An open trial of VA CONNECT: Caring for Our Nation's Needs Electronically during the COVID-19 Transition

Abstract

The COVID-19 pandemic impacted emotional well-being due to safety concerns, grief, employment impacts, and social interaction limitations. Face-to-face mental health treatment restrictions were especially impactful to veterans who often gain social enrichment from Veterans Health Administration (VHA) care. We present results from a novel group-based telehealth intervention, VA Caring for Our Nation's Needs Electronically during the COVID-19 Transition (VA CONNECT), which integrates skills training and social support to develop a COVID-19 Safety & Resilience Plan. Veterans (n ​= ​29) experiencing COVID-related stress participated in an open trial of this 10-session, manualized group VHA telehealth intervention. We examined whether COVID-19-related stress, adjustment disorder symptoms, and loneliness decreased, and coping strategy use increased after participation in VA CONNECT. Between baseline and two-month follow-up, participants reported a significant reduction in perceived stress and adjustment disorder symptoms, and an increase in planning coping skills use. Significant changes were not observed in loneliness or other specific coping strategies. Findings may support the utility of VA CONNECT as an intervention for pandemic-related stress and improving certain coping skills. Future research should explore group-based telehealth interventions like VA CONNECT with other populations within and outside of the VA, which have value during major disruptions to face-to-face mental healthcare access.

1. Introduction

The Coronavirus Disease 2019 (COVID-19) pandemic dramatically transformed daily functioning with physical distancing, travel bans, and “sheltering in place” orders. Along with the pandemic's direct health crisis and profound loss of life, citizens were devastated by loss of employment, decreased economic security, and monumental mental health concerns (e.g., excessive fear, uncertainty, anger, and anxiety; Hossain et al., 2020). In the United States, the outbreak and progression of the COVID-19 pandemic was associated with an alarming psychological toll coupled with reductions in access and availability of mental health providers (Centers for Disease Control and Prevention, 2020). Furthermore, 71% of Americans reported concern that physical distancing would negatively impact mental health (Benenson Strategy Group, 2020). These may be particularly salient for veterans who rely on few social contacts for support, such as those offered through the Veterans Health Administration (VHA; Benenson Strategy Group, 2020; Drebing et al., 2018; Kayman et al., 2016).

1.1. Telehealth response

In light of such concerns, it is imperative that mental health needs are met remotely when access to in-person care is limited. Previous research suggests e-mental health services, like internet-delivered interventions, are often as effective as face-to-face treatment and can be rapidly developed (Andersson, 2016; Wind et al., 2020). The VHA has been a particularly strong advocate of telehealth options even prior to the COVID-19 pandemic and provides existing infrastructure to better understand contexts most appropriate for telehealth services (Chen et al., 2019; Luci et al., 2020; Moreau et al., 2018). For example, VHA individual tele-psychotherapy is regularly conducted, especially in rural areas (Bumgarner et al., 2017; Egede et al., 2015; Lu et al., 2014). However, there is far less guidance for group-based interventions. Emerging literature supports the feasibility, acceptability, and efficacy of telehealth-adapted VA group interventions for various mental health needs during the COVID-19 pandemic (Fletcher et al., 2021; Fortier et al., 2022; Pernicano et al., 2022; Peterson et al., 2022; Possemato et al., 2022). However, to date, investigation into the potential utility of e-mental health interventions targeting COVID-19-specific concerns has been limited. A recent scoping review (Chennapragada et al., 2022) demonstrated positive effects of mental health treatments to address depression during the pandemic. These included interventions leveraging cognitive behavior therapy (CBT), psychoeducation, problem solving training, mindfulness, heliotherapy, strengthening social support, and muscle relaxation. Mode and schedule of delivery varied depending on study setting, though nearly all were delivered during a short window (i.e., between 1 and 35 days), several by telehealth, and few in a group format. While these interventions offer relevant insights into potentially valuable approaches for managing patient needs, they were exclusively designed for patients who had contracted COVID-19. There is, therefore, still a need to understand the value of telehealth adapted group interventions to address the broader consequences of the pandemic, even among those who have not contracted the coronavirus.

Such research could provide insights into how to best mitigate mental health consequences under pandemic conditions, especially for groups like veterans who may be more susceptible to the impact of COVID-19. For example, military veterans generally experience higher rates of mental health conditions compared to the general population (Hankin et al., 1999; Thompson et al., 2016; Williamson et al., 2018), as well as socioeconomic issues like homelessness (Tsai and Rosenheck, 2015). Additionally, military personnel have higher rates of exposure to potentially traumatic events than civilian adult counterparts (Frans et al., 2005). Coupled with values and beliefs developed during military service that may discourage disclosure and help-seeking (Zinzow et al., 2013), these factors may make veterans more susceptible to adverse experiences in response to stressors, including those associated with a global pandemic.

