“I wish people could come together like we have,” Patient and Provider Perspectives on VA Residential PTSD Treatment

Abstract

Although most PTSD care in the Veterans Health Administration (VHA) is provided on an outpatient basis, the VHA has 40 residential rehabilitation treatment programs (RRTPs) designed to treat Veterans who require more intensive and closely monitored care. Unfortunately, the clinical outcomes of these programs are modest, and previous attempts to identify key drivers of outcomes have uncovered few modifiable factors. The present study, informed by the model of Resources, Life Events and Changes in Psychological State, was designed to identify factors associated with treatment response among RRTP patients and providers. Semi-structured interviews were conducted with 24 patients and 12 providers at three regional RRTPs, using interview guides based on the theoretical model. Data were analyzed using rapid analysis. Results showed that patients and providers agreed on several factors critical to RRTP success. These factors included the provision of evidence-based psychotherapy, support and understanding from fellow patients, and skill and support from providers. Patients and providers also noted the importance of psychological flexibility, openness, and willingness to change. Patients who experienced less symptom improvement over the course of treatment were more likely to report poor therapeutic alliance. These findings underscore the importance of continued emphasis on evidence-based psychotherapy delivery but also suggest that RRTPs might find additional ways to capitalize on the residential milieu to encourage engagement in treatment and a focus on therapeutic change.

Current estimates suggest that 12% of Veterans (Harpaz-Rotem & Hoff, 2020) suffer from posttraumatic stress disorder (PTSD), a disease characterized by intrusive memories, avoidance, negative alterations in cognition and mood, and hyperarousal (APA, 2022). PTSD is associated with decreased quality of life (Schnurr et al., 2006) and increased risk for suicide (Conner et al., 2014; Ilgen et al., 2010, 2012). Although most Veterans with PTSD are treated on an outpatient basis, the Veterans Health Administration (VHA) has 40 residential rehabilitation treatment programs (RRTPs) designed to treat Veterans who require more intensive and closely monitored care. These programs cost an average of $20,497 per episode of care (NEPEC, 2014). They treat less than 5% of VHA PTSD patients but account for 33% of VHA’s budget for specialized PTSD care (NEPEC, 2014), due to the high cost of residential treatment as compared with outpatient treatment. Unfortunately, despite these costs and intensive efforts, many RRTP patients experience only modest symptom improvement (Grau et al., under review; Sripada et al., 2019). Due to the high cost and suboptimal outcomes of these programs, the National Academies of Sciences has called for strategies to improve treatment outcomes of VHA PTSD treatment programs, particularly in RRTPs (IOM, 2014).

Because RRTP programming varies substantially by site, several investigations have attempted to identify factors that distinguish high-performing from low-performing programs. A key factor in RRTP performance is presumed to be the availability of gold-standard treatments for PTSD, including Prolonged Exposure (PE) and Cognitive Processing Therapy (CPT) (Cook et al., 2014). These treatments are both evidence-based psychotherapies (EBPs) that are recommended as first-line treatments by VA/DoD, the American Psychological Association, and the International Society for Traumatic Stress studies (APA, 2017; ISTSS, 2018; VA/DoD, 2017), but are not used consistently in RRTPs (Shiner et al., 2018). To assess the impact of EBP availability on RRTP performance, Cook and colleagues examined the association between EBP implementation and program-level performance across RRTPs, but found that there was little association between EBP penetration and program level PTSD outcomes (Cook et al., 2019). Moreover, Grau and colleagues conducted a system-wide analysis of individual response trajectories for patients in PTSD RRTP treatment and found that EBP receipt was not associated with trajectory membership (Grau et al., under review). Thus, additional work is needed to elucidate the factors aside from EBP availability that contribute to treatment response in RRTPs.

One way to examine treatment response is through the lens of the Resources, Life Events and Changes in Psychological States model (Murrell & Norris, 1983). In this model, psychological, social, material, and physical health resources buffer the effects of undesirable events and boost the effects of desirable events on psychological states (Murrell & Norris, 1983). For example, greater personal resources can buffer the effect of spousal loss on subsequent bereavement (Norris & Murrell, 1990) or reduce the likelihood of developing depression (Phifer & Murrell, 1986). We previously used this framework in a system-wide analysis of RRTP survey data and found that all four resource domains were associated with patient-level improvement (Sripada et al., 2019). However, the specific ways in which these resource domains helped or hindered treatment response could not be assessed at the level of administrative data.

Present Study

The goal of the current study was to build on these quantitative findings in an explanatory sequential design (Fetters et al., 2013). We aimed to identify patient- and program-level factors associated with degree of treatment response, using a qualitative approach informed by the findings of our previous quantitative work. A qualitative approach can provide in-depth, descriptive, discovery-oriented assessment of contributors to treatment response that are not available at the level of administrative data. Furthermore, this analysis can reveal new insights on PTSD treatment outcomes, as patients may have different views from providers on the experience of psychotherapy (Hodgetts & Wright, 2007) or different criteria for judging success of therapy (Hill et al., 2013). Guided by theory of Resources, Life Events and Changes in Psychological States (Murrell & Norris, 1983), we queried patients and providers about each of the model’s domains: a) psychological resources and risk factors, b) social/contextual resources, c) material resources, and d) physical health resources. The rationale was that interviewing patients and providers would provide additional perspectives and facilitate greater understanding of how individual-level factors may influence outcomes. Identifying factors associated with poor treatment response in RRTPs may enhance VHA’s ability to target and improve treatment approaches for the high-risk individuals requiring intensive care.

