Understanding barriers and facilitators to implementation of psychosocial care within orthopedic trauma centers: a qualitative study with multidisciplinary stakeholders from geographically diverse settings

Background Psychosocial factors are pivotal in recovery after acute orthopedic traumatic injuries. Addressing psychosocial factors is an important opportunity for preventing persistent pain and disability. We aim to identify barriers and facilitators to the implementation of psychosocial care within outpatient orthopedic trauma settings using the Consolidated Framework for Implementation Research (CFIR) and Proctor’s taxonomy of implementation outcomes, and to provide implementation strategies derived from qualitative data and supplemented by the Expert Recommendations for Implementing Change. Methods We conducted live video qualitative focus groups, exit interviews and individual interviews with stakeholders within 3 geographically diverse level 1 trauma settings (N = 79; 20 attendings, 28 residents, 10 nurses, 13 medical assistants, 5 physical therapists/social workers, and 3 fellows) at 3 trauma centers in Texas, Kentucky, and Massachusetts. We used directed and conventional content analyses to derive information on barriers, facilitators, and implementation strategies within 26 CFIR constructs nested within 3 relevant Proctor outcomes of acceptability, appropriateness, and feasibility. Results Stakeholders noted that implementing psychosocial care within their practice can be acceptable, appropriate, and feasible. Many perceived integrated psychosocial care as crucial for preventing persistent pain and reducing provider burden, noting they lack the time and specialized training to address patients’ psychosocial needs. Providers suggested strategies for integrating psychosocial care within orthopedic settings, including obtaining buy-in from leadership, providing concise and data-driven education to providers, bypassing stigma, and flexibly adapting to fast-paced clinics. Conclusions Results provide a blueprint for successful implementation of psychosocial care in orthopedic trauma settings, with important implications for prevention of persistent pain and disability. Supplementary Information The online version contains supplementary material available at 10.1186/s43058-021-00208-8.


Contributions to the literature
In this multi-site qualitative study of orthopedic trauma providers, we identified key strategies to facilitate implementation of psychosocial care into orthopedic trauma practices.
We integrate the Consolidated Framework for Implementation Research with Proctor's implementation outcomes as a novel approach to comprehensively characterize the barriers and facilitators among implementation determinants that could impact implementation outcomes.
These data serve as a blueprint for maximizing successful implementation of psychosocial care and aligning orthopedic trauma practices with evidence-based biopsychosocial models of care.