1.2. VA CONNECT

In response to the growing need to deliver telehealth interventions to a larger veteran audience, the authors developed VA Caring for Our Nation's Needs Electronically during the COVID-19 Transition (VA CONNECT) at the onset of the COVID-19 pandemic. VA CONNECT expands upon prior research highlighting the efficacy of building structured safety and wellness plans on symptoms related to suicide, depression, and anxiety to support at-risk individuals (Cook et al., 2012; Goodman et al., 2020a, 2022; Kayman et al., 2016; Stanley and Brown, 2012). Throughout their participation in VA CONNECT, participants develop a personalized action plan, a written compilation of therapeutic skills and resources to increase coping ability and manage future stress during the COVID-19 pandemic. Participants also learn strategies for better communication and problem solving, skills that are particularly relevant for those living in close proximity to others and struggling with interpersonal conflict. Given research suggesting group-format interventions often facilitate feelings of interpersonal connection and support among group members (Sullivan et al., 2021), VA CONNECT was also expected to facilitate connection among veterans by building an essential sense of community for those who are otherwise lonely and isolated.

1.2.1. Session protocol and content

VA CONNECT was delivered via telehealth across ten 60–75-min sessions twice per week (5 weeks total). The manualized intervention combines cognitive behavior therapy (CBT) and dialectical behavior therapy (DBT) skills, psychoeducation, and peer support to teach coping, emotion management, and stress relief while preparing a personalized Safety & Resilience Plan (Goodman et al., 2014, 2016, 2020b; Linehan et al., 1991; Stanley and Brown, 2012). Safety plans are intended to provide a written summary of skills and tools that are easily accessible for future use, especially when experiencing strong emotions that impact recall. VA CONNECT contains designated content for each session, enabling veterans to learn immediately helpful coping strategies, share concerns, exchange information, and support one another. Veterans added personalized skills and relevant information from group sessions to their individualized Safety and Resilience Plan, which consisted of nine distinct categories: 1) establishing and maintaining daily routine during COVID; 2) crisis resources; 3) breathing, mediation and relaxation techniques, 4) development of internal coping strategies, 5) recognizing and coping with emotions; 6) managing relationships during physical distancing mandates, 7) maximizing mental and physical health treatments; 8) improving physical well-being (e.g., diet, exercise); 9) finding meaning and gratitude (See Appendix A).

Each session topic introduced strategies targeting unique COVID-19 stressors (Table 1 ). These included: breathing and mindfulness exercises; internal coping strategies; emotion identification; managing interpersonal relationships; maximizing mental and physical health treatment; improving physical well-being; building positive emotions; and coping with COVID-19-related loss. Optional weekly booster sessions (maximum of 4) were offered to graduates for additional review and practice of session skills and techniques, continued peer support, and refinement of the Safety & Resilience Plan.

Table 1.

VA CONNECT treatment overview.

Session	Topic	Content	Homework
1	Introductions and Orientation	Psychoeducation about Acute Stress and COVID-19 Pandemic Review of Group Rules Crisis Management Strategies (e.g., Veterans Crisis Line) VA CONNECT Safety & Resilience Plan Introduction	Make environment more safe or comfortable
2	Meditation, Mindfulness, Relaxation techniques	Breathing Exercises Meditation Visualization	Practice one strategy reviewed
3	Internal Coping Strategies	Distress Tolerance Self-Soothing	Practice one of each type of strategy
4	Recognizing and Coping with Emotions	Emotion Labeling Components of Emotional Experience (e.g., warning signs, vulnerabilities) Emotion Mapping Activity	Label emotion Identify two coping strategies for components
5	Managing Interpersonal Conflicts	Current Relationships Strategies to Safely Socialize Communication and Problem-Solving	Share plan with loved one OR social activity
6	Maximizing Mental and Physical Health Treatment	Obstacles to Health Management During COVID-19 Medication Management Managing Sleep	Practice strategies to improve sleep
7	Physical Health Management	Nutritional Eating and Water Intake Access to Food Resources Physical Activity/Exercise	Apply strategy to diet and physical activity
8	Dealing with Loss during COVID-19	Emotional Reactions to Loss Strategies for Grieving and Alternative Commemorations	None
9	Building Positive Emotion	Meaning and Sources of Hope Ways to Practice Gratitude	Practice gratitude exercise
10	Recap	Review Safety & Resilience Plan Explore Sharing of Plan with Others Compile Resource List	None

Participants were provided a copy of the manual by email or postal service, which was projected on the screen during each session. Additionally, participants were provided an instructional document guiding them through use of the telehealth platform, VA WebEx. Admission to the group occurred on a rolling basis.

