Posttraumatic stress disorder and interpersonal process in homeless veterans participating in a peer mentoring intervention: Associations with program benefit

Abstract

Homelessness among veterans has dropped dramatically since the expansion of services for homeless veterans in 2009, and now engaging homeless veterans in existing programs will be important to continuing to make progress. While one promising approach for engaging homeless veterans in care is involving peer mentors in integrated services, posttraumatic stress disorder (PTSD) may diminish the effects of peer mentorship. This mixed methods study examined how interpersonal and emotional processes in homeless veterans with and without PTSD impacted their capacity to engage in relationships with peer mentors. Four focus groups of 5-8 homeless male veterans (N=22) were drawn from a larger multisite randomized trial. Qualitative analysis identified five primary themes: Disconnectedness; Anger, hostility, or resentment; Connecting with others; Positive view of self; and Feeling like an outsider. Thematic comparisons between participants with and without a self-reported PTSD diagnosis, and between those who did and did not benefit from the peer mentor program, were validated using quantitative methods. Disconnectedness was associated with self-reported PTSD diagnosis and with lack of program benefit; Feeling like an outsider was associated with program benefit. Results suggest that disruption to the capacity to develop and maintain social bonds in PTSD may interfere with the capacity to benefit from peer mentorship. Social rules and basic strategies for navigating interpersonal relationships may differ somewhat within the homeless community and outside of it; for veterans who feel disconnected from the domiciled community, a formerly homeless veteran peer may serve as a critical “bridge” between the two social worlds.

In 2009 former Secretary of the Department of Veterans Affairs (VA) Eric Shinseki initiated an expansion of VA services for homeless veterans (Secretary Shinseki details plan to end homelessness for veterans, 2009), and between 2010 and 2016 homelessness among veterans dropped by 47% (Henry, Watt, Rosenthal, & Shivji, 2016). Now that the foundation of these services has been built, engaging homeless veterans in programs that are available to them will be an important element of continuing to make progress in this area (Ellison et al., 2016). Research suggests that one promising approach for increasing engagement of homeless veterans in care is involving peer mentors in integrated services (Resnik, Ekerholm, Johnson, Ellison, & O’Toole, 2016; Veterans Health Administration, 2004). High rates of Posttraumatic Stress Disorder (PTSD) among homeless veterans, however, may present a unique challenge for engagement, because a diagnosis of PTSD has been found to diminish the positive effects of peer mentorship in this population (Resnik et al., 2016). Understanding the mechanisms by which PTSD attenuates the benefits of peer mentorship may be important to efforts to expand the reach and impact of peer support programs in this population (Ellison et al., 2016).

The definitions of “peer”, “peer provider”, and “peer mentor” vary by setting. Within the VA peers are other veterans. In mental health care settings peers are commonly understood to be individuals with similar psychiatric or substance abuse diagnoses, or in some cases similar life circumstances such as homelessness. Peer providers or mentors are those who are further along in the recovery process and are able to draw from their “lived experience” to serve as role models and to offer empathy, hope, and practical knowledge (Chinman, Salzer, & O’Brien-Mazza, 2012). Preliminary evidence suggests that when added to existing programs, peer support services can contribute to improvements in a range of recovery outcomes (Chinman et al., 2014) including better social functioning and fewer day homeless (van Vugt, Kroon, Delespaul, & Mulder, 2012). In the parent study for this investigation, homeless veteran participants generally reported social, emotional, and practical benefits from their experiences with peer mentors, but they also identified limitations of the peer mentors in terms of competencies and knowledge about services (Resnik et al., 2016).

Rates of PTSD have been found to be extremely high in homeless populations (Carlson, Garvert, Macia, Ruzek, & Burling, 2013; North, Smith, & Spitznagel, 1994; Tsai, Link, Rosenheck, & Pietrzak, 2016; Tsai & Rosenheck, 2015), with the presence of PTSD being associated with a five-fold increase in the risk for homelessness in a cohort of Vietnam veterans (Rosenheck & Fontana, 1994). While homelessness increases vulnerability to trauma exposure, most studies have found that PTSD is present years before an individual becomes homeless (Metraux, Clegg, Daigh, Culhane, & Kane, 2013; North & Smith, 1992; North et al., 1994; Rosenheck & Fontana, 1994; Tsai & Rosenheck, 2015). PTSD is characterized by isolation and avoidance, reflecting a fundamental rupture in the capacity for interpersonal trust and connection (Charuvastra & Cloitre, 2008). In Vietnam veterans, the link between warzone trauma and homelessness was found to be mediated by social isolation and mental illness (Rosenheck & Fontana, 1994), suggesting that damage to the capacity to access and benefit from social support may be a pathway by which PTSD contributes to homelessness.

In summarizing the “critical ingredients” of peer support programs like peer mentoring, Mead and McNeil (2006) identified the formation of a safe, mutually supportive relationship with a person who shares similar experiences as fundamental to the peer support process. It is within this relationship that hope, self-awareness, and a sense of personal responsibility and empowerment can develop, and from which practical skills and a sense of community can be cultivated (Mead & MacNeil, 2006). While little formal research has been done on the effectiveness of peer support interventions specifically for PTSD, clinical providers have suggested that veterans with PTSD are less likely to engage with peer support services (Joseph, Hernandez, & Jain, 2015). In two recent qualitative investigations, veterans who had received therapy for PTSD in the VA suggested that connecting with peers who had experienced success in recovery from PTSD might provide hope, and may help veterans become more comfortable trusting others and forming relationships (Hundt, Robinson, Arney, Stanley, & Cully, 2015; Jain, McLean, & Rosen, 2012). Yet these veterans also indicated that social anxiety, difficulty with trust, and fear of judgement could prevent some veterans with PTSD from attending sessions or engaging with peer support services in the first place.

The purpose of the current study was to build upon previous work (Resnik et al., 2016) by exploring how interpersonal and emotional processes in homeless veterans with and without PTSD impacted their capacity to engage in effective relationships with peer mentors. Specifically, we sought to identify interpersonal and emotional processes that distinguished between veterans with and without a self-reported diagnosis of PTSD, and between those who did and did not benefit from peer mentorship.

