Leveraging multi-stakeholder engagement to develop an implementation blueprint for a brief trauma-focused cognitive behavioral therapy in primary care

Abstract

Objective:

The implementation of evidence-based treatments (EBTs) to address posttraumatic stress disorder (PTSD) is a public health priority. Successful EBT implementation requires effective collaboration between multiple stakeholder groups, including hospital leaders, providers, and patients, to build buy-in for this effort.

Method:

We describe our implementation science approach to meaningful stakeholder engagement, as part of a hybrid type 1 effectiveness-implementation trial of Skills Training in Affective and Interpersonal Regulation for PTSD treatment in primary care (STAIR-PC) at a large safety net hospital. We used primary care and patient community advisory boards (CABs) to interpret key informant interviews and identify strategies to adapt the intervention to ensure fit with the primary care setting. We documented our stakeholder engagement methodology through comprehensive field notes and minutes from CAB meetings, detailing the focus of meetings, suggestions for intervention and delivery adaptations, decision-making processes, and how disagreements about adaptations between stakeholders were resolved. To support replicability, we specify and operationalize implementation strategies to be used across each implementation phase of the trial.

Results:

Key strategies involved a) ensuring that research questions are relevant to both patients and clinical providers; b) tailoring interventions that are flexible and adaptable to the needs of the local setting; c) continuous engagement of patients and providers throughout the implementation process; and d) building mutual respect, trust, and credibility between the research team, various provider groups, and patients.

Conclusions:

Our approach to engaging stakeholders informed an implementation blueprint to guide implementation of EBTs for PTSD in safety net hospital primary care clinics.

Posttraumatic stress disorder (PTSD) is a significant public health challenge that disproportionately affects individuals receiving care in safety net hospitals, defined as facilities in which care is provided to a significant number of patients who are uninsured, underinsured, or insured through Medicaid (Sutton et al., 2014). In these settings, lifetime exposure to interpersonal violence in primary care and behavioral health clinics may be as high as 94% with current diagnosis rates of PTSD ranging from 33% to 46% (Alim et al., 2006; Roberts et al., 2011; Gillespie et al., 2009). Despite high instances of trauma and PTSD in these settings, PTSD is vastly under-recognized and undertreated, with only 13% of individuals with PTSD receiving any treatment (Alim et al., 2006). Only 33% of those with PTSD who access treatment receive a minimally adequate dose of therapy (4 hours; Wang et al., 2005). These unmet treatment needs drive large racial, ethnic, and income disparities in mental health (Dinwiddie et al., 2013; Cook et al., 2012; Alegría et al., 2007).

Poor access to evidence-based treatments (EBTs) for PTSD in safety-net settings is foremost due to a shortage of trained mental health specialists (Dinwiddie et al., 2013; Blanco et al., 2007; Lê Cook et al., 2013; Garland et al, 2005). Providers working in these settings often have inadequate time or financial support for specialized training in EBTs for PTSD (Foa et al., 2013) and may have misperceptions that EBTs are inappropriate or even harmful to their clients (Frueh et al., 2006), thereby rendering EBTs largely unavailable. Racial and ethnic minority clients themselves often share these providers’ skepticism and negative beliefs about manualized treatments (Williams, Domanico, et al., 2012; Williams, Malcoun et al., 2014; Abdullah & Brown, 2011).

The overall shortage of trained specialists, lack of racially and ethnically diverse specialists, and perceived low quality of care contributes to inefficient engagement in care (Lê Cook et al., 2014; Roberts et al., 2011). Underutilization of outpatient specialty care, based on systemic, psychological, and economic factors, has led to concerning inequities (Alegría et al., 2018; Institute of Medicine, 2003). Improving quality of outpatient care by increasing the availability of EBTs in safety-net settings is necessary to address PTSD disparities. This includes expanding availability of EBTs to integrated behavioral health settings where treatment may be less stigmatized and may be more acceptable to patients (Kaltman et al., 2014; Fox et al., 2012).

There is a strong public health case for system reorganization of safety-net hospitals to address unmet PTSD treatment needs. However, this effort requires effective collaboration between multiple stakeholder groups, including hospital leadership, providers, patients, and implementation scientists. Stakeholder engagement can help researchers understand complex systems, differences in organization, provider, and patient culture, and areas of need or strength (Craig et al., 2018; Ekirapa-Kiracho et al., 2017; Mallery et al., 2012). This contextual information is critical to understanding the need to tailor EBTs and plan for implementation.

