Guidelines for the treatment of eating disorders (EDs) recommend that individuals with the most severe eating disorder symptoms and co-occurring emotional disorders receive treatment in intensive care settings such as residential or day-hospital programs (American Psychiatric Association, 2006; Yager et al., 2014). Given the difficulty of implementing cohesive evidence-based practices in these complex hospital systems—with multimodal, multidisciplinary treatment and complex, severe patients—minimal data exist concerning empirically supported therapies (ESTs) or evidence-based practices (EBPs) in these settings (Frisch, Herzog, & Franko, 2006; Goode, 2016; see also Lilienfeld, Ritschel, Lynn, Cautin, & Latzman, 2013; Spring, 2007, for the distinction between ESTs and EBPs). Although no randomized controlled trials (RCTs) have compared specified comprehensive residential treatments for EDs, there is a general call for increased evidence-based care from insurance companies, oversight organizations, consumers, and researchers (Friedman et al., 2016; Goode, 2016). Fortunately, several existing theoretical frameworks have been developed from implementation research in other health-care settings, which may be well suited to inform EBP implementation in ED treatment programs. The goals of this study are to outline an implementation science model, describe research supporting this model, and illustrate its application to ED settings through description of a case example of successful EBP implementation across a national network of residential and day-hospital treatment programs.
Multiple psychological treatments for EDs have amassed evidence to support their efficacy in outpatient RCTs. Cognitive-behavioral therapy (CBT) and interpersonal psychotherapy (IPT) have strong evidence to support their use with outpatient adults with bulimia nervosa (BN) and binge eating disorder (BED; Agras, Walsh, Fairburn, Wilson, & Kraemer, 2000; Wilfley et al., 2002), while family-based therapy (FBT) has strong support for use with adolescents with anorexia nervosa (AN; e.g., Lock et al., 2010). Other treatment approaches—such as dialectical-behavioral therapy (DBT) for BN and BED (Bankoff, Karpel, Forbes, & Pantalone, 2012), specialist supportive clinical management (SSCM) for AN (e.g., Schmidt et al., 2015), and transdiagnostic CBT-enhanced (e.g., Fairburn et al., 2015)—have more moderate research support with EDs. However, none of these approaches have been tested in RCTs as the basis of residential or day-hospital programs.
Residential programs have many key differences from outpatient psychotherapy settings. Psychotherapy takes place most often in groups, multiple times per day, multiple days per week. The population in treatment in residential programs has severe symptoms, including typically severe symptoms of an ED as well as severe symptoms of a co-occurring mood, anxiety, posttraumatic, or obsessive-compulsive disorder (Twohig, Bluett, Torgesen, Lensegrav-Benson, & Quakenbush-Roberts, 2015). Patients with different severe EDs and different severe comorbidities are, by necessity, treated together in groups. Furthermore, many of the key targets of outpatient treatment—such as regulation of eating and reduction in behavioral symptoms—are reduced or eliminated in the highly structured residential setting. For example, interventions targeting the reduction in binge eating, or increasing food intake, are less relevant in a setting where patients eat three prescribed and supervised meals a day, and have no access to food outside of mealtime. For these reasons, residential and day-hospital programs for EDs have struggled to implement existing empirically supported protocols for outpatient therapy as the comprehensive basis for their treatment approach in their settings (e.g., Lowe, Bunnell, Neeren, Chernyak, & Greberman, 2011).
In the past 12 years, one private, national provider of residential and intensive outpatient programs for severe EDs undertook a multifaceted effort to improve the overall consistency and quality of clinical care and clinical research, culminating in the implementation of an adapted EST for emotional disorders across its two residential and 14 intensive outpatient programs. Between 2005 and 2013, the Renfrew Center grew substantially and undertook multiple steps to evolve its treatment model and enhance quality of care. In 2013, the administration decided to consider the Unified Protocol for transdiagnostic treatment of emotional disorders (UP; Barlow, Ellard, et al., 2011; Barlow, Farchione, et al., 2011) as a foundational EST for possible implementation. Between 2013 and the end of 2016, the implementation team accomplished the following: developed adapted protocols for each level of care (residential, day hospital, and intensive outpatient); defined UP-consistent approaches for multimodal treatment (e.g., group and individual therapy, nutrition, nursing); provided training to over 400 clinicians and staff; established adherence and competency among residential therapists; and implemented the full program across all treatment sites (see Figure 1 for timeline). In addition, the team collected outcome data at the two residential sites before and after implementation, allowing for continuous evaluation of patient improvements across the implementation timeline. The adapted, intensive, multimodal treatment is now referred to as the “Unified Treatment Model” (UTM) by the organizational stakeholders.
