Providers’ use and perceived benefits and barriers of delivering Cognitive Behavioral Therapy for Nightmares (CBT-N) to Veterans

Abstract

Training and consultation in cognitive behavioral therapy for nightmares (CBT-N) were introduced for Veterans Health Administration (VHA) providers to reduce the frequency and severity of nightmares and improve functioning of veterans experiencing chronic nightmares. This study aimed to evaluate providers’ sustained delivery of CBT-N and their perceptions of benefits and barriers to implementation to improve the successful adoption and long-term use of CBT-N throughout the VHA. VHA providers (N = 100) who completed the CBT-N training program at least six months prior to the study were sent an online survey about their continued use, feedback, and perceptions of CBT-N. The survey received a 65% response rate. Overall, perceptions of CBT-N were favorable. Ninety-five percent of providers reported continued delivery of CBT-N since completing the training program, with 100% of responders intending to use CBT-N in the future. Among CBT-N users, treatment was perceived to reduce nightmare frequency and/or severity and improve other domains of functioning (i.e., sleep health, trauma symptoms, quality of life). A third of the CBT-N users denied any barriers to using CBT-N. Primary barriers to use were related to providers’ facility inner setting (e.g., unable to accommodate required session length or frequency, veteran population not appropriate for CBT-N) or veteran factors (e.g., schedule constraints, patient decision in shared decision-making). Results encourage ongoing CBT-N training initiatives within the VHA. Directly tracking veteran outcomes and identifying necessary modifications to adapt CBT-N within the constraints of the inner setting, while preserving treatment integrity, are warranted to ensure and promote intervention effectiveness.

Introduction

Nightmares pose a significant health concern among veterans. A recent study of a nationally representative sample of veterans found that approximately 1 in 6 veterans reported a lifetime history of trauma-related nightmares (Worley et al., 2025). Nightmares are associated with increased nocturnal awakenings, reduced sleep efficiency, and elevated motor activity during sleep (Germain & Nielsen, 2003). Beyond their sleep impact, nightmares negatively affect daytime functioning, contributing to a poorer perception of health, heightened psychological distress, substance use, and increased risk of suicidal ideation and behaviors (Maguen et al., 2009; Miller et al., 2015; Tanskanen et al., 2001; Vandrey et al., 2014). Importantly, these effects persist even when controlling for mental health conditions. Once considered secondary concerns, nightmares are now recognized as symptoms requiring targeted intervention (Babson & Feldner, 2010; Spoormaker & Montgomery, 2008).

A 2018 American Academy of Sleep Medicine (AASM) position paper recommends imagery rehearsal therapy (IRT), a cognitive behavioral approach, as the primary treatment of nightmare disorder (Morgenthaler et al., 2018). Broadly, this short-term treatment targets a nightmare by modifying the aversive content into a new dream script that is repeatedly imagined before sleep. Multiple IRT and other cognitive behavioral approaches exist that vary slightly in procedure and share a similar name (e.g., imagery rehearsal therapy; exposure, relaxation and rescripting therapy); therefore, the term cognitive behavioral therapy for nightmares (CBT-N) serves as an umbrella term encompassing the variation of multi-component cognitive and/or behavioral treatments for nightmares in adults. Meta-analyses find that CBT-N is efficacious in treating nightmares and can also improve symptoms of insomnia, posttraumatic stress disorder (PTSD), and depression (Yücel et al., 2019; Zhang et al., 2022).

Despite its efficacy, few training opportunities exist to enable delivery of high-quality CBT-N within the Veterans Health Administration (VHA). In response to this need, grassroot CBT-N dissemination efforts were initiated. In 2014, expert VHA clinicians and investigators, with previous training in delivering and evaluating CBT-N, began providing local trainings at their facilities and to other facilities upon request. In 2020, these experts formed a CBT-N workgroup and expanded their efforts to include the first national trainings for VHA providers with organized consultation and training requirements. Currently, the training program includes clinical workshops and four-months of expert consultation on a five- to six-session multi-component manualized protocol (Davis, 2009; Pruiksma et al., 2023). The CBT-N protocol used in this effort has nightmare exposure, rescription, and imagery rehearsal at its core. This protocol also includes sleep behavior modification, relaxation training, and relapse prevention planning. Between session practice assignments and tracking on daily sleep logs is included. Individual sessions typically are 60 to 90 minutes, and groups sessions are approximately 120 minutes.