Method

Qualitative Approach and Research Paradigm

We purposively sampled RRTP programs with good outcomes and poor outcomes, using an explanatory sequential design and connecting qualitative data to previously collected quantitative data (Fetters et al., 2013; Palinkas et al., 2011). First, we collected and analyzed quantitative data from national RRTP surveys collected at the time of program intake and discharge (Sripada et al., 2019). We analyzed average program performance according to average PTSD symptom change within the program and rank ordered programs on this metric (Sripada et al., 2019). We then based our sampling frame on this quantitative data by selecting three regional sites. Of the four PTSD programs within our Veterans Integrated Services Network (VISN), one fell in the top third, one in the middle third, and two in the bottom third of PTSD symptoms. We selected the site in the top third, the site in the middle third, and the site in the bottom third that was geographically closest to us. We also built on quantitative findings by including in our interview guide the patient-level factors that we identified in our quantitative analysis. Finally, we compared the outcomes of our current analysis with findings from our previous quantitative analysis (Sripada et al., 2019).

Setting

All three sites were Midwestern PTSD RRTPs offering intensive treatment to Veterans with military-related PTSD and co-occurring difficulties including substance use disorders. Programs were comprised of interdisciplinary teams made up of social workers, psychologists, peer support specialists, psychiatrists, and nursing staff. In addition, staff from the chaplain service, dieticians, and recreational therapists were actively involved in programming. In all three sites, programming was offered seven days a week with a focus on EBP delivery. RRTP patients participated in additional psychoeducational and psychotherapeutic groups including anger management, relapse prevention, recovery planning, sleep enhancement, leisure education, and physical conditioning.

Sites differed in certain aspects of programming. The high and low performing sites utilized a cohort model, in which groups of Veterans entered the program together and were discharged together. In contrast, the mid-performing site used a rolling admission model in which Veterans entered the program and discharged from the program at different times. Another difference was that the high-performing program had separate specialized tracks/cohorts, including a women’s track and a track for Veterans who had experienced a traumatic brain injury. A third difference was that the low- and high-performing sites initiated EBPs at the onset of care, whereas the mid-performing site used a step-wise approach in which EBPs were initiated after three weeks. At the mid-performing site, Veterans were initially enrolled into a 22-day treatment program that focused on developing skills to manage symptoms associated with PTSD. The focus of this phase was to establish a stable foundation for recovery by building basic coping skills (emotion regulation, assertiveness skills, relaxation training, mindfulness, and relapse prevention). RRTP patients were then provided the option of discharging from care or participating in an EBP for six weeks.

Location and Participants

In-person interviews were conducted with patients and providers at three regional RRTPs in order to better understand their perceptions of the causes and factors associated with good and poor treatment response. We interviewed between three and 13 patients and three to five providers/administrators per site. Psychotherapy providers (such as psychologists and social workers) and administrators (such as program leads and Administrative Officers) were identified via service rosters. Interviews were conducted in December 2017 (high and mid-performing sites) and September 2018 (low-performing site).

Sampling Strategy

Providers were emailed an informational sheet about the study to request their participation in an interview with the written permission of their supervisors if the interview was scheduled during their regular tour of duty. Patients were recruited by flyer and through presentations at regular RRTP group meetings. All procedures were approved by the Institutional Review Boards of the primary site and the study sites.

Participant Characteristics

At the low-performing site, we interviewed three patients and three providers. At the mid-performing site, we interviewed 13 patients and four providers. At the high-performing site, we interviewed eight patients and five providers. In total, we interviewed 24 patients and 12 providers/administrators. Twenty-five percent of patients and 58% of providers were female.

Interview Procedures

Interviews were conducted in person at the RRTPs by trained qualitative staff and postdoctoral research fellows. Each interview lasted 30–45 minutes. Veteran participants received a $20 gift card for participation in the interviews. Interviews took place at the time of consent. Patients were informed that the interviewer would not divulge any content from the patient’s interview to their provider or treatment team. Participants consented to medical record review, which enabled assessment of symptom change over the course of residential treatment.

Interview Topic Development

The interview included discussion of patient-level factors that may affect treatment response. Providers were also interviewed about organizational and programmatic factors contributing to treatment response, as well as system factors that may affect intervention selection and implementation. Interviews were semi-structured to allow participants to express themselves in their own words while covering specific topics (Weiss, 1995). However, to better facilitate a rapid analysis approach, topic areas were covered in a fixed order as much as possible without sacrificing quality of the data. The interviews were conducted using a topic guide that was developed through discussion with RRTP providers, qualitative researchers, and PTSD intervention researchers. Interview questions inquired about psychological resources, social resources, material resources (e.g., program- and treatment-related factors), and physical health problems that may enhance or interfere with the benefits of treatment, with specific follow-up questions and probes that focused on predictors of treatment response identified in prior quantitative analyses (Sripada et al., 2019).

Data Processing

Interviews were audiotaped and transcribed verbatim for content. We conducted analyses concurrently with ongoing interviews so that insights from earlier interviews could inform subsequent data collection and analysis. For example, after the initial round of interviews, we further specified questions about outside stressors, psychological resources, and physical health resources. We made these changes in order to probe more deeply into the constructs of interest. Because of the changes we made, it is possible that we gleaned more specific information on these constructs from the mid- and low-performing sites (second and third rounds of interviews) than we did from the high-performing site (initial round of interviews).