Background
Musculoskeletal traumatic injuries are a major public health problem [1]. The impact of traumatic injuries extends beyond immediate physical health, as approximately 20-50% of patients go on to develop persistent (chronic) pain and disability [2,3], disproportionate to residual pathophysiology. Patients with greater pain and more functional limitations are likely to pursue additional surgeries and medical procedures with questionable potential for benefit, resulting in increased health care costs and a significant public health burden [4,5].
Recovery after a traumatic injury is a complex process that extends beyond the severity of the physical injury itself. The biopsychosocial model [6] recognizes that biological, social, and psychological factors are interrelated and contribute together to the recovery process and long-term outcomes. Mounting evidence shows that misconceptions and distress (e.g., catastrophic thinking, fear of movement, depression, and posttraumatic stress) are important modifiable risk factors for persistent pain and functional limitations after traumatic musculoskeletal injuries, regardless of the injury severity [3,7,8], location [9,10], and type [11,12]. Recognizing these modifiable risk factors early creates an opportunity to intervene with patients who are at risk for persistent pain and disability in the acute post-injury phase, when psychosocial treatments are most effective [13,14].
Despite the strong evidence for the role of psychosocial factors in recovery after traumatic injury, these factors are untreated or undertreated in most patients [15]. The Lower Extremity Assessment Project, a large prospective study of patients with orthopedic trauma, showed that while 50% endorsed psychological distress 3 months post-injury and 42% 2 years later, only 12% had received any mental health care early post-injury with numbers increasing only to 22% by the 2 year mark [15]. In 2019, the American Association of Orthopedic Surgery in partnership with the Major Extremity Trauma Research Consortium [16] developed clinical practice guidelines that recommended accounting for psychosocial factors when caring for people with traumatic injuries. Further, existing evidence supports the costeffectiveness of the integration of psychosocial care within orthopedic trauma care at both societal and organizational levels [17][18][19]. However, implementation of these guidelines into orthopedic trauma settings has been severely limited [18].
Multiple factors known to prevent the successful implementation of evidence-based clinical practice guidelines in medical practices have been documented, including providers' resistance, negative attitudes and lack of knowledge, skills, and organizational management support and resources [20]. Within the general surgical field, prior research has shown opposition to innovation from surgeons [21,22]. Orthopedic surgeons may be particularly resistant to implementation of new clinical guidelines because they tend to prefer to retain substantial autonomy over their work practices and challenge external interventions [23]. In a survey of 350 orthopedic surgeons, 90% were "somewhat" or "very likely" to notice psychological factors, but only 60% were "somewhat" or "very likely" to refer their patients to psychological treatment [24]. Surgeons noted lack of time, mental health stigma, and feeling uncomfortable making referrals as barriers. Qualitative research among orthopedic trauma surgeons and staff is needed to gain a nuanced understanding of setting-specific barriers, facilitators, and implementation strategies to allow for the successful integration of psychosocial care in orthopedic trauma settings, consistent with AAOS guidelines.
The Consolidated Framework for Implementation Research (CFIR) [25] provides a framework for identifying and reporting on implementation determinants from the perspectives of stakeholders that would be impacted by incorporation of psychosocial care within orthopedic trauma settings. Additionally, Proctor's taxonomy of implementation outcomes [26] provide a framework for measuring the success of implementation processes across multiple implementation domains including (1) acceptability (how tolerated psychosocial interventions would be within orthopedic trauma settings), (2) appropriateness (how relevant implementing psychosocial interventions would be within orthopedic trauma settings), and (3) feasibility (the extent to which psychosocial interventions could be successfully implemented within orthopedic trauma settings).
Integration of these two frameworks provides a novel approach to comprehensively characterize the barriers and facilitators among implementation determinants (CFIR) that could directly impact the specific implementation outcomes (Proctor) that would be used to determine the success of the implementation process, before engaging in concerted efforts toward implementation of a clinical innovation. For this particular study, we were interested in psychosocial care in general, rather than a specific treatment modality or care model (e.g., psychotherapy referrals versus care delivered within the orthopedic trauma setting, psychologist versus social worker delivered care) given prior survey data showing general challenges of orthopedic surgeons with psychosocial aspects of recovery.
We aimed to conduct a qualitative study at three geographically diverse outpatient orthopedic trauma centers to understand multidisciplinary stakeholders' perceptions of barriers and facilitators to the implementation of psychosocial care. We also sought to identify potential implementation strategies to overcome barriers and capitalize on facilitators, using both our qualitative data and the Expert Recommendations for Implementing Change (ERIC) [27][28][29], a taxonomy of implementation strategies. Results will inform implementation of psychosocial care within orthopedic trauma settings to maximize outcomes for patients, surgeons, staff, and the larger health care system.

Setting
Sites A, B, and C (anonymized) are level I trauma centers in Austin, Texas; Lexington, Kentucky; and Boston, Massachusetts. Human subject oversight was provided by the Institutional Review Board of Site C. We followed the Consolidated Criteria for Reporting Qualitative Research [30] guidelines in study presentation (Additional File 1).

Participants
Participants were outpatient orthopedic trauma providers across the three sites. Recruitment was facilitated through presentations to departments by "surgeon champions," representing a purposive sampling approach. Orthopedic providers were eligible for study inclusion if they were directly involved in the care of outpatients with acute musculoskeletal injuries (e.g., fracture, dislocation, rupture) within any of the three level 1 trauma centers. Completion of an eligibility screening survey emailed to participants constituted implied consent for focus group participation.
The screening survey was distributed to 94 providers, of which 88 (94%) completed the survey and consented to participation. Of those consented, 79 (90%) participated in qualitative data collection (20 attending surgeons, 28 residents, 10 nurse practitioners/registered nurses/physician assistants, 13 medical assistants, 5 physical therapists/social workers, and 3 clinical research fellows). Nine providers (10%) consented but did not attend a focus groups due to planned or unexpected scheduling conflicts. Table 1 displays participant characteristics.