A long-term goal of these CBT-N dissemination efforts is for VHA providers to integrate this treatment in their routine clinical practice across settings. However, dynamic contextual factors of these real-world settings, especially within large, complex healthcare systems, may pose barriers to sustained use of these newly acquired skills. Rigid clinical structures (e.g., scheduling grids), high productivity requirements, and generally higher risk clinical populations may require nuanced training and implementation of new treatment protocols. Therefore, a deeper understanding of barriers and facilitators of implementing CBT-N into practice is needed for scaling this bottom-up CBT-N training program initiative effectively. The Consolidated Framework for Implementation Research (CFIR; Damschroder et al., 2022) provides a framework for comprehensively evaluating barriers and facilitators for implementation of innovations. For this stage of CBT-N implementation, the following CFIR domains are relevant: innovation (CBT-N), inner setting (where CBT-N is being implemented), and individuals involved (CBT-N providers and veteran patients). This framework has been applied to implementation of other sleep-focused psychotherapies within VHA (e.g., CBT for Insomnia [CBT-I]; Koffel & Hagedorn, 2020), allowing for appropriate comparison.

Therefore, this study aimed to understand VHA-trained CBT-N providers’ ongoing use and perceptions of facilitators and barriers to the implementation of CBT-N following training completion. Unlike top-down rollouts, grassroot initiatives rely on clinician-driven adoption without formal mandates, raising questions about whether they face similar implementation challenges or distinct barriers. This study also allows for exploration of whether nightmare-focused treatments present unique obstacles to be sustainably integrated into real-world settings where formal implementation supports may be limited.

Method

Overall design

Study procedures were determined to be non-research by the Minneapolis VA Health Care System Research and Development committee and therefore do not require oversight by the Institutional Review Board. All VHA-trained CBT-N providers who had completed the training program requirements at least six months prior to the time of the evaluation were emailed a description of the evaluation and invited to participate by accessing the web-based survey through an embedded hyperlink (VA REDCap; Harris et al., 2019; Harris et al., 2009). The survey was voluntary, anonymous, and non-incentivized. Follow-up reminder invitations were sent two and four weeks later.

Participants

Invited participants were 100 VHA mental health providers who completed requirements of the CBT-N training program at least six months prior to the survey. These providers are VHA mental health staff; deliver psychotherapy services on a regular basis; and work in settings where nightmares are a presenting issue and CBT-N can be implemented. To apply to the CBT-N training program, interested providers responded to email announcements and calls for applications that were sent to relevant listservs (e.g., behavioral sleep medicine and PTSD providers), evidence-based practice coordinators, or word of mouth. For selection into the program, applications were automatically rejected from providers who did not complete the application, did not receive supervisory approval, worked in settings that cannot deliver CBT-N as intended without significant modifications (e.g., inpatient units, polytrauma rehabilitation), and/or provided clinical services <25% of their effort. The remaining applications were ranked by three members of the CBT-N workgroup with priority going to clinicians who indicated prior experience with evidence-based practice and worked within regions that have fewer trained CBT-N providers. Any conflicts were disambiguated by the CBT-N workgroup. The highest ranked providers were then assigned into consultation groups until the available spots were filled. To successfully complete the training program, these providers met the following requirements: (a) attended a virtual training workshop; (b) attended a minimum of 10 weekly consultation calls; (c) completed CBT-N with at least three individual veteran patients, two groups, or combination of group and individual during consultation, and (d) demonstrated fidelity, as determined by their CBT-N consultant, to the main treatment components. At the time of this survey distribution, approximately 74% of all providers who initiated consultation successfully completed the program. The reasons for non-completion were most frequently due to changes in job responsibilities, not receiving CBT-N-appropriate referrals, veteran patient attrition, or personal matters.

