or too vague. One explained, “If one more person tells me to just look at the [website] I will scream. There is too much on [there], it is overwhelming to sort through, a lot of it is outdated, and what is there is not helpful in explaining the day to day operations details.” Finally, clinical supervisors who were unqualified, inexperienced, from another program/clinic, or uncomfortable with the self-disclosure and boundary issues unique to peer support were cited as problematic.

Role confusion on the part of both PSSs and staff was also cited as a major barrier. Ambiguous or inappropriate roles with poorly defined boundaries were problematic. Another concern was PSSs being asked to practice above (e.g., therapy that a clinician should do) or below (e.g., administrative tasks that a scheduling clerk should do) their scope of practice.

Regarding negative stakeholder attitudes, participants, including two PCPs, reported that primary care staff may not value or be open to peer support because they either did not understand its potential contributions to patient care or had negative attitudes due to misconceptions about PSSs. Three PCPs had concerns that patients may not be receptive to peer support in PC-MHI due to not understanding the concept or not accepting help in general.

Characteristics of PSSs that were reported as potentially problematic included having skill deficits that do not improve with remedial training, being “too independent” (e.g., not seeking supervision enough), and not being the right fit with the PSS role or primary care setting.

Participants raised several issues related to team cohesion as potential barriers. PSSs “seem to be discounted or not given the same weight as other staff because they don’t have a degree” [PC-MHI provider]. Lack of respect for PSSs may impede their receiving trust and acceptance from the team. Participants raised boundary concerns, noting the “fine line” of how much of the PSS’s mental health history staff truly needed to know. Another potential barrier, raised by two PSS supervisors, one PC-MHI provider, and one PCP, was mental health clinicians being defensive or territorial if PSSs were “seen as competition” who may replace therapists.

One PSS and one PCP perceived no barriers for integration of peer support into PC-MHI.

Perceived Initial and Long-term Facilitators

There was high overlap between factors identified as initial versus long-term facilitators of implementation, so these categories are discussed together, but unique findings are noted. Perceived facilitators of initial integration and maintenance of peer support in PC-MHI were (see Table 2): administrative support, team cohesion, program functioning, stakeholder buy-in, peer characteristics, and role clarity. Access and visibility was a facilitator of initially establishing the program, and evidence of success was an important facilitator of long-term maintenance.

Administrative support was the most commonly reported initial and long-term facilitator. This included funding for PSS positions, logistical support (e.g., private space in primary care with a desk, computer, and telephone), and access to the electronic medical record, scheduling support, and group rooms or vehicles. Access to appropriate training for PSSs was emphasized when initially integrating peer support into PC-MHI, as PSS training needs span a range of areas, including general facility onboarding, peer support certification, mental health and primary care specific knowledge, and relevant evidence-based peer-delivered interventions. Training on boundaries, self-disclosure, and ethics was highlighted as essential given the unique PSS role.

Access to continuing education and professional development was emphasized for long-term maintenance. A PSS supervisor and PSS suggested having a lead PSS position to supervise PSSs. PSSs and PSS supervisors in particular encouraged the availability of Peer networks to allow for “co-reflection or co-supervision, time to debrief regarding obstacles, barriers, pitfalls” as well as “guidance from other experienced peers on how to do peer support in real life.”

Good clinical supervision for PSSs was viewed as essential at program startup and over the long-term. A PSS suggested using “somebody kind of in between mental health and primary care,” such as a PC-MHI provider. Clinical supervision is needed to help PSSs process any patient matters that may be triggering and navigate challenging boundary and ethical issues. Supportive administrative supervisors who understand the need for PSSs to take sick leave during times of exacerbated mental health symptoms were advised. PSS supervisors expressed need for helpful, easily accessible, efficient training resources for supervisors. Finally, the need for guidelines and national policies was recognized, but participants urged national and local leadership to allow “autonomy for the team and the local program to fit the Peer’s services with the needs of the team… based on what the population needs” [PSS supervisor].

Participants identified several aspects of program functioning that would support initial implementation. Staff first need general education “about what a peer is and what their job is, what they do, for the team and for Veterans” [PSS], then education on program specifics (e.g., appropriate referrals, workflows). Two PCPs requested diversity in PSSs to allow matching patients based on similar demographics, background, or condition/diagnosis. Acknowledging the “big difference between the directive or policy and how you actually do it in real life,” PSS supervisors conveyed the need for “nuts and bolts” guidance for administrators and supervisors developing the program, including real-world examples of how PSSs are used in PC-MHI.