Procedure
We conducted 18 focus groups (7,8, and 3 at sites A, B, and C, respectively) with 76 participants (42,21, and 13 at each site). We combined providers of several roles (e.g., nurse practitioners with physician assistants) to create groups within the target range of 4 to 8 participants. Department chiefs participated in individual interviews (N = 3; 30 min). Focus groups (60 min) were facilitated by trained staff via Zoom and were followed by optional (10 min) exit interviews using "breakout rooms".
Our semi-structured qualitative script ( Table 2) was developed iteratively by a multidisciplinary team including psychologists, orthopedic surgeons, and an implementation science expert. The script was designed to generate data related to strategies to maximize the relevant implementation outcomes among those delineated by Proctor [26] and overcome inner and outer setting implementation challenges when integrating psychosocial care, as delineated by CFIR [25] (Table 3). For this study, we were specifically interested in the most widely used Proctor outcomes [26,[31][32][33], namely (1) acceptability (tolerability of psychosocial interventions in this setting), (2) appropriateness (relevance of psychosocial interventions in this setting), and (3) feasibility (viability of implementation in this setting). Focus groups were facilitated by predoctoral and postdoctoral research fellows in psychology with training by the multidisciplinary team and no prior relationship with participants (AMV, JB, JD, RAM). Focus groups, exit interviews, and individual interviews were audio recorded and transcribed verbatim by research assistants.

Data analysis
Our data analysis involved two types of content analysis: directed, to identify implementation determinants and outcomes, and conventional, to identify implementation strategies [34]. For the directed content analysis approach, we developed a coding framework by combining all 39 CFIR implementation determinants and the three Proctor implementation outcomes (acceptability, appropriateness, feasibility), thereby selecting codes a priori based on these conceptual frameworks. Given the different orientations of these two frameworks, with CFIR focusing on determinants and Proctor framework focusing on the success of implementation strategies, we decided to integrate both frameworks in order to achieve a more comprehensive understanding of different aspects of psychosocial care integration in orthopedic settings. We believe this expanded theoretical coverage of barriers and facilitators as related to their determinants (CFIR constructs) as well as their respective implementation outcomes (Proctor constructs) will provide for more efficient implementation planning [35]. We also allowed for new codes to emerge during the coding process, but no new codes emerged within the scope of our research questions. Using NVivo software as a data management tool, three coders systematically applied the coding framework to transcripts. Each transcript was independently coded by two coders. Coders met to discuss discrepancies in coding and reach resolution. Coder agreement was excellent (Kappa = 0.93).
We took a collaborative approach to data interpretation. Four team members (JB, JD, MR, RAM) looked at the charted data within each code (CFIR determinants and Proctor outcomes) and identified emerging barriers and facilitators to implementation. We aimed to comprehensively capture all barriers and facilitators that emerged, without concern for the frequency with which barriers and facilitators were raised. We then sought to        Emphasize importance of psychosocial services for pain, recovery, and overall health with a focus on a mind-body framework Intervene with patients to enhance uptake and adherence* Involve patients and family members* "Well, you know, we still live in a world where, you know, unfortunately, psychosocial issues are still considered taboo, and so I would say that when you start to say anything about treatments that involve anything related to the mental capacity, mental space, they assume you mean that you think they're crazy and they're not going to get-you know, that it's all in their head." -Nurse Practitioner, site A align barriers and facilitators with specific implementation determinants and outcomes. When consensus was not reached as to which framework constructs aligned with identified barriers or facilitators, the coding team engaged in discussion to arrive at a consensus-based, collaborative decision to categorize them under the constructs that were most relevant, after consulting the definitions of these dimensions and with considerations for all options [34]. At times, we also observed multifaceted barriers [36] related to different aspects of each framework, and in these cases, we coded the identified barrier to multiple constructs of each relevant framework. Similarly, we intentionally allowed some barriers and facilitators to contradict and to represent the breadth and diversity of opinions expressed by participants. For the identification and selection of implementation strategies, we used conventional content analysis, allowing descriptions of implementation strategies to emerge from participants' own words regarding how to overcome barriers and capitalize on facilitators. We supplemented implementation strategies extracted from the qualitative data with implementation strategies selected from ERIC using the CFIR-ERIC matching tool [37], to ensure a comprehensive, data-driven approach to implementation strategy identification and selection.