Data Analysis

We used a rapid data analysis process (Beebe, 2001; Gale et al., 2019), a technique used when there is a need for targeted, actionable information. To conduct systematic analysis, we created a qualitative codebook, using domains from the Murrell and Norris (1983) model. The template also had an “other” category to capture any new domains, as well as sections that captured perspectives on existing interventions. The analysis team, consisting of the principal investigator and a senior qualitative analyst, underwent training in the use of the qualitative codebook, including domain definitions and abstraction instructions. Coding was conducted using NVivo. Analysis team members read and summarized early transcripts independently, using the developed template, then compared summaries and refined domains, definitions, and summary procedures in an iterative process until consensus was reached. Once consistency was established across the analysis team, we divided the transcripts and completed the summaries, meeting regularly to review summaries, refine domain definitions, add new domain categories or sub-domains as agreed to by the team, and memo insights gleaned from the data. Once summarized, data were organized using matrices, which streamlined the process of noting similarities, differences, and response trends, making the summary and synthesis of important findings compact and accessible (Averill, 2002). We compared patient responses to provider responses in each domain.

Subgroup analysis. We identified patterns of similarities and differences in the responses of treatment responders versus nonresponders, using a matrix that drew material from the transcript summaries on individual patient factors that affect treatment response. Treatment response was assessed via the 20-item PTSD checklist for DSM-5 (PCL-5; Weathers et al., 2013), which is administered regularly during PTSD RRTP stays. PCL-5 scores were retrieved from the VA Corporate Data Warehouse. Good treatment responders were patients who experienced at least an 18-point reduction on the PCL-5 over the course of residential treatment. A reduction of 15–18 points is considered to be reliable change on the PCL-5 (Marx et al., 2021).

Hypothesized Themes.

Previous qualitative studies have identified factors associated with psychotherapy treatment engagement and response, including therapeutic alliance factors (Hundt et al., 2020; Kehle-Forbes et al., 2022; McLaughlin et al., 2014; Nilsson et al., 2007), distress tolerance/avoidance of negative emotion (Crowe & Luty, 2005), or lack of social support (Price et al., 2013). Previous quantitative investigations suggest that RRTP providers perceive the use of EBPs and the cohesion of the residential milieu to be important contributors to RRTP treatment effectiveness (Cook et al., 2014). In our previous quantitative analysis (Sripada et al., 2019), psychological resources including self-efficacy and self-help behaviors predicted better treatment response, whereas comorbid personality disorders and greater pain severity predicted worse treatment response. Additional factors that may affect the outcomes of residential PTSD treatment include cooccurring substance use and overall functioning (Szafranski et al., 2014, 2016). Our interview guide and analysis codebook included each of these domains.

Results

Psychological Resources

Patients and providers tended to agree on the psychological resources that bolstered treatment, including openness, willingness to change, motivation, persistence, resilience, hope, and flexibility. When asked about traits that were central to treatment success, several patients and providers mentioned the importance of being “honest, open, and willing.” For example, one patient described the importance of being open to change:

… I had that mindset when I made the choice, the decision to enter this treatment, to humble myself and just go with the possibilities and that’s the trust and part of my illness is I do not trust. So, it’s just that concept, even before I came to the program, it was something that I had to relinquish. So that was, to me, the way I look at it, that was the part of the program before the program is to just allow myself to be open.

For this patient, being open and trusting was key. Patients and providers also highlighted the importance of motivation. As one provider said:

… there are also people that come here from the beginning and from Day 1, they put in 100% and they work really, really hard, and develop, you tell them to do the 3 worksheets and they do 6, you know and you know these are the people that you know you see come in and just really turn around. You know their symptoms really improve.

As this provider noted, patients who worked very hard in therapy experienced substantial benefit. However, several common barriers were noted in the domain of psychological resources, including guilt and shame. As one patient stated,

I mean I have so much guilt and shame from the past that I can’t, I’m doing better at it but I still can’t leave the past in the past. And if something could be going very you know great and the back of my mind I’ll say oh well you know I don’t deserve to be happy, I don’t deserve to be getting this excellent treatment that I’m getting because I lied to my kids for 20 years about my drinking or I lied to other people or I quit jobs and stuff and yeah, the past is my biggest issue.

For this patient, guilt and shame made it difficult to move forward in therapy.

Some providers mentioned that substance use disorders and emotion dysregulation were treatment limiting factors. Other providers noted that patients were less likely to improve when they viewed PTSD as a stable trait instead of an illness with a recovery trajectory. Additional representative quotes can be found in Table 1.

Table 1.