Method

This was a mixed methods study that involved a secondary analysis of transcripts of four focus groups of homeless veterans who had been randomized to receive a peer mentor intervention.

Peer Mentors

Peer mentors who were formerly homeless veterans were recruited through the recommendations of the Homeless Veteran Supported Employment Program (HVSEP) Supported Employment Specialists in the Veteran Resources and Recovery Center (VREC) of the VA, and they were hired as term VA employees in the Research service. Efforts were made to recruit peers with experience in recovery from substance abuse, but there were no specifications regarding the peer’s history of PTSD diagnosis or treatment. Peer mentors were trained using the MISSION-Vet manual (Smelson, Sawh, Kane, Kuhn, & Ziedonis, 2011), a structured peer support intervention developed by the VA and the Substance Abuse and Mental Health Services Administration (SAMHSA). This training was specific to the experimental design of this study, and was not the same curriculum provided to peer specialists who are hired as permanent employees by the VA.

Prior to being assigned clients, peer mentors underwent extensive training in a face-to-face intensive two day meeting. During this training the MISSION-Vet modules most relevant to a primary care setting were emphasized, including motivational interviewing, patient empowerment, enhancing self-efficacy, and negotiating for behavior change. The training included sessions on how to interact with members of the primary care team; how to navigate within the VA system; how to place limits and define the goals of supportive interactions with veteran clients; and what to do in case of emergencies. Peers mentors were supervised by research staff for fidelity to the treatment protocol, and for the presence of adverse events.

Peer mentors participated in the Patient Aligned Care Team (PACT) meetings as liaisons between providers and veteran participants. They also met with veteran participants for routine scheduled visits to reinforce treatment care plans, to identify unaddressed needs, and to help mediate with the care team about any unresolved issues. Intensity of contact was suggested to be twice weekly for the first month, gradually declining over the next 6 months, with ad hoc contacts as needed. Actual contact was dictated by the care plan and veteran preference.

Sample

Participants were a convenience sample randomized to receive peer mentor services in one of two multi-site randomized controlled trials (http://clinicaltrials.gov, NCT01550757). The sample included homeless veterans enrolled in VA primary care PACT, or a population-based, homeless-oriented primary care adaptation known as H-PACT. Veterans were considered homeless if they were unsheltered; staying in an emergency shelter; doubled-up with family or friends; or in transitional housing (Stewart B. McKinney Homeless Assistance Act of 1987). Participants were excluded if they were enrolled in Mental Health Intensive Case Management or other VA-based case/care managed program, as simultaneous participation in another intensive intervention program could potentially cofound findings in the parent study. Finally, participants were excluded if they had plans to leave the area within 6 months, or were unable or unwilling to provide informed consent.

Presence or absence of self-reported PTSD diagnosis was determined at study baseline via a modified version of the Basic Shelter Inventory (BSI) (Robertson, Ropers, & Boyer, 1985), a measure that has been used in several sentinel research studies on homelessness (O’Toole, Gibbon, Hanusa, & Fine, 1999; O’Toole, Johnson, Borgia, & Rose, 2015; Robertson et al., 1985). The BSI includes self-report of physical and mental health problems, causes and duration of homelessness, current sheltering arrangements, prior health utilization patterns, self-reported needs, and overall self-reported states of health.

“Benefit” from the peer mentoring experience was defined based upon results of prior analyses. Briefly, veterans were classified by two study investigators as having benefitted or not from the peer mentoring experiences based on the classification of their responses to questions about their understanding of the role of the peer mentor (i.e. “What is your understanding of the purpose of the peer mentor?”), and whether they felt they benefitted from having a mentor (i.e. “How would you describe the impact of being assigned to a peer-mentor on your team? Do you think your experience in the past year would have been different if you had not had a peer mentor? How so?”). Cross-case analysis was used to compare those who did and did not benefit on outcomes. To validate this classification methodology, we compared participant responses to questions administered at the end of the study and used ROC analyses, described elsewhere (Resnik et al., 2016). The best response category threshold of a single end of study item, “having a peer helped me better navigate the VA system” was identified for classification of peer benefit. Data from this parallel classification method was then applied to the full 102 person sample randomized to a peer mentor, and findings from the earlier qualitative analyses were validated (Resnik et al., 2016).

Data Collection

Four 90-minute focus groups of 5-8 participants each were conducted. Groups were stratified by location and primary care setting (PACT vs. H-PACT). Focus groups had a semi-guided format that utilized open-ended questions and allowed the moderator flexibility in question wording and additional probing to clarify participant remarks. Participants were asked to discuss their experience with VA healthcare. They were also asked about their understanding of the role and purpose of the peer mentor, their interactions with their mentor, perceived benefits and limitations of their mentor, and the help their mentor provided. Finally, they were asked to the identify persons within the VA who were most important to them, and most trusted, and to describe their relationship with members of the primary care team. Focus group audiotapes were transcribed verbatim.

Data Analysis

The qualitative analysis was an iterative process of examining each transcript, coding, and revising coding schemes using a constant comparison approach (Lincoln & Guba, 1985). The following three questions provided the framework for our analyses: 1) What interpersonal and emotional processes distinguished veterans with a self-reported diagnosis of PTSD from those without? 2) What interpersonal and emotional processes distinguished veterans who benefitted from peer mentorship from those who did not? 3) To what extent do these processes overlap? Two authors (EV and LR) discussed the coding, and refined the coding scheme. The coding categories were grouped into larger categories by theme, and the organization and hierarchy of codes discussed and refined. Comparison between sub-groups (those with self-reported PTSD and those without; those who benefitted from the peer intervention and those who did not) were facilitated by construction of a comparison matrix (Lincoln & Guba, 1985) that visually displayed the coding categories by participant.