Survey

A brief (approximately 5 – 10 minute) survey was created to assess providers’ profession and contextual factors (e.g., type of VHA facility, primary domain of clinical practice, Veterans Integrated Service Network [VISN] location, and average weekly patient contacts), and current CBT-N use with veteran patients. Those who denied use of CBT-N since completing the training program were asked what factors contributed to their decision and whether they intended to use CBT-N with their patients in the future. Participants who reported using CBT-N in their care delivery within the past six months were asked to estimate the number of individual cases for which they delivered a full or partial course of CBT-N. They also were asked to rate their perceptions of CBT-N benefits focused on nightmare frequency or severity, overall sleep quality, trauma and/or depression symptoms, improved quality of life, changes in how the veteran perceived their nightmares, and overall satisfaction with the treatment (all items on a Likert scale 0 – 5, with 0 = no improvement or no satisfaction to 5 = significant improvement or very satisfied). These providers also were asked to endorse any potential barriers to the use of CBT-N from a list guided by the CFIR framework (see Table 1 for list). These barriers fit within the following CFIR domains: Innovation Characteristics (complexity of the intervention, relative advantage); Inner Setting (e.g., structural characteristics, compatibility, available resources), and Individuals (e.g., patient needs for participation; provider characteristics). Outer setting factors (e.g., financing, sociocultural influence) were not specifically assessed at this stage of investigation due to their difficulty to directly influence and the focus of identifying pragmatic changes. Process factors (e.g., the activities and strategies used to implement) also were not explicitly assessed. However, open-ended space was provided for all participants to share additional feedback on their experience with CBT-N (“Other barriers, please specify”, “Feel free to provide comments about any of the above barriers”; “Please share any other feedback/comments about your experience with CBT-N”).

Table 1.

List of potential barriers to use of CBT-N that responders could select.

Domain	Potential Barriers
	No barriers.
	Other barriers? Please specify.
Innovation
	I do not find CBT-N useful for my patients (e.g., little symptom improvement).
	This treatment is not culturally appropriate/sensitive for my Veterans.
	It is too complicated for my Veterans.
	CBT-N is not a match for my patient population or needs.
	This treatment has too much work outside of session for preparation and documentation.
	I do not believe CBT-N is suited for telehealth.
Individuals
	My patients are not interested in receiving CBT-N.
	My Veterans don’t like manualized treatment.
	I do not feel confident in delivering CBT-N.
	I need additional support (e.g., consultation).
Inner Setting
	My case load is too large to offer this treatment.
	I am unable to accommodate 60- or 90-minute appointments for this treatment in my schedule.
	I am unable to schedule weekly appointments for this treatment.
	My clinical setting limits my ability to see patients for the duration of this treatment (5 or 6 sessions).
	I conduct groups and it’s not valuable by VA metrics (not enough RVUs, long sessions) .

Finally, to estimate additional impact of these dissemination efforts, providers were asked whether they offered this treatment to patients in their private practice settings, if applicable, and if they have trained or supervised other VHA providers at their site in delivering this treatment modality.

Analytical Plan

Data were analyzed using R version 4.2.3. Descriptive statistics were used to report summary outcomes. Open-ended responses on perceptions of benefits and barriers to use were subjected to thematic analysis. Categories of clinician responses were identified by one of the authors (K.E.M.), identifying specific meaning units for further coding (Braun & Clarke, 2006). These categories were characterized and formatted as a codebook (available upon request). Along with the codebook developer, two additional authors (C.J.B. and J.A.C.) independently rated each response to fit within one or two thematic code(s). Interrater agreement was calculated using Krippendorf’s alpha, with alphas of 0.80 being considered satisfactory (Hayes & Krippendorff, 2007). There was substantial agreement between raters across each of the open-ended response categories (alpha = 0.94 for perceived benefits; alpha = 0.91 for perceived barriers). Any conflicting codes were disambiguated through the use of majority consensus between raters to determine final codes for each response.

Results

An overall response rate to outreach efforts for this survey was 65% (N = 65) over five weeks. The survey responders were psychologists (n = 49, 75.4%), social workers (n = 14, 21.5%), or other mental health providers (n = 2, 3%). All VISNs, except VISN 9, were represented, within the following U.S. geographic regions: Southeast (n = 11, 16.9%), Continental (n = 11, 16.9%), Pacific (n = 14, 21.5%), Midwest (n = 11, 16.9%), and North Atlantic (n = 18, 27.7%). Most responders completed consultation within the past 12 months of when the survey was distributed [Median = year 2023; range = years 2015 – 2024]. Participants reported primarily working in PTSD care teams (n = 25, 38.5%), general outpatient mental health clinics (n = 16, 24.6%), primary care-mental health integration (n = 11, 16.9%), behavioral health integrated program (n = 6, 9.2%), or clinical resource hubs (i.e., networks that combine in-person care and telehealth to support underserved medical facilities, n = 6, 9.2%). On average, participants had 24.6 (SD = 10.5) patient contacts per week. Table 2 provides additional details of the sample’s professional characteristics.