Results
Data best fit within 26 of the 39 CFIR constructs. The remaining 13 constructs either did not have any pertinent data or they had little data that also fit within one of the 26 constructs. These 26 constructs most pertinent to implementation of psychosocial care within orthopedic settings span all 5 CFIR domains. We also identified determinants corresponding to each of the three Proctor implementation outcomes (acceptability, appropriateness, feasibility) in our coding framework (Table 3; Fig. 1). Below, we discuss implementation determinants according to the CFIR constructs (italicized in paragraph) they best represent, nested within Proctor implementation outcomes (section headings). We also present key implementation strategies to address these determinants derived directly from the qualitative data. Table 3 also presents additional more general implementation strategies identified from ERIC.

Acceptability
Regarding culture, many participants highlighted valuebased approaches to care that focused on patients' wellbeing, including mental health, over efficiency, or financial profit. On the other hand, some noted that the culture in their clinics values efficiency and that any innovation that may decrease efficiency would not be viewed favorably. Participants proposed educating providers on the importance of psychosocial care as well as ensuring a streamlined referral process. In terms of implementation climate, most participants conveyed openness to clinical innovations broadly and psychosocial interventions specifically. However, some expressed resistance to clinical innovations. Regarding access to information and knowledge, participants voiced concerns about orthopedic providers' lack of knowledge about the importance of psychosocial factors in patient care and the belief that a systemic barrier in medical education contributes to this knowledge gap. As one surgeon described, "You know, I have very rudimentary knowledge of psychiatry from medical school and that's all I resort to" (site C). Participants proposed providing providers with concise, data-driven psychoeducational resources in virtual formats. With respect to knowledge and beliefs about psychosocial interventions, participants noted that orthopedic providers tend to understand the psychological toll of traumatic injuries. Some reported previous experience with psychology, while a few reported general bias against the relevance of mental health factors. In terms of self-efficacy, participants noted the wide range of variability among orthopedic providers regarding ability to discuss psychosocial factors, suggesting that psychosocial interventions may be more easily integrated into some surgeons' clinics than others. Regarding engaging of opinion leaders, participants highlighted that buy-in from leadership within the orthopedics department is key for engaging the rest of the providers in supporting psychosocial care implementation. One research staff member described, "If you have leadership within the orthopedic trauma department to say, 'This is a priority. We want you guys to start implementing this into your patient visits' … That's probably the path to success" (site B).

Appropriateness
Participants expressed concerns about the adaptability of psychosocial interventions to racially/ethnically diverse patients, including non-English speakers. They suggested that psychosocial care should be tailored to patients' sociocultural identities, provided in multiple languages, with on-site translation and racially/ethnically diverse clinicians. Regarding evidence strength or quality, several participants presented skepticism regarding the relevance of addressing psychosocial factors to improve patient outcomes given the abundance of effective medical options. However, many participants noted that surgeons and other providers are highly receptive to empirically supported interventions to improve patient functional outcomes. They suggested highlighting the evidence base for psychosocial interventions to generate provider buy-in. As one surgeon noted, "We're in the age of evidence-based medicine… no one can refute evidence" (site B). In terms of complexity, some participants expressed concerns about whether psychosocial interventions are appropriate for patient with lower levels of education and health literacy and encouraged the use of "lay language" with illustrations.
Related to patient needs and resources, some participants noted skepticism that participants would followup with outpatient mental health care (e.g., due to transportation barriers or homelessness). To circumvent these barriers, participants suggested flexibly adapting psychosocial treatment pacing and duration to patients' circumstances and prioritizing psychoeducation to enhance motivation and adherence. As one resident described, "I think one thing I noticed on my psych rotation is that a lot of these folks are living a very teetering life where one unfortunate circumstance can have their life spin out of balance … So, getting appropriate resources for them is really important." (site B). Related to external policy and incentives, participants mentioned that increased uptake of telehealth practices due to COVID-19 can facilitate access to psychosocial care and enhance integration.
Regarding structural characteristics, participants noted that fast-paced clinic flow and high patient volume challenge the integration of psychosocial care into orthopedic trauma clinics. They suggested streamlining the referral process and soliciting feedback throughout the implementation process. Some reported that the multidisciplinary nature of their clinics might facilitate integration, as patients and providers already view their clinic as a "one-stop shop" for multiple forms of care (e.g., physical therapy, dietetics). Regarding compatibility, providers perceived a need for additional support to help patients develop healthy pain coping strategies. They acknowledged that integrated psychosocial care would reduce this burden on orthopedic providers.