Early, multiphasic engagement with stakeholders may promote community investment in the success of the intervention and lead to improved effectiveness and implementation outcomes (Boaz et al., 2018; Concannon et al., 2014; Israel et al., 2001). However, in practice, researchers often involve stakeholders minimally or in name only (Bowen et al., 2017; Israel et al., 1998). This is problematic given that researchers who fail to foster healthy partnerships may erode trust in the research process and EBTs, and may even reinforce mistrust based on historical exploitation of vulnerable groups by researchers and the medical system (Durant et al., 2011; Scharff et al., 2010). We argue that, despite the significant time and effort required to engineer meaningful stakeholder engagement (Pandi-Perumal et al., 2019), this work is essential to the conduct of ethical clinical research that produces mutual benefit.

Truly meaningful researcher-stakeholder partnerships are built on central tenets of collaboration, reciprocity, and trust. Ethical and effective collaboration relies on combining the naturally diverse skillsets, knowledge bases, experiences, and expertise of various stakeholders to address the complexities of conducting interventions in a particular setting (Israel et al., 2001; Minkler & Wallerstein, 2008; Schensul, 1985). Collaborating across all stages of research, including planning, data collection, and dissemination of results, allows for provisions to be built at each step that mutually benefit both researchers and stakeholders in the process of reciprocity (Maiter et al., 2008). Stakeholders in reciprocal relationships will be more committed to project goals and more disposed to integrate research findings into their own lives and practices (Hamilton et al., 2017). However, given personal and collective histories of maltreatment in medical research, community members may doubt their ability to form truly equal partnerships with researchers or providers (George et al., 2014; Davis et al., 2012; Wilkins, 2018). Researchers must serve as a bridge between levels of hierarchy in medical systems by relaying ideas that place equal value on various perspectives (Israel et al., 2001, 2003; Lord et al., 1998).

Studies that include comprehensive stakeholder engagement strategies often do not report or publish on their process of involving stakeholders in implementation efforts (Bowen et al., 2017). An important exception is a recent paper that described a five-step researcher-stakeholder engagement process driven by an implementation science framework, which consisted of researchers a) collaborating with health system leaders to plan a shared research agenda; b) engaging with leaders to discuss early study results, receive feedback, and establish buy-in for continued research efforts; c) using evidence to create an implementation blueprint to spread the evidence-based practice to additional healthcare facilities; d) evaluating the implementation toolkit in real-time clinical practice; and e) continuing to consult with the health system leaders and clinical providers throughout the implementation process (Elwy et al., in press). The Veteran’s Administration (VA) has also published on “partnered research” approaches to improving care (AuYoung et al., 2017; Yano, 2015; Zulman et al., 2014). Their approach relies heavily on the structure and culture of the VA, which may limit generalizability to other settings. Though useful guides, stakeholder engagement in these studies focus mainly on organizational culture and engaging healthcare system leadership and providers as key stakeholders, rather than directly addressing patient barriers to care such as mistrust, stigma, and mental health literacy—as well as trauma-specific patient barriers such as avoidance of trauma memory, reminders, and negative affect. More research is needed on the process of engaging system/organization, provider, and patient stakeholders to support the optimization of EBTs for PTSD, as well as selecting strategies for implementation in safety-net settings. This is especially pertinent for engaging patients who have a history of mistrust of the healthcare system (Norris et al., 2017).

The current study aims to provide a detailed and replicable approach to meaningful stakeholder engagement to support the development and execution of an implementation blueprint for an EBT—Skills Training in Affective and Interpersonal Regulation (STAIR) for PTSD, adapted for primary care (STAIR-PC). The term implementation blueprint refers to a plan for implementation that includes goals and strategies, scope and timescale of change, and progress measures (Lewis et al., 2018). The implementation blueprint includes aspects of intervention adaptation and system redesign, as well as strategies to support the research process (e.g., support with recruitment and retention of providers and patient participants). This report is part of a larger hybrid type I effectiveness-implementation trial (Curran et al., 2012), wherein we will test the effectiveness of STAIR-PC in a safety net primary care setting, while gathering data on implementation. The study phases follow the Replicating Effective Programs (REP) framework (Kilbourne et al., 2007), which was developed to facilitate the translation of academic or research-based programs into community practice settings by offering an implementation protocol. This paper describes stakeholder engagement—which includes key informants (interview participants) and community advisory board (CAB) members—across REP phases to gather information on the local setting and patient population, contextualize findings pertaining to implementation barriers and facilitators, and specify an implementation blueprint.