To facilitate interpretation, we also counted the number of veterans for whom we had coded at least one statement in each of the identified sub-themes. We then conducted separate Fisher’s exact tests (Preacher & Briggs, 2001) comparing the percentage of veterans with and without self-reported PTSD for each theme and sub-theme, as well as the percentage of veterans who did and did not benefit from peer mentorship for each theme and sub-theme. Given the small sample size, we did not correct for multiple comparisons and we accepted significance levels of p<.10 as suggesting that the theme or sub-theme may be worthy of further investigation.

Results

The sample consisted of 22 male veterans. Mean age of the sample was 55.2 years, 41% percent identified as white, 41% African American, and 18% identified as “other” (Table 1). Half of the sample reported having more than a high school education; 23% reported ever having been incarcerated. Thirty-six percent of the sample reported a diagnosis of PTSD, 77% reported a diagnosis of depression, 55% reported a history of substance abuse, and 68% reported having another mental health diagnosis. Data analysis resulted in the identification of five major themes: Disconnectedness; Anger, hostility or resentment; Connecting with others; Positive view of self; and Feeling like an outsider (Figure 1). Each them is described below.

Table 1.

Participant demographic/clinical characteristics

ID*	PTSD diagnosis?	Age	Race	Months Homeless
Did not benefit from peer
01		51	Asian	24
02	√	56	White	60+
03		59	White	4
04	√	52	African American	60+
05	√	54	White	10
06	√	62	White	6
07	√	65	White	48
08		70	White	24
Benefited from peer
09	√	62	African American	3
10		52	White	2
11		58	African American/Hispanic	60+
12	√	59	African American	12
13		59	African American	60+
14		31	African American	48
15		64	refused	9
16		35	African American	60+
17		48	White	24
18	√	62	White/Native American/Alaska Native	12
19		57	African American	60+
20		54	African American	6
21		54	African American	60+
22		50	African American	24

Note.

^*ID number assigned is a “pseudo-id” used for this analysis only

Figure 1.

Themes and Subthemes

Disconnectedness

The theme “Disconnectedness” reflects the idea that connection with other people or with social institutions is dangerous, undesirable, or of limited utility. The first two sub-themes included statements reflecting a broadly negative view of the self (“Feelings of shame or of being damaged”) or of others (“Suspiciousness”) that appeared to justify distancing oneself from others. The sub-theme “Difficulty engaging or avoidance of health care providers” included comments conveying rejection, or at least ambivalence, of the idea that interpersonal relationships with providers (including peer mentors) could be a vehicle for healing or recovery. Statements were coded in “Resignation” if they suggested that the veteran had given up on accessing important resources from other people or from the system. Each sub-theme is described below.

The sub-theme “Feelings of shame or of being damaged” captures statements reflecting veterans’ feelings that they were damaged, a “bother”, or unworthy of connection with others. Some statements suggested that veterans felt that it would be frightening or painful if others were to become aware of their problems. One compared his “embarrassment” about being homeless to what he believed it would feel like to admit he was homosexual:

For myself to admit to anyone that I was homeless was a major embarrassment. I think it would have been a lot easier admitting that you were gay or something. Not that there is anything wrong…but, it was difficult to admit to anyone that you were homeless… (05, no benefit, PTSD, homeless 10 mos.).

Another lost his peer’s phone number, and his shame prevented him from taking steps to recover it:

… I am the original absent-minded professor… I lost his phone number and I was not going to admit that to you [the research assistant] … and so I obviously never called him and never used it. And it just irritated the hell out of me that it was another episode where my absent-mindedness caused me to leave something behind or to lose something… (08, no benefit, no PTSD, homeless 24 mos.).

The sub-theme “Suspiciousness” includes statements reflecting the belief that connections with or proximity to other people was potentially dangerous, or that such connections left the veteran vulnerable to exploitation. One described how this fear affected his willingness to stay in homeless shelters:

The homeless shelters…. I stayed there a couple nights then went to the woods, I couldn’t stay there. The people in there, I don’t know (02, no benefit, PTSD, homeless 60+ mos.).

Some veterans’ comments suggested that they believe they were being systematically exploited. One stated that the VA was trying to “sweep under the rug” important issues that kept veterans homeless. Another reported that he felt like “we [veterans] are guinea pigs” when it came to PTSD research and pain management treatment, and that his suffering was being used both for others’ profit and as a means of controlling him:

Here you got some people coming in here to teach you how to control your pain. And 95% of us guys have been through almost everything you can think about as far as pain goes. You can’t teach me how to control pain… If you don’t do these curriculums they got, you don’t get your [pain] medicine unless you listen to this person…. (12, benefit, PTSD, homeless 12 mos.).

Some veterans expressed the belief that the disability claims system was designed to deny them access to what they had earned through their military service. One veteran described his suspicions about the disabled veterans’ support services provided in the VA:

I have a social worker over in the hospital. Like I say, it is simply, like I say I want to go and for my claim and they say go through the disabled veterans but the disabled veterans is in the VA hospital!! So the government is who pays the veterans! (09, benefit, PTSD, homeless 3 mos.).

The sub-theme “Difficulty engaging or avoidance of health care providers” reflects veterans’ wariness about entering into trusting relationship with providers, including peer mentors, that might provide a context for identifying and addressing physical or emotional problems. In some cases, veterans expressed tension between their need for help and their discomfort engaging with others. During a discussion of the VA emergency room, for example, one veteran stated, “I have never had to use the emergency room,” but then described a situation where he drove to the VA hospital “when I had a series of migraine headaches, two of which lasted for 48 hours over a 5-day period.” When he arrived at the hospital he refused care despite the provider’s efforts to engage him, “She tried to get me into the hospital, I didn’t want to go in and I told her no….” This veteran described similar ambivalence about his relationship with his primary care provider of 10 years (a nurse practitioner), alternatively indicating that he trusted her to take care of him and then describing his relationship with her as “less than friendly” and conveying his belief that she did not listen to him:

…I didn’t feel she was listening to me when I told her I was having a reaction to a drug …… And the drug turned on me. I printed out a thing from the internet, and I had every single one of the things…illusions of grandeur, right down to the…suicide…She wouldn’t even look at the paper, no no it is not the drug (08, no benefit, no PTSD, homeless 24 mos.).