Representative Quotes by Theme

Theme	Example Quotes
Psychol. Resources
Openness	… being open-minded, being willing to listen to other options other than you know your own belief system, challenging the belief system, going to the handbooks and doing the worksheets, um, yeah, I would have to say the number one thing is actually the open-mindedness and willingness to want to do something about it rather than just you know saying a big F you and I’m just going to do it my way and my way didn’t work obviously (patient).
Openness	I’m a recovering alcoholic so I’ve been in AA for almost 17 years, so AA’s outlook is very similar but I’ve never had PTSD or depression actually dealt with directly, and but the philosophies and things with AA and the attitude you have to have in AA about being, well they use the three letters, HOW, honest, open mind, and willing, got me to come into this program with those three things in mind (patient).
Emotion Dysregulation	Well emotional regulation usually prevents a Veteran from being able to really complete our program. So if they don’t have decent emotional regulation, or interpersonal skills, we usually can’t get them all the way through, it usually implodes. Other ones we can weather, we can weather substance abuse, we can weather antisocial, we can weather them going up against the rules of things, we can even weather some psychosis … but when they’re really dysregulated, that really interferes (provider).
PTSD viewed as stable trait	…some people, they are invested in the idea that you know “I have PTSD and there’s nothing you can do about that, you know I’m going to have PTSD until the day I die and this is just what it looks like.” Sometimes they have their own self-imposed like idea of what PTSD is as part of their identity (provider).
Social and Contextual Resources
Not alone	… nobody understands my trauma, nobody understands what I’ve been through in the military or previous to that and having (crying) other women here who are also scared to come forward, who are also scared to talk about it, has been a relief because I know that I’m not the only one, like I know that it’s just not me and that I’m not being an ass for coming forward or talking about it, so it’s helped. It’s helped a lot (patient).
Cohesion of residential milieu	When someone’s struggling, you know often times, I’ve seen this happen again and again, somebody will reach out and somebody in the group will offer them support, they’ll go out and have dinner and then the next thing you know they’re doing better and they’re re-engaged. It can have a very positive effect in the same way you know if you have a bad group, sometimes that can rub off on people too. It’s usually not the case, I think usually it’s the other way around, we get more where there’s, you know, no soldier left behind where they’re trying to help each other (provider).
Families and family programs	Families, major family stuff going on, which they use that as avoidance, so family stressors going on and they use that as a reason to bail. Probably more often than not it’s an avoidance of doing treatment and gee, you really really had to go deal with a family emergency, and sometimes the family emergencies are legit, I mean somebody literally does get sick or bad things really do happen and it’s not avoidance that’s causing them to need to return to that, but often times it’s avoidance that gets in the way (provider).
Families and family programs	We do have some family members that are very supportive and appropriately supportive. Unfortunately, we also have a lot of family members that are supportive in very unhealthy ways, so they really cater to the Veteran’s avoidance and really kind of help the Veteran take on the sick role and we are seeing a little bit more of that I think with the caretaker program and having so many caretakers for Veterans with PTSD really reinforces a lot of unhealthy relationship patterns. So that’s been pretty frustrating to see that, that happening (provider).
Social connection	It helps relieve some of the pressure I put on myself with how I expect interactions to go, I’m less skeptical. I better understand where people are coming from, I’ve been kind of a pessimist and rather negative about people because I felt they didn’t have any experiences or things that I did, but everybody’s got their own story and I’m more accepting of that now (patient).
Post-discharge care	… we have a lot of folks who aren’t local and so I think one thing that would really help with that transition is to have more of their support system involved but also like when they’re going back to their referring providers, like we really try to do a good job kind of communicating but some of them are more receptive to that than others, some VAs they’re going back to have more resources than others. I think that, you know that’s a really sensitive and vulnerable time for them and so I just wish that we could kind of do more to support them (provider).
Material Resources
Therapy modalities	…just giving you knowledge about what you have, symptoms that can arise, gives you a lot of peace in your mind because a lot of times you’re just kind of lost, you have no idea what’s going on with you, so the information aspect of the treatment is quite helpful (patient).
Therapy modalities	The way they had it like spread out as far as us having our trauma sessions, but then we would you know try to soften it a little bit by going on out on the outside … and I loved it. It was like a good balance between really getting in depth and hardcore with it and then you know kind of just lightening it up and breathing a little by going out to you know like horse farms or you know group sessions, bowling and stuff like that (patient).
Structure	Structure, it’s definitely been structure. It’s waking up at this time you got to go to this class, this class, this class, and before I came here, it would be laying bed until 2:00 in the afternoon, get up, not do anything, just feel sorry for myself. So I think that helped out a tremendous amount (patient).
Structure	…and what I’ve learned over time working with people that for instance had borderline personality disorder, that’s a pretty common one, that structures and boundaries are very important and comforting to people with borderline personality disorder and for people with PTSD because my definition of trauma has no boundaries, no limits, it’s kind of you know it’s a violation of boundaries and so when folks with borderline personality disorder tendencies bump up against structure, their initial response is to push it, but then once they realize the wall is not moving or that that’s actually what you say is what you mean, then they settle into it and feel more comfortable because it’s like, “Okay, now I know where the line is, I know where the ground is underneath me,” and so managing those kinds of issues in that way has been really helpful (provider).
Cohort model	I love that we start together and end together because the other treatment program that I was in kind of was a cluster. You know you only come in with 1 or 2 people, you only, you don’t get really that retention, you know that safety in numbers type feeling because every other day there’s people leaving, you know. Here, we start at the same time and leave at the same time and so it kind of gives us that extra little bond, you know, it’s kind of like going through the war time again, you know, we make friends with people and work on literally ourselves without having to try, so it’s pretty good (patient).
Rolling admission	… it’s hard to develop treatment that’s a rolling admission treatment that you get a lot of success with or a lot of change I think it’s much easier when you have a cohort with a protocol that you follow from beginning to end to see change because you get a build up of earlier stuff. So yeah, I think as far as having that limitation of having a rolling cohort, which allows folks access to care, results in maybe not as much change as we’d like to see but we still see change (provider).
Rolling admission	… and I think also the rolling admission part of our treatment is sometimes kind of nice because you’ll have someone nearing the end as well as someone at the beginning and the one at the end often gives hope to those at the beginning that, “You can do this, things will change, kind of have faith in this process,” and that gives them a little bit more kind of, a bolster of, “Maybe I can be like that when I leave here,” which often, just having that hope that things can be different will create that change (provider).
Physical Health Resources
Availability of medical care	…we’ve a really great nursing staff that takes care of making sure people end up at the right appointments. They are also aware of like what groups are okay to schedule during, like you wouldn’t want someone to miss an individual CPT session or CPT group because they’re going to miss the homework, they’re going to miss a lot of, so what do we do, we, the nurses know, “Okay, we can send them during this period of time that way they’re not [going] to miss so much programming.” We also have physical therapy downstairs which is kind of cool because then they can get to physical therapy and come back up, they don’t have to go to the main hospital for that. A lot of the medical appointments they have to go to urgent care, they have to go to the main hospital which knocks them out of a lot of programming (provider).