Table 2.

Sample professional characteristics.

Characteristic	M or n	SD or %
Highest degree
PhD	36	55.4
PsyD	14	21.5
Master’s	15	23.1
Years since earning degree	12.1	6.8
Profession
Psychologist	49	75.4
Social Worker	14	21.5
Marital family therapist	1	1.5
Other	1	1.5
Primary Domain of Clinical Practice
PTSD clinical team	25	38.5
General mental health	16	24.6
Primary care-mental health integration	11	16.9
Behavioral health integrated program	6	9.2
Clinical resource hub	6	9.2
Sleep center	3	4.6
Serious mental health clinic	1	1.5
Inpatient	1	1.5
Residential	1	1.5
Domiciliary	1	1.5
Integrated in specialty clinic	1	1.5
Other	2	3.1
Type of VA facility
Medical center	30	46.2
Outpatient clinic	11	16.9
Community-based outpatient clinic	11	16.9
Vet center	3	4.6
Clinical resource hub	8	12.3
Other	2	3.1
Patient contacts per week	24.6	10.48

Continued CBT-N Use

Nearly all responders reported continuing to deliver CBT-N in their local clinical setting since completing the training program (n = 62, 95.4%), and all responders (N = 65, 100%) reported they intended to use CBT-N with future VHA patients. Most providers estimated delivering a full treatment course to 1 – 5 patients (n = 43, 69.4%), followed by 6 – 10 patients (n = 15, 24.2%), and 11 – 20 patients (n = 4, 6.5%) in the past six months. Similar frequencies were observed for delivering partial courses of CBT-N: 1 – 5 patients (n = 48, 77.7%), 6 – 10 patients (n = 9, 14.5%), 11 – 20 patients (n = 3, 4.8%), and 21 – 30 patients (n = 2, 3.2%). Five providers (8.1%) reported using CBT-N in a group format, with most delivering 1 – 5 groups (n = 4) and one provider delivering 6 – 10 groups in the past 6 months. The average number of veterans per group was 3.2 (SD = 0.8, range = 2 – 4). The three providers who have not used CBT-N since completing the training program reported that it was due to a lack of referrals for CBT-N (n = 3) and/or being unable to accommodate at least 60-minute sessions in their clinical setting (n = 1).

Facilitators and Barriers to CBT-N Use

Providers identified facilitators and barriers for CBT-N use since completing the training program. These factors fell under three broad CFIR constructs: intervention characteristics, inner setting, and individuals. These facilitators and barriers are described below and outlined in Figure 1.

Figure 1.

Identified Consolidated Framework for Implementation Research (CFIR; Damschroder et al., 2022) facilitators and barriers, as applied to CBT-N implementation in VHA.

Intervention Characteristics

Perceived benefits.

Overall, providers who have continued to use CBT-N since completing the training program reported generally favorable perceptions of the treatment benefitting their veterans, as indicated by the means in the following areas (see also Table 3): improving nightmare frequency and/or severity (M = 4.0, SD = 0.8), shifting how the veteran relates to their nightmares (M = 4.0, SD = 0.9), improving overall sleep quality/sleep health (M = 3.8, SD = 0.9), improving quality of life and/or daytime functioning (M = 3.8, SD = 0.9), and improving trauma and/or depression symptoms (M = 3.6 , SD = 1.0). These providers also generally perceived that their veterans were satisfied with the treatment (M = 4.0, SD = 0.8). In open-response, 11 providers noted the following additional perceived benefits: CBT-N increased veterans’ readiness for other treatment, particularly trauma or exposure-based psychotherapies (n = 5), treatment improved other domains of functioning (social interactions, daily behaviors that influence sleep; n = 4), delivering the treatment contributed to provider satisfaction (n = 2), and CBT-N offered more efficient access to care by reducing the need for referral to specialty care (n = 1).