Sessions followed the following format: check-in, homework review, session outline review, presentation of new material in manual chapter, discussion questions, in-session exercises (i.e., practice of techniques and/or role plays), assignment of homework specific to individual group members, and barriers to homework completion. Session content was modeled on a structure similar to a traditional safety plan. For example, session one oriented participants to the Veterans Crisis Line and other emergency resources, as well as strategies to combat vulnerability factors for strong emotions which encouraged gaining/maintaining structure during COVID (e.g., waking up at the same time every day, brushing teeth, changing clothes). Due to the rolling basis enrollment of the study, sessions were not necessarily completed in chronological order. However, following the introductory session, content is not necessarily additive and thus did not require following the session order listed. For example, strategies to maximize physical health could have been reviewed prior to discussing managing interpersonal difficulties if thought to be beneficial to most group participants. If a participant entered treatment after the first session, interventionists outreached to individually orient veterans to the group format, identify personal goals for treatment, and review group rules. Notably, a tenth session focused on coping with loss and grief during COVID (e.g., death of a loved one, loss of employment, closure without typical services) was developed and added in response to patient-identified needs and the overwhelming degree of loss experienced by clinical patients at the James J. Peters VAMC in the Bronx, NY.

1.3. Present study

The present study expands on prior findings supporting the feasibility and acceptability of VA CONNECT (Gromatsky et al., 2021), the first group-based telehealth VA intervention specifically developed to address acute stress and related psychological concerns secondary to the COVID-19 pandemic. We specifically explored its utility for addressing the following treatment targets: decreasing perceived stress (primary), adjustment disorder symptoms (primary), and loneliness (secondary) as well as increasing effective coping skills (secondary). Evaluation and refinement of such interventions are pivotal for addressing current and future pandemic-related mental health needs as the virus continues to affect the global population.

2. Method

2.1. Participants

A total of 33 veterans were referred and enrolled, 29 of whom completed a baseline assessment. Four participants were lost between enrollment and initiation of study protocol. All 29 participants met inclusion criteria, which included: English proficiency; 18 years of age or older; ≥14 (“moderate” or higher) on the Perceived Stress Scale (PSS-10; Cohen and Williamson, 1988); approval from the veteran's mental health provider for their participation; access to internet and technology for participation in telehealth. Exclusion criteria included: untreated schizophrenia diagnosis or active psychotic disorder; inability to tolerate group therapy; or inability to connect to telehealth format. The trial followed a non-stratified trans-cohort model in which veterans of different demographics (e.g., rural and urban, young and elderly) may have been in the same group, allowing veterans of varying backgrounds to discuss shared experiences of COVID-19 stress.

2.2. Measures

Demographic data were collected via self-report. Clinical diagnoses were self-reported and confirmed via electronic health record review. COVID-19-specific items were included in consultation with co-Investigators working on large-scale COVID-19 projects at the Icahn School of Medicine at Mount Sinai (e.g., assessing diagnosed or suspected COVID-19 infection; living situation; sleep issues related to COVID-19).

Aspects of group cohesion were assessed with the Group Cohesion Scale-Revised (Treadwell et al., 2001) and are reported descriptively. This reliable and valid 25-item self-report measure assesses participant experience of communication, interaction, member retention, decision making, vulnerability among group members, and consistency between group and individual goals. Each item is rated on a four-point scale from 1 (“Strongly Disagree”) to 4 (“Strongly Agree”).

2.2.1. Primary outcomes

Perceived stress was measured using the Perceived Stress Scale-10 (Cohen and Williamson, 1988). This 10-item self-report measure assesses an individual's perceptions of life stressors experienced in the last month on a scale of 0 (“Never”) to 4 (“Very Often”). Scores range from 0 to 40 with higher scores indicating higher perceived stress. Thresholds have been defined as low stress (scores ranging from 0 to 13), moderate stress (scores ranging from 14 to 26), and high stress (scores ranging from 27 to 40). Past research on the Perceived Stress Scale supports its psychometric reliability and validity (Cohen and Williamson, 1988; Lee, 2012).

Adjustment disorder symptoms were measured using the Adjustment Disorder New Module – 8 (ADNM; Ben-Ezra et al., 2018). This 8-item self-report questionnaire assesses the frequency of adjustment disorder symptoms experienced within the past six months in relation to stressful life events. Symptoms are rated on a 4-point Likert scale ranging from 1 (“Never”) to 4 (“Often”) with higher summed scores indicating more adjustment difficulties. The 8-item version of this scale evaluates preoccupation with the stressor and failure to adapt. A cut-off score of 18.5 is recommended to identify adjustment disorder for clinical uses. This scale has been psychometrically validated as a measure of adjustment disorder symptoms in help-seeking patients (Kazlauskas et al., 2018).