In this study, we aim to review the implementation research that informed this project, describe implementation research frameworks that help to organize this research, and illustrate each set of research conclusions with examples of their application in this case.
1.1 Implementation research review and theoretical framework
Several decades of implementation research across diverse fields, including health-science implementation research, have yielded converging data regarding effective practices for EST implementation. Multiple reviews have already usefully synthesized this research for the health-care community (e.g., Greenhalgh, Robert, Macfarlane, Bate, & Kyriakidou, 2004). In 2009, the available research data were incorporated into a major review and analysis of implementation research across fields, yielding the influential Consolidated Framework for Implementation Research (CFIR; Damschroder et al., 2009). In the review provided below, we aim to provide a “review of reviews” specifically relevant to the outcome of psychotherapy implementation processes. We will link our observations regarding evidence-based implementation practices to the CFIR framework, as well as the National Implementation Research Network's (NIRN) synthesis of implementation literature (Fixsen, Naoom, Blase, & Friedman, 2005). Our review and description will be grouped by the NIRN implementation phases: “exploration,” “adoption,” “planning,” “initial implementation,” “full implementation,” and “sustainability.” Within these phases, we will include variables from the CFIR that are cited as predictive of successful implementation (listed and described in Table S1), as well as the conclusions from other reviews of implementation literature more specific to the health-care field (e.g., Greenhalgh et al., 2004). Variables predictive of implementation success taken from these prior reviews are denoted by italicized font in the text below.
1.2 The empirically supported treatment
The UP (Barlow, Ellard, et al., 2011; Barlow, Farchione, et al., 2011) is a modular, emotion-focused treatment designed to be applicable to mental health conditions that involve a prominent emotion component (e.g., mood, anxiety, personality, eating disorders). The UP is based on psychopathology and emotion science research indicating that emotional disorders share underlying patterns of negative affectivity and emotion avoidance, which maintain and exacerbate varying symptoms (Barlow, Allen, & Choate, 2004). The UP aims to promote transdiagnostic symptom improvement by targeting higher-order mechanisms of psychopathology through a variety of emotion-focused CBT techniques (Barlow, Ellard, et al., 2011; Barlow, Farchione, et al., 2011; see also Table S2). The higher-order mechanisms targeted by the UP are observed in psychopathology research to precede the development and expression of ED symptoms, and include negative and labile affectivity (Haedt-Matt & Keel, 2011; Lavender et al., 2013; Levine & Marcus, 1997; Stice, 2001), emotional avoidance and alexithymia (Danner, Sternheim, & Evers, 2014; Fulton et al., 2012; Nordbø, Espeset, Gulliksen, Skårderud, & Holte, 2006; Svaldi, Griepenstroh, Tuschen-Caffier, & Ehring, 2012; Wildes, Ringham, & Marcus, 2010), and maladaptive emotion-driven behaviors (Wildes et al., 2010; Wonderlich et al., 2008). Prior to selection for adaptation and implementation in this context, the transdiagnostic UP had demonstrated effectiveness in multiple open trials (Ellard, Deckersbach, Sylvia, Nierenberg, & Barlow, 2012; Ellard, Fairholme, Boisseau, Farchione, & Barlow, 2010) and efficacy in treating anxiety and depressive disorders in an RCT (Farchione et al., 2012), although it had not been tested in an RCT specifically for the treatment of primary EDs.
The UP was intentionally designed to have maximum facility for dissemination and implementation. Transdiagnostic, modular treatments hold promise for wide-scale dissemination because they apply to a wide array of patients, have flexible and individualizable components, and are palatable to treatment providers (McHugh, Murray, & Barlow, 2009). Furthermore, distilling empirically supported intervention elements (e.g., exposure, cognitive reappraisal; see Gonzalez, Butler, & England, 2015) in a single overarching and cohesive model decreases the burden on clinicians to become proficient in dozens of disorder-specific manuals. The UP for outpatient, individual therapy includes eight modules, each of which includes multiple basic and applied research-supported interventions. These are familiar across cognitive-behavioral and third-wave psychotherapy approaches, and include psychoeducation about emotion, awareness of the components of emotion and the unfolding of emotional experience over time, the development of cognitive flexibility, analyzing how symptoms function as emotion-driven behaviors or efforts to avoid emotion, the development of new behaviors in response to emotion, and exposure therapy practices (see Table S2).