This veteran avoided interactions with VA mental health clinicians, explaining that he received his “mental” therapy from a family member. As a result, he had not been seen by VA mental health, and had no diagnoses of psychiatric illness.

For some participants who had sought psychiatric treatment, their connection with the provider was not necessarily comfortable. When asked about his comfort level in the clinic, one veteran reflected that “In the past I have been a little unavailable due to a lot of problems on my side”. Later when he was asked about someone at the VA he felt comfortable with, he elaborated on his struggle:

Veteran: I am just now willing to start open up a little more, I will have to get back to you on details as far as who will be…

Interviewer: What about today?

Veteran: I have nobody. (04, no benefit, PTSD, homeless 60+ mos.).

Another veteran reported that he was unwilling to disclose anything about himself other than what was absolutely necessary to get his needs met:

Veteran: I am really not comfortable just saying anything I want to say …. (01, no benefit, no PTSD, homeless 24 mos.).

Other reasons veterans described for not engaging with providers included: belief that the providers were just trying “to make a job for” for themselves; wanting to keep a “low profile” and being “too independent”; or liking to do “a lot of studying on my own” .One veteran described his difficulty engaging with his peer mentor this way:

Veteran: I liked the guy and his outlook was a bit more brighter than mine was. The reason I didn’t continue through with the relationship with him is because I am not a people person… (03, PTSD, no benefit, homeless 4 mos.).

The sub-theme “Resignation” describes veterans’ willingness to accept difficult circumstances, often justifying these by citing limited system resources or comparing the situation to one they could imagine would be worse. Two veterans reported “understanding” being denied housing because they were unwilling to sacrifice relationships with a pet or the well-being of those they cared about:

…I didn’t get no housing because of the dog … But I understand the situation with my dog, but I wasn’t about to give her up (05, no benefit, PTSD, homeless 10 mos.).

And,

“I got an 11 year old daughter and my priorities are her. Eventually I will get a place … I am the one that put me in this situation—I got my daughter and my Ex and put up in a high rise and they only take a third of your income and all that. But I can’t do that, I can’t do that to my daughter. So I am just paddling along. Just treading water” (01, no benefit, PTSD, homeless 24 mos.).

Anger, Hostility or Resentment

The theme “Anger, hostility, or resentment” reflects patterns of thinking or behavior that are consistent with either suppressed or expressed anger. The sub-theme “Aggression” includes statements describing situations where a veteran reported that he had responded to a conflict or problem with aggression that was considered by others to be inappropriate and which led, in some cases, to negative consequences. One veteran described having had a sense that aggression was not appropriate to the context in which he used it:

I guess I had anger issues throughout my life. I was working at Home Depot at the time and I went after a customer which was not a smart thing to do but he disrespected one of the employees so…I ended up losing that job and became homeless … (05, no benefit, PTSD, homeless 10 mos.).

Later, this veteran described using aggression successfully where it could also have led to a more negative consequence. He had become increasingly angry after making several unsuccessful attempts to resolve an issue with provider turnover in a VA mental health clinic. He resorted to threatening violence:

… when I first started coming to mental health I would tell my situation and several months later another doctor stepped in. This continued for the longest time and it happened so many times. The last doctor was [Doctor X] and I had already explained to him about things and so I went in and I put a gun on the table and I said am I going to get the help I need or what? It wasn’t the smartest thing I could have done, but I got the help I needed. … (05, no benefit, PTSD, homeless 10 mos.).

When focus group members were later asked about their confidence in getting needed mental health services, this veteran rated his confidence very highly, “probably because of my situation, I brought the gun in.”

The sub-theme “Hostile attribution bias” reflects the tendency to view ambiguous social situations as reflecting hostility or malicious intent, and responding accordingly. This sub-theme includes statements falling into two areas: those that appeared to reflect a degree of confidence in the self, coupled with very little trust of others; and those that reflected a sense of dependence upon others who were nonetheless viewed as hostile, untrustworthy or ill-intentioned. Comments by one veteran describing his first meeting with his peer mentor exemplified the first area:

…. I am not much for people telling me what to do anyhow, I have been on the streets for 25 years. And [peer], when I first met [peer], and I was trying to talk and he said, hey I am talking shut up! And so I shut up. And then when he said, what do you have to say? And I said—oh, can I speak now? Master? That is how we first started.

This veteran appeared to have had sufficient flexibility in his hostile cognitive schema that he was eventually able to shift his perspective and make a connection with his peer:

But then when he realized where I was coming from—now we are cool. His job is to supposed to try and get you back, I mean… they were trying to get me in the system.…. I had been on the streets a long time, and he just came too strong (13, benefit, no PTSD, homeless 60+ mos.).

In contrast, another veteran described his relationship with his brother’s family as vacillating between dependence and resentment, being “grateful” for a place to stay, but wanting to “strangle” his 3 year old nephew for constant whining. He reported feeling dependent on providers, but also said that he told his primary provider not to “prescribe me any more medications because I won’t take it…”, and he believed that trusting his providers had put him in danger because he had “been to the hospital twice with reactions to drugs…” (08, no benefit, no PTSD, homeless 24 mos.).

The sub-theme “Sense of betrayal by the military/VA systems” reflects veterans’ beliefs that they had been betrayed by the military, or by the VA system that was supposed to care for them after military service. One participant explained his frustration with the long wait for his disability claim to be resolved:

I have been waiting on my disability claim for 5 years and the only thing they did was blow up a bunch of bombs around me. They are taking too long and it is hard to live out here. Everyday someone is out there dying for this country (22, benefit, no PTSD, homeless 24 mos.).

In most cases veterans expressed this feeling of betrayal with confusion and frustration. For example:

I have been coming here for 25 years and they haven’t fixed my problem yet. I am service-connected for my left knee, torn meniscus, and they don’t fix it I don’t know why! (17, benefit, no PTSD, homeless 24 mos.).