Participants highlighted different aspects of program functioning to facilitate program maintenance. PSSs should seek out referrals, provide regular reminders that peer support is available and be visible and easily accessible. PSSs should maintain an active feedback loop (e.g., verbal report or progress note sent to PCP) because “closing the circle following the referral increases the likelihood of future referrals” [PC-MHI provider]. Ongoing efforts to improve the program (e.g., soliciting feedback from PSSs and primary care staff) were advised.

Team cohesion was an important facilitator of successful implementation initially and over time. Participants highlighted the need for respect, support, and trust for the PSS by the team. For true collaboration, a PCP and a PSS supervisor explained that the PSS should be treated as an equal and viewed as “an integral part of the team,” “not just an accessory” who “doesn’t really count.” Participants identified several strategies by which to facilitate integration into the team, including PSSs introducing themselves and sharing their background with staff, presenting at staff meetings, regularly attending team meetings, actively seeking to collaborate with staff, and providing feedback on referrals. A PSS supervisor suggested “program building as a team” to foster a sense of teamwork while also driving continuous improvement efforts. To maintain peer support over time, participants emphasized acceptance of the PSS by the team, including the PSS in team meetings, and good communication among team members.

Stakeholder buy-in was discussed primarily as a facilitator of initial implementation. The most straightforward strategy to increase buy-in among primary care staff was to educate staff about how peer support can uniquely contribute to patient care. Participants noted this education may be different in primary care versus mental health settings, as primary care staff may be less familiar with recovery or how PSSs could assist patients with behavioral health concerns. Participants suggested having a PCP as champion to vocally support the program and engender enthusiasm, and simply sharing examples of how peer support helps patients (e.g., testimonials from patients). Several participants, including a PCP, noted increased buy-in when staff witness PSS-patient interactions for themselves. One PSS supervisor recommended coming up “with something that will make the primary care team’s lives easier right away” to show how PSSs can help. Participants noted that sharing empirical evidence from the literature, particularly regarding the utility of peer support for engaging “ambivalent” patients, can also help gain staff buy-in.

Access and visibility was identified as a facilitator of initial integration of peer support in PC-MHI. Participants stressed the need for PSSs to be readily available (e.g., open to warm hand-offs) and easily accessible (e.g., nearby in primary care) for referring providers. PCPs and PC-MHI providers reported that frequent reminders (e.g., emails, internal instant messages), even on a daily basis initially, would facilitate utilization. Participants reported that the PSS being highly visible (e.g., circulating through halls) and having a consistent, full-time presence in primary care would help. Participants recommended distributing educational materials (e.g., handouts, flyers) to both primary care staff and patients to increase their understanding of peer support as well as their awareness of its availability and applicability in PC-MHI.

Role clarity was discussed more so with respect to initial implementation. Participants (across all provider types) emphasized that training was needed to ensure that everyone on the team understood “what the Peer does and doesn’t do, what peer support looks like, what is the [PSS] role….” Participants recommended as much specificity as possible, ideally with “a list of what the Peer can and can’t do” [PSS supervisor] to clarify expectations and role boundaries. While other providers were believed to be most in need of clarification, PSSs also discussed the value in understanding their own role and unique contributions to the team. A PSS supervisor expressed a need for “education on where the Peer role starts and stops and where the clinician role starts and stops” to distinguish the PSS role from those of other team members, such as PC-MHI providers, with whom there could be overlap in types of patients or presenting problems.

Participants identified peer characteristics as a facilitator of initial implementation, but did not discuss this in regard to long-term maintenance. In VHA, PSSs must be Veterans, and as such are viewed by patients to have “more credibility… due to their… knowledge of military lifestyle, deployments, the VA system” [PSS]. Participants reported that PSSs working in PC-MHI settings should have relevant knowledge and skills for the primary care setting, specifically a working knowledge of mental/behavioral health and medical concerns that are prevalent in primary care and skills including rapport building, active listening, goal setting, and problem solving. In PC-MHI, PSSs need to have the right personality for both the PSS role and the primary care setting, which includes being outgoing, confident, flexible, resourceful, comfortable with self-disclosure, eager to learn, well connected in the community, and able to communicate well. Overall, participants stressed the need “to get the right Peer in the right place” [PSS supervisor] for a good fit with the team and the goals of its specific peer support program.