2.2.2. Secondary outcomes

The Brief COPE (Carver et al., 1989) is a 28-item self-report measure of coping strategy utilization. This scale includes 14 two-item subscales which are each analyzed distinctly: self-distraction, active coping, denial, substance use, use of emotional support, use of instrumental support, behavioral disengagement, venting, positive reframing, planning, humor, acceptance, religion, and self-blame. Each item assesses the frequency with which each coping skill is currently used by participants on a scale from 1 (“I haven't been doing this at all”) to 4 (“I've been doing this a lot”). In addition to an overall score for the measure, higher subscale scores indicate greater engagement in a particular coping mechanism. Psychometric evaluation of the Brief COPE supports its reliability and validity (Carver, 1997).

Loneliness was assessed using the UCLA 3-item Loneliness Scale (Hughes et al., 2004) and evaluates: relational connectedness, social connectedness, and self-perceived isolation. Responses are provided on a 3-point scale indicating the frequency that the participant is currently experiencing (1, “Hardly Ever”; 2, “Some of the Time”; and 3, “Often”). Total scores range from 3 to 9 with higher scores indicating higher levels of loneliness. Past research has verified the validity and reliability of the short-form assessment of loneliness that also lessens participant burden (Arimoto and Tadaka, 2019).

2.3. Procedure

This open trial was conducted between June and December 2020 ​at the James J. Peters VAMC . All study procedures were approved by the local Institutional Review Board (IRB). Veterans self-reporting acute stress in response to COVID-19 were referred to VA CONNECT by a VHA specialty care/mental health service provider. One group was run consistently during the course of the trial. Admission to the group was rolling, allowing veterans immediate access to group participation following completion of the baseline assessment. Informed consent was obtained from all participants via phone and documented in the veteran's VHA electronic medical chart.

Self-report assessments of outcome variables were conducted at 3 time points: baseline (T0, prior to beginning VA CONNECT), after completion of the intervention (T1; approximately five weeks after beginning VA CONNECT), and follow-up (T2; approximately two months after beginning VA CONNECT, following completion of optional booster sessions). Assessments were completed by veterans independently via the web-based survey tool, Qualtrics, at each assessment period. When necessary (e.g., due to technological issues), participants were provided an option to complete the assessment battery via phone or mail. Basic demographic information and COVID-19-related items generated collaboratively with expert Co-Investigators were obtained at baseline.

As New York City rapidly became the epicenter of the COVID-19 crisis in March and April 2020, the VHA increased existing efforts to conduct tele-mental health visits. Due to physical distancing regulations and limited access to face-to-face mental health care for the foreseeable future, VA CONNECT was delivered via VA WebEx – an encrypted, HIPAA-compliant videoconferencing platform for VHA clinical care/research.

2.3.1. Interventionists, competence, and adherence

Study interventionists had extensive training in CBT, DBT, and suicide risk assessment. Two interventionists conducted each session to allow one to address emergencies that may arise with individual group members. One provider (SS) was present for all sessions to ensure continuity. The other interventionists rotated after the conclusion of the 10-session cycle. Sessions were never run without a licensed provider present. All interventionists participated in a one-day training to review treatment manual components, instruction in group therapy principles and group didactics, and suicide risk assessment. Virtual group supervision continued after each session. All interventionists followed the 82-page manual inclusive of session handouts and worksheets. The treatment was adjunctive to ongoing care, allowing veterans to continue receiving services from existing providers.

2.4. Data analytic approach

To examine whether participants in VA CONNECT experienced improvements in primary outcomes of interest (i.e., reduced perceived stress and symptoms of adjustment disorder) and secondary outcomes (i.e., coping strategies and loneliness) during the study period, we compared data collected at baseline (T0) to data collected at the final follow-up (T2, approximately two months after baseline). We used random intercept coefficient models to control for subject-to-subject variability and used a 5% level of significance when interpreting results. Estimates from models are theoretically adjusted by including data collected at the end of treatment (T1). Models were also adjusted using the number of sessions attended by each participant. We used complete case analysis and did not impute missing data.

3. Results

3.1. Sample characteristics

Demographic information for the sample is presented in Table 2 . Participants were predominantly male (86.21%) whose age varied greatly, ranging from 27 to 72 years old (M ​= ​50.52, SD ​= ​14.39 years). The sample was racially diverse with nearly half (41.38%) identifying as Hispanic; most identified as either African American (31.03%), multiracial (27.59%), or another non-Caucasian race (13.79%). Level of highest completed education also varied; however, most participants had completed some higher education (62.77%). Nearly half (44.83%) identified as being retired and nearly a third were either unemployed (24.14%) or laid off due to COVID-19 (10.34%). Additionally, almost half (44.83%) reported living alone and most veterans either rented or owned their own residence (72.41%). Most (82.76%) reported they were not formally diagnosed or suspected of having COVID-19 at the time of their baseline appointment. Nonetheless, over half reported experiencing worry (58.52%) and nightmares (65.52%) since the onset of the COVID-19 pandemic.