Method

Local Setting

This study took place at a large academic medical center, including the largest safety net hospital in New England. Primary care practices spanning two departments (General Internal Medicine and Family Medicine) serve approximately 40,000 and 10,000 patients, respectively. Approximately one third of patients do not speak English as their primary language and more than half represent underserved populations, such as the low-income and elderly, who rely on government payers for their coverage. An internal community health needs assessment conducted in 2019 revealed that many patients are experiencing issues related to housing, food insecurity, transportation, healthcare access and utilization, chronic disease management, mental health or substance use, and exposure to violence. The majority of primary care patients are insured through Medicaid (70%). Over half (56%) of Medicaid-insured patients at the hospital have a behavioral health diagnosis (with 40% having a mood or anxiety disorder). In response to such high behavioral health needs, both primary care practices have adopted an integrated behavioral health (IBH) model which relies on collaboration, coordination, and colocation of primary care physicians, psychiatrists, psychologists, nurse practitioners, and social workers to enhance overall patient health. A population-based, stepped care approach is employed in both primary care practices, whereby patients are initially offered a low-intensity intervention, symptoms are monitored, and the level of care is adjusted depending on progress. Both practices have routine screening for depression and substance use every 6 months to assist primary care providers in identifying behavioral health needs. Patients are referred to the IBH team via a warm handoff (same-day referral during the medical visit) or through a referral order in the Electronic Medical Record (EMR). The referral volume is relative to the size of the practices, with the IBH team in General Internal Medicine receiving approximately 400 referrals per month and the IBH team in Family Medicine receiving approximately 80 referrals per month. The IBH teams include 6.0 FTE of social work and psychologist time and 1.0 FTE of Psychiatrist time in General Internal Medicine, and 2.0 FTE of social work and 0.5 FTE of Psychiatrist time in Family Medicine. The disparity between the sizes of these practices and the available IBH clinical staff underscores the need for low-intensity, time-limited interventions and the availability of easily accessible higher levels of care when needed.

Selected Intervention

Due to its brevity and low-intensity, Skills Training in Affective and Interpersonal Regulation adapted for Primary Care (STAIR-PC) was selected in consultation with IBH leadership, who specified that appropriate interventions for their stepped care model needed to be brief (5 sessions or less) and low-intensity, while also having demonstrated effectiveness in primary care and/or safety net settings. (Jain et al., 2020; Smith et al. 2020). IBH leadership was also presented with well-established first line EBTs for PTSD including Prolonged Exposure, Cognitive Processing Therapy, and the original version of STAIR. STAIR is an empirically supported cognitive behavioral therapy for PTSD that has demonstrated effectiveness in a variety of contexts, including safety-net settings (Cloitre, Koenen et al., 2002; Cloitre, Stovall-McClough, et al., 2010; Trappler & Newville, 2007). Jain and colleagues (2020) adapted STAIR for primary care (STAIR-PC), condensing it to five sessions that teach coping skills to address emotion regulation and interpersonal effectiveness difficulties common in trauma survivors without necessitating recounting of the trauma memory. STAIR-PC was further adapted and pilot tested by our team for a similar patient population (Smith et al., 2020). See Ametaj and colleagues (2021) for full detail on adaptation to the local patient population (note: our version of manual is now used by the treatment developer in national dissemination efforts and does not deviate in core components). The hybrid type I trial aims to establish effectiveness of a further adapted STAIR-PC while also gathering data on implementation.

REP Framework

The REP framework is a multiphasic approach for implementing EBTs in real-world settings. The approach specifies steps needed to maximize fidelity to the EBT core components while allowing for local tailoring and adaptations to treatment delivery in order to maximize intervention uptake and sustainability (Kilbourne et al., 2007). Full details on stakeholder engagement and main research activities by REP phase can be found in Table 1.

Table 1.