1.3 The organization
The Renfrew Center, founded in 1985, was one of the first comprehensive residential ED treatment facilities for women in the United States. It is one of the largest private ED treatment networks in the United States and currently includes two residential and 14 nonresidential (day-hospital and intensive outpatient) sites, with thousands of patient admissions per year. When the organization was founded, ESTs for EDs were only in early stages of development. The organization's clinicians evolved their own treatment approach for EDs aimed at supporting patients’ relational connectedness, empowerment, emotion expression, and self-directedness, which they described as “feminist-relational.” As the organization's treatment network expanded, new and diverse treatment elements were developed by clinical staff at different sites, without formal inclusion of empirically supported interventions.
2 “EXPLORATION” PHASE: DEVELOPMENT OF THE INNER SETTING AND READINESS
The exploratory initial phase of any implementation process may take place over multiple years prior to the selection of an innovation for implementation. Review of the implementation research indicates that there are elements of the inner setting of the organization that develop during the exploration phase that set the stage for success. Importantly, different facets of the organization should reflect readiness; readiness, in turn, is positively affected by the relative size and maturity of the organization, the leadership engagement (Damschroder et al., 2009; Fixsen et al., 2005; Greenhalgh et al., 2004), the nature and quality of the organization's social networks, and the organization's ties to the outside world (i.e., cosmopolitanism).
Health-service organizations that successfully implement innovations tend to be mature and relatively large because these organizations are more likely to have differentiated administrative departments and roles, resources available to devote to the implementation, effective systems of data collection, and sophisticated social structures within the organization (Greenhalgh et al., 2004). Within the CFIR, these characteristics are labeled structural and cultural characteristics of the inner setting that can promote success in implementation. Research in implementation science emphasizes the importance of strong interpersonal relationships and communication networks within an organization, to promote and distribute an innovation.
3 KEY IMPLEMENTATION CONSTRUCTS FROM THE “EXPLORATION” PHASE
In the years prior to 2013, the Renfrew Center organization underwent many changes in preparation for the implementation of an evidence-based clinical approach. These changes synergized with existing characteristics of the organization to facilitate movement toward selection of the UP for implementation. Such organizational characteristics and change processes are described below in relation to the essential components of the exploration phase.
3.1 Structural maturity and leadership engagement
As the organization expanded and matured, the clinical administration was increasingly motivated to coordinate clinical care. Clinical administrators wanted to identify effective clinical practices and make them consistent across therapists, disciplines, and sites. To support this effort, clinicians were promoted to positions of administrative importance (e.g., “Chief Clinical Officer”) with increased agency-wide influence. A “Clinical Excellence Board” was created, including the organization's primary clinical, business, and research leaders. The Clinical Excellence Board—made up of individuals within the organization—was the group that collectively decided that evidence-based practice was necessary, chose the EST to implement, and selected the implementation team, who were empowered to plan the implementation and communicate required steps to clinicians and staff. In this sense, the implementation was top-down, although it was collaboratively initiated by multiple internal clinical and business leaders who could be considered “stakeholders.” This group later provided the forum for important figures to discuss and support each step of implementation, promoting leadership engagement across all sites—which is identified as one key factor contributing to readiness at the inner setting level (Damschroder et al., 2009).
3.2 Research and training resources
A formal Research Department existed internally in the organization for many years. Over time, the Research Department improved the infrastructure and scientific possibility of the research, adding validated measures of psychopathology and improving response rates. An outside research consultant was hired, who supported these efforts and published scientific articles using the available data (Lowe, Davis, Annunziato, & Lucks, 2003). These publications brought positive attention to the organization, which led to further investments in the Research Department. Later in the implementation process, the existence of this system allowed patient outcome data to be quickly collected, analyzed, and presented to organizational leadership and staff, garnering essential support for the UTM.