Table 2.

Demographic characteristics of VA CONNECT sample.

Variable	M (SD)/N (%)
Age	50.52 (14.39)
Sex
Male	25 (86.21)
Female	4 (13.79)
Race
Caucasian	8 (27.59)
African American	9 (31.03)
Multiracial	8 (27.59)
Other	4 (13.79)
Hispanic	12 (41.38)
Highest Education
High School Graduate or GED	5 (17.24)
Some College/Associate's	14 (48.28)
Bachelor's Degree	8 (27.59)
Post Graduate Work or Degree	2 (6.9)
Employment Status
Full-Time	2 (6.90)
Unemployed	7 (24.14)
Laid Off Due to COVID-19	3 (10.34)
Student	2 (6.90)
Retired	13 (44.83)
Self-Employed	2 (6.90)
Number of Adult Offspring (18+)	1.38 (1.40)
Living Alone	13 (44.83)
Living Situation
Own Residence	21 (72.41)
Staying with Others	4 (13.79)
Transitional/Temporary Housing	2 (6.9)
Street Homeless	2 (6.9)
Number of People in Veteran's Home
1 person	14 (48.28)
2 people	5 (17.24)
3 people	2 (6.90)
4 people	1 (3.45)
5 people	3 (10.34)
6 or more people	4 (13.79)
COVID-19 Diagnosed/Suspected	5 (17.24)
Worry Since Pandemic Onset (Y)	17 (58.62)
Nightmares Since Pandemic Onset (Y)	19 (65.52)

Note: n ​= ​29.

The sample reported a range of clinical diagnoses which were confirmed through electronic health record review. On average, each participant reported 2.31 diagnoses (SD ​= ​0.93, range: 1–5). These included borderline personality disorder (55.2%), major depressive disorder (48.3%), post-traumatic stress disorder (41.4%), bipolar disorder (27.6%), generalized anxiety disorder (20.7%), alcohol use disorder (13.8%), substance use disorder (10.3%), intermittent explosive disorder (6.9%) and schizoaffective disorder (n ​= ​1, 3.5%).

Groups were well attended with an average of 3.8 participants per manualized session (SD ​= ​1.56; range: 2–8). Participants attended, on average, 7.3 manualized sessions (SD ​= ​3.95) and 2.3 supportive booster sessions (SD ​= ​1.98); approximately half of participants attended one or more supportive booster session (n ​= ​17, 58.6%). Participants reported an average group cohesion score of 75.9 (n ​= ​23, SD ​= ​12.12, range: 49–96, possible range: 25–100) at T1 after attending manualized sessions.

3.2. Outcome analyses

Descriptive information for study primary and secondary outcomes by timepoint is presented in Table 3 . Results of random-intercept models measuring changes in study outcomes between baseline and follow-up timepoints are presented in Table 4 and graphically in Fig. 1 . Results were adjusted for the number of manualized sessions attended by each veteran (M ​= ​7.34, SD ​= ​3.95 sessions). However, results of tests of significance did not differ between unadjusted and adjusted estimates and statistics detailed below are based on unadjusted means.

Table 3.

Descriptive characteristics of outcome variables by timepoint.

Outcome Variable	Adjusteda	Time 0 ​M (SE)	Time 1 ​M (SE)	Time 2 ​M (SE)
PSS	Unadjusted	24.86 (1.32)	23.23 (1.73)	21.48 (1.91)
	Adjusted	24.61 (1.52)	23.24 (1.56)	21.49 (1.57)
ADNM	Unadjusted	26.90 (0.74)	26.50 (1.08)	24.80 (1.16)
	Adjusted	26.78 (0.94)	26.56 (0.97)	24.90 (0.98)
COPE Denial	Unadjusted	3.90 (0.30)	3.65 (0.34)	3.64 (0.38)
	Adjusted	3.87 (0.34)	3.72 (0.34)	3.74 (0.35)
COPE Substance	Unadjusted	3.38 (0.38)	3.08 (0.32)	3.08 (0.37)
	Adjusted	3.30 (0.33)	3.12 (0.35)	3.15 (0.35)
COPE Instrument	Unadjusted	5.10 (0.35)	5.50 (0.37)	5.08 (0.38)
	Adjusted	5.17 (0.34)	5.50 (0.35)	5.15 (0.36)
COPE Behavioral	Unadjusted	4.17 (0.37)	4.19 (0.41)	3.88 (0.37)
	Adjusted	4.12 (0.37)	4.25 (0.38)	3.93 (0.38)
COPE Reframing	Unadjusted	4.83 (0.35)	5.12 (0.37)	4.88 (0.42)
	Adjusted	4.85 (0.37)	5.09 (0.38)	4.95 (0.38)
COPE Planning	Unadjusted	5.14 (0.37)	5.46 (0.41)	5.88 (0.33)
	Adjusted	5.17 (0.36)	5.42 (0.37)	5.91 (0.38)
Loneliness	Unadjusted	7.31 (0.35)	6.88 (0.38)	6.76 (0.39)
	Adjusted	7.27 (0.35)	6.91 (0.36)	6.76 (0.36)