Social and Contextual Resources

Patients and providers identified several key factors in the domain of social and contextual resources, including camaraderie among Veterans, therapeutic alliance with providers, and decreased stigma after hearing the stories of others going through similar experiences. Some patients and providers also noted that friendships and the development of social skills were important. Several participants mentioned the impact of family support and family involvement.

An emergent theme from patient and provider interviews was the cohesion of residential milieu. For patients, the group setting was generally perceived as a very positive aspect of treatment. Many patients noted that it helped them feel less alone and that they felt a kinship with the other Veterans:

…I think the brotherhood, I mean we were all in the military and we’re all, it’s amazing how 15, 17 whatever people are here don’t know each other, never met each other, but put them in a group and it’s like a big family. It’s amazing how it works and I wish things out in the world would, people could come together like we have.

In addition to providing social support, the residential milieu provided Veterans with a forum to share traumatic experiences in way they previously had not been able to do. Being around other Veterans with similar traumatic experiences was also noted to help reduce stigma. One patient specifically noted the benefit of the cohort model (as opposed to rolling admission) and how it helped form a bond with the other Veterans in the program (see Table 1).

Staff in all three programs echoed the importance of the milieu format. Several said it was one of the most important parts of treatment:

I think some of the relationships that they form within the program are definitely some of the most therapeutic … I think being able to share things that they … haven’t shared you know with anybody or very limited people and to be able to talk about those things in a group is incredibly healing.

From this provider’s perspective, sharing traumatic experiences with trusted peers was an invaluable aspect of treatment. However, several downsides to the residential milieu were also noted, including conflict between patients and attributing therapeutic change to being around other Veterans instead of attributing it to confronting one’s trauma. One staff member emphasized the importance of helping patients generalize their newly attained social skills:

It’s probably one of the more common things we work on with them is for them to realize they could do the same thing with civilians too because they’ll get the notion that, “Oh, the only reason I’m doing better up here is because I’m around other Veterans, only other Veterans understand me, this is the only place I can do well,” and that’s just a bunch of, more stuck points. So we try to help them realize, “What are you doing different up here? What are you doing to engage with these people that you’re most likely doing at home.” You’ll engage with your neighbors and your family and your friends and if they can then realize, “Oh yeah, it’s what I’m putting into it that’s making a difference more than just the magic and the fact that we’re all Veterans.”

As this provider noted, helping patients generalize skills was a key task.

Several additional observations emerged from the data in this domain. Veterans whose cohorts were viewed as very cohesive took pride in this accomplishment. Staff noted that when cohorts got along well, the cohorts were powerfully therapeutic, but when cohorts did not get along well or when individual Veterans exhibited disruptive behavior, this dynamic had a significant negative impact on the group.

Therapeutic alliance emerged as another important theme. Most Veterans had very positive things to say about their providers. They noted that the open-door policy, the team approach, the insight therapists provided, and the way therapists pushed them were particularly helpful. As one patient said,

My individual therapist was great. In the beginning, I didn’t like her because she called me on my crap, you know but like I said that’s exactly what I needed. She remembered everything, everything that I talked about, even things that I thought she wouldn’t remember, she would bring them back up and help me, not force me to talk about it, but help me along the way in talking about it. So she really helped me in my progression in the program.

For this patient, therapist skill was viewed as a key component. Similarly, providers stated that therapeutic alliance was a particularly important part of positive treatment response.

So I think it’s very important that you have a good relationship between the patient and therapist. Sometimes I think that might be more important than what you’re actually doing or saying.

Patients and providers agreed that little could be accomplished without the foundation of a good therapeutic relationship.

Families were described as both a barrier and facilitator to recovery. Some families were described as supportive, and providers talked about the merits of family programming to help sustain positive change. However, some families were described as unsupportive or disruptive to treatment, and programs struggled to incorporate family programming due to both patient hesitancy and logistical difficulties. In terms of factors that negatively impacted outcomes, Veterans noted that their families didn’t always understand their problems:

My family’s supportive, but at the same time they don’t really understand my particular issue. Their views and you know God bless them, I love them dearly, is you know sort of “Take an aspirin and you know, it’ll be alright,” and that’s not the case at all.

This patient’s family did not fully understand their experiences and struggles. This sentiment was echoed by several other patients.

