Suicide is a serious public health issue in the USA (Fazel & Runeson, 2020), specifically affecting Veterans (McCarthy et al., 2015). Recent data suggests that 17 Veterans die by suicide daily (U.S. Department of Veterans Affairs [VA], 2021). Among the empirically supported interventions for suicidal thoughts and behaviors (Jobes, 2006; Linehan, 1993; Stanley et al., 2018), the Safety Planning Intervention (SPI) has been associated with a reduction in suicidal behavior among suicidal individuals following discharge from acute care settings (Stanley et al., 2018).

Considered a best practice, the SPI is a brief intervention in which the clinician and client work together to develop an individually tailored safety plan to reduce the short-term suicide risk. The safety plan comprises a written, prioritized list of personal suicide warning signs; internal coping strategies; social contacts or settings offering support and distraction from suicidal thoughts; contact information for VA professionals, the crisis line, and emergency services; and specific steps for how to make the immediate environment safer (Stanley & Brown, 2012; Stanley et al., 2008).

Multiple augmentations of the safety plan, such as its use in a smartphone application, also suggest statistically significant decreases in suicidal thoughts and behavior (Melvin et al., 2019). However, there are few recommended guidelines or mechanisms for refinement of the SPI in Veteran populations beyond its initial development (Stanley et al., 2018).

In the VA, the SPI has been expanded to a multi-session group intervention called Project Life Force (PLF; Goodman et al., 2020, 2021) to maximize SPI effectiveness and leverage the benefits of group treatment including (1) diminishing social isolation and increasing social support, a protective factor against suicide (Rings et al., 2012); (2) its cost-effectiveness and maximizing staff time (McDermut et al., 2001); (3) the peer movement among those who have experienced suicidal crises is strong and growing (Bowersox et al., 2021); and (4) because unit cohesion in Veterans and military service members can positively impact mental health (Whitesell & Owens, 2012).

PLF is a manualized 10-session safety planning group that leverages Dialectical Behavioral Therapy (DBT) and psychoeducational approaches to bolster Veterans’ distress tolerance, emotion regulation, and interpersonal skills. In addition to addressing all six steps of the safety plan, PLF also focuses on augmenting physical well-being and strategies for sharing the plan with family/significant others, and reasons for living (Goodman et al., 2020, 2021). Veterans iteratively refine their safety plans as they learn these new skills throughout treatment. Importantly, the group format mitigates loneliness and fosters increased belongingness (Hatcher et al., 2013; Van Orden et al., 2012; Griffith, 2015; Johnson et al., 2019), both key risk factors for suicide prevention. The group cohort model facilitates connection among Veterans and aims to build a sense of community and social net, which is essential for those who are otherwise isolated, particularly relevant during COVID-19 when physical distancing orders were in place.

With the rapid rise of COVID-19, health care services transitioned to telehealth video delivery. In March 2020, a multi-year PLF randomized control trial (RCT) was fully converted to a telehealth intervention, Project Life Force-Telehealth (PLF-T) as in-person PLF groups were no longer allowed. To date, over 110 PLF-T sessions have been conducted with Veterans in the Bronx, NY, and Philadelphia, PA. Following procedures outlined by Sullivan and colleagues (2021a), the PLF research team focused on implementing telehealth procedures (e.g., mailing headphones for privacy, delivering group via telephone or videoconference platform, providing support for technical issues, emailing and mailing PLF intervention materials) that would make treatment accessible across racial, ethnic, and income groups.

Minimal evidence exists for suicide-specific group treatment for high-risk individuals offered over telehealth (Gentry et al., 2019; Lopez et al., 2020; Sullivan et al., 2021b). As a first step to understanding PLF-T, qualitative interviews were conducted to assess acceptability, feasibility, impact, and implementation (adaptations, barriers, and facilitators) while quantitative outcome data are being collected. Funded by a Clinical Science Research and Development supplement, this qualitative study focused on answering two key questions about PLF-T: (1) What are key stakeholder experiences of this group safety planning telehealth-delivered intervention?; and (2) What can we learn about adaptations, barriers, and facilitators to prepare for future implementation of PLF-T?

Methods

Setting and Participants

This study was conducted at two VA Medical Centers: James J. Peters VA (JJPVA) Medical Center in the Bronx, NY, and Michael J. Crescenz VA Medical Center in Philadelphia, PA. To deliver PLF via telehealth, the research team utilized either VA Video Connect (VVC) or VA WebEx software. PLF-T group facilitators included a psychiatrist and clinical psychologist, both experienced DBT and SPI clinicians. Training in PLF was performed by the treatment developer with ongoing fidelity measures and audiotape reviews by an outside rater. Approval was obtained from Institutional Review Boards (IRBs) at both sites.