Note: n ​= ​29. M ​= ​Least Squares Means. SE = Standard Errors.

PSS = Perceived Stress Scale. Scores range from 0 to 40 with higher scores indicating higher levels of perceived stress. Scores between 0 and 13 are considered low; scores between 14 and 26 are considered moderate; and scores between 27 and 40 are considered high.

ADNM ​= ​Adjustment Disorder – New Module. Scores range from 8 to 32 with higher scores indicating more adjustment issues. A cut-off score of 18.5 is recommended to identify adjustment disorder for clinical uses.

COPE = The Brief COPE questionnaire individual subscales. Subscale scores (Denial, Substance, Instrument, Behavioral, Reframing, and Planning) each range from 2 to 8 with higher scores indicating greater use of coping skill type.

Loneliness ​= ​The UCLA Loneliness scale. Scores range from 3 to 9.

Time 0 data was collected at baseline, Time 1 data was collected after approximately five weeks, and Time 2 data was collected at the final two-month follow-up, after approximately four weeks of booster sessions.

^aEstimates are adjusted by number of sessions attended by each participant.

Table 4.

Treatment outcome changes: Baseline to follow-up using random intercept models.

Variable	Adjusteda	Estimate	SE	t	p
Primary Outcomes
PSS	Unadjusted	−3.12	1.17	−2.67	.01
	Adjusted	−3.12	1.17	−2.67	.01
ADNM	Unadjusted	−1.96	0.87	−2.25	.03
	Adjusted	−1.88	0.87	−2.16	.04
Secondary Outcomes
COPE Denial	Unadjusted	−0.15	0.29	−0.51	.61
	Adjusted	−0.14	0.29	−0.46	.65
COPE Substance	Unadjusted	−0.21	0.37	−0.58	.56
	Adjusted	−0.15	0.36	−0.41	.69
COPE Instrument	Unadjusted	0.03	0.34	0.08	.94
	Adjusted	−0.02	0.34	−0.06	.95
COPE Behavioral	Unadjusted	−0.23	0.33	−0.68	.50
	Adjusted	−0.20	0.33	−0.59	.56
COPE Reframing	Unadjusted	0.11	0.33	0.34	.74
	Adjusted	0.09	0.33	0.29	.77
COPE Planning	Unadjusted	0.77	0.34	2.27	.03
	Adjusted	0.75	0.34	2.21	.03
Loneliness	Unadjusted	−0.53	0.32	−1.69	.10
	Adjusted	−0.51	0.32	−1.60	.12

Note: df ​= ​49. SE = Standard Error. ADNM ​= ​Adjustment Disorder – New Module. COPE = Brief COPE questionnaire individual subscales. PSS = Perceived Stress Scale.

Results represent changes in primary and secondary outcome variables from baseline (timepoint 0) to final two-month follow-up (timepoint 2), after approximately four weeks of booster sessions.

^aEstimates are adjusted by number of sessions attended by each participant.

Fig. 1.

Box plot representation of primary outcomes by timepoint. Mean scores are plotted as horizontal lines, medians as a diamond, with minimum, maximum, interquartile range. T0 data was collected at baseline, T1 data was collected after approximately five weeks, and T2 data was collected at the final two-month follow-up after approximately four weeks of booster sessions.

PSS = Perceived Stress Scale. Scores range from 0 to 40 with higher scores indicating higher levels of perceived stress. Scores between 0 and 13 are considered low; scores between 14 and 26 are considered moderate; and scores between 27 and 40 are considered high.

ADNM ​= ​Adjustment Disorder – New Module. Scores range from 8 to 32 with higher scores indicating more adjustment issues.

A cut-off score of 18.5 is recommended to identify adjustment disorder for clinical uses.