In other cases, the sense of betrayal seemed to take on a more personal tone, and the idea that services were being intentionally withheld was more explicit:

They never inform you, with the housing list that we got vouchers. It is like sssshhhh, hush hush…. You have to come down here every day. A friend of mine told me he was here 90 days. …. He wasn’t here when they put the list up, he had been there everyday for 90 days except when they put that list up. Nobody tells you (22, benefit, no PTSD, homeless 24 mos.).

For this veteran, the sense of personal betrayal by the VA system appeared to have generalized to an inability to trust any of the individuals who worked within the system. When asked if he had a person at the VA that he felt most comfortable talking to about his life he responded, “no-one”.

Connecting with Others

The theme “Connecting with others” reflects openness to the possibility of accepting help from others. The sub-theme “Accepting help when it is offered” expresses veterans’ acceptance of help and respect they received from providers. These statements often reflect a degree of passivity or vulnerability. A veteran might convey his belief that providers “care more about my health than I do,” or that he viewed himself as somewhat disempowered in the context of a paternal (or maternal) health care system:

They mother me, sometime hen-peck me as a matter of fact but I know it is with my best interest at heart (03, no benefit, no PTSD, homeless 4 mos.).

Such statements could sometimes reflect the sense that receiving help required transferring some degree of agency to the caregiver. For example, another veteran commented that his peer mentor “put me where I needed to be put, straightened out” (11, benefit, no PTSD, homeless 60+ mos.), suggesting that he placed a significant part of the responsibility and the credit for his recent recovery with the peer. Similarly, another veteran described his helplessness in dealing with his mental health issues, conveying his belief that without the help of his psychiatrist and social worker, “I would probably still be lost” (09, benefit, PTSD, homeless 3 mos.).

The sub-theme “Trusting others” reflects a veteran’s capacity to trust peers or providers enough to be able to share difficulties with them, or to express vulnerability. One veteran described being able to talk about his problems with a provider because, “he knows my private business and yet is able to conduct business as usual without prejudice which is very helpful and admirable” (01, no benefit, no PTSD, homeless 24 mos.). Another veteran expressed surprise more than once at finding himself being able to trust providers, suggesting that it was a new experience for him to feel like he could trust them:

…Dr X. It really seems like he treats you as an equal. And he listens. A lot of doctors just want to go ahead and tell you what is what, don’t ask you how you are doing and how you are feeling—what do YOU think you should take. Not just take this or take that. You always end up being a guinea pig trying to get the medications and stuff (02, no benefit, PTSD, homeless 60+ mos.).

Comments reflecting the ability to trust peers and other health care providers were often associated with descriptions of mutual respect. One veteran drew a connection between the “care and concern” he felt from his “care worker” and his sense that “I just get some feeling of understanding. They don’t think I am in there trying to run some game or ulterior motives…” Later this veteran described more explicitly how mutual respect was important to him in his relationship with his peer mentor:

The first meeting…he told me stuff like I was going to be treated with dignity and respect, and not let my rights be violated, and if I felt like they weren’t working with me, let him know if I had complaints to file. And my responsibility and what I should expect from him (21, benefit, no PTSD, homeless 60+ mos.).

Comments reflecting openness to trusting peers as opposed to other providers often appeared to take on a different tone. Veterans described feeling understood because the peers were part of their own social networks:

If you are having a rough time and need someone to talk to, he is gonna be there and understand what is going on, he actually understands. … He actually understands these guys (12, benefit, PTSD, homeless 12 mos.).

One veteran described his relationship with his peer as being like having a family member he could trust to be on his side:

We used to get together once every two weeks. I would come in and we would go out and have coffee and talk. How to stay strong. [In working with] HUD-VASH, you got to fight for yourself. That helped me out a lot because I don’t have family here (11, benefit, no PTSD, homeless 60+ mos.).

The relevance of the peers’ shared experiences of the participants’ social worlds was conveyed by one veteran when he talked about how his peer mentor had been able to help him gain supported housing. The veteran described that he learned to trust his peer mentor after he was successful in advocating for him in a situation where he believed there could be no second chances:

… HUD-VASH - I was on it one time and lost it, I got I got it back the second time—and that isn’t supposed to happen (17, benefit, no PTSD, homeless 24 mos.).

Positive View of Self

The theme “Positive view of self” reflects participants’ positive views of their own ability to cope with challenges, stress, and adversity. The sub-theme “Empowerment” reflects a degree of self-efficacy that made it possible for veterans to effectively collaborate with providers. Some focus group members saw their doctors as “approachable” enough for them to take an active role in their health care and recovery by “ask[ing] them for things you need” (01, no benefit, PTSD, homeless 24 mos.). One veteran described how his satisfaction with his primary care team was directly related to his active collaboration with the providers:

…they do wonderful with me because I let them know. I tell them what I need and I make sure they listen to what I got to say and they do listen to me… We see eye to eye on everything (11, benefit, no PTSD, homeless 60+ mos.).

In most cases, statements coded here reflected willingness to ask for help directly, whether it was from health care providers, “If you ask the right questions you can always find somebody…” (18, benefit, PTSD, homeless 12 mos.); or other patients, “Most of the time I am asking and it is usually other vets I get pertinent information from” (03, no benefit, no PTSD, homeless 4 mos.).

This sense of empowerment was described by one veteran who reported that when he felt one of his former primary care providers was “more about getting her paperwork done than listening to my situation,” he took action “to change her right away. So now I have a primary care who I trust” (20, benefit, no PTSD, homeless 6 mos.).