Feasibility
Participants expressed concerns about the cost of psychosocial care integration. Some noted that patients' insurance might not cover psychosocial care and departmental funds may be required to make psychosocial intervention part of standard care. They suggested presenting data to departmental and organizational leadership demonstrating the cost-effectiveness of psychosocial care integration. Regarding organizational networks and communication, some participants expressed concerns about effectively communicating patients' psychosocial needs within their large multidisciplinary team and suggested integrating psychosocial care information within electronic health records. Some participants mentioned that their clinics do have existing channels for communication about patient needs across providers, which would facilitate communication from medical staff to surgeons regarding patient psychosocial concerns.
In terms of available resources, participants expressed concerns about insufficient time and human resources to feasibly integrate psychosocial care. As one medical assistant shared, "Adding another job responsibility on to the trauma clinic… in the trauma clinic specifically it's extremely hard already trying to do the job at hand" (site A). To circumvent these barriers, they highlighted the importance of ensuring a clear division of responsibilities and ensuring adequate staff to facilitate referrals. Relevant to tension for change, participants expressed differing opinions regarding the urgency of addressing psychosocial needs in orthopedic patients. Some expressed an urgent need while others noted that the current situation does not seem intolerable enough to require innovation.
Participants reported barriers to the process of engaging orthopedic providers in the process of integrating psychosocial care, including heterogeneity of provider preferences for psychoeducational materials. Participants suggested engaging departmental chiefs and opinion leaders as a key strategy for increasing buy-in. Relevant to engaging patients in psychosocial care, participants raised concerns regarding mental health stigma and limited willingness of patients to "open up" in the context of fast-paced orthopedic visits. As one medical assistant described, "I think people, when they see the words 'anxiety' and 'depression,' 'mental health,' they get scared and they think 'Oh that's, you're getting too private now'" (site A). Strategies to circumvent these barriers included hiring psychosocial care providers who share racial and cultural identities with patients and emphasizing the importance of psychosocial care for pain and recovery to patients.