Stakeholder engagement and research activities by REP Phase

	Pre-Conditions (Oct 2018 - Feb 2019)	Pre-Implementation-A (Mar 2019 – Dec 2019)	Pre-Implementation-B (Jan 2020- Present)	Implementation (Next Steps)	Maintenance and Evolution (Next Steps)
Stakeholder Engagement	Key Informants			Key Informants
		Primary Care CAB
			Patient CAB
Primary Research Activities	Identify barriers to implementation (organizational needs assessment, usual care; i.e. integrated primary care context) Distinguish core elements, menu options Assemble Primary Care Community Advisory Board (CAB)	Refine package, core elements, and menu options (refined based on CAB input, adjudicated by intervention developers) Refine training, technical assistance, approach strategies per CAB input Select expert recommended implementation strategies Assemble Patient CAB	Finalize package based on Patient CAB input, prepare for trial start Operationalize implementation strategies and finalize blueprint Pilot implementation package (intervention and implementation strategies) in primary care (CURRENT)	Ongoing training Technical assistance Full-scale evaluation	Organizational, financial changes National dissemination Re-customize delivery as needed

Note. REP = Replicating Effective Programs (Kilbourne et al., 2007)

We have organized our report based on three REP phases: Pre-Conditions, Pre-Implementation (Part A), and Pre-Implementation (Part B). The Pre-Conditions phase focuses on understanding the local context and patient population to identify the need for a tailored intervention and implementation strategies. In Pre-Conditions (Oct. 2018 – Feb. 2019), we administered surveys and conducted semi-structured interviews with key informants in the primary care setting to understand key barriers to implementation (N=22; primary care physicians=6, primary care leaders and/or administrators=5, IBH therapists=8; community health workers=3). Content analysis of key informant interviews will be published in a forthcoming manuscript. Interview findings support that training gaps as well as system and workflow challenges were substantial implementation barriers, and thus should be the initial focus of implementation blueprint development. The next phase, Pre-Implementation, involved the use of two Community Advisory Boards (CABs) to develop an implementation blueprint and further augment the intervention for the setting (see Stakeholders, below). In Pre-Implementation Part A (Mar. 2019 – Dec. 2019), we relied on the CABs to interpret, contextualize, and address barriers to implementation identified in the key informant interviews. In Pre-Implementation Part B (Jan. 2020 – Mar. 2020), the CABs assisted in the selection and execution of expert-recommended implementation strategies specified by Powell and colleagues (2015) such as changing record systems and developing educational materials. Pre-Implementation Part B also includes the pilot trial, currently underway. The final two REP stages, Implementation and Maintenance and Evolution, will be the focus of future collaborations that will fully integrate patient/trial participant perspectives to support a large-scale evaluation and organizational changes (e.g., adoption of standardized approaches screening and stepped care for PTSD; resource allocation).

Stakeholders

The research team assembled two Community Advisory Boards (CABs), one comprised of representatives from primary care (referred to as the PC CAB) and one comprised of patients (Patient CAB). CAB members represented a wide range of perspectives, and the focus of each CAB differed slightly. The PC CAB concentrated on specifying an implementation blueprint for organizational, provider, and patient fit. The Patient CAB focused on ensuring that the proposed implementation blueprint developed by the PC CAB and the selected intervention fit the needs of patients. These CABs were held separately due to power differentials, ensuring that patients could speak freely and critically about their care while knowing that their perceptions would be relayed anonymously to providers.

The PC CAB (N=9 serving at a time) included professionals representing medicine, behavioral health, and administration, all of whom were selected due to their reputation as “champions” (those with highly relevant expertise) and/or “gatekeepers” (those who have pertinent decision-making authority in the practice). The PC CAB included 13 representatives over time (due to turnover; 2 left the organization; 2 transitioned out of leadership role), and included psychologists, social workers, psychiatrists, primary care physicians and nurse practitioners, and operations managers. These members also reflected integrated behavioral health leadership, primary care clinician champions for trauma-specific treatment, and specialty behavioral health leadership. Representatives for the Patient CAB (N=6) were nominated by PC CAB representatives based on these patients’ familiarity with various mental health services at the hospital and their own histories of PTSD. PC CAB representatives gained verbal permission from patients prior to researcher contact. Researchers organized feedback and field notes into themes to synthesize and relay feedback between groups, ensuring equal power across patients, providers, and organizational leadership. This strategy allowed for engaged and collaborative decision-making during CAB meetings.