The sub-theme “Worthiness and responsibility” reflects evidence of a veteran’s belief that he was worthy of respect and access to resources, and that he followed through on this belief by taking personal responsibility for securing them. One veteran reported that whereas before he had “little esteem” and wasn’t “ready to hear” what others had to say, because of his peer mentor’s support he was now able to “get over the fear of trying to do the right thing” (11, benefit, no PTSD, homeless 60+ mos.). Another veteran stated that he now recognized his own resilience and was “being receptive” and “striving” to take responsibility to cope with current distress, solve problems, and make constructive choices for his own sake and for the sake of others:

“… I have a whole new outlook on trying to keep better care of myself today. A few months back I was still in the throes of not caring enough. And I have been taking better health concerns for myself and being cautious, I tend to be a little more health conscious myself” (04, no benefit, PTSD, homeless 60+ mos.).

This emerging sense of self-worth and self-sufficiency in the context of (cautious) connectedness with others was illustrated by two veterans’ responses to a question about who they thought was most important in their accessing healthcare:

“Me. I can’t pick one person at the VA, they have all contributed, also I was at a positive point …where once you start realizing you have to think it, and they helped me but I really kinda had to take that step myself. So yeah, me” (14, benefit, no PTSD, homeless 48 mos.).

In some cases, this sense of worthiness allowed veterans to stand up for themselves in situations where they felt they were being mistreated or misunderstood by others who were in relative positions of power. One veteran described his situation:

“I had a mental health intern that wanted me to get someone to take care of my money. I shut down right then and there. I mean, she was supposed to be for my mental health and she was giving me a suicidal cause right there. I brought it to the regional office and I won, I can be in charge of my own money.” (18, benefit, PTSD, homeless 12 mos.).

From this experience the veteran concluded that it was as important for him to understand the constraints and limitations of his providers in receiving appropriate care, as it was for him to understand his own needs: “You got to know what is going on with them as much as what is going on with you” (18, benefit, PTSD, homeless 12 mos.).

The sub-theme “Belonging” reflects veterans’ beliefs that connection through mutual relationships could be beneficial. We classified text into this sub-theme when there were data suggesting that a veteran saw himself as an important part of a relationship with at least one other person, or that he viewed himself as belonging to a larger group that he valued. One veteran reported that his 12-year-old son lived with him (19, benefit, no PTSD, homeless 60+ mos.), and another stated that he was taking care of his “grandkids and daughter” (12, benefit, PTSD, homeless 12 mos.). Other veterans made comments suggesting that that they belonged to a community of veterans: “…Thank god I am a veteran because we have services they don’t have!” (20, benefit, no PTSD, homeless 6 mos.,), and “the VA help their own” (19, benefit, no PTSD, homeless 60+ mos.).

Developing a greater sense of mutuality in the context of belonging was expressed directly by one veteran when he was asked what he understood to be the purpose of the peer mentor. This veteran described how opening himself up to support from the peer mentor allowed him to feel connected to the other veterans:

It was also to help other veterans go through the process of getting through the same thing they were going through. Like, my peer mentor [peer] was in [shelter], the same place I was. And we bonded together because of that. And I understood that we were helping other veterans and that is what got me through the program (18, benefit, PTSD, homeless 12 mos.).

Feeling Like an Outsider

The theme “Feeling like an outsider” reflects veterans’ beliefs that as homeless persons they were not part of mainstream society, but rather that they identified as a part of a homeless “outgroup”. In some cases veterans rejected the imposition of an “outsider” identity, responding with anger. For example, one participant asserted that even though he was “an outstanding member of the community where I live”, he felt that he was seen by staff at the homeless clinic as a member of a “homeless” group he neither respected nor identified with:

… there are derelicts here [at the VA clinic location], the homeless people here. So they treat everyone in the same status, and I wasn’t really homeless. I utilized the system to get to where I was going (12, PTSD, benefit; homeless 12 mos.).

Another veteran appeared to be more comfortable acknowledging his homelessness, viewing his status as fluid and the peer mentor as “a person that had information that I needed to help me get back to start being a positive member of society again” (15, no PTSD, benefit, homeless 9 mos.).

Some veterans made statements that seemed to suggest that this “outsider” identity was acceptable or even preferable to being part of mainstream society. One veteran’s acceptance of his identity as a homeless person, and his parallel lack of interest in finding stable housing, was met with confusion by both the interviewer and by another veteran participant.

Veteran: I haven’t gotten involved with it [looking for housing] beyond using the shelter system.

Interviewer: Is that something you have talked to a case manager about?

Veteran: Actually, they have talked to me about it because I am not utilizing it. There is one social worker in particular that asks me every time she sees me over the last 3 or 4 years asks if I am interested in housing, and I have to tell her no I am not. That is not something I want to pursue (16, benefit, no PTSD, homeless 60+ mos.).

This veteran went on to explain that the problem resided not with him but with society: the police who wanted to call him “crazy”, and people who were pressuring him to take medication he did not want.

Participants who made statements reflecting the greatest acceptance of their identities as members of a homeless “out-group” had been homeless for at least 3 years. Some appeared to maintain sufficient flexibility in their identities, however, that there existed a possibility that they could be touched by social changes in the mainstream society from which they defined themselves as outsiders. One veteran told this story:

I was living on the streets since 1988 and 10 months ago I was pushing my shopping cart with my dog down the street and the police came and they happened to have one of the social workers from here with them, because the Obama Administration had released funding for homeless veterans. And so that is how I got into the HUD-VASH, that was 10 months ago (13, benefit, no PTSD; homeless 60+ mos.).

Comparisons Among Themes and Sub-themes

Veterans with self-reported PTSD and those who did not benefit from the peer mentoring program were more likely to make statements coded within the theme of “Disconnectedness” than were veterans without self-reported PTSD or those who benefitted. Within the theme of “Disconnectedness”, the sub-theme of “Resignation” was particularly evident in those with self-reported PTSD and in those who did not benefit from peer mentorship. Members of both groups made statements suggesting they had become resigned to their circumstances, expressing neither hope nor distress when situations arose that demonstrated that system priorities were not aligned with their values or responsive to their needs. The broad themes of “Anger, hostility or resentment” and “Positive view of self” were similar in participants with or without self-reported PTSD, and in those who did or did not benefit from the peer mentorship program.