Participants reported that evidence of success was a facilitator of maintaining, rather than establishing, peer support in PC-MHI. Evidence could take the form of provider and/or patient satisfaction and was reflected in terms of both individual success stories and program evaluation data. Participants noted the importance of a feedback loop in which anecdotal examples of success stories could be shared. In addition to empirical evidence from the literature, promising local outcome data would help demonstrate program utility. Participants noted a need to measure impact on quantitative metrics, as such data can help make a compelling case for sustainability.

Leadership Support

Participants identified three primary aspects of leadership support: stakeholder buy-in, administrative support, and evidence of success. Support from a range of key stakeholders was perceived as essential to successful integration of peer support into PC-MHI, including local staff (PCPs, PC-MHI providers, nurses), local primary care and mental health leadership (e.g., chief of primary care), local hospital administrators (e.g., chief of staff, medical center director), and national administrators (e.g., national program office). Participants felt that leadership should have (and ensure everyone else from the top down has) a clear understanding of the concept of peer support and how PSSs uniquely contribute to primary care teams, respect and value PSSs with “awareness that peer support is an equal part of the team and is just as important as others on the team” [PSS], and publicly endorse and convey a strong commitment to having peer support (e.g., by sending hospital-wide emails). Several participants suggested that leadership should require the use of peer support in PC-MHI to facilitate uptake and use of the program.

As discussed previously, administrative support in various forms was viewed as integral to successful implementation. Other valued aspects were patience to allow for hiring the “right match” for a PSS position, willingness to advertise peer support to patients (e.g., on hospital website), administrative release time for ongoing program development, open communication with program leads, and provision of appropriate, readily available clinical supervision for PSSs.

Finally, participants acknowledged that evidence of success, such as improved health outcomes or reduced utilization, would be necessary to ensure continued leadership support.

Discussion

As peer support continues to expand into VHA PC-MHI and other integrated primary care settings (Chinman et al., 2017a; Daaleman & Fisher, 2015), practical guidance regarding implementation is needed to help facilitate successful programs. We assessed factors relevant to the Adoption, Implementation, and Maintenance dimensions of the RE-AIM framework (Glasgow et al., 1999), and the themes raised by participants in the current study can inform future implementation efforts to maximize success of peer support in PC-MHI settings (see Table 3 for summary of key recommendations). In this study, key stakeholders stressed the importance of administrative and leadership support and described several potential barriers as well as facilitators of initial and long-term implementation to mitigate barriers. Many of the findings were consistent with prior research from specialty mental health settings, but we also identified numerous implementation considerations that are specific to the PC-MHI setting.

Table 3.

Recommendations to Help Facilitate Successful Implementation of Peer Support in Primary Care-Mental Health Integration Settings

Administrative support

Ensure adequate logistical support for PSSs including access to private meeting spaces in primary care, a dedicated phone line, a computer, and the electronic medical record Provide PSSs with initial training, continuing education, and opportunities for peer networking and professional development Provide PSSs with a qualified clinical supervisor, ideally a PC-MHI provider with familiarity with peer support Ensure PSS supervisors have training specific to peer support issues (especially boundaries, self-disclosure, and ethics) and access to user-friendly resources Design system-wide guidelines to allow flexibility for individual sites to tailor their peer support programs to suit local needs

Program functioning

Provide primary staff with general education about the concept of peer support as well as education specific to peer support program, including appropriate referrals and workflows Establish a feedback loop between PSSs and referring providers

Role clarity

Educate PSSs and primary care staff on unique PSS role and scope of practice Define a clear scope of work to distinguish PSS role from that of other providers

Team cohesion

Facilitate regular communication between PSSs and primary care team, including staff meetings, team meetings, verbal feedback, co-signed progress notes Involve PSSs and primary care staff in program development and continuous improvement

Stakeholder buy-in

Share testimonials from patients who benefitted from peer support Enlist a physician champion to engender enthusiasm for peer support Arrange for PSSs to shadow PCPs and PC-MHI providers initially and engage with patients in session as appropriate to allow providers to witness first-hand the PSS-patient connection as well as what PSSs actually do with patients Identify a group, clinical pathway, subpopulation that PSSs could assist with to quickly demonstrate the value of peer support to PCPs

Access and visibility

Ensure that PSSs are easily accessible (by phone, email, in-person) and highly visible (circulate through clinic halls periodically) within the primary care clinic Send frequent reminders of PSS availability (emails, instant messages) to primary care staff Distribute marketing materials for primary care staff (when to refer) and patients (what it is)

Peer characteristics

Hire PSSs who demonstrate flexibility, confidence, resourcefulness, and ability to communicate and network well

Evidence of success

Collect local program evaluation data as well as anecdotal reports of success to share periodically with local leadership and primary care staff

Note. PC-MHI = primary care-mental health integration; PCP = primary care provider; PSS = peer support specialist.