When asked about family involvement in treatment, providers reported mixed feelings on family programming. They said it had varying degrees of uptake and that family members could be both helpful and a hindrance, depending on how supportive they were (see Table 1). Providers noted that patients’ relationships with their families were complicated and involving families as a way to bolster treatment was often easier said than done. Nevertheless, some providers said that family programming could be formalized or expanded upon. They also mentioned that part of post-discharge planning involved trying to connect patients to social supports at home. Thus, families were viewed as an importance influence on patients.

Some patients noted that the skills they learned would improve their social connectedness with friends and family back home:

Like I said I isolated for so long, cut off from my friends, and here … you’re forced to be around people and it’s helped me, now that I’m seeing what it’s like to be around friends again because I’ve made friends here it has caused me now to reach out to friends that I had cut off, that missed me and I had missed them, so yeah, we’re all going to the same things, so it builds a great friendship, even when a guy is isolating we’ll notice and then we take it upon ourselves to go knock on the door and make sure the guy’s doing okay.

As this patient noted, residential treatment helped rehabilitate social skills as well as PTSD symptoms. A few patients talked about what they expected their lives to look like when they returned home. However, no patients described concrete transition plans, even though they were explicitly asked how the RRTP program prepared them for the transition back home. Furthermore, some Veterans expressed apprehension about discharge: “I’m afraid, I’m totally afraid. I’m cool here, but I’m wondering when I do go back home, am I going back in that same situation? Am I strong enough to get myself up out of it?”

Providers mentioned that part of post-discharge planning was trying to connect patients to outpatient appointments. Some expressed discouragement at the lack of availability of adequate post-discharge care. However, several providers from the low performing program mentioned their aftercare program as an important part of post-discharge care. Patients from that program expressed satisfaction with the availability of the aftercare program but said they would like to have it integrated into the RRTP program during RRTP program participation.

Material Resources

Therapy activities, specifically EBPs, were cited by patients and providers as the program elements that contributed most to therapeutic gains. Several patients also mentioned psychoeducation, normalization, and having a toolbox of skills to draw upon to deal with stressors. Patients and providers noted the benefits of the structured RRTP environment and providers discussed the advantages of the cohort model versus rolling admission model.

When asked about the most helpful part of treatment, many patients and all providers mentioned some aspect of EBPs for PTSD, and in particular, many referenced CPT. Over half of patients mentioned components of CPT, including the worksheets, writing, or working on stuck points: “I would say writing about the traumas, doing the worksheets because it really gets to the core of our problems…” As another noted, “…I never imagined all these years to look at it another way I’m looking at what I see in how I feel … went out and used it and it’s great.” For these patients, engaging in cognitive restructuring and challenging stuck points was very beneficial.

Several patients commented on the value of psychoeducation and normalization. Other patients mentioned additional therapy modalities, including in vivo exposure or “classes” more broadly. Some said it was nice to have a balance between trauma-focused work and non-trauma focused or recreational work. Several patients said the extra structure of the residential program was very helpful for them, and similar to the structure provided by the military.

…being in the military you’re always told where to go and what to do … they kind of figure things out for you and that’s what we’re used to that’s what I’m used to anyway. So, you can do it, you just need to be pointed in the right direction and that’s what’s nice about being in the VA facility, they point you in the right direction.

The defined structure was perceived to be conducive to recovery.

When asked about the least helpful components of treatment, patients most frequently mentioned the facilities within the residential environment. For example, some patients noted that the sleeping arrangements were problematic because the beds were uncomfortable, they were disrupted by staff check-ins, or they were bothered by other residents. Other patients mentioned that they did not like the food. These concerns highlight the impact of basic amenities on the patient’s experience of treatment.

Providers unanimously said that EBPs were the most important component of residential treatment. They also said that having a highly structured program was therapeutic for patients with PTSD or personality disorders. In terms of program admission structure, opinions were mixed. Some providers preferred cohort models whereas others preferred rolling admission (see Table 1). Other factors were viewed more consistently as weaknesses. For example, lack of wrap-around services for dormitories was mentioned by several individuals at the low-performing site. One provider stated:

You know when people have nightmares in the middle of the night we need to make sure we have staff that can respond in the most therapeutic way because that’s really where we start to see things bleed out … I would love for us operate on a unit where there’s the group room, the patients that are there, like the rooms are right there, … and then the staff that are there, like the milieu staff also working within the program so we can kind of wrap around some of that…we can do everything we can within our power here to just really be therapeutic, hold accountability, and really help them work on it, but … we’re having a lot of tests outside of the therapy room.

For this provider, lack of wrap-around services and having dorms too far from the residential treatment area were perceived to be treatment-limiting factors. This sentiment was echoed by some of the patients at this site.

Physical Health Resources

Patients and providers noted that medical issues were common, including chronic pain, physical deconditioning, sleep disorders, and other medical comorbidities. Some patients stated that their poor sleep was due to the disruptions of the residential environment whereas providers tended to attribute poor sleep to PTSD and other mental health problems. Patients were split on perceptions of medical care. Some patients believed that medical needs would be met if they requested medical care, whereas other patients thought that medical care was lacking. Providers mentioned that more embedded medical care would be beneficial.

Patients were also divided on whether medical conditions interfered in their ability to focus on PTSD treatment. As one patient noted:

… I’ve got a pretty serious knee injury going on right now, I’ve got three tears in ligaments and tendons and I’ve kind of taken a self-stance on this that I need the PTSD [treatment] before I really need to worry about that … because I want to have this before I go into something else, and then I would have the other option of having surgery and coming back after I recover from surgery, so, but I’d like to be here, do it now.