The theme “Connecting with others” also was similar in veterans with and without self-reported PTSD, and in those who did and did not benefit from the program on the broad theme. Within this theme, however, the sub-theme “Accepting help when it is offered” appeared to be more common among veterans without self-reported PTSD and those who benefitted from the program. While the sub-theme “Trusting others” was similar between self-reported PTSD and program benefit groups, we did find differences in program benefit between those who made statements about trusting the peer mentor vs. professional health care providers. Specifically, none of the veterans who made statements exclusively reflecting trust in professional providers benefitted from the peer-mentoring program, whereas veterans who made statements about trusting the peer all benefited, regardless of whether or not they also trusted the provider.

Finally, veterans who made statements consistent with the theme “Feeling like an outsider” were most likely to benefit from the peer intervention program, but they were no more or less likely to report a diagnosis of PTSD.

Results of the Fisher’s exact tests support the findings of the qualitative comparisons (Table 2). The theme of “Disconnectedness” was associated with both self-reported PTSD diagnosis (p=.02) and lack of benefit from the peer mentoring program (p=.02). Among the sub-themes, “Resignation” was positively associated with self-reported PTSD (p=.002) and negatively associated with program benefit (p=.04), and “Accepting help when it is offered” was negatively associated with self-reported PTSD (p=.08) and positively associated with program benefit (p=.08). “Feeling like an outsider” was not associated with self-reported PTSD diagnosis, but was associated positively with benefitting from the peer mentoring program (p=.002).

Table 2.

Quantitative comparisons of themes and sub-themes.

	PTSD			Program Benefit
	PTSD n=8	No PTSD n=14	Fisher’s exact test p-value	Benefit n=14	No benefit n=8	Fisher’s exact test p-value
Disconnectedness	8 (100%)	6 (43%)	.02*	6 (43%)	8 (100%)	.02*
Shame	1 (13%)	4 (29%)	.61	2 (14%)	3 (37%)	.31
Difficulty Engaging	4 (50%)	5 (36%)	.66	4 (29%)	5 (62%)	.19
Suspiciousness	4 (50%)	4 (29%)	.39	4 (29%)	4 (50%)	.39
Resignation	5 (62%)	0 (0%)	.002**	1 (7%)	4 (50%)	.04*
Anger, hostility, resent.	5 (62%)	7 (50%)	.67	7 (50%)	5 (63%)	.67
Aggressive	1 (13%)	2 (14%)	1.0	2 (14%)	1 (13%)	1.0
Hostile attribution bias	3 (37%)	3 (21%)	.62	3 (21%)	3 (37%)	.62
Feeling betrayed	1 (13%)	3 (21%)	1.0	3 (21%)	1 (13%)	1.0
Connecting	6 (75%)	11 (78%)	1	12 (86%)	5 (63%)	.31
Accept help	3 (37%)	11 (78%)	.08†	11 (78%)	3 (37%)	.08†
Trust others	4 (50%)	11 (78%)	.34	11 (78%)	4 (50%)	.34
Outsider	1 (13%)	4 (29%)	.61	5 (36%)	0 (0%)	.002**

Note: Number (percentage) of participants with evidence of theme/sub-theme for each group.

^†= p < .10,

^*= p < .05,

^**= p < .01

Discussion

We found that veterans with self-reported PTSD were more likely to express the theme of “Disconnectedness”, and that “Disconnectedness” was negatively associated with benefit from the peer mentoring program. This finding is consistent with the concerns about peer support for PTSD articulated by veterans in treatment for PTSD in both of the previously cited qualitative studies (Hundt et al., 2015; Jain et al., 2012) as well as with research in social bonds and PTSD (Charuvastra & Cloitre, 2008). Similarly, the sub-theme “Accepting help” was negatively associated with self-reported PTSD and positively associated with program benefit, perhaps reflecting the difficulty individuals with PTSD have in opening themselves up to the degree that is necessary to establish a mutual and effective helping relationship.

Though generally the sub-theme “Trusting others” was similar in benefit/no-benefit groups, differences did emerge with respect to whether the veteran described being able to trust a peer vs. a professional provider. Specifically, the capacity to establish trust with the peer, but not necessarily with professional providers, was associated with program benefit. Research in the area of access to care suggests that fear of stigma and of ill-treatment by professional health care providers may prevent homeless veterans from seeking care when they need it (O’Toole, Johnson, Redihan, Borgia, & Rose, 2015). Peer mentorship may be a way of overcoming this obstacle to receiving VA homeless services. Facilitating trust-building skills among peer mentors could also increase the capacity for veterans to develop the sense of safety necessary to develop and benefit from mutual helping relationships (Mead & MacNeil, 2006).

Within the broad theme “Disconnectedness”, the sub-theme “Resignation” distinguished both between the subsets of veterans with and without self-reported PTSD, and between those who did not benefit from the program. We found this observation particularly compelling. On the one hand, this sub-theme reflects an element of perspective-taking, and statements could be interpreted as reflections of veterans’ positive capacity to consider their needs in the context of limited system resources and the needs of others. From this vantage point it seemed reasonable for two of the focus group members to assume responsibility for their continued homelessness. Yet in the context of having to choose between two fundamental needs, e.g. stable shelter and critical relationships (Irvine, 2013), placid acceptance of the need to keep “treading water” in an untenable situation may be seen as reflecting the emotional numbing and loss of hope that underlies much of the devastation of PTSD. Similarly, accepting ongoing homelessness could reflect an unconscious internalization of the stigma that justifies the existing social structure. Consistent with these interpretations, social and political psychologists have pointed out that individuals who are most harmed by the status quo can also be its strongest defenders (Jost, Banaji, & Nosek, 2004).