Overview of the PC CAB and Patient CAB Meeting Process

PC CAB meetings were structured to make efficient use of providers’ time and to allow adequate space to gather input and facilitate group discussion. At each meeting, we presented an agenda with an overview of specific requests for information and objectives of the meeting. When needed, PC CAB members added items to the agenda either prior to or at the beginning of the meeting. This allowed us to be responsive to both research and PC CAB needs and objectives. The intervention materials and the implementation blueprint were repeatedly presented for comment and review, with iterative revisions being made after each meeting. Once the PC CAB was satisfied with the intervention package and implementation blueprint, we held a meeting to collaboratively develop questions to ask the Patient CAB. The Patient CAB was presented with the planned intervention and implementation blueprint and asked to provide feedback. We also asked the Patient CAB a series of questions regarding the need for additional modification to the intervention or implementation blueprint, and then opened the floor for general commentary on personal experiences of PTSD treatment and the current state of PTSD care at the hospital.

Key informants and Patient CAB members were paid $20/hour for their time and contribution to the research, though PC CAB members were incentivized solely through personal desire to improve PTSD treatment access in the clinic. This study received an exempt determination from the Institutional Review Board.

Findings

Key Tasks and Decisions: Evolution of the implementation blueprint over time

Full detail of the focus and key tasks and decisions by meeting are presented in Table 2. We have provided the researcher-set agenda items in the “focus” section. The “key tasks and decisions” section reflects CAB members’ responses to “focus” items as well as additional items and priorities introduced by the CAB members. In the first PC CAB, key informant interview findings from the Pre-Conditions phase were presented with the aim of addressing systems and operations, provider, and patient barriers to implementation. This first meeting helped to confirm that a) the research question was relevant to providers and patients, and b) that the selected intervention would need to be to be flexible and adaptable to accommodate the unique needs of the setting. The second PC CAB meeting focused on options for addressing the workforce shortage, training gaps, and technological supports for screening, documentation, care coordination, and delivery. The third PC CAB meeting solidified plans for training, consultation, recruitment, and cooperation with the Patient CAB. The Patient CAB highlighted the additional importance of reducing discrimination and stigma in the patient experiences, and provided feedback on issues such as the misdiagnosis of PTSD resulting from biased screening practices, the lack of access to non-pharmaceutical treatment, and a perceived lower quality of care for racial minorities. The Patient CAB also noted enthusiasm about expanded access to behavioral treatments for PTSD in primary care and appreciated the ability to influence their care options. Researchers synthesized Patient CAB feedback (i.e., themes identified through review of comprehensive meeting minutes and field notes) and presented it to the PC CAB prior to jointly finalizing the intervention and implementation blueprint. Patient CAB feedback suggested the importance of cross-provider training in screening, assessment, treatment options, and referral processes for PTSD treatment. Additionally, patient CAB feedback highlighted the need to review and adjust current practices and promote organizational change to address health equity concerns. For pragmatic reasons, the timescale of organizational changes was beyond the timeline for this project. Items raised by the Patient CAB that have not been addressed in the immediate are the explicit focus of ongoing and future efforts to provide higher quality and equitable care for PTSD. We have integrated these concepts into provider trainings.

Table 2.

Evolution of the intervention and implementation blueprint over time

	Focus	Key Tasks and Decisions
Key Informant Interviews (Oct 2018 to Feb 2019)	Assess potential barriers and facilitators to implementation Assess need to tailor to local context	Mixed-methods formative evaluation (published elsewhere) Prepare summary and proposed plan to present to CAB
PC CAB 1 (Mar 2019)	PTSD screening procedure Revise scheduling blocks to accommodate therapy visits Care coordination between primary care physician and IBH clinicians Fidelity monitoring procedures Training needs Telehealth option*	No universal screen, but screening for PTSD once referred to IBH clinician Revise scheduling templates Creation of SmartText notes to assist with documentation and coordination Audio and chart review for fidelity monitoring Tailored trainings by provider type to facilitate identification of patients, referral, treatment, and engagement
PC CAB 2 (May 2019)	Plan for selection of interventionists, and addressing potential for turnover Expand on plan for cross-training in PTSD diagnosis and treatment Standardize communication process across the practice Reduce clinical jargon in manual Telehealth option*	Explore the use of CHWs (in addition to IBH clinicians) as potential interventionists* Build tools into medical record, including assessments and treatment documentation Revise treatment manual for preferred terminology
PC CAB 3 (Nov 2019)	Finalize manual Plan for training and consultation Plan for study therapist in-person recruitment Identify key questions that should be posed to the upcoming Patient CAB	Cross-train IBH clinicians (primary interventionists) and CHWs (supportive role) in STAIR-PC Twice monthly virtual group consultation, with additional “office hours” Develop research protocols that do not add to existing space limitations in the clinic (in-person informed consent; remote survey assessments) Approved streamlined intervention [30 minute block; allow flexible scheduling, specified goal of treatment completion by week 12]
Patient CAB 4 (Feb 2020)	Review plan to increase therapy capacity for PTSD Review screening procedure and referral process Gather feedback on the implementation blueprint and the intervention manual	Advocated for universal screening for PTSD, since current practices were perceived as discriminatory (race-based)* Enthusiastic about the potential for increased access to therapy, as their past experiences have been medication-focused Advocated that initial engagement should focus on mental health stigma reduction Prepared summary to present to PC CAB Advocated for use of people from the community with lived experiences as interventionists (e.g., CHWs)*
PC CAB 5 (Feb 2020)	Finalize training and consultation plan Implement scheduling blocks to permit capacity for therapy (add more NEW slots) Prepare for trial start (review final list of interventionists, screening and referral procedure, research protocol, and manual materials) Agreed on the importance of reducing mental health stigma as integral to engagement	Schedule various presentations and training (e.g., Grand Rounds) Practice changes implemented Study protocol and manual finalized Schedule training and consultation meetings