Table 3.

Provider perceptions of CBT-N benefits (N = 62).

Items	M (SD)	No benefit/Not satisfied (0 – 2) %	Neutral (3) %	Improvement/Satisfied (4 – 5) %
Nightmare frequency and/or severity	4.0 (0.8)	5	16	79
Shifts in how the veteran relates to the nightmare	4.0 (0.9)	5	19	76
Overall sleep quality/sleep health	3.8 (0.9)	6	23	71
Trauma and/or depression symptoms	3.6 (1.02)	10	32	58
Quality of life and/or daytime functioning	3.8 (0.9)	6	26	68
Veteran satisfaction with treatment	4.0 (0.8)	2	21	77

Perceived barriers.

Over a third of the providers who continue to use CBT-N reported no barriers to use with veterans (n = 23, 37.1%). Among the providers who did cite barriers to continued CBT-N use, concerns about CBT-N’s complexity (the treatment has too much administrative work outside of session, n = 4); lack of relative advantage (little symptom improvement (n = 3), and cultural appropriateness (treatment is not culturally appropriate/sensitive for veterans, n = 1) were cited. Six providers elaborated on barriers related to complexity and relative advantage, see Table 4 for details.

Table 4.

Provider open-response feedback on barriers of CBT-N use (n = 27).

CFIR Domain	n (%)	Response Example
Innovation
Complexity	5 (19%)	“Some veterans struggle with the workload outside of sessions.”
Relative Advantage	1 (4%)	“I would say 25% of patients have a really good response to the treatment with significant reduction in nightmares. The other 75% either a) don’t complete, b) are too activated by the therapy, or c) it doesn’t work.”
Inner Setting
Compatibility	2 (7%)	“It’s been a bit difficult implementing CBT-N in a residential settings given that it’s in a hospital setting, bed is not comfortable.”
Relative Priority	2 (7%)	“We don’t take as many cases for CBT-N and encouraged to do more trauma-focused treatments (i.e., CPT, PE, EMDR).”
Resources	9 (33%)	“The only thing stopping me from offering it more frequently is the demands on scheduling”; “It is hard for me to offer the treatment as much of as quickly as I would like because of caseload issues”;“I can do 60-minute appts but do not have support for a 90 minute. It is hard to get all the content in 60 minutes.”
Executing Treatment	5 (19%)	“Need a co-facilitator for the amount of work. It is hard to get a large group of referrals to do the group and then we have dropout or scheduling issues”
Individual
Patient Opportunity or Need	10 (37%)	“Veteran work schedules”; “Some patients are not proficient with technology, making it difficult to get sleep data efficiently.”
Patient Motivation	6 (22%)	“Patient avoidance of writing nightmares”; “While engaging in shared decision making, veterans pick CBT-N the least often.”

Note. CFIR = Consolidated Framework for Implementation Research.

Inner Setting

Compatibility and relative priority.

The following barriers were endorsed as incompatibilities of CBT-N with the providers’ clinical setting: unable to accommodate 60- or 90-minute appointments in schedule (n = 11), current caseload is too large to offer this treatment (n = 7), unable to schedule weekly appointments (n = 6), the clinical setting limits the provider’s ability to see veterans for the duration of the treatment (i.e., 5 or 6 sessions; n = 3), and delivering CBT-N in group is not valuable by VHA metrics (n = 1). In open responses on Inner Setting barriers (n = 13) organizational structure such as demands on schedules and inability to accommodate the recommended session lengths, requirements to offer other treatments (e.g., trauma-focused therapies), or the specific clinic setting not being a good fit for the treatment (e.g., residential setting) were reported.

Resources.

In open-responses, five providers described challenges with materials and equipment in treatment delivery, including difficulties navigating virtual sessions in a group format and sharing between session work (e.g., worksheets, sleep diaries) through secure messaging.

Individual Characteristics

Patient needs and resources.

While CBT-N was generally perceived favorably as a treatment option, some providers indicated that, when presented the option of CBT-N, veterans were not interested in receiving the treatment (n = 9), CBT-N was not a match for their veteran population or needs (n = 4), or that veterans do not like manualized treatment (n = 1). In open responses, providers (n = 16) reported additional barriers related to trauma-related avoidance, and logistical barriers to receiving care (e.g., veterans unable to schedule regular appointments, difficulty using technology for virtual appointments).