Participants were recruited from a parent study, an RCT comparing PLF to treatment as usual among high-risk suicidal Veterans at the aforementioned two VAs (Goodman et al., 2020). The parent study recruited subjects from inpatient psychiatric units, VA high-risk lists managed by Suicide Prevention Coordinators, and outpatient mental health clinics.

Inclusion criteria included 18–89 years of age, and heightened suicide risk defined as a current suicide-related inpatient admission; presence on the high-risk list; and for outpatients, either a suicide attempt in the past year or active suicidal ideation with at least some intent within the past month using the Columbia Suicide Severity Rating Scale (C-SSRS; Posner et al., 2011). Exclusion criteria included unable to speak English or consent, unable to tolerate group intervention format, medically supervised withdrawal for substance use, schizophrenia diagnosis, or current participation in another intervention trial. Full inclusion and exclusion criteria are found in a previous publication (Goodman et al., 2020).

Veteran participants who were randomized to the PLF-T video group in the parent study were invited to participate in this qualitative study at the end of their participation. Before participation in the qualitative interviews, all Veterans provided informed consent.

Procedures

Due to the COVID-19 pandemic, all data collection was conducted via telephone or Webex, a HIPAA-compliant videoconference platform. Interviewers had at least a Master’s degree in clinical psychology or mental health counseling and received training in qualitative interviewing (e.g., role-playing) and biweekly supervision (e.g., audio recording review and feedback) by a researcher with expertise in qualitative methods and implementation science (SRP). All interviews were conducted using interview guides and were audio-recorded and transcribed verbatim by a team of research assistants, de-identified, and reviewed for accuracy by the project coordinator. Before the interviews, participants completed a demographic questionnaire and several clinical self-report assessments (post-PLF group participation) including the Beck Depression Inventory (BDI-II; Beck et al., 1996), Beck Hopelessness Scale (BHS; Beck, 1988), and the C-SSRS. Data collection took place between July 2020 and July 2021.

Qualitative approaches have been successfully applied in previous studies of suicidal individuals’ views (Awenat et al., 2017; Biddle et al., 2013; Hunter et al., 2013; Winter et al., 2013) and selected to investigate group participants’ views on acceptability, feasibility, and impact of the PLF-T intervention. Participant interview guides were developed using a theoretical framework of acceptability and feasibility of health care interventions (Sekhon et al, 2018) and revised based on feedback from the PLF-T research team. Specifically, we focused on the acceptability (e.g., experience with PLF-T, burden, and satisfaction) and feasibility (e.g., telehealth access, reasons for choosing phone or videoconference, experience with telehealth). Based on the literature on the effectiveness of SPIs (Goodman et al., 2021; Rogers et al., 2022), impact of PLF-T was assessed by asking questions about group participation and its impact on isolation, suicidal thoughts and behaviors, suicide-related coping (e.g., talking to your provider openly about suicide, calling a hotline or other supports, distraction), and creating/using a safety plan. Participants completed the Acceptability of Intervention Measure (AIM), Intervention Appropriateness Measure (IAM), and Feasibility of Intervention Measure (FIM) at the end of the qualitative interviews (Weiner et al., 2017). The AIM, IAM, and FIM measures yield sum scores from four items rated on a Likert scale of 1–5, for a total range from 5 to 20. On scales from 1 to 20, higher scores indicate higher levels of acceptability and appropriateness and feasibility.

Given the support for telehealth at the VA (Ferguson et al., 2021) and to prepare for future implementation, the PLF-T team (PLF-T group facilitators [MG, SJH] and project coordinator [SS]) were interviewed about adaptations to deliver PLF-T and implementation barriers and facilitators. Interview guides were developed using the Consolidated Framework for Implementation Research (CFIR; Damschroder et al., 2009) and reviewed by the PLF-T research team. The CFIR is an implementation framework to guide understanding of key determinants that impact the implementation of evidence-based practices. We used four CFIR domains relevant to this early stage of PLF-T implementation: (1) intervention characteristics (e.g., adaptability), (2) characteristics of individuals (e.g., facilitator beliefs and knowledge about PLF-T), (3) the outer setting (e.g., patient-level barriers, patient needs), and (4) the implementation process (e.g., planning and executing the implementation of PLF-T).