For this patient, it was important to focus on PTSD and try to delay his knee surgery until RRTP treatment was complete.

Providers noted that some patients were able to benefit despite serious medical conditions, whereas medical conditions and medical care got in the way for other patients. Patients and providers noted that having medical treatment onsite/nearby was a treatment facilitator, because it allowed patients to get medical treatment in conjunction with PTSD treatment (see Table 1). In contrast, when medical treatment was at a distance or unavailable at the local site, this was perceived as a barrier:

You know people who are not stable medically, they shouldn’t be here, that sometimes causes a problem because we don’t have on-site emergency services, or other medical services so … if they have medical issues come up they have to go to the main VA to get those medical appointments and that takes them out of our program and it could be disruptive.

The dichotomy between onsite resources facilitating treatment and inaccessible resources impeding treatment echoed the comments about dormitories, insofar as having dormitories at a distance or disconnected from the program was perceived as detracting from the effectiveness of the program.

Patients and providers noted that patients tended to see improvements in physical health due to regular routines of sleep, eating, and exercise. As one patient said,

I’m pushing myself kind of back to that military lifestyle, going, working out and eating 3 meals a day and you know actually getting 8 hours of sleep and now I’m seeing literally from this time 6 months ago, a completely different person. I don’t look the same, I don’t talk the same, walk the same, I don’t act the same. I mean I’ve even noticed, it was so bad to the point, I started mumbling a lot, so when I’m talking to people, they’re like, “Huh? What?” you know, now, I’m carrying all the conversation, people are like, “You’re very articulate, you know you sound normal, you sound good.”

For this patient, RRTP treatment was physically transformative in addition to being psychologically beneficial.

Patients and provider noted several ways to improve physical health care. Veterans suggested extending open gym hours. Providers suggested expanding treatment options for chronic pain, which many providers perceived to be the biggest barrier to PTSD treatment.

Responders versus Nonresponders Analysis

Eleven patients were categorized as responders and nine patients as nonresponders. Four patients did not have PCL-5 data and so were excluded from the analysis. As described above, we created a matrix of treatment responders versus nonresponders, divided by resource domain.

Psychological Resources: Responders versus Nonresponders.

Both responders and nonresponders mentioned determination, persistence, resilience, openness, acceptance, and hope. Interestingly, three responders mentioned sleep improvements whereas two nonresponders mentioned continued insomnia or sleep problems.

Social and Contextual Resources: Responders versus Nonresponders.

Both responders and nonresponders mentioned the camaraderie and social support provided by fellow RRTP participants. Most patients were very positive about their therapists, but three of the nine nonresponders said they did not get along very well with their therapists, lending support to the notion that therapeutic alliance is a key component to RRTP treatment effectiveness.

Material Resources: Responders versus Nonresponders.

CPT was universally acknowledged as the most helpful component of treatment, by responders and nonresponders alike. Many patients talked about other classes as well. Several patients, including responders and nonresponders, voiced that they wished the program was longer.

Physical Health Resources: Responders versus Nonresponders.

Medical problems were common in both responders and nonresponders. Veterans in both groups mentioned that they would like more gym hours and closer proximity for medical appointments.

Discussion

Patients and providers agreed on several factors critical to RRTP success. These included EBPs (particularly CPT), support and understanding from fellow patients, and skill and support from providers. Each of the five queried domains was deemed important to treatment response. In the domain of psychological resources, patients and providers described the importance of psychological flexibility, openness, and willingness to change. These factors echo the findings of a recent qualitative study of PTSD treatment participants in which motivation and hard work were perceived as key to success (St. Cyr et al., 2022). A few providers noted that substance misuse and emotion dysregulation were barriers to improvement. In the domain of social and contextual resources, patients and providers emphasized the importance of social support, normalization, and shared learning experiences, echoing the findings of other interview studies (Hundt et al., 2017; Kehle-Forbes et al., 2022; St. Cyr et al., 2022). Providers noted that the transition to outpatient care was a weak point of their programming. In the domain of material resources, patients and providers discussed the positive impact of EBPs but had variable views about other types of programming. Enhancements to the residential environment and the availability of wraparound services were viewed as very beneficial. In the domain of physical health resources, patients discussed the utility of the gym and providers noted the ways that chronic pain limited treatment effectiveness. These findings deepen our understanding of the factors promoting treatment response in the residential setting.

We also examined similarities and differences between patients’ and providers’ views of each domain. In general, patients and providers agreed on barriers and facilitators in each domain. However, certain factors were viewed differently. Patients frequently noted the importance of being amongst other Veterans, whereas providers sometimes downplayed the significance of this element. Providers were sometimes quick to attribute barriers to be related to patient avoidance, whereas patients perceived all concerns to be legitimate. Patients consistently noted the value of social support from their cohorts, whereas providers noted that some cohorts were more cohesive and therapeutic than others. Patients often described the arduousness of their journey to recovery and relief from PTSD symptoms. Providers noted the wide variability in patients’ level of engagement and degree of benefit in care. On the whole, patients and providers agreed that residential PTSD treatment was an excellent program that was often transformative for patients.