Note. PTSD=posttraumatic stress disorder; CAB=community advisory board; PC=primary care representatives; IBH=integrated behavioral health; CHW=community health workers. The “focus” section reflects researcher-set agenda items. “Key tasks and decisions” reflect CAB members’ responses to the “focus” items and additional items and priorities introduced by the CAB members.

^*not addressed at pre-implementation stage of the trial based on CAB-determined priorities

Of note, the need for intervention augmentation did not emerge as central importance during key informant interviews or over the course of CAB engagement. We think this is attributable to our previous work in adapting the intervention locally (Smith et al., 2020; Ametaj et al., 2021), and overall positive attitudes about the fit of this EBT for their setting and population. We did not revise core components of the intervention, but did work diligently to streamline the administration of the protocol to fit the 30-minute therapy block permitted in IBH. We imagine that there may be a need for further modification to the intervention that will be informed more directly by the pilot study (i.e., patient and provider perspectives of actually receiving or delivering this treatment).

Selected Implementation Strategies for the Implementation Blueprint

Table 3 provides detail on implementation strategies by REP phase, selected from a compilation of expert-recommended strategies detailed by Powell and colleagues (2015; “The Expert Recommendations for Implementing Change [ERIC] Project”). In our initial study design, the research team selected three overarching implementation strategies that would span all REP phases: the use of advisory boards and work groups, identifying and preparing champions, and obtaining and using patients’ feedback. The research team also selected assessing the setting for readiness, identifying organizational barriers and facilitators, and conducting consensus discussions as strategies for the Pre-Conditions stage (e.g., key informant interviews). During the Pre-Implementation stage, we mapped feedback and suggestions from primary care and patient CABs onto strategies outlined in the ERIC Project. In Table 3, we specify how we operationalized each of these strategies for our own implementation blueprint; for example, formalizing the education and training processes of IBH providers, ensuring technological capability of each clinic/practice, and promoting adaptability and sustainability of the STAIR-PC intervention.

Table 3.