Provider needs and resources.

No respondents indicated a lack of confidence in CBT-N delivery; however, two providers cited a need for additional support via consultation (n = 2).

Additional Impact of Dissemination Efforts

Regarding the potential increased capacity of the CBT-N dissemination efforts, fourteen providers (21.5%) reported training or supervising other providers and/or health care trainees in delivering CBT-N at their site since completing consultation. Those participants reported training an average of 2.6 (SD = 1.9) additional providers/trainees in CBT-N. Regarding the potential additional reach of the dissemination efforts on patient care outside of VHA, five of 12 providers reported delivering CBT-N in their private practices.

Discussion

The present study examined trained CBT-N providers’ continued use of CBT-N, perceived benefits, and implementation barriers. Nearly all respondents reported continued use of CBT-N, and all participants intended to use CBT-N with veterans in the future. This high continued use of CBT-N is promising for uptake. While most providers are delivering the treatment on a limited basis (ranging from 1 – 5 patients in the past six months), these findings are similar to other dissemination efforts of evidence-based psychotherapies (EBPs) within VHA (e.g., Ruzek et al., 2017). It is unknown how these rates compare to CBT-I training efforts in VHA, as there is no systematic tracking of the number of trained providers who continue to deliver CBT-I (Koffel et al., 2018). A recent paper by Pfeiffer and colleagues (2023), which examined the guideline-concordant use of CBT-I in the VHA, observed, based on standardized note templates, that only 46% of trained providers continued to actively offer CBT-I in FY2021, suggesting implementation barriers even with top-down EBP training initiatives. The evidence of continued use of CBT-N begins to tackle the previously observed underutilization of nightmare treatment (Nadorff et al., 2015; Youngren et al., 2019), yet, additional efforts to increase treatment uptake will likely be necessary.

Clinicians endorsed perceived benefits from CBT-N on their veterans’ well-being, including reduced nightmare frequency and/or severity, improved veteran perspectives on their nightmares, as well as improvement across other symptom domains. While this feedback shows promise for the benefits of CBT-N, it reflects clinicians’ perceptions rather than direct feedback from veterans. Identifying specific, measurable benefits directly from veterans could further shape clinician attitudes toward using CBT-N in practice.

Despite the perceived benefits of CBT-N, barriers to implementation exist for some but not all providers (~58%). These barriers primarily align with Inner Setting constraints, such as caseload management, compatibility of session length and frequency to clinic settings, and mismatches of priorities between CBT-N and specific veteran needs. Individual characteristic factors, such as veterans opting for other treatments or logistical barriers, also were reported. Many of the identified barriers are not isolated to CBT-N and may be relevant to other protocol-driven interventions (see Koffel et al., 2018; Koffel & Hagedorn, 2020). Although it may be clinically appropriate that veterans choose interventions other than CBT-N (e.g., trauma-focused evidence-based protocols for PTSD), this outcome highlights the need to examine why veterans may not choose CBT-N. Additionally, clinicians, who may not frequently work with trauma-related symptoms, may benefit from enhanced training on presenting the treatment rationale and addressing potential veteran avoidance to improve patient understanding and motivation. Pruiksma and colleagues (2025) provide a helpful discussion of the pros and cons of first treating PTSD versus sleep, and highlight the importance of case formulation and patient involvement in the decision process.

The Inner Setting barriers highlight a potential need for clinic setting changes to structure providers’ schedules to protect time for weekly, 60-min sessions for the provision of CBT-N. Group delivery of CBT-N, which has evidence to support (e.g., Balliett et al., 2015; Krakow et al., 2001; Long et al., 2011), may offer another opportunity to reach more veterans with fewer resources. Alternatively, modifications to the protocol may be necessary for clinical settings unable to accommodate this schedule change (e.g., primary care); however, research would be needed to explore the effectiveness of shorter or less frequent sessions. Other empirically supported treatments have been modified for the primary care setting (Bramoweth et al., 2020; Rauch et al., 2017). Modifications for this setting may increase access to this short-term treatment, in line with primary care mental health integration’s (PCMHI) foundational goal to provide timely evidence-based treatments to individuals, particularly those with lower distress and/or those who may be reluctant to engage in longer-term mental health treatment (Possemato et al., 2018). Together these results support the continued efforts to disseminate CBT-N to a variety of clinical teams (e.g., primary care, general mental health, and PTSD clinical teams), while ensuring treatment compatibility and relative priority for such settings.