Administrative and Leadership Support

Administration support was consistently described as essential to aid implementation of peer support in PC-MHI. This includes logistical resources that may be difficult to secure, such as private meeting space in primary care and funding for PSS positions (Chinman et al., 2008; Mayer et al., 2016). Training and professional development opportunities for PSSs were emphasized by PSSs and PSS supervisors as integral to long-term success, and research has shown that a significant minority of PSSs perceive unmet training needs (Cronise et al., 2016; Chinman, Salzer, & O’Brien-Mazza, 2012). In PC-MHI training should cover topics relevant to PSSs working in any setting (e.g., recovery orientation; Chinman, Shoai, & Cohen, 2010), and setting-specific topics such as primary care culture and common chronic medical conditions.

Participants stressed the importance of appropriate clinical supervision for PSSs in PC-MHI, meaning supervisors ideally would have experience in the PC-MHI setting and familiarity with peer support. Supervisors who are new to working with PSSs report challenges due to not understanding PSSs’ capabilities and PSSs having unspecific performance goals (Cabral et al., 2014). Although VHA provides extensive resources for PSSs and PSS supervisors (e.g., training materials, listservs, consultation; Chinman et al., 2012), PSS supervisors in this study reported difficulty accessing and/or using these resources due to the overwhelming amount of materials and their inability to discern the day-to-day “nuts and bolts” of practice from policy documents.

Participants conveyed the need for administrators to permit flexibility in interpretation of national guidelines on peer support services to ensure that individual programs can be tailored to address the most pressing needs of the local population. Across all provider types, there was a strong sense that peer support is “not one size fits all.” At the same time, fidelity to the peer support model is important to ensure efficacy (Chinman et al., 2017b). Given this tension, Fisher et al. (2015) advocate for ensuring fidelity with the key functions that define peer support (e.g., assistance in daily self-management), rather than focusing on specific roles or protocols that PSSs may use. For large healthcare organizations such as VHA, national policy and guidelines could require fidelity to key functions of peer support (see Chinman et al., 2016), but allow flexibility in how local programs choose to implement it (e.g., focus on target subpopulations).

Participants noted that Adoption and Maintenance of peer support in VHA PC-MHI requires support from leaders in the program/clinic, local hospital, and national organization. Prior research has also shown the importance of support from key stakeholders at all levels of the organization (Chinman et al., 2006, 2016; Davidson, Bellamy, Guy, & Miller, 2012). Participants called for top down communication from leadership in which peer support was clearly endorsed and recognized as a valuable component of VHA healthcare that offers unique benefits to Veterans. One way to obtain leadership support is to demonstrate positive patient- or system-level outcomes of interest to PC-MHI and primary care leadership via program evaluation.

Perceived Barriers and Facilitators of Implementing Peer Support

Several perceived barriers to implementation of peer support in PC-MHI were consistent with barriers identified in prior research from specialty mental health settings, including PSS role confusion and insufficient stakeholder buy-in, especially among referring providers. Although VHA employs approximately 1,100 PSSs, very few have worked in the PC-MHI/primary care setting (Chinman et al., 2017a); thus, primary care staff are especially unlikely to be familiar with peer support. Role confusion is a consistent challenge in integrated primary care (Mayer et al., 2016). Consistent with the recommendations of Cabral et al. (2014) and Mayer et al. (2016), staff education on the goals and scope of peer support, roles of PSSs, and benefits of peer support for patients and providers, respectively, was emphasized as a way to address concerns.