In 2013, at the urging of the Chief Clinical Officer, the organization created an internal Training Department with its own budget and staff. This was crucial to the later success of the implementation, as the Training Department formed the core of the implementation team, and assumed primary responsibility for carrying out the highly effortful process of implementation. These organizational developments can be understood as co-reinforcing structural changes (e.g., creating new departments) and cultural shifts (e.g., overall shift toward empiricism; Damschroder et al., 2009). To be fully ready for implementation, concrete available resources were also required. For example, new technology and equipment were necessary at each site to conduct research and deliver training, such as digital audio recorders, secure digital storage software, videoconferencing software, and flat-screen TVs with Internet access.
3.3 Quality and structure of social networks
The relationships among the clinical leadership, Research Department, Training Department, and implementation consultants were the foundation of the UTM implementation. The Renfrew Center already had a strong informal tradition of emphasizing and supporting interpersonal relationships, in keeping with their feminist-relational philosophy. In the period prior to the implementation, the clinical administration created additional formal networks of communication to coordinate clinical care. A complex formal network of in-person and conference call meetings was established; for example, during initial implementation, the Chief Clinical Officer and other members of the Training Department each had over 12 hr of regularly scheduled meetings each week, during which they discussed the implementation with each other, the implementation consultants, research consultants, and clinical leaders from every discipline and site.
The organization's engagement with research and practice standards outside their walls led to an increased emphasis on the importance of ESTs in general, as well as to their introduction to the UP itself. Participation in national advocacy organizations, practitioner networks, and research conferences led to an increased appreciation for the importance of utilizing ESTs in regular practice. Engagement in research conferences and publications led specifically to their familiarity with the UP and its potential application to EDs (Barlow, 2013; Barlow & Boisseau, 2011), as well as a personal introduction to the UP experts who would support the implementation effort and join the implementation team.
4 “ADOPTION” PHASE: RELATIVE ADVANTAGE, EVIDENCE, COMPATIBILITY, AND ADAPTABILITY
The “adoption” phase of implementation begins with selection of a particular innovation, and includes detailed planning and the creation of a receptive environment (Fixsen et al., 2005). Implementation research suggests that the specific intervention characteristics associated with its successful implementation include the perceived relative advantage of the intervention over the current practices and the perceived strength and quality of the evidence for the intervention (Damschroder et al., 2009). Equally or more important—and perhaps less appreciated in the psychotherapy research community—are the intervention's compatibility with the current culture and practices, and adaptability to the specific needs of the setting and the population (Damschroder et al., 2009; Greenhalgh et al., 2004). According to the CFIR, health-science innovations that are likely to be chosen for adoption are those that are perceived as adaptable, trialable (i.e., suited to limited trials to examine feasibility and impact), and optimally simple to implement (Table S1 and Damschroder et al., 2009; Greenhalgh et al., 2004). Implementation studies observe a strong relationship between the time and resources devoted to adapt an innovation for the local context and success (Damschroder et al., 2009).
During the exploration phase, the clinical leadership team met regularly to consider the benefits and limitations of different ESTs for application across their entire continuum of care. They particularly focused on the adaptability of the ESTs to their two large residential programs, where the treatment setting and population most differed from the context and population for RCTs for EDs. The available RCT data had been collected in the outpatient setting, where it was appropriate to prioritize food regulation and ED behavioral symptoms as targets of treatment. In the residential setting, however, food intake is highly regulated by supervised meals structured by the nutrition department, and ED behavioral symptoms (e.g., purging, overexercising, and binge eating) are actively discouraged through intensive staff oversight. Structural limitations also limit opportunities for patients to engage in these behaviors (e.g., no access to food outside of meals, supervised bathroom visits). Three of the available ESTs for EDs—CBT-BN, CBT-E, and FBT—were perceived by the clinical leadership to have limited application in the residential setting because food and behavioral symptoms were already so regulated in that environment.