While not directly assessed in the present survey, Outside Setting factors, such as national budgets, staffing allocations, and mandated performance metrics, likely contribute significantly to shaping the Inner Setting conditions that affect CBT-N implementation. Limits with budgets or staffing may constrain the protected time for training or delivery of CBT-N. Similarly, because CBT-N is currently not formally recognized in VHA performance metrics and lacks standardized documentation fields in the electronic health record (EHR), it may reduce leadership engagement and suggest low organizational priority. Unfortunately, that may in turn weaken providers’ motivation to deliver CBT-N. A focus on strategic alignment with leadership and enhancement of practical infrastructure, such as standardized note templates, could elevate CBT-N’s uptake. However, recent evidence by Bramoweth and colleagues (2025), which found that training alone significantly improved delivery of brief insomnia treatment across sites even without specific implementation strategies, reinforces that training investments and appropriate modifications to the protocol could achieve feasible and impactful implementation results rather than requiring large-scale systemic overhauls.

Finally, regarding the treatment complexity, streamlining the screening processes and documentation could reduce the challenges of implementing CBT-N. One strategy to do this may be to develop a CBT-N treatment companion mobile application (app), analogous to the CBT-I Coach app for CBT-I, which could include auto-calculation of data from the sleep and nightmare log and weekly symptom measures, and tools to reduce the burden of between session work for veterans (e.g., easily accessible relaxation and guided imagery exercises, reminders to complete exercises or logs, mobile access to the sleep and nightmare log). Another approach for additional provider support could be to provide specialty add-on workshop or consultation components for those providing treatment in telehealth or group formats.

Limitations

Several limitations of this study should be noted. First, while the response rate (65%) was acceptable (Burns et al., 2008), there may be a potential selection bias for those using CBT-N or perceiving benefit from the treatment. The use of CBT-N provided by clinicians who did not complete the survey, did not complete the training program requirements, or who received training outside of the training program is unknown. While the values presented here are likely a best-case estimate from the trained providers who adopted the treatment, it may not reflect the true total use or perceptions of CBT-N within the VHA system. A future initiative is to create a competency checkout procedure to incorporate CBT-N providers trained elsewhere. Relatedly, while similar themes emerged from responders who provided feedback to open-response opportunities, it is unknown whether these responses reflect the true thematic saturation that may have occurred with a 100% response rate. Another limitation is that findings of treatment benefit are based on provider self-report; thus, it is unknown how well providers’ impressions of CBT-N benefit reflect actual improvements in veteran outcomes. Future steps for the program evaluation will require standardizing documentation of CBT-N use in the EHR for additional tracking of reach, evaluating change in provider competency in delivery of CBT-N over the course of the training program, and identifying veterans’ symptom change via daily sleep and nightmare logs and self-report measures of PTSD, depression, and quality of life to evaluate effectiveness of CBT-N in the real-world VHA setting. Finally, this study also lacked perspectives of other key stakeholders, including veteran patients and hospital leadership and administrators. An exploration of these constructs from these perspectives will be essential to garner a comprehensive understanding of CBT-N implementation barriers.

Conclusion

The study outcomes suggest that the current grassroots model of CBT-N training has led to the successful continued delivery of CBT-N by VHA providers, with perceived positive outcomes, as well as areas for growth. Additional support of these efforts will allow broader implementation of CBT-N and the ability to address necessary modifications for greater reach across settings, while providing insight and guidance for implementation of EBPs across complex healthcare systems.

Impact Statement.

Grassroots dissemination efforts are underway to train Veterans Health Administration mental health providers in the delivery of cognitive behavioral therapy for nightmares (CBT-N). This study assessed providers’ continued use of CBT-N and the factors that facilitate or create barriers to implementation following training and consultation. Results indicated high uptake and positive perceptions of CBT-N for improving symptoms, but treatment complexity, inner setting, and individual characteristics may create barriers. The findings emphasize the need to make CBT-N adaptable to diverse clinical settings while monitoring patient outcomes.