Participants raised concerns regarding team cohesion, especially if staff were unfamiliar with or closed off to the idea of peer support or held negative attitudes about PSSs due to stigma. These findings are consistent with prior research showing that many PSSs often do not feel fully accepted or respected by the team (Cronise et al., 2016; Vandewalle et al., 2016) and face negative attitudes from colleagues (Chinman et al., 2008; Davidson et al., 2012; Walker & Bryant, 2013). Integrating PSSs into VHA primary care teams may be challenging due to the newness of peer support in this setting, typical staff misconceptions about PSSs, and the large nature of the care team (see Kearney et al., 2011). PSSs have to not only get acquainted with staff from a range of professional backgrounds, but also learn how to collaborate with a number of teams, each with their own team culture. Participants discussed several strategies to enhance team cohesion, which were consistent with recommendations from Mayer et al. (2016) for community PC-MHI clinics, including high visibility at staff meetings, frequent communication (e.g., curbside consultations), marketing of peer support, and enlisting a physician champion.

Implementation Barriers and Facilitators Unique to Peer Support in PC-MHI

Due to the wide range of patients and presenting problems seen in PC-MHI (Kearney et al., 2011) combined with the broad array of services PSSs are capable of providing (Jacobsen et al., 2012), there is great potential for peer support programs in PC-MHI to be under-articulated, which could cause confusion and impede efficient utilization. Participants recommended taking time to assess local needs and provide clear guidance on appropriate referrals, which is consistent with prior research from community PC-MHI clinics (Mayer et al., 2016) and best practice recommendations (Money et al., 2011). Regarding program scope and role clarity in the PC-MHI setting, participants noted a need to distinguish between the purview of PSSs and PC-MHI providers in particular, given the high degree of overlap between patients who may be referred for either service. For example, a patient reluctant to seek specialty mental health care could be referred to a PC-MHI provider for brief intervention or a PSS for navigation support.

Participants specified that the scope of peer support services and the specific roles that PSSs will undertake should be determined in advance by a team including representation from primary care, PC-MHI, and peer support. Including stakeholders from all three areas will ensure that the key principles of the different cultures and models of care are considered. PSSs, their supervisors, PC-MHI providers, PCPs and other primary care staff should then receive education when programs are starting, as well as over time to prevent drift. Inadequate supervision is an established barrier to implementation of peer support (Vandewalle et al., 2016), but knowing that appropriate supervision is provided to PSSs bolsters primary care staff confidence (Mayer et al., 2016). Participants recommended that PSSs in primary care be supervised by PC-MHI providers, who are an ideal choice given their ability to assist PSSs navigating the unique PC-MHI setting.

Another challenge with peer support in PC-MHI has to do with matching patients with PSSs. Several PCPs expressed a desire to refer patients to specific PSSs based on a perceived match. In most VHA specialty mental health settings, PSSs are hired for certain clinics/programs due to their lived experience with a particular condition; for example, a Veteran in recovery from substance use disorder would be hired as a PSS in a substance use treatment clinic. In contrast, PSSs cannot be as close a “match” with primary care patients, as this population does not have one predominant condition, but rather, a wide variety of mental health, health behavior, stress-related, and physical health symptoms and conditions. No one PSS or group of PSSs can possibly have lived experience with all the possible mental and physical health struggles that primary care patients will need assistance with (Fisher et al., 2014). In primary care, peer-ness will thus be based on “shar[ing] a common cultural history or identity with the individuals being served, rather than a shared experience of mental illness” (Siantz et al., 2016, p. 240, 242). While not as precise a match as having the same exact condition, mutual identification as Veterans and primary care patients dealing with chronic health problems has considerable value (Brownstein, Hirsch, Rosenthal, & Rush, 2011) and facilitates rapport and engagement.

Obtaining buy-in for integrating peer support from relevant stakeholders is essential in any new setting, but it may be more challenging in primary care. Unlike specialty mental health settings, in which peer support has a much longer history in VHA, PCPs and other primary care staff are likely to be unfamiliar with peer support. Thus, it may take longer for primary care staff to grasp the concept of peer support. To increase buy-in, participants endorsed strategies with a compelling experiential component to make peer support and its value more “real,” such as having PCPs and PC-MHI providers witness patient-PSS interactions. Given the many demands placed on PCPs, finding ways for PSSs to contribute to the team right away also helps.