Discussion
We conducted a qualitative study with orthopedic trauma providers at 3 geographically diverse level 1 outpatient trauma clinics to understand barriers and facilitators to integrating psychosocial care within usual outpatient orthopedic trauma care and identify implementation strategies to overcome barriers and capitalize on facilitators from the perspectives of stakeholders as well as from ERIC [29]. By providing information on CFIR determinants organized by Proctor implementation outcomes, we demonstrate the value of integrating these two frameworks for the analysis of qualitative data, to provide a more complete picture of the challenges to implementing psychosocial care within orthopedic settings. We observed high enthusiasm for this qualitative study within participating orthopedic trauma departments; 94% of individuals approached consented and 90% of these individuals participated in the focus groups.
Overall, providers appreciated the role of psychosocial factors in recovery after orthopedic trauma and noted that implementing psychosocial care within their practice can be acceptable, appropriate, and feasible. To be acceptable, psychosocial screening and treatment must be seamlessly integrated within the fast-paced clinic flow, with clear delineation of each provider's role.
Because readiness for implementation is heterogenous, it is important to provide tailored education (e.g., brief videos or presentations) on the process and scientific evidence for psychosocial care to surgeons and staff, in addition to patients. Early adopters can serve as "champions" for these effortscatalyzing cultural change and correcting any negative biases. Providers with greater self-efficacy regarding psychosocial care could lead trainings and offer shadowing experiences. Early and sustained support from leadership is key.
Results suggest that to be appropriate, interventions must be tailored for content and delivery (e.g., lay language that normalizes challenges to decrease mental health stigma). Stepped care models have been successful in other settings [38] and may provide useful in triaging patients to appropriate levels of care including outside referrals for those with complex psychosocial needs (e.g., homelessness, severe psychopathology, substance use). Because orthopedic providers may not know about evidence-based psychosocial treatments, it is important to provide brief education including how they differ from the typical surgical protocol. For example, psychological treatments have largely moved away from treating one discrete condition (e.g., depression) toward process-based psychosocial interventions [39,40]. These interventions target core constructs (e.g., pain catastrophizing) that cut across a variety of medical (types of orthopedic trauma injuries) and psychological (depression, anxiety, and posttraumatic stress) conditions. Although empirical evidence for the role of these interventions for orthopedic trauma is emerging [41,42], these approaches may be counterintuitive to surgeons trained to perform specific surgeries (e.g., extramedullary fixation device) for specific diagnoses (e.g., hip fracture). Refining protocols for psychosocial intervention implementation over time based on lessons learned and flexibly tailoring them to the resources already available in each clinic will help circumvent barriers related to heterogeneity of provider buy-in and resource availability.
Results show that increasing feasibility of psychosocial care for orthopedic trauma patients will require ensuring that psychosocial treatment is provided regardless of patients' insurance status. When possible, efforts should be made to reduce costs, access new funding sources, or develop resource sharing agreements to reduce patient fees. Indeed, early psychosocial care can decrease long-term healthcare costs for orthopedic patients. Educating leadership on the long-term cost-effectiveness of psychosocial care in terms of both reducing medical care utilization over time and reducing surgeon burden, while also ensuring a streamlined process with enough staff support is important. The use of telehealth can increase accessibility.
The current study has several strengths and limitations. First, we conducted the largest qualitative study on this topic to our knowledge. Second, we captured diverse experiences by including diverse stakeholders across 3 level 1 trauma settings. Third, we used evidence-based implementation frameworks and combined CFIR with Proctor and ERIC to more thoroughly understand barriers and facilitators to implementation, yield as many implementation strategies as possible across levels of the organizations (e.g., individual providers, clinic culture, patient needs and resources), and organize implementation strategies to guide future work. Notably, we also derived implementation strategies directly from our qualitative data, which were generally consistent with the strategies suggested by ERIC, increasing confidence in our findings. A challenge that we encountered in the analysis was how best to make decisions about where barriers and facilitators identified from the data fitted best, particularly when they could be mapped onto more than 1 CFIR construct. In such cases, we mapped information onto the construct that we considered to be the best 'fit'. Our sample was primarily young, White and Non-Hispanic/Latino, which could impact the transferability of our findings to older and non-White populations. The percent of women surgeon participants was also low, although we enrolled all available women surgeons. Future qualitative studies should aim to use more diverse samples. Lastly, while our goal was to explore 3 of the Proctor outcomes (acceptability, appropriateness, and feasibility), future studies should also qualitatively explore information on outcomes of cost and sustainability. Toward this end, qualitative interviews with administrative leaders and external stakeholders (e.g., payers) would provide valuable, in-depth information to further maximize overall success of efforts toward implementation of psychosocial care within orthopedic trauma settings.

Conclusions
We found widespread support for psychosocial care integration within orthopedic trauma settings. Multidisciplinary providers perceived psychosocial care as crucial for optimizing patient outcomes and reducing provider burden, noting they lack the time and specialized training to fully address patients' psychosocial needs. Providers also perceived that psychosocial care integration would be challenging due to fast-paced clinical flow. By integrating CFIR, Proctor, and ERIC frameworks, we identified actionable strategies for integrating psychosocial care, including obtaining buy-in from department leadership, succinctly communicating the importance of psychosocial care to providers, tailoring interventions to patients from diverse backgrounds, bypassing stigma, and flexibly adapting to fast-paced clinical flow. Mental health practitioners, clinical researchers, and implementation scientists can use these data as a blueprint for maximizing successful implementation of psychosocial care and aligning orthopedic trauma practices with evidence-based biopsychosocial models of care.
Abbreviations CFIR: Consolidated Framework for Implementation Research; ERIC: Expert Recommendations for Implementing Change
Additional file 1. COREQ (COnsolidated criteria for REporting Qualitative research) Checklist