Implementation blueprint: Selected implementation strategies by REP Phase

REP Phase	Strategies	Description of how this strategy was operationalized in our implementation blueprint
All Phases	Use advisory boards and workgroups Identify and prepare champions Obtain and use patients/consumers and family feedback	We assembled a Primary Care community advisory board (PC CAB) consisting of various stakeholders in primary care and behavioral health leadership (N=9). Members of the PC CAB also serve as local champions of the intervention, and this group directly informs all phases of the study, including intervention selection and adaptation, development of implementation blueprint, organizational readiness for implementation, and oversight of the trial. We also assembled a Patient CAB comprised of patients with PTSD who had good knowledge of the local context (N=6). The group was presented with the proposed PTSD treatment and our implementation blueprint; then feedback on various components were elicited. Feedback from the Patient CAB was relayed back to the PC CAB to make final adjustments prior to pilot trial. We will convene the Patient CAB periodically during the pilot trial, to assist with monitoring and need for further modification to intervention or implementation plan.
Pre-Conditions	Assess for readiness and identify barriers and facilitators Conduct local consensus discussions	We completed a mixed-methods contextual analysis to understand barriers and facilitators to implementation, guided by the CFIR, which included surveys and interviews with key stakeholders in primary care and integrated behavioral health specialists (N=22). We met with primary care leadership to select the PTSD intervention prior to grant submission. Initial contextual analysis with key stakeholders in primary care and integrated behavioral health specialists (N=22) included assessment of the importance of treating PTSD in primary care, and whether the selected treatment was appropriate for the local setting and patient population.
Pre-Implementation (A)	Develop a formal implementation blueprint Promote adaptability Change record systems	We have worked with our PC CAB to (1) set goals for recruitment (enroll 1 new patient per week); (2) set goals for PTSD symptom reduction (linked to 10 point changes in PCL-5) and functional improvement (e.g., social functioning, quality of life); and (3) assess appropriateness of the intervention (need for further refinement of intervention and implementation plan) once 30 patients have enrolled in the treatment. Several modifications to the intervention were made based on feedback from the contextual analysis, including changes to design and packaging, condensing content to fit to 30-minute therapy appointments, and increasing flexibility of delivery (deeming some components as optional; allowing for less than weekly sessions) We ensured that our PTSD screener (PC-PTSD-5) was accurately embedded in the electronic medical record. We developed SmartText to allow for consistent documentation of engagement in the therapy. We started the process of adding a full symptom PTSD scale (PCL-5) to the electronic medical recorded for ongoing symptom monitoring.
Pre-Implementation (B)	Develop educational materials Conduct educational meetings Distribute educational materials Conduct ongoing training Provide clinical supervision Organize clinician implementation team meetings Remind clinicians	We prepared tailored education materials for various audiences within primary care, including primary care “all staff,” physicians and resident physicians, behavioral health specialists, and community health workers. Materials range from education on screening, assessment, and referral for PTSD treatment, overview of PTSD treatments, how to support engagement in PTSD treatment, and basic tenets of trauma-informed care. Training materials will be delivered during various ongoing staff meetings, and video-recorded materials will be available on shared drive, and email will be used to distribute information on how to locate educational materials. We developed a plan for therapist training, which includes initial 1-day didactic training and twice monthly group consultation. The initial didactic training will be video recorded, so new-hires will be able to complete the training during onboarding process, then join in on ongoing consultation for further training. Consultation will involve near real-time audio review of 20% of therapy sessions to assist with fidelity (adherence and competence). We will also offer additional consultation and technical assistance “office hour” style, with two hours of clinician trainer’s time blocked each week for as needed consultation. We will repeat tailored educational meetings for each new cohort of physician residents. Study research assistant (RA) will provide weekly reminders to clinicians regarding appointments with any patients currently enrolled in the study. Reminders will also include links to session materials on the shared drive. The study RA will also check in with clinicians during twice monthly consultation meetings to identify if there were any “missed” screenings for PTSD symptoms that the RA could complete via telephone.

Note. PTSD=posttraumatic stress disorder. Strategies selected from the Expert Recommendations for Implementing Change Project (ERIC: Powell et al., 2015)

Discussion

Implementation of EBTs in healthcare settings is often complex because diverse individuals from across an organization and among different managerial levels must be involved in all aspects of the process (Fisher et al., 2016). Key strategies identified for successfully engaging stakeholders in the development of an implementation blueprint to guide implementation of STAIR-PC involved a) understanding different time horizons for addressing important clinical problems between the research team and clinicians; b) ensuring that research questions are relevant to both patients and clinical providers; c) designing or tailoring interventions that are flexible and adaptable, taking into account the needs of the local setting; d) continuous engagement of patients and providers throughout the implementation process, as guided by the REP framework; and e) building relationships of mutual respect, trust and credibility between the research team, various provider groups, and patients.

Additionally, it is important to consider that healthcare organizations are often not centered around the implementation of EBTs for behavioral health, and that these interventions may need to be adapted to fit organizational context and need. There is also a case to be made that healthcare systems could benefit from reorganization around behavioral health concerns, especially in safety net hospitals where behavioral health needs are high. Even in hospitals that have embraced IBH models, the implementation of PTSD treatment (versus depression care management) can pose additional challenges, given that first line treatments for PTSD often place high burden on patients, therapist time, and systems. Thus, intervention adaptation and implementation strategies should focus centrally at reducing time and resource burdens associated with these treatments.