Characteristics of individual PSSs can also serve as barriers or facilitators to success in the PC-MHI setting. Prior research has established desired characteristics of PSSs regardless of setting, such as stability in recovery and lived experience (Chinman et al., 2006, 2010; Jacobsen et al., 2012). Finding qualified PSSs for PC-MHI settings can be challenging (Siantz et al., 2016), but taking time to find the right person with the right skillset and personality is worth it. Many of the desired characteristics for PSSs in PC-MHI could be taught (e.g., knowledge of mental health and medical conditions, communication skills), whereas some were personality traits. In fast-paced primary care, PSSs need to be flexible and adaptable (e.g., willing to accept warm hand-offs), outgoing and approachable (e.g., able to quickly gain trust), and resourceful and motivated (e.g., willing to engage in outreach to develop a network of resources). Hiring a PSS who is a good match for the PSS role in general, and the PC-MHI setting in particular, will help to ensure implementation with fidelity to key pillars of peer support and PC-MHI.

Access and visibility were important facilitators of initial Implementation of peer support in PC-MHI, as they promote utilization by providers and access for patients. PCPs in particular, who juggle diverse competing demands and serve as gatekeepers to countless clinical programs and services, called for ongoing, repeated reminders that peer support is available (e.g., daily email notification of PSS’s hours). Many suggestions for increasing visibility of peer support were consistent with strategies recommended for PC-MHI providers initiating a PC-MHI program in primary care (Robinson & Reiter, 2016), such as distributing marketing materials.

Limitations

Participants were employees of VHA, which is a unique healthcare system, so the results may not be generalizable to other healthcare settings. The majority of participants were from a single VHA medical center, and it is unclear whether findings are generalizable to other VHA sites, which may differ in culture, experience with peer support, and demographics of Veterans and providers. All participants were volunteers, and individuals who declined to volunteer may have different views on peer support. We did not assess demographics beyond gender, but future work should examine whether diversity in sociodemographic characteristics and specific lived experience impacts the effectiveness of peer support (Hundt, Robinson, Arney, Stanley, & Cully, 2015; Oh & Rufener, 2017; Siantz et al., 2016). During the course of the study, the study site began offering peer support in PC-MHI; as a result, some providers had a basic understanding of peer support. Future research should investigate this topic during contemporaneous implementation of peer support in PC-MHI to best capture the process.

Although a range of stakeholders were involved, PSSs and PSS supervisors currently working in PC-MHI settings, as well as local VHA administrators and leadership, were not included. While PSSs and PSS supervisors were not from PC-MHI settings, they were able to speak at length on general peer support concerns. Regardless of the setting, integration of peer support appears to generate similar concerns (Chinman et al., 2008; Mayer et al., 2016; Vandewalle et al., 2016). We were able to tap into the unique aspects of the PC-MHI setting through the perspective of PCPs and PC-MHI providers. Capturing the perspectives of key stakeholders in local and national VHA leadership positions is a critical next step. Finally, our interview guide was informed by the RE-AIM framework (Glasgow et al., 1999) and was designed to capture factors relevant to the Adoption, Implementation, and Maintenance domains. Although these domains are highly relevant and helped to organize themes found in this study, future research grounded in implementation science frameworks, such as the consolidated framework for implementation research (CFIR; (Damschroder et al., 2009) is needed to build off the current findings to develop and evaluate specific implementation strategies.

Implications and Future Directions

Findings from this study point to various aspects of VHA culture, resources, and primary care staff that could be leveraged to facilitate successful implementation of peer support in PC-MHI. Engaging national and local leadership to promote the program and provide funding and resources is critical. Access to good clinical supervision, training, and continuing education is essential for PSS success. Little research has explored best practices in PSS supervision and workforce development (Coufal et al., 2014; Fisher et al., 2014; Silver & Nemec, 2016), but future work is needed given the rapid proliferation of peer support (Rogers & Swarbrick, 2016). Ensuring that PSS supervisors have adequate training and resources is necessary, but simply making resources available is not sufficient. Sharing examples of high functioning programs and ways PSSs are being used in PC-MHI clinics, as well as best practices for training, supervision, and program development would allow for efficient dissemination while providing practical guidance and inspiration. Flexibility and autonomy for local programs is necessary to ensure that peer support services yield maximum impact and address the needs of the local population.

At the provider level, knowledge of the role and benefits of peer support, and cohesion among primary care teams, were identified as paramount. Participants offered numerous specific suggestions (see Table 3) related to program functioning, team cohesion, and stakeholder buy-in that could easily be implemented in real-world practice. Future research should explore strategies for facilitating integration of PSSs into existing teams (Silver & Nemec, 2016). As noted, assessing VHA administrators’ and leadership perspectives is a necessary next step in this line of research. Equally important will be gathering feedback from Veterans as to how the VHA can best utilize PSSs to increase engagement in mental health and medical care.