Data Analysis

Descriptive statistics were used to summarize available demographic and clinical information for Veteran participants. Veteran participant and PLF-T group facilitator interview data were coded by two independent coders (SRP and SS) and the PLF-T project coordinator interview was coded by two independent coders (SRP and EM because SS was the PLF-T project coordinator). A summary template and matrix analysis approach, a pragmatic and rigorous, deductive approach to synthesizing interview data into a template of key themes (Averill, 2002) was used to categorize participant responses along key topics (e.g., acceptability, feasibility, impact). First, the interviewers developed draft summaries of participant interviews. This involved summarizing the content discussed along key interview topics. Second, a researcher with expertise in implementation science and qualitative methods (SRP) reviewed each summary with each interviewer. Third, the researcher developed a draft template for documenting content emerging from each interview. Column headings represented key topics with some additional category breakdowns (e.g., experience with PLF-T, experience with telehealth, disclosure, loneliness/isolation, use of safety plan). Lastly, the two coders (SRP and SS for the participant and PLF-T facilitator interviews; SRP and EM for the PLF-T project coordinator interview) systematically extracted information from each summary and entered it into the draft matrix using short phrases (i.e., charting) (e.g., positive telehealth: convenient, easy access; barriers: lack of privacy). The coders met regularly after extracting information from 1 to 2 summaries, to review the matrix for areas of disagreement and to discuss and refine the draft matrix. The final matrix was then evaluated by a research team member (EM) based on a final review of transcripts.

We used several strategies for maximizing the rigor of our qualitative approach including progressively reducing the data using a series of defined steps (e.g., transcribing, summarizing, charting); using multiple team members at each step in data analysis; conducting frequent debriefing meetings throughout data analysis; and keeping an audit trail (Creswell & Creswell, 2017). We verified our results using a member-checking activity, wherein the researcher presented findings from the participant qualitative data at a PLF-T research team meeting to solicit their feedback (Birt et al., 2016).

Results

Veteran Participant Characteristics

Across sites, 19 Veterans were approached, and 17 agreed to participate. Study participants attended an average number of 7.1 telehealth sessions. PLF treatment completers are defined as attending 7 or more sessions (Goodman et al., 2020) with 13 participants meeting this designation. The demographic and clinical characteristics of Veterans are described in Table 1. Participants were mostly male (88%), age 50 (SD = 15.6), ethnically diverse, and either divorced or separated (54%). Suicide symptoms upon study entry included past month ideation with methods consideration (100%) and past year aborted, interrupted, or actual suicide attempt (59%).

Table 1 Demographic and clinical characteristics of Veteran participants (N = 17)

On scales from 1 to 20 with higher scores indicating higher levels of acceptability, appropriateness, and feasibility, Veterans rated PLF-T as highly acceptable (M = 17.50; SD = 2.92), appropriate (M = 17.25; SD = 3.59), and feasible (M = 18; SD = 2.45). Themes and subthemes from interviews with Veteran participants and the PLF-T team are described below.

Veterans: Acceptability

There were several themes related to the acceptability of PLF-T including the convenience and low cost of accessing the group by video, group facilitator efforts to be inclusive and ensure equitable participation, and the opportunity for reliable and consistent care.

Convenience

All participants cited the advantages of joining the group via telehealth including reduced financial strain, the ability to attend group, and balance competing demands like childcare or other appointments (medical or legal) and eliminated travel.

I really like the telehealth, and that is just me personally, I like the video chat because it gives me convenience of doing it from home. I was walking down the street and you know with my head buds on and was in the group and I had to take care of some important business and I did not want to miss the group; I had to go and see my parole officer. (P4)

A few participants explained that telehealth was preferred because of the lack of travel lessened the burden of physical limitations or disability, decreased the risk of getting COVID, and reduced their anxiety being triggered by commuting via public transportation.

Having to go out to a meeting now that could have been a burden because sometimes, I am not always at my best. I have PTSD, you know, sometime being on the bus, it’s crowded, it can be quite an experience for me. You know so, um just being able to do it in the comfort of my home is better. (P7)

Inclusive and Equitable Participation

Many participants commented that group facilitators tried to ensure equitable participation and create space for each group member to speak even though they were not in person. Participants attributed this to the facilitators’ ability to invite all group members via telehealth to participate, even those on the phone.