With respect to our negative findings, we were surprised that the overall theme “Anger, hostility, or resentment” and its sub-themes did not appear to differ in PTSD/no-PTSD or benefit/no-benefit groups. It may be that for stigmatized individuals who live day to day in dangerous and disempowering circumstances, anger, hostility, and resentment serve somewhat adaptive functions both in terms of physical safety and identity maintenance. For example, hypervigilance and hostile attribution bias may at times be protective for those who live in a physical and social context of chronic threat and disempowerment (Anderson, 1999; Richardson, Brown, & Van Brakle, 2013). Although aggressive behavior carries a risk of injury or severe social sanctions and is generally viewed a “last resort” to problem solving by members of the status quo, it may be viewed as a viable tool for disempowered people who have limited access to alternative, socially-sanctioned avenues for needs fulfillment (Anderson, 1999).

The different set of social assumptions held by homeless veterans and professional health care providers may underlie one of the most powerful elements of peer mentoring. Mead and MacNeil (2006) commented that unlike in typical provider-patient relationships, in peer support relationships “Power structures are always on the table and negotiated”, and “Both people figure out the rules of the relationship” (Mead & MacNeil, 2006, p.33). This flexibility may allow for the rules governing peer interactions to be tested and revised according to the needs of the individuals involved, within a lived social context that is more clearly understood by both. For example, it may be that language we coded as reflecting “hostile attribution bias” would not necessarily be viewed negatively by homeless veterans in the context of building peer relationships. Our attempts as researchers to understand interactions may have been limited by our position within the “dominant paradigm” (Mead & MacNeil, 2006, p. 31), underscoring Mead’s and MacNeil’s (2006) caution against using this standard as the metric for judging the usefulness of peer support.

Perhaps one of the most important findings of this study was that veterans who made statements suggesting they felt like outsiders were the most likely to benefit from the peer mentoring program. As touched upon above, social rules and basic strategies for navigating interpersonal relationships may differ somewhat within the homeless community and outside of it. If an individual has learned to survive and function as a homeless person, he has likely had to develop an alternative way of interacting both with peers and with mainstream society; if he has been homeless for many years, he may also have re-defined himself, his value system, and his loyalties within this very different context (Farrington & Robinsion, 1999; Snow & Anderson, 1987). In programs that included peer support for combat amputees (Pasquina, 2004) and for returning combat veterans (Greden et al., 2010), veterans have emphasized the importance of “shared experience” as being among the most powerful elements of the programs, commenting that “if you haven’t been there, you don’t get it”, and “another veteran who has been there may make it easier to get help” (Greden et al., 2010). For homeless veterans, the “culture gap between military and civilian worlds” (Caddick, Phoenix, & Smith, 2015, p. 287) may make the disconnection with mainstream civilian society even more isolating. In such instances a formerly homeless veteran peer who has made progress in this transition may serve as a critical “bridge” between these worlds.

Limitations

Our study has several limitations. First, our categorization of diagnosis of PTSD was based upon self-report. While relying on veterans’ self-disclosure of PTSD diagnosis is not unprecedented and has been used in other recent investigations of homelessness in veterans (Metraux, Cusack, Byrne, Hunt-Johnson, & True, 2017), the approach runs the risk of misclassifying subjects as not having PTSD when in fact they did have it. Disengagement from mainstream society and its institutions might have made it less likely for some of the participants to have had sufficient encounters with medical or psychiatric professionals to have received the diagnosis. Consistent with this interpretation, one veteran described having a history of psychiatric symptoms during the focus group, yet he denied having ever been diagnosed with a psychiatric disorder when asked specifically about this on the Basic Shelter Inventory. On the other hand, service connection for disability carries with it financial compensation, and it is also possible that veteran participants may have reported the presence of a disorder for which they may not actually have meet diagnostic criteria. This possibility is made less likely, however, by the emphasis during the informed consent process that information provided for the research study would have no bearing on eligibility for benefits. Nonetheless, in this study the construct of PTSD may be understood more reflect veterans’ perceptions of having experience a trauma or traumas that significantly impacted life and functioning, rather than a formal diagnosis of the disorder consistent with DSM 5 diagnostic criteria.

A second limitation of this study is that the sample was comprised entirely of men, and therefore the findings are not likely to reflect the experiences of homeless female veterans. Third, our sample was small, and we made no adjustments in our statistical analyses for multiple comparisons. These findings should be considered exploratory and hypothesis-generating, and should be interpreted with appropriate caution. Finally, as discussed by Mead and MacNeil (2006), as researchers operating outside of the homeless community we are limited by our own assumptions and biases, most of which we are likely unaware. It is hoped that this presentation of our findings can open discussions among researchers, health care professionals, and peer support providers as to how to collaboratively develop a constructive approach to building, improving and evaluating peer support services for homeless veterans.

Conclusions

This study provides insights into the ways that PTSD may interfere with homeless veterans’ capacity to engage in the kind of mutually supportive, safe, and trusting relationships necessary to benefit from peer mentorship. Though benefiting from the program and reporting a diagnosis of PTSD were not mutually exclusive, in general both individuals with self-reported PTSD and those who did not benefit expressed greater disconnectedness with others, less willingness to accept help from others, and a greater sense of resignation to their current circumstances. These findings may have implications for treatment matching within the homeless peer support program. Specifically, further research could investigate whether formerly homeless veterans who are relatively advanced in their recovery from PTSD might be appropriate peer mentors to currently homeless veterans with PTSD. The presence of another veteran with PTSD who has “been there” but has been able to manage isolation and mistrust that comes from trauma may be a powerful tool for instilling hope, and for helping veterans PTSD overcome the anxiety and fear of judgment necessary to build a connection that will allow them to benefit from the peer mentoring program.

Regardless of PTSD status, veteran participants who made statements suggesting they felt like outsiders benefitted from the peer mentoring program. Although the sub-theme “Trusting others” generally did not distinguish between veterans with and without self-reported PTSD or those who did or did not benefit from the program, within this sub-theme the capacity to develop trust with the peer mentor was associated with benefit, regardless of trust in professional providers. Peer mentoring may offer an important bridge between homelessness and engagement with homeless services in the VA, and the capacity for peer mentors to establish trust with homeless veterans may be a critical factor in veteran peer mentoring program engagement.