Knowing which implementation strategies to apply, when to apply them, and at what level requires careful coordination across managerial levels and often negotiation with various stakeholders. In this project, some implementation strategies discussed with the CABs were tabled, but are the focus of ongoing and future stakeholder-engaged efforts. For example, we discussed the expansion of which providers in primary care may deliver the intervention, but then deemed it secondary to the current study scope. For our planned STAIR-PC pilot trial, we will focus on IBH therapists who currently provide the majority of therapy, yet there is interest based on the Patient CAB feedback and among some providers to have community health workers involved in intervention delivery. Additionally, IBH has had great success with telehealth delivery of depression care management (and a recent shift to telehealth in light of Covid-19 has made this a more feasible option). There may be opportunities to similarly expand treatment delivery options for trauma-focused interventions in the future. Finally, pilot trial data and Patient CAB feedback may support the need for universal screening for PTSD in primary care—a strategy that was not feasible given organizational priorities around depression care management at the time of this work.

The CABs have been essential to ensuring that the intervention is feasible and acceptable and, ultimately, effective in the local setting. Ongoing CAB meetings are planned to support the pilot trial, then full integration of STAIR-PC into usual care in IBH. Stating clear goals for the CABs at the outset and at the start of each meeting allowed for focused work in addressing major barriers to implementation. This approach was respectful of the CAB members’ limited time and helped to minimize CAB member work between meetings. CAB members showed appreciation for the structure and transparency of the meetings (i.e., the research team presented exactly how stakeholder feedback was incorporated into revisions). This made participation meaningful and reciprocal, and built rapport between researchers, providers, and patients.

Over the course of this project, the reciprocity of the research-practice partnership was carefully fostered. For example, the principal investigator provided didactic trainings and attended IBH meetings to form rapport between the researchers and interventionists prior to implementation. This initial relationship building helped to smooth over initial worries expressed by some providers around fidelity monitoring via review of audio-recorded therapy sessions, and concerns about researcher investment in the long-term sustainability of new interventions. Second, interventionists will become another set of stakeholders to add to the project, and will provide direct feedback on further refinement over the course of the pilot study. Conducting this research in an academic medical center means that the interventionists will be able to extend their new knowledge by cross-training the trainees they supervise. Third, by engaging clinical and administrative leadership in the CAB, the importance of addressing PTSD in primary care was highlighted while also acknowledging the competing demands and priorities that primary care practices face. The final implementation blueprint includes explicit ways that researchers will support various providers, ranging from technical support on the intervention and general training on PTSD diagnosis and treatment, to assistance with screening and documentation. We hope that, through our reciprocal and mutually beneficial approach to the partnership, we can improve screening and detection of PTSD in primary care and increase access to PTSD treatment while minimizing therapist and system burden. Opportunities for interventionists to learn more about the research process and contribute to publications may also increase engagement and participation in future research studies.

Next steps in this work include the pilot testing and subsequent integration of STAIR-PC into usual care at the hospital. Together, we will be submitting for funding to support the next phases of REP, which includes full implementation, maintenance, and evolution of the intervention.

Conclusion

This paper describes our use of provider and patient CABs to respond to anticipated barriers to the implementation of an evidence-based treatment for PTSD in primary care. This researcher-provider-patient collaboration yielded an adapted intervention that meets the needs of patients and providers and a formal implementation blueprint to guide the integration of the intervention into usual care at the hospital. This partnership, three years in the making, will continue to support hospital-wide intervention rollout, uptake, and sustainability through planned meetings over the next three years (and likely beyond). The partnership has provided a strong foundation for efforts to build capacity for delivering evidence-based treatment across a large primary care practice in a low-resource safety net hospital. Our next steps are to assess if these pre-implementation efforts will lead to successful implementation and sustainability of an evidence-based treatment for PTSD in primary care, and to work diligently to better understand and address issues regarding health equity raised by the Patient CAB that were not addressable in the timescale of this initial work.

Clinical Impact Statement.

Organizing healthcare systems around posttraumatic stress disorder (PTSD) treatment is important, especial in safety net hospital settings. System alignment—to support delivery of evidence-based treatments (EBTs) for PTSD—requires effective collaboration between multiple stakeholder groups, including hospital leadership, providers, and patients. The current study describes the use of a multi-phase stakeholder engagement process to tailor and develop an implementation blueprint for an EBT for PTSD in safety net hospital primary care clinics. Our approach may be helpful to others aiming to adopt a stakeholder engagement strategy to increase PTSD treatment availability in real-world clinic settings.