Furthermore, the clinical leadership perceived that the residential program setting and population had particular needs that were not met by the available ESTs for EDs. Severe comorbidity is the norm in residential treatment, and extensive group session material that attended to both the ED and serious comorbidities—such as posttraumatic stress disorder, substance use disorder, self-harm, panic disorder, social phobia, and obsessive-compulsive disorder—was felt to be necessary. Several of the available ESTs—CBT-BN, CBT-E, FBT, and IPT—were not perceived to adequately address all these common and severe co-occurring disorders. In addition, the organization already had the experience of piloting some of the core interventions from CBT-BN—such as “regular eating”—in one of their nonresidential programs and had found it was difficult to implement due to the minimal attention to comorbidity and emotion regulation, and their sense that the intervention lacked flexibility (Lowe et al., 2011). Finally, the clinical leadership attempted to pilot interventions from other ESTs—such as DBT for BED/BN and ACT—and found it difficult to apply the model to the full range of ED symptom presentation, as more limited ED clinical examples and data were available for these approaches. The clinical leadership contemplated using different components from different ESTs to create a more comprehensive, composite approach, but they feared this would feel unwieldy and confusing to the patients, that it would be impossible to train all the staff to competency in so many different approaches, and that this approach would likely fail to unify treatment across their programs and levels of care.
When the clinical leadership became aware of the UP, although the approach had no RCT demonstrating efficacy with primary EDs, it was perceived to be adaptable, trialable, and relatively simple to implement because it was principle-based, modular, and transdiagnostic. The principle-based treatment lent itself to adaptation (Fixsen et al., 2005); the modules lent themselves to manageable trials; and the transdiagnostic application made it relatively simple to apply across diverse patient groups. The UP was also judged to be compatible with the existing culture of the organization because it was emotion-focused: The modules of the UP are parts of a coherent process of identifying, approaching, and learning new lessons about emotion (see Table S2). From the earliest conversations with Renfrew, UP consultants emphasized their willingness to adapt the protocol for the population, setting, and organization.
5 KEY IMPLEMENTATION CONSTRUCTS FROM THE “ADOPTION” PHASE
5.1 Relative advantage
In the exploration and consideration stage of implementation, prior to selecting a foundational EST, multiple factors led toward increased emphasis on the scientific literature to help define the intervention model. Clinical leadership acknowledged that manualized psychotherapy interventions were demonstrating superior outcomes relative to treatment-as-usual, and they expressed clinical and ethical obligations to consider them. Clinical leaders understood a new EST-derived model may provide a relative advantage for patient outcomes as well as for overall cohesiveness for the treatment and for the training. The business leadership realized that research-supported interventions might have economic benefits; demonstrating maximal treatment benefit to oversight agencies and insurance companies was becoming more important. Furthermore, because all the interventions of the UP were “unified” by their approach to emotion, the UP was seen to have the relative advantage of bringing a cohesive set of principles to coordinate interdisciplinary, transdiagnostic treatment and training.
5.2 Evidence strength and quality
Although the evidence strength and quality for an intervention are clearly important, practitioners tend to perceive evidence differently from researchers. Crucially, the leadership was aware that no EST had been fully implemented and assessed in an RCT specifically for residential and day hospitals, and therefore, all existing clinical evidence had potentially questionable generalizability to their setting and population. One criticism of the UP implementation at Renfrew, voiced primarily by those outside the organization, was that the protocol had not been tested in outpatient therapy for individuals with principal EDs in an RCT. However, the UP consultants and others in the research field argued that a relevant research evidence base spans therapeutic efficacy, therapeutic effectiveness, and basic psychological processes (Lilienfeld et al., 2013). For example, the extensive emotion research on which the treatment was based was shown to be relevant to EDs as well (Boswell, Anderson, & Anderson, 2015; Stice, 2001; Svaldi et al., 2012; Wildes et al., 2010). Extensive research with EDs indicates that negative and labile affect, emotional avoidance, and core negative beliefs about emotion are risk factors and maintaining mechanisms. Furthermore, similar treatment elements to those in the UP modules (e.g., mindfulness, cognitive restructuring, and exposure therapy) had each demonstrated effectiveness for individuals with EDs in separate studies. The basic research to support each module as applicable to psychopathology present in EDs and the treatment outcome research that reflects similar interventions in existing ESTs and EBPs for EDs are presented in Table S2. UP development was anchored by the goal of identifying and distilling empirically tested intervention principles and strategies associated with positive treatment outcomes across diverse problem areas (Farchione et al., 2012), and therefore, it is not surprising that it shares interventions with existing ESTs for EDs. Therefore, the research evidence, broadly defined, was considered to come from a reliable source, and to sufficiently support the concept of EBP (Damschroder et al., 2009).
The clinical leadership sought an approach that would be compatible for integration with their existing approach. Their “feminist-relational” approach encouraged emotional awareness and expression, in the context of therapeutic relationships that were collaborative and nonpathologizing (Maine, 2006; Nardozzi & Hranicka, 2006). The leadership team felt that the UP philosophy and strategies had important points of connection with the original feminist-relational model. Like the existing model, the UP emphasized that avoidance and suppression of emotion were central to the development and maintenance of EDs, and UP strategies promoted emotion awareness and acceptance. In the UP, symptoms are understood as attempts to regulate emotion—a rationale that was seen as nonpathologizing. Like the existing philosophy (see Lowe et al., 2011), the UP had a cohesive approach to treating different EDs (e.g., anorexia nervosa, binge eating disorder) as well as co-occurring emotional disorders, with an emphasis on all aspects of emotional functioning. These points of connection promoted the perception of compatibility and adaptability, and therefore organizational readiness.
The aspects of the UP that were found to be less compatible with the existing treatment culture were in areas of specific technique, and therefore not fully appreciated or addressed until implementation was underway. Feminist-relational therapy had been conducted in “process groups,” to allow room for self-expression, and patients were encouraged to find the route to recovery “within themselves.” The UTM, in contrast, had a set of psychoeducational concepts and structured exercises for each group within a structured group schedule. Some clinicians thought using a manual was contrary to nonhierarchical belief systems (see Lowe et al., 2011). Furthermore, the UP manual is predominately “intrapsychic” in focus, meaning it directs more attention to the processes and functions of emotion within an individual; there is relatively less focus on interpersonal emotional processes or relational skills. The resolution of values and techniques took time and effort to resolve. Following extensive training (described below), clinicians were required to try the manual, record their sessions, compare impressions with the trainers, and consider data on patient satisfaction and outcome. Clinicians often found the manual to be more compatible with their values, more effective, and more flexible than anticipated. In turn, the implementation team was responsive to the feedback from clinicians that the first iteration of the UTM manual was not sufficiently “relational,” meaning that there were both not enough examples of relational triggers to emotion—which are common in eating disorders—and also that the semi-didactic, group treatment format promoted a degree of emotional disengagement or distance between the patients, the therapists, and the material. The implementation team undertook training in relational-cultural therapy (Jordan, 2010) and further adapted the training program and treatment manual. Relational training now includes clinicians’ identification of their own emotional responses, development of mutual empathy in treatment, and awareness of hierarchical dimensions of therapy situations. Many examples of relational situations provoking emotion were also added to the manual and homework, and we introduced a brief emotional/relational “check-in” at the beginning of each group to promote emotional connection between group members and with the therapist. It was key to identify which core components of the UP had to be maintained unchanged when building the UTM, and which elements could be considered part of the “adaptable periphery” (Damschroder et al., 2009, p. 52) that could be shaped and reworked to meet the local needs of the organization. The expert consultants considered these changes to enhance, rather than detract from or conflict with, the core aims and principles of the UP.
The process of adapting the EST to the needs of the local environment was a complex process with multiple steps, and we provide multiple examples of this process here. Substantial preliminary steps were taken to co-educate the UP consultants and implementation team prior to the adaptation. One consultant conducted multiday site visits to both residential programs, including intake assessment, group therapy, individual therapy, nutrition counseling, meals, after-meal support group, nursing and weight check-ins, treatment-team meetings, community meetings, and milieu crisis counseling—referred to by the team as “embedding” the consultant in the organization. After presenting an overview of the UP to the clinical staff, the consultant collected attitude questionnaires and conducted clinician focus groups as well.
The adaptations of the UP and the residential treatment schedules were extensive, iterative, and comprehensive. The original UP was modular and principle-based, but it was also progressive; later modules (e.g., exposure therapy) built upon the basis of earlier modules (e.g., nonjudgmental emotional awareness; see Table S2). Changes were made to the UP treatment process and treatment materials to accommodate many aspects of intensive treatment, such as (a) therapy primarily administered in groups; (b) intensive treatment administered in programs of varying average length-of-stay and treatment hours-per-week at different levels of care; (c) the treatment of patients admitted in a compromised state who may take weeks to become medically stable, adequately motivated, or cognitively rehabilitated; (d) a roster of staff who could accommodate limited options at different times of the day; (e) the incorporation of specialized groups for subpopulations with particular issues (e.g., substance abuse, trauma); (f) adequate emphasis on principles of relational therapy key to the philosophy, brand, and identity of the Renfrew Center; and (g) the adaptation of the treatment rationale and principles to various activities and disciplines (e.g., nutrition and mealtime, psychiatry and medication, nursing and weight stabilization, milieu counseling).
We will provide some examples of how these issues were addressed here, and the adaptations are described in fuller detail in Table S2. Rather than providing psychoeducation in “homework,” which under-motivated patients may not complete, the psychoeducation was provided in the group itself, interspersed with exercises encouraging application of the psychoeducation to individual experience. Eating disorder examples (restriction, food rules, body image, binge eating and purging) were created and presented in every group and each homework exercise along with examples regarding depression and anxiety. Highly engaging exercises were developed, including media (clips from movies) and role-play exercises to increase engagement of under-motivated patients. Interpersonal exercises (e.g., relational check-in at the beginning of group) were introduced to help patients become aware of and share current emotions, and to reduce their tendency to “check out.” Instead of progressive modules 1–8, we created three stages of early, middle, and late stage interventions, with assessments and markers developed to define when individuals were ready to progress to the next stage. All UP groups were treated as “mandatory” in the patients’ schedule, so patients were not scheduled for individual therapy or pulled out for psychiatry appointments during that allotted time (see also Table S2).
The fact that the Renfrew Center offered treatment at multiple levels of care for EDs of varying severity levels posed another challenge for treatment adaptation. Some patients were admitted to residential or day-hospital programs from within the network's own sites and presumably had exposure to UP concepts prior to admission. Others entered treatment from other programs and had little to no familiarity with UP concepts. To ensure adequate exposure to the core treatment concepts of the UP without excessive repetition for patients with prior exposure, multiple variants of group exercises were developed and offered in the manual. Furthermore, therapist training and supervision focused on skills to bring the material “to life” using the personal experiences and current emotions of group participants.
Further, challenges were encountered for patients stepping down from the residential to less structured day-hospital or intensive outpatient settings, where their behavior was less closely monitored and patients were required to monitor their own restriction, purging, and exercise symptoms outside of program hours. To address the transition, patients were provided with opportunities to practice applying UP skills to real-world settings through off-campus passes (e.g., time away from the program to eat with family members or friends) and therapeutic exposures (e.g., shopping trips and restaurant outings) while still in residential treatment. Any incidence of eating disorder symptom use that did arise in lower levels of care would be addressed therapeutically using the UP framework. Specifically, during group and individual sessions, a patient would evaluate the antecedent triggers, emotional responses, and consequences of her behavior. Plans to choose alternative actions (i.e., more approach-oriented behaviors) the next time a similar emotion was triggered were then reviewed.
6 “PLANNING & PREPARATION” PHASE: READINESS ASSESSMENT AND INTERVENTION
Extensive research suggests that the “readiness” of the organization should be assessed and any issues with readiness addressed prior to attempting implementation. Aspects of the implementation climate may reflect more or less readiness, and may facilitate or obstruct adoption of an EST (Damschroder et al., 2009). The readiness of the implementation climate includes the clinicians’ attitudes toward the EST and implementation project, the overall tension for change in the clinical community (and therefore motivation for the implementation), the distribution of goals and accountability to goals, and efforts to create a facilitative learning climate. The readiness assessment was carried out by the UP consultants, and other activities were carried out by the organization's internal champions (i.e., respected individuals in the organization who take the lead in promoting and modeling the new practices as well as maintaining a positive implementation climate).
7 KEY IMPLEMENTATION CONSTRUCTS FROM THE “PLANNING & PREPARATION” PHASE
7.1 Readiness assessment and intervention
Following the presentations and clinician focus groups at the site visits in 2014, in which the compatibility of the UP rationale and the organization's philosophy were highlighted, the residential clinicians reported positive views of the UP and a relatively high level of tension for change: They agreed that treatment should be more cohesive, evidence-based, and effective, and they reported optimism that the UP would help meet these goals. Questionnaires were used to assess clinicians’ “readiness” to implement EBPs, based on their attitudes toward the UP and the strength of their adherence to a particular theoretical orientation. In general, staff reported positive attitudes toward evidence-based practice overall. However, more positive attitudes were correlated with stronger CBT orientation, and stronger positive attitudes were reported at one of the two residential sites.
Self-report data were collected from N = 159 clinical staff (93.7% female; Mage = 37.81 years, SD = 12.51) following the initial intensive training in the UP. We briefly summarize a few key results from t tests and correlational analyses here. Clinicians endorsed a variety of theoretical orientations, with one of the two residential sites more strongly endorsing a CBT (p < .01) orientation, and the other more strongly endorsing a feminist orientation (p < .01). On average, the sample reported finding the UP “logical” (M = 4.04, SD = 1.57 on a 0 [not at all] to 6 [very much] scale). Staff also generally endorsed the belief that the UP would be successful in helping patients with their eating problems (M = 4.09, SD = 1.46, same scale) and co-occurring emotional problems (M = 4.12, SD = 1.47, same scale). Analysis of the clinician attitudes revealed some important patterns to address. On the EBP Attitude Scale (EBPAS; Aarons, 2004), the perceived usefulness of research-supported interventions was positively correlated with level of CBT self-identification (r = .32, p < .05). Level of CBT identification was also associated with more overall positive beliefs about exposure (r = .27, p < .05) on the Therapist Beliefs about Exposure Scale (TBES; Deacon et al., 2013). Site differences emerged on the EBPAS items as well. For example, one site endorsed a higher degree of overall willingness to adopt a new therapy and was also significantly more willing to adopt a new therapy if required by a supervisor or their agency, or if sufficient training was provided (ps < .05). Conversely, the alternate site more strongly endorsed the beliefs that they “know better than academic researchers” [about what is effective practice], and “research based interventions are not useful,” and negative overall beliefs about exposure therapy (ps < .05).
Given the anticipated difficulties in the initial implementation for the site with relatively less observed readiness, additional support was provided to that site by the Training Department during the initial implementation period, further described below. This support was tailored to help address the points of compatibility and incompatibility between feminist/psychodynamic practice and the UP; additional training in the flexible use of a manual to non-CBT clinicians; and additional support for highly experienced clinicians. The Training Department observed that it seemed particularly difficult for these highly experienced clinicians, who were themselves less familiar with CBT practices (due to the lesser training in CBT during their own education), but were often in supervisory roles. They found it was important to identify and address the particular emotions involved in feeling “de-skilled,” and supervising younger clinicians who seemed to feel more comfortable with the new practices than the supervisors themselves.
7.2 Internal champions
The essential function of internal champions to implementation processes is widely acknowledged (Damschroder et al., 2009; Fixsen et al., 2005). In the early stages of adoption and planning, the clinical leadership at Renfrew played multiple champion roles, such as transformational leader, network facilitator, and organizational maverick (Greenhalgh et al., 2004). The transformational leader is the early visionary who embraces the model and “harnesses support from other members of the organization” (Greenhalgh et al., 2004, p. 603), and the network facilitator “develops cross-functional coalitions within the organization” (Greenhalgh et al., 2004, p. 603). Renfrew's Chief Clinical Officer arranged many hours of conversation with key administrative stakeholders across departments and clinical staff to garner support; she directed the UP consultants and members of the Training Department to create collaborative working groups with disparate clinical leaders, including those who had difficulty applying the UP to their discipline. These meetings led to new applications of the UP, as well as the creation of new champions specifically among groups who might have opposed the implementation otherwise (Li et al., 2009; Ranmuthugala et al., 2011). Later in the process, members of the Training Department who had participated in the adaptation and planning process, and had achieved certification in the UP, became invaluable internal champions, called upon to explain and defend the model repeatedly. As the implementation was rolled out to nonresidential sites, supervisors who expressed particular enthusiasm about the model were given additional training and received the title “Expert Champion.” The Expert Champions subsequently provided training, implementation support, and supervision to other sites, in a “cascading” training and supervision design (Martino et al., 2011).
7.3 Learning climate
Worth a deeper look:
Test scores at this school are about the same as the state average, though still below the top-performing schools in the state. Because test scores in some states are so low, many students at this school may not be performing at grade level.
Notice something missing or confusing?