Perceived stress reported by participants significantly decreased from baseline to follow-up (t[49] ​= ​−2.67, p ​= ​.01) as did adjustment disorder symptoms (t[49] ​= ​−2.25, p ​= ​.03). Additionally, Brief COPE Planning subscale scores were significantly higher at follow-up compared to baseline (t[49] ​= ​2.27, p ​= ​.03). Changes in ways of coping measured by all other Brief COPE subscales were not significant between baseline and follow-up. Change in self-reported loneliness was marginal but not statistically significant (t[49] ​= ​−1.69, p ​= ​.10).

4. Discussion

VA CONNECT is the first manualized, tele-mental health group delivered to veterans for the specific purpose of addressing COVID-19-related concerns. Findings suggest veterans experienced a statistically significant reduction in stress and adjustment difficulties after completing VA CONNECT, an intervention designed to provide skills and support during a global pandemic that dramatically changed lives and contributed to adversity. Improvements in outcome measures persisted following treatment completion, perhaps due to continued practice of the skillset developed. Group members had an opportunity to gain knowledge of a variety of strategies, especially planning ability, that may mitigate stress – including how to minimize COVID-19 exposure, make the environment safe and comfortable, and emergency resources. Additionally, changes may have been contributed to by guided in-session exercises and encouragement to practice skills between sessions (e.g., paced breathing, mindful meditation, and visualization).

During the study period, participants specifically reported increased use of coping skills related to strategizing for the future, as measured by significant increases in the Brief COPE Planning subscale. This may be related to the intervention's focus on building a resilience and safety plan, which are individually tailored with tangible examples of everyday activities to engage in. This proactively developed resource may have encouraged subsequent successful management of COVID-related fears, stressors, and adjustment difficulties. Furthermore, VA CONNECT combines guidance for gaining and maintaining structure during the pandemic (e.g., a regular sleep schedule, brushing teeth, changing clothes) with the development of specific internal coping strategies and emotional regulation skills (e.g., mindfulness, distraction, self-soothing, relaxation techniques, emotion mapping). The dual focus on behavioral and emotional strategies may have encouraged greater engagement with skill use and future-oriented thinking. Therefore, efforts to augment coping during times of acute stress and uncertainty may benefit from focusing on planning skills.

Later sessions of VA CONNECT were aimed at combatting interpersonal difficulties, addressing loneliness, and decreasing isolation. However, contrary to expectations, loneliness did not decrease during treatment despite participation in twice weekly psychotherapy groups with other veterans. This may be related to barriers of VA CONNECT's group telehealth format noted during the feasibility pilot (Gromatsky et al., 2021) that hindered treatment engagement. For example, during the feasibility pilot, some participants discussed technological and connectivity issues whereas a smaller proportion of veterans (25%) noted that participating in a virtual group with strangers was a barrier. Nonetheless, pilot study findings suggested participants found VA CONNECT to be acceptable and feasible, and present analyses support strong attendance and retention. Notably, additional qualitative pilot findings suggested 75% of veterans reported increased interpersonal connection and reduced isolation from participating (Gromatsky et al., 2021). Further work is needed to increase our understanding of treating isolation and loneliness among veterans.

4.1. Clinical implications

Results have significant implications for mental health treatment both within and outside the VA. Findings add support to the limited existing literature detailing mental health interventions which target emotional needs of individuals impacted by the COVID-19 pandemic (Chennapragada et al., 2022). Furthermore, interventions developed during the pandemic for individuals with depression and loneliness, rather than a COVID-19 diagnosis, had focused exclusively on adolescents (Cruwys et al., 2021). Notably, there was a relative paucity of research among adults which the present study sought to address. Results further attest to the potential utility of delivering e-mental health services to veterans who were socially and/or physically isolated due to the COVID-19 pandemic. Participation in VA CONNECT resulted in the development and refinement of a personalized Resilience and Safety Plan, adapted from evidence-based safety planning (Goodman et al., 2020a, 2022; Stanley et al., 2008, 2018). As with other action plans, the Resilience and Safety Plan created for VA CONNECT was collaboratively developed by each veteran with the support of study interventionists and other group members with the aim of mitigating subsequent stress related to COVID-19. Based on these results, VA CONNECT's group telehealth format may be a feasible treatment which can be easily implemented throughout the VA with potential for widespread adoption.

This intervention was developed and implemented early in the COVID-19 pandemic when there was a confluence of factors increasing stress among citizens. These included a limited understanding of the illness and diminished access to healthcare, testing, vaccines, life-saving medical equipment, and personal protective equipment. Stressors were further exacerbated by food shortages and the inability to visit loved ones in inpatient or long-term care facilities. In light of continued uncertainty of the COVID-19 pandemic over two years after its onset and whether physical distancing guidelines will be extended, there is great value in accessibility of a concrete plan for managing future episodes of stress and symptom exacerbation. Furthermore, availability of evidence-based telehealth interventions like VA CONNECT are especially valuable for other times of emergency (e.g., natural disasters, future pandemics; Marani et al., 2021) that prevent face-to-face psychotherapy, as well as among previously under-studied and difficult to access populations (e.g., rural, homeless populations).