Very pleasant, very amicable, very sociable. I guess, inclusive, always felt like they were always rotating properly as far as getting enough feedback and, or enough involvement from each veteran, equally distributed throughout the session. So I really appreciate that about the instructors. (P5)

Reliable and Consistent Care

Participants commented on the reliability and consistency that PLF-T offered. Many commented on the reduced burden of making and rescheduling appointments and long waitlists.

With the telehealth when it comes to dealing with the group especially. I get undivided attention. Because I know that there’s a guarantee that I’ll have a session that Wednesday, number one. Then number two, I had a struggle with Bronx VA of trying to not only get a psychiatrist but to get my therapy, therapy going. And it took them months. At least with this I know every Wednesday, there is group. (P6)

Veterans: Feasibility

Participants shared the main challenges related to internet connectivity, operating the videoconference technology, and lack of privacy. With technical support from the communication coordinator (e.g., calling participants and connecting them to the group by phone, providing headphones, conducting an orientation on how to use Webex) and VA-issued iPads, these challenges were easier to address. While many reported that they participated by telephone and this made it easier to express themselves, others felt that they may be optimally engaged by video. For future PLF-T implementation, participants recommended providing training for participants on how to use the video platform, providing mobile phone cards for those who do not have unlimited minutes, and Webex apps pre-installed on the iPads for ease of use.

Lack of Privacy

A few participants raised privacy concerns before joining PLF-T. Although this did not impede their attendance, they found ways to secure private space for the group.

There were sometimes where my girlfriend was in house and certain things, I didn’t want her to hear, so I had to have her go to other room and I had to, you know, be quiet or careful. So, I guess because, you know, sometimes when there’s people around, you know, you don’t have that privacy. (P3)

Veterans: Impact

Themes related to the impact of PLF-T focused on increased awareness of suicidal thoughts and behaviors, increased connection and social support, willingness to disclose suicidal thoughts and behaviors, and how the safety plan provided coping strategies when experiencing suicidal thoughts and urges.

Increased Awareness of Suicidal Thoughts and Behaviors

A majority of participants commented on the group and its impact on building awareness of their thoughts and feelings in a time of crisis. Participants reported that the groups helped them learn to identify warning signs, and understanding the connection between their depression, PTSD and substance use disorder and suicidal thoughts, urges, and plans.

So warning signs. If I have felt something that I don’t know in the moment how I am feeling, I look at my safety plan, look on my warning signs and try to identify how I am feeling and try to do something to distract myself. (P1)

Increased Connection and Social Support

Participants reported that in the context of physical distancing, lockdown with families, and concerns about COVID-19, they often felt lonely and isolated. Many reported that the ability to connect with other Veterans having similar experiences and receiving input from others helped them feel supported. Many reported that they felt that fellow Veterans were like brothers and sisters who heal together. Additionally, PLF-T created a safe space to open up to others, to connect through facial expressions, and to give and receive words of support from group members.

Well, we are Veterans and we are brothers and sisters. So, we wanna heal because you might have something that, that I have and you might have some treatment that make him feel better. We help each other out. (P10)

And um just you know having other people in this group with similar situations gave us an opportunity to brainstorm um ways to deal with it better, you feel me. Just having different people’s ideas and mind working at the same time. I thought that was a really great. (P14)

Willingness to Disclose Suicidal Thoughts and Behaviors

A majority of participants reported that they were able to open up and disclose their suicidal thoughts to the group or other support persons, share details about emotionally challenging periods during COVID and isolation, and ask for group input on problems that wanted help in solving.

I mean, let’s face it like, you know, bringing up a conversation about how I feel suicidal is not, is not something you would do at dinner with your friends. So just being in a group with people who were talking about it and sharing those thoughts and those experiences they were having, I mean I just opened up right away. For me it was just good right away, it was just, it was an instant connection to the telehealth group. (P11)

I know that if I reach a 4 or 5, and you know what I think a 4 and a 5 is for me um in terms of dangers of getting to a suicidal idea. It made it more open; it helped me be more open in terms of, you know, people outside of group I think I would be more open to um calling someone. (P2)

Coping Strategies

Participants report that the safety plan helped them learn how to use distraction to put time between thoughts and actions and identify the need to speak to someone when in a time of crisis. Many commented that they placed the safety plan in different locations for easy access and revised it to their changing environment or supports, making it a living document.

So it helped prepare me a little bit, helped prepare me a little bit more. So in other words you know if, if, if 10 is the highest for the place I most don’t want to be, and 1 is being in this good place, you know uh it kind of helped me to evaluate some things in a way that I can address the issue at 4, 5, or 6 opposed to waiting to get to 8 or 9 to try address it. So I guess having those rules helped me to connect the dots a little better and to come up with um a game plan on how to um deal with things, cope with things a little better. (P5)

PLF-T Team: Adaptations to PLF for Telehealth Delivery

PLF-T team members identified several adaptations to deliver PLF via telehealth (Table 2). Adaptations included using a communications coordinator to conduct assertive outreach to facilitate engagement, providing a telehealth orientation session, restructuring sessions to review suicidal ideation severity, and emailing and screen sharing the PLF manual and safety plans to maximize learning. Compared to in-person groups when suicide risk assessment was conducted immediately before the group meeting, via telehealth it is conducted when participants join the meeting and asked if they want to speak to a group facilitator before joining the group. PLF-T research team members identified the need to further operationalize the assessment of suicide risk before weekly group sessions via telehealth as a future needed adaptation.

Table 2 Adaptations to PLF for telehealth delivery

PLF-T Team: Barriers and Facilitators to PLF-T Implementation

CFIR-informed example interview questions, barriers, and facilitators to implementing PLF-T reported by the PLF-T team are described in Table 3. Barriers to PLF-T implementation exist at the participant, provider, and system levels. Participant access and preference to join by phone, provider level of comfort with engaging high-risk individuals in a telehealth modality, and lack of infrastructure to facilitate access to technology and intervention materials were common barriers described. Specific roles (communication coordinator) and competencies (facilitators skilled in group therapy and technology) of the PLF-T research team allowed for flexibility and facilitated a rapid transition to telehealth.

Table 3 Barriers and facilitators to PLF-T implementation

Discussion

In the context of the COVID-19 crisis, Project Life Force was rapidly converted to telehealth delivery and continued to provide a suicide-specific safety planning group intervention to high-risk suicidal Veterans at rates similar to the in-person format. This study provides initial qualitative and quantitative evidence that a suicide-specific safety planning group delivered via telehealth to high-risk Veterans is feasible and acceptable. Veterans reported that PLF-T provided convenient and reliable access to care. Challenges associated with telehealth interventions, such as access to technology, internet connectivity, and privacy, were mitigated by adaptations the PLF-T team made including providing headphones to ensure privacy, sharing ground rules about confidentiality, and creating a safe space, a technology support orientation, and providing VA-issued tablets and flexibility to join groups via phone. Participants shared that the PLF-T group help them increase their awareness and insight into suicidal thoughts and behaviors, increase their connection to others having similar experiences as well as receive and provide social support to other group members, increase their willingness to disclose suicidality, and learn about coping strategies when experiencing suicidal thoughts and impulsive urges.

Information about barriers and facilitators implementing PLF-T with this small sample of high-risk Veterans provides an opportunity to develop implementation strategies to support future implementation and scale-up of PLF-T. Strategies may include creating an easy-to-access repository for all intervention materials and sharing best practices for telehealth and establishing a consortium of support for research and clinical programs navigating regulations and policies for telehealth in the VA. In addition, it will be important to evaluate the criteria for enrollment into PLF-T based on access to technology and comfort with telehealth group therapy.

There are several limitations. First, this study included a small size and focus on Veterans in two urban sites who mostly had access to a telephone or videoconference which may limit the generalizability of our findings to other Veteran populations in other locales and non-Veteran patients. The VA has been at the forefront of many telemedicine projects and therefore was better placed to expand during the pandemic. As a unified federal system without the state-to-state regulatory discrepancies that plagued those in other systems, it is likely to be able to integrate this type of program more easily. Second, acceptability and feasibility were assessed after participation in PLF-T and did not assess for pre-post group changes in these outcomes. Third, only 17 of the 19 eligible PLF-T participants agreed to this qualitative sub-study. It is possible that the two individuals who opted out may have disliked the telehealth format. Lastly, attrition for the PLF-T intervention exists (4 of 17 were not treatment completers) and does not appear to be related to the telehealth format; however, a full understanding of their reasons for dropping out requires further study.

Conclusion

The study provides evidence that, with a nimble and person-centered approach, telehealth groups are a viable and satisfactory option to support high-risk suicidal Veterans who have access to technology. Research with larger more diverse samples is needed to appreciate the multiple factors that can impact telehealth acceptance, feasibility and future implementation.