Another study aim was to identify differences between patient responders and patient nonresponders. Few differences emerged. However, three of the nine nonresponders expressed dissatisfaction with their therapists. Given that patients and providers consistently described the importance of therapeutic alliance in facilitating therapeutic gains, this is one modifiable factor that merits focused attention. Notably, a recent meta-analysis found that therapeutic alliance was a consistent predictor of PTSD outcomes in both in-person and remotely delivered treatment (Howard et al., 2022). Beyond this factor, we were unable to identify any differences between the responder and nonresponder groups. This lack of pronounced differences between responders and nonresponders perhaps reflects the fact that individuals with PTSD view recovery differently from providers (St. Cyr et al., 2022), and so degree of symptom change on the PCL-5 may not map on well to patients’ perceptions of improvement. A qualitative study of Canadian Veterans who underwent outpatient PTSD treatment found that Veterans often perceived recovery to be ongoing and nonlinear, and that numerical thresholds of clinically significant symptomatology on self-reported measures of PTSD did not always correspond to Veterans’ perceptions of improvement and recovery (St. Cyr et al., 2022). Another reason we may not have uncovered differences between patient responders and nonresponders is that we primarily focused on program characteristics and did not delve into patient contextual factors or social determinants of health. However, our findings highlight the fact that even Veterans who did not meet criteria for reliable change on the PCL-5 still had meaningful experiences in residential treatment that gave them new insights into their illness and changed their perspective.

Our analysis also allowed us to compare findings from the low-performing, mid-performing, and high-performing sites. We found that patient and provider views were generally consistent across facilities. However, some differences emerged. At the high performing site, many patients and providers noted the cohort model as a strength. At the mid performing site, providers had differing views on the rolling admission model. At the high performing site, most women in the women’s program appreciated the community provided by the program. At the low performing site, patients and providers noted the therapeutic disruption caused by having the dorms in a separate facility and not staffed by RRTP staff. These findings echo those of Sayer and colleagues (2017) that the practice environment, including infrastructure outside the clinic, influence the ability of a clinic to perform its core mission. We were not able to systematically investigate the impact of program design on outcomes. Additional work is needed to assess the impact of delivering EBPs following a skill-building phase versus delivering EBPs at the onset of treatment, concurrent with skill-building programs (e.g., Menefee et al., 2016).

The identified themes expand on and provide context for some of the findings of our previous quantitative analysis (Sripada et al., 2019). We previously found that greater psychological resources predicted better treatment response. In the current study, we expanded on this finding by identifying the importance of motivation, persistence, resilience, hope, and flexibility, as well as acceptance and willingness. In our quantitative study, we found that comorbid personality disorders and greater pain severity predicted worse treatment response. In the current study, providers noted that personality disorders were not problematic per se, but that the specific construct of emotion dysregulation was a trait that could interfere with treatment response and disrupt the group dynamic. Patients and providers both noted the significance of pain and other medical comorbidities in limiting treatment response. In fact, some providers stated that this was one of the biggest barriers to program engagement. Thus, finding additional ways to integrate medical or physical care into PTSD RRTP programming may reap significant benefits.

Our study has several limitations. Qualitative studies necessarily use small, non-random samples, which limits the ability to generalize findings from a study sample to a population. However, the goal of qualitative studies is to examine complex phenomena in depth to provide deeper understanding, uncover hidden variables, and start the process of building plausible explanations (Agius, 2013). An alternative strategy would have been to use survey techniques to assess patient and provider perceptions of reasons for poor response. However, this method could potentially be insufficient to address complex, personalized accounts or explanations of why treatment was not effective. Another limitation is that we clarified some questions after the initial round of interviews but did not return to the initial participants to ask our refined questions. Thus, it is possible that we missed some nuances in our initial round of interviews pertaining to outside stressors, psychological resources, and physical health resources.

Our study findings point to several modifications that may enhance treatment response. One is to prioritize the delivery of EBPs, which patients and providers consistently noted to be the most beneficial aspect of treatment. Notably, the VA recently put out a call for PTSD RRTPs to participate in a pilot program to provide EBPs in a massed delivery format, i.e. at least 4 EBP sessions per week (see FY 2022 VA Office of Mental Health and Suicide Prevention VHA 11 Clinical Services Memorandum VIEWS 7312721, Implementation and Expansion of Massed Evidence Based Psychotherapy). Providing EBPs in a massed format may improve access to and engagement with these first-line treatment modalities. Another potential modification is to find ways to capitalize on the residential milieu to encourage engagement in treatment and focus on change. Focusing on the factors that patients and providers noted to be most important, including openness, flexibility, and motivation, may help patients benefit more from treatment. Additionally, patients expressed uncertainty about the transition back home. It is possible that more information and resources were provided at the time of discharge, after the interviews were conducted. However, this lack of discussion could also point to a need to start building a discharge and transition plan early in care, especially given the fact that RRTP patients often lose their therapeutic gains after discharge (Grau et al., under review). Discharge plans may need to be strengthened in order to ease the transition away from residential treatment and help patients continue the recovery process. In addition, improving the residential environment to the extent possible may help enhance the cohesiveness of care and prevent disruptions. Thinking of creative ways to bolster social resources following discharge may also be helpful. Despite these challenges, all patients and providers described RRTP treatment as important and beneficial. Overall, patients and providers viewed residential treatment as a powerful experience that could make a substantial impact on Veterans’ lives.

Impact Statement.

When queried about factors that contribute to treatment response in VA residential rehabilitation treatment programs for posttraumatic stress disorder, patients and providers agreed that first-line evidence-based treatments, social support from Veterans and providers, and patient openness and willingness to change were all critical components to success. Policy makers and program administrators could consider ways to enhance these elements in residential programs.