4.2. Limitations and future directions

While VA CONNECT is a novel intervention with several notable strengths, there are limitations to this study worth noting. While promising and statistically significant, changes in stress and adjustment symptom scores were small, with subjective stress remaining “moderate” and adjustment disorder cut-off scores being exceeded across all timepoints. In the absence of validated cut-offs for change scores, additional research is necessary to understand their clinical significance. All participants were recruited from the James J. Peters VAMC in the NYC metro area – an epicenter early during the COVID-19 pandemic – limiting the generalizability of intervention effects for veterans living in other geographic areas. Although recruitment was limited to this area, participants lived across five states throughout the study. Furthermore, participants were predominantly older, non-Caucasian men. While diagnosis of a mental health disorder was not an inclusion criteria and VA CONNECT was not designed specifically for a clinical population, identification of participants relied on referral from a VHA mental health provider and all veterans completing the intervention had one or more mental health diagnosis, the most common being borderline personality disorder. Thus, future efforts to apply VA CONNECT within other VAs would be beneficial for understanding its utility among female veterans, those living in other geographical locations, and among community samples or those recruiting from the broader VHA population.

We also did not assess the degree of loss of life experienced by participants at baseline or follow-up timepoints and this topic was not included in the original treatment manual. Many participants identified as racial minorities and/or lived in lower socioeconomic areas disproportionately affected by COVID-19 death. This disparity was so apparent that a tenth session focused on bereavement and grief was added to the protocol.

The rolling basis of admission to the group facilitated immediate access to care but meant that content was not necessarily delivered in the same order for participants. Nonetheless, each veteran had the opportunity to complete each session and content was not duplicated to prevent redundancy. Additionally, while one of the interventionists (SS) was present at all sessions, other interventionists rotated during the course of the trial which could have impacted group dynamics. Group cohesion was, nonetheless, high and this dynamic resembles the dynamics of VA training settings where mental health students, especially, conduct rotations in different settings. Finally, despite promising findings, the present study did not have a control group limiting the ability to determine whether reported decreases in study outcomes were due to the intervention. Further large-scale efforts (e.g., a randomized control trial with a comparison group) are therefore necessary to fully evaluate the efficacy of VA CONNECT. This might include monitoring a separate group of veterans experiencing subjective stress who do not receive an additional clinical intervention, or who receive psychoeducation resources (e.g.) in the absence of a skills and support group.

4.3. Conclusions

The present study investigated the effects of participation in a novel, telehealth group intervention (VA CONNECT) among a sample of veterans experiencing COVID-19 related stress. Skills and tools developed across ten sessions were applied to create a personalized Resilience and Safety Plan to manage ongoing stress during the COVID-19 pandemic. We found VA CONNECT was effective for reducing perceived stress and adjustment disorder symptoms as well as improving coping abilities related to future planning. Findings support the use of VA CONNECT as an effective means of delivering psychotherapy to isolated veterans which should be expanded on with other populations within and outside of the VA, and when limits to in-person care are in place (e.g., future pandemics and national emergencies).

Author statement

Contributors Molly Gromatsky, PhD – Conceptualization, Investigation, Methodology, Writing (Original Draft, Review, and Editing)

Sarah R. Sullivan, MS – Conceptualization, Data Curation, Methodology, Investigation, Project Administration, Writing (Original Draft, Review and Editing)

Ashley L. Greene, PhD – Investigation, Writing (Original Draft, Review and Editing)

Usha Govindarajulu, PhD, MS – Data Curation, Formal Analysis, Writing (Original Draft, Review and Editing)

Emily L. Mitchell, BA – Data Curation, Methodology, Project Administration, Writing (Review and Editing)

Emily R. Edwards, PhD – Investigation, Methodology, Writing (Review and Editing)

Robert Lane, PhD – Investigation, Writing (Review and Editing)

Kyra K. Hamerling-Potts, BA – Data Curation, Methodology, Project Administration, Writing (Review and Editing)

Angela Page Spears, BS – Methodology, Writing (Review and Editing)

Marianne Goodman, MD – Conceptualization, Funding Acquisition, Investigation, Methodology, Supervision, Writing (Review and Editing)

Dr. Gromatsky and Ms. Sullivan served as co-first authors who wrote the first draft of the manuscript and all authors contributed to and have approved the final revised version of the manuscript being resubmitted for consideration.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Handling Editor: Dr. Leonardo Fontenelle

Appendix A. Supplementary data

The following is the Supplementary data to this article: