_________ Mohan Vinjamuri, LMSW, PhD is an Assistant Professor, Department of Social Work, Lehman College - City University of New York, Bronx, NY 10468. Lydia P. Ogden, MSW, PhD is an Assistant Professor, Simmons University School of Social Work, Boston, MA 02115. Jessica M. Kahn, LMSW, PhD is an Associate Professor, Department of Social Work, Lehman College - City University of New York, Bronx, NY 10468. Copyright © 2018 Authors, Vol. 18 No. 4 (Fall 2018), 1218-1236, DOI: 10.18060/22333 This work is licensed under a Creative Commons Attribution 4.0 International License. Student, Faculty, and Field Instructor Approaches to SBIRT Implementation: Implications for Model Fidelity Mohan Vinjamuri Lydia P. Ogden Jessica M. Kahn Abstract: Informed by an empirically-based implementation model, this study examined how social work faculty, student, and fieldwork instructor approaches to using the evidence-based SBIRT protocol affected implementation and model fidelity. Data were obtained from two rounds of focus groups with three groups of stakeholders (faculty, students, and fieldwork instructors) about their experiences teaching, learning, using, and supervising SBIRT and were analyzed using a hybrid inductive and deductive process. Analyses yielded three main categories of approaches: those that impeded implementation and model fidelity; those that supported implementation but were not congruent with model fidelity; and those that supported both implementation and model fidelity. Lack of consciousness about model fidelity was an issue across groups. Efforts to find a fit between the protocol, settings, and professional approaches to social work often led to implementation but questionable model fidelity. Repeated exposure to new material and opportunities to engage with it, having specific tools, and supporting learners’ efforts to uphold social work values can promote faithful implementation. Keywords: Evidence-based practice; implementation science; model fidelity; SBIRT Implementation science and evidence-based practice (EBP) have been a central concern of social work for decades, leading to significant changes in the profession as evidence increasingly informs social work practice, administration, and policy. An ongoing challenge of the profession has been delivering evidence-based interventions within agencies to clients in ways faithful to the protocols upon which the scientific evidence was established. In 2005, Mullen, Shlonsky, Bledsoe, and Bellamy asserted that social work lacked literature that empirically examined and addressed barriers and facilitators to implementation. Since then, knowledge and information developed by social work and other professions have been used to address these issues (Acri et al., 2017; Atkins & Frederico, 2017; Cabassa, 2016; Gray & Schubert, 2012; Kerner & Hall, 2009; Otto, Polutta, & Ziegler, 2009). As a result, protocols are now being designed to maximize diffusion so that, while establishing the evidence, researchers consider how the intervention will be adapted and integrated into routine practice (Kerner & Hall, 2009). Nonetheless, much work remains to be done to establish evidence-based practices in real world settings. In particular, understanding of what happens between development of the scientific evidence and implementation of the evidence-based protocol in practice remains incomplete (Gray & Schubert, 2012). https://creativecommons.org/licenses/by/4.0/ ADVANCES IN SOCIAL WORK, Fall 2018, 18(4) 1219 One aspect of implementation involves transmission of knowledge across social work faculty, fieldwork instructors, and students. How these three groups of stakeholders approach scientific information and how their respective approaches might affect the role of evidence in social work have received limited attention to date. This focus group-based study addresses this important ongoing issue and contributes to knowledge about how an evidence-based protocol reaches clients and whether it reaches clients in a way faithful to the original model. Informed by the authors’ (Ogden, Vinjamuri, & Kahn, 2016) empirically-based model of barriers and facilitators to implementing an EBP in student fieldwork placements, this study addressed the question: What were the approaches of faculty, students, and field instructors to implementing SBIRT that impeded or promoted model fidelity? For the purpose of this project we defined “approaches" as the combination of self-reported perspectives, attitudes, and actions around SBIRT implementation. Background In 2005, Fixsen, Naoom, Blase, Friedman, and Wallace found “the science related to implementing EBPs and programs with fidelity and good outcomes for consumers lags far behind the development of them” (p. vi). Arguably, this remains the case in social work today (Cabassa, 2016). Aarons, Hurlburt, and Horwitz (2011) wrote about implementation science as a quickly growing discipline with lessons learned from business and medical settings being applied in social service settings. However, they cautioned that it is unclear how well results from other types of organizations translate to settings with different historical origins and customs, such as public mental health, social service, and substance misuse sectors. Within social work there are no agreed-upon standards or steps to implementation or universal definition of what the science of implementation involves (Atkins & Frederico, 2017). Varying depictions of implementation science have emerged in the literature. Palinkas, He, Choy-Brown, and Hertel (2017) defined implementation science as the “generation and application of models and conceptual frameworks that identify potential barriers, facilitators, the process, and outcomes of program, practice, and policy implementation” (p. 182). Implementation of a new EBP can happen at various levels from the “paper” level with new policies and procedures; to the “process” level with trainings, supervision, and different reporting forms; and through the “performance” level with real, functional changes to operational impact with good effects for clients (Fixsen et al., 2005). However, implementation must be distinguished from adoption, which is merely a decision to use an evidence-based intervention (Mitchell, 2011). Implementation aims to achieve regular use of evidence-based interventions. Sustaining the practice is key. Delivering complex social interventions requires a comprehensive implementation strategy, including specific actions within a planned, long-term implementation and maintenance process (Mildona & Shlonsky, 2011). Thus, in real-world settings, stages of implementation and maintenance are not necessarily linear but are, rather, dynamic. Key Features and Impediments of Successful Implementation Implementation and model fidelity can be promoted successfully. Existing lists of factors needed for successful implementation are invariably lengthy and complex, Vinjamuri et al./IDENTIFYING HOW SOCIAL WORKERS’ 1220 suggesting implementation requires more than training practitioners (Aarons et al., 2011; Bhattacharyya, Reeves, & Zwarenstein, 2009; Fixsen et al., 2005; Kerner & Hall, 2009; Mildona & Shlonsky, 2011; Mitchell, 2011; Palinkas et al., 2017). Nuanced, dynamic interactions affect implementation, and ideal circumstances are not always possible to attain. Implementation can be impeded at any stage of the process, with impediments coming from individual practitioners or agency and organizational factors. Using a multi-systemic lens, Palinkas et al. (2017) identified several barriers to implementation, including the limited time and resources of practitioners, insufficient training, lack of access to peer-reviewed research journals, lack of feedback and incentives for use of EBPs, assumptions behind the design of research trials, and inadequate infrastructure and systems to support implementation. Focusing on organizational factors, Mitchell (2011) cited culture, climate, structure, mission, and philosophy of the organization plus leadership and network connectedness. Agencies may also struggle to provide training and supervision, incentives for practitioners, material resources, and administrative support. In terms of practitioner-specific barriers, Gray and Schubert (2012) described resistance to change, especially when new ideas are inconsistent with organizational beliefs, and Mitchell (2011) identified the attitudes of providers, such as skepticism about the clinical value of EBPs. Acceptance of a new protocol can be particularly challenging if it was developed for and tested with client populations with relatively simple problems, homogeneous groups, or when perceptions of inconsistencies between protocol and recognized characteristics of effective programs appear. Real or perceived mismatches to client populations and their complex needs can also impede implementation. Practitioners may also be resistant if they are concerned about clinical freedom and the ability to respond to individual client needs. While these factors have not all been exhaustively or empirically examined, they serve as a guide for this study. To date, data-derived, specific measurement instruments to guide the process of implementing and evaluating the implementation of an evidence-based protocol are lacking. The current study aimed to add information about the implementation process, focusing upon factors that impede or promote model fidelity as a protocol is implemented. An SBIRT-Based Implementation Model Previously, the authors (Ogden et al., 2016) used focus group data to develop a model (herein referred to as “the Implementation Model”) that identified barriers and facilitators to implementing SBIRT in social work student fieldwork placements (see Figure 1). Development of the Implementation Model revealed issues related to model fidelity. Subsequent focus groups affirmed further model fidelity issues, which became the focus of the present study. The following is a basic overview of the Implementation Model, which is provided to contextualize this inquiry and findings. ADVANCES IN SOCIAL WORK, Fall 2018, 18(4) 1221 (Ogden et al., 2016) ADVANCES IN SOCIAL WORK, Fall 2018, 18(4) 1222 The Implementation Model (Ogden et al., 2016) identified multiple interacting factors influencing students, faculty members, and fieldwork instructors. These three stakeholder groups can also be considered representative of practitioners, trainers, and supervisors, respectively. All of the identified factors can serve as either barriers or facilitators to implementing an evidence-based practice. The factors in the Implementation Model include intrapersonal characteristics, as well as mezzo- and macro-level factors. Intrapersonal characteristics of receptivity, comfort, and flexibility affect an individual’s approach to new knowledge. Also at the individual level are the practitioner’s perception of the fit between the protocol and professional values, responsibilities, and practices plus one’s sense of mastery of the material, authenticity in implementing it, and confidence in doing so. Mezzo- and macro-level factors that can affect implementation include agency factors, such as the population served, the agency’s theoretical orientation, established approaches to treatment, consequences of the protocol for clients, bureaucratic processes, agency flexibility, agency support, and supervisor support. The Implementation Model shows how these factors work in dynamic ways, often compounding or counteracting each other and leading to an undetermined net effect. How the factors connect to model fidelity is of particular concern to the present study because data used to develop the Implementation Model suggested wide variance. The current study is a first step towards understanding the dynamic interplay between factors that promote the use of SBIRT and those that contribute to model fidelity. To date, this implementation model appears to be the only one grounded in SBIRT-implementation data directly connecting social work education to implementation and practice. The implementation model provided the central analytic frame but was also critically examined throughout the analytic process. Design The present study used a train-the-trainer model. Faculty with expertise in the Screening, Brief-Intervention, and Referral to Treatment (SBIRT) protocol used materials provided by the Substance Abuse and Mental Health Services Administration (SAMHSA, 2011) to train all social work faculty in an urban college in the Northeastern United States. The nine hours of training included a specially created SBIRT manual, role-plays, and videos as well as dialectic and lecture-based instruction. In turn, those faculty then trained all graduating bachelor- and master-level social work students to use SBIRT. Student training occurred over several weeks during required core curricula courses. The training was the same for undergraduate and graduate students. The students received at least four hours of training in the classroom, which included lecture, discussion, role-play, and videos. All students received the SBIRT manual, which they could bring to their fieldwork placements, plus visual aids, which are an integral part of the SBIRT implementation process. Training was completed by the end of the first semester, and students were charged with using SBIRT in their fieldwork agencies for one semester. Fieldwork instructors were charged with supervising students’ use of SBIRT. All fieldwork instructors received the same basic information about SBIRT: Through emails and telephone conversations, faculty advisors and field education staff informed fieldwork instructors that students were receiving SBIRT training and were expected to use it in field. Fieldwork instructors also received an electronic version of the SBIRT manual and were ADVANCES IN SOCIAL WORK, Fall 2018, 18(4) 1223 asked to review it. The social work faculty advisors, who had been trained in SBIRT and were teaching it to the students, conducted on-site visits with every fieldwork instructor and student dyad. At those meetings, the faculty advisors provided basic introduction to the SBIRT protocol, answered questions, and told the fieldwork instructors what to look for when supervising SBIRT. For additional training, fieldwork instructors were invited to participate in voluntary three-hour trainings at the college provided by project faculty. Furthermore, the social work department offered voluntary advanced SBIRT training and specialized training for SBIRT with older adults. Although not all fieldwork instructors attended the additional voluntary trainings, all received the SBIRT manual and engaged in in-person discussions with trained faculty members about SBIRT and its role in their agencies. Some fieldwork instructors did not allow their students to use SBIRT in their fieldwork placements. At the end of each academic year of the study, faculty, students, and fieldwork instructors participated in focus groups that examined the factors that impeded or promoted the implementation of SBIRT. Methods Sample and Recruitment This article reports on second and third year focus group evaluation data of a three- year SBIRT training grant. The training model is described more thoroughly elsewhere (Ogden et al., 2016). The present study examined data from end-of-year focus groups with students, faculty, and fieldwork instructors, with one of each group held in 2015 and 2016, totaling six focus groups. Student recruitment. For student recruitment, all faculty members teaching SBIRT used a script to ask their SBIRT-infused course sections for one volunteer to participate in the focus groups, making clear that participation would have no effect on their grades or academic standing. Undergraduate- and masters-level social work students were recruited for the same focus groups. In Year 2, there were 222 students (90 MSW and 132 BA) in total with 15–25 students per SBIRT-infused course section and 12 sections. In Year 3, there were 226 (73 MSW and 153 BA) students in total with 15–25 students per section and 12 sections. In the project’s second year, there were 12 students (6 MSW and 6 BA) in the student focus group. In the third year, there were 13 students (8 MSW and 5 BA). Students were provided with a $25 gift card incentive for participating. Faculty recruitment. All faculty teaching SBIRT were asked to join the focus group with the exception of the three who had designed and facilitated the SBIRT training in order to decrease social desirability bias in focus group discussions. As compensation for participating in the project, faculty received two weeks of summer salary. This compensation was provided to pay for their time attending SBIRT trainings and retreats, as well as the additional work of learning and incorporating new material into their courses. There were nine social work faculty participants in the second-year faculty focus group and nine participants in the third year. Fieldwork instructor recruitment. All current fieldwork instructors (186 in Year 2 and 192 in Year 3) were sent an e-mail requesting their voluntary participation in a focus Vinjamuri et al./IDENTIFYING HOW SOCIAL WORKERS’ 1224 group. Six participated in the second-year focus group. Of these participants, two received the additional voluntary SBIRT training. Fifteen fieldwork instructors participated in the third year focus group. Of these participants, ten received the additional voluntary SBIRT training. Fieldwork instructors who participated in the focus group received a $25 gift card. Research Participant Protections Several steps promoted protection of research participants. Focus group participation was voluntary for all participants. Because providing detailed demographics could have easily led to identification of the participants, particularly in the faculty focus group that drew from a very small population, demographic information was not collected from any focus group participants. Confidentiality was maintained throughout focus group participation, recording, and transcription. All participants provided informed consent to be audio-recorded. During focus groups, members received random numbers as identifiers, which ensured anonymity in the discussion and recording. Identifying information regarding individual interviewees was, therefore, not available in transcripts, which promoted the integrity of data analysis as well as protected confidentiality. Focus group audio recordings were professionally transcribed. All study data were stored on password- protected computers in locked offices. Institutional Review Board approval was granted by the host college. Focus Group Procedure The focus groups were held each year within one month of completion of the academic year that included the SBIRT curriculum. Focus groups lasted 60-90 minutes with moderation provided by faculty involved in the SBIRT project evaluation team. All six groups involved in-depth discussions guided by moderators using semi-structured, open- ended interview questions informed by the Implementation Model. The discussions addressed topics such as participants’ experiences learning, using, teaching, and/or supervising SBIRT; difficulties and barriers encountered in implementing SBIRT; the role of the agencies and academic institutions in implementation; the impact of the protocol on perceptions towards people who use substances; and recommendations for improving the training. Data Analysis Verbatim transcripts of the six focus groups were the source of data for the present study. Analysis involved thematic coding of the transcripts, guided by the hybrid process of inductive and deductive thematic analysis described by Fereday and Muir-Cochrane (2006). A codebook was developed using the authors’ Implementation Model, which originated from focus group analysis of the same project’s first year (Ogden et al., 2016). Using this codebook, each author reviewed transcripts identifying the presence of factors from the Implementation Model while remaining open to new codes that expanded on existing concepts or illustrated new phenomena or processes. The rigor of the study was enhanced by the input of multiple researchers throughout the analytical process. Each focus group was coded by at least two authors. Creation of audit trails delineated clear pathways ADVANCES IN SOCIAL WORK, Fall 2018, 18(4) 1225 from codes to analytical and process memos. Analytical memos included examples of existing codes and new codes that emerged, their definitions and illustrative quotes, and the story of the relationships among the codes. The relationships between new codes and existing codes were discussed amongst the authors and documented in analytic and process memos. Constant comparison was the central analytical process leading to the organization of initial codes into categories of inductively-identified themes. Transcript data were repeatedly analyzed, each time using codes that emerged in the previous stage of the analysis as a temporary conceptual framework, while the authors remained open to discarding and creating new concepts and categories. The authors engaged in reflexive discussion wherein ideas and assumptions were considered, challenged, and debated, ensuring that individual researcher biases and opinions were moderated so that the emerging analysis was grounded in the data. Results Analysis of focus group data from six focus groups, collected over a two-year period, led to an in-depth understanding of approaches that supported or presented barriers to SBIRT implementation and model fidelity from the perspective of three groups of stakeholders. These approaches fell into three categories: approaches that impeded implementation and model fidelity; approaches that supported implementation but were not congruent with model fidelity; and approaches that supported both implementation and model fidelity. While some approaches appear to be transferable to the implementation of EBPs more generally, some appear to be specific to SBIRT and/or the SBIRT project from which the data were drawn. Approaches that Impeded Implementation and Model Fidelity Approaches that impeded implementation and model fidelity of the SBIRT protocol had two common elements. First, some participants expressed viewing the universal screening principle as optional, which immediately meant that model fidelity was not followed since universal screening is a key element of faithful implementation of SBIRT. Thus, lack of universal screening was a primary indicator of model infidelity. Second, participants identified barriers to fully using SBIRT in fieldwork agencies related to the degree to which SBIRT fit with existing policies, practices, and systems and the level of support provided by fieldwork instructors. Viewing the universal screening principle as optional. The SBIRT for substance use protocol begins with the premise that almost anybody can be misusing substances, including alcohol, and, therefore, the brief screening for substance use should be used with every client. This is central to the model’s public health approach of targeting individuals without severe substance use disorders who would benefit from reducing their substance use (SAMHSA, 2011). However, across all stakeholder groups, data emerged delineating times when the students did not use the universal screening principle. Several factors contributed to this breach in model fidelity. Fieldwork instructors sometimes viewed asking about substance use as inappropriate for their client populations and therefore, did not permit the universal screening. Other Vinjamuri et al./IDENTIFYING HOW SOCIAL WORKERS’ 1226 times, they chose clients within their agencies who they thought would be appropriate. For example, one fieldwork instructor stated: My student was very excited to use the SBIRT but she had three components of her field placement, and it was only appropriate to be used in one of the components because she was in one component where they had home visits where they had a questionnaire where they asked questions about drug and alcohol use so it’s appropriate in that one. But the other two components, one was interviewing parents for a Head Start program and the other was interviewing parents for a parenting program. And I had requested that for those two parts that she not use it. It is unclear how this fieldwork instructor decided that the screening was acceptable in some circumstances and not in others, as substance use can certainly be a contributing factor to problems in parenting. However, this approach seems to have been typical, as several field instructors identified populations, or sub-populations, for whom they felt using SBIRT was inappropriate. Therefore, model fidelity was lacking in those cases as the result of ideas about who should or should not be screened for substance misuse. Stereotypes about substance users contributed to the belief amongst fieldwork instructors and students that some clients should not be screened for substance misuse. As in the above example where the fieldwork instructor believed some parents did not need to be screened, stereotypes included beliefs about what substance users looked like, how they behaved, and whom one should screen for substance misuse, as well as what non-substance users looked like. For example, one student explained not using SBIRT because, “I work with older adults, so, sixty and older. I pretty much didn’t encounter any client who would currently be using alcohol.” Because of the clients’ older ages and the services provided, that student did not even ask if the clients ever drank alcohol. Some faculty accepted students’ perceived inabilities to implement SBIRT, colluding with and enforcing student preconceived notions about substance users and where substance use screening can occur. Other faculty did address students’ preconceived notions, but these could be so entrenched that even faculty feedback did not change students’ fixed beliefs or affect their client interactions: I had a handful of students who seemed to not want to use it in the field, one who didn’t even sign the agreements and was absolutely not interested and said, “We’re not comfortable using it with kids.” . . . I consistently pointed out, having had experience with kids, that they start engaging in these behaviors a lot younger than we think, so that bringing it up in a way that they can understand . . . in a way that’s comfortable, is important, because it needs to be addressed. Some faculty did successfully address students’ preconceived notions, and their students developed an understanding of the reality of not being able to “tell” if someone used substances. Those students then employed the universal screening protocol. In other words, their stereotypes were successfully challenged: My professor, she always told us, “No matter what population you’re in, it’s important to ask this question.” . . . Because you never know, what—what the ADVANCES IN SOCIAL WORK, Fall 2018, 18(4) 1227 client’s been through or—or if they’ve used that and how that could’ve affected them . . . We have to know everything about the person . . . From this training, I’ll know that [substance use screening] is something important and no matter where I’m at, I should ask those questions ‘cause it could be beneficial. Encountering and responding to agency barriers. Practices are not introduced in a vacuum; rather they are introduced in the context of existing practices, policies, and systems. As in the present study, it is often left to individual practitioners to determine how a new practice will or will not fit in. Fieldwork instructors were particularly conscious of contextual factors and how they affected staff buy-in to new EBPs. For example, some described intake forms they were required to use and that could not be changed easily. Others were positioned to respond to such barriers more actively: We have a form that was part of the intake process which I intend to revamp and substitute it with . . . the SBIRT [forms] because I find it more humanistic and that’s where I am. And that’s probably going to take place this summer, with my social work staff, because I want them all to buy into it. . . . The substance abuse counselors are not using the model which is very upsetting for me. Being positioned to address agency barriers and encourage buy-in was an important factor affecting implementation. The fieldwork settings also influenced students’ experiences using SBIRT due to the level of support fieldwork instructors provided. A student described this phenomenon: My supervisor didn’t know SBIRT . . . so that was one problem and, you know, many of our [client] interactions were very quick and . . . and the only question they want you to ask is if . . . they have a current problem or if they have a history of alcohol abuse or substance abuse. Students often faced the problem of knowing more about the intervention than their fieldwork instructors. This may represent a common barrier: If a practitioner receives training and wants to bring it to an agency, that practitioner would need to educate both superiors and peers, possibly creating complicated workplace dynamics. Agency contexts affected the messages students received about the fit of SBIRT with their work, which may have been contrary to the student’s perceptions and plans. For example, one faculty member described students who were eager to implement SBIRT but who were told by their fieldwork instructors and/or other agency personnel that SBIRT did not fit with the agency’s work: Some of [the students] had very short-term interactions with clients and so then if you're only going to talk to somebody once then you're not going to do a substance abuse screening . . . that’s what they were told . . . The response I would say was disappointing overall, but that’s not the student’s fault. It’s not because they weren’t enthusiastic . . . it was just, [SBIRT] didn’t fit. While the SBIRT model posits that a brief, single time interaction is in fact an ideal place for conversations about substance use, beliefs and stereotypes about substance use, and, Vinjamuri et al./IDENTIFYING HOW SOCIAL WORKERS’ 1228 relatedly, perceptions about what substance use screening and intervention is and where they fit inhibited their use in practice. In some cases, existing protocols impeded implementation by portraying a new practice as burdensome. One fieldwork instructor said: I think the biggest barrier though is the fact . . . that the agencies usually have their other tools that [they are] using . . . like the assessment package is so huge that to incorporate something else kind of seemed like a burden. Perception of burdensomeness of an intervention by fieldwork instructors and other agency personnel thus emerged as a significant barrier to implementation. Approaches Congruent with Implementation but Not Model Fidelity Two themes emerged that supported implementation of SBIRT but appeared to compromise model fidelity. The first theme is described as “finding a fit.” Finding a fit between SBIRT and existing practice allowed components of SBIRT to be implemented; however, it likely compromised model fidelity because pieces of the protocol were altered or omitted so SBIRT would fit better with usual practice. As a result, participants considered their practice to be consistent with SBIRT, but they were not using SBIRT in ways true to the evidence-based protocol. In line with the Implementation Model, flexibility that promotes implementation is desirable; however, straying from what the research indicates is effective to engage in selective or modified activities might be just as ineffective as not using any elements of an established protocol. The second theme, “being thrown off by client reactions,” concerns reactions to clients who responded to SBIRT in a negative way. Client reactions could move a student from using SBIRT as taught to ending the protocol prematurely or altering the intervention, which in both cases compromises model fidelity. Finding a fit. Some participants showed receptivity and flexibility, searching for a fit between SBIRT and practice-as-usual while demonstrating their desire to use SBIRT. However, it became clear from the data that some participants were picking and choosing aspects of SBIRT based on what worked with their existing practices. Furthermore, as participants grew more comfortable with the material, they wanted to make it their own, fitting it into their existing practice approaches so that it felt authentic. However, this approach may have compromised model fidelity. For faculty, finding a fit meant adding in their own exercises as they were teaching. Although this was encouraged by the project directors to improve faculty buy-in, it also meant that a single model for teaching was not followed. As one faculty member shared: I got more comfortable with the material and the format of it that I too wanted to do more with it. So, I tried doing some role plays toward the end but I found that it was hard to balance whether or not I was going too much outside of the evidence- based model versus being very prescriptive in terms of what is supposed to be there . . . But as I got more familiar with it and felt more comfortable, I wanted to do more interactive activities. ADVANCES IN SOCIAL WORK, Fall 2018, 18(4) 1229 Significantly, faculty expressed adding to material but never described skipping any pieces of it, and, aware of the issue of model fidelity, appeared to have adhered to a level of model fidelity despite instructional innovations. Fieldwork instructor approaches to SBIRT emphasizing authenticity were more problematic from a model fidelity perspective. One field instructor described, “You don’t get everything, but you take away the most important points. What’s important to you.” This type of approach was common among fieldwork instructors, whose discussions never addressed model fidelity or the importance of “getting everything,” which is central to model fidelity. A “take away the most important points” approach inherently compromises model fidelity. Students described a similar “take away the most important points” approach to SBIRT: I use SBIRT in a very loose, unstructured way; I used it in a way that just suited me based on the placement that I was at; on top of it, my placement did not want to turn the school into a drug rehab center ‘cause like all the kids were like smokin’ weed and drinkin’ on the weekend. While the student seemed to be implementing SBIRT, it was not in a way that was faithful to the full SBIRT protocol. Participants used their preconceived notions to pick and choose not just “if” they would use it but also when and how to do so. Being thrown off by client reactions. The SBIRT training delivered as part of the present study could not cover all material necessary to respond to clients in real life situations, including the more complicated emotional responses many people will have in discussing how substances are affecting their lives. As a result, not knowing how to handle client reactions was a barrier to fully implementing SBIRT, and evidence suggests that it led to a breakdown in model fidelity. One fieldwork instructor described a student’s use of SBIRT as follows: She did the SBIRT with that person and the person wasn’t a drug or alcohol user, but apparently there was a family member that was a drug and alcohol user, and the minute she brought up the topic she wasn’t able to continue with the rest of the interview . . . because the parent just fell apart and started talking about this family member and all these issues. And then my student felt like she wasn’t prepared, like, how to react to that happening. Students who got “stuck,” not knowing how to respond, often did not fully implement the protocol. For example, one student working with high school students explained, “I didn’t know where to go when someone says, “I don’t have a problem with my cocaine; I don’t— I don’t care to change it . . .” Another student experienced clients who became “aggravated” or “angered” and wanted additional training on how to “move on” in those situations. Limits of training is another aspect of implementation not identified in the Implementation Model but that was clearly significant. Vinjamuri et al./IDENTIFYING HOW SOCIAL WORKERS’ 1230 Approaches Congruent with Implementation and Model Fidelity Several themes emerged from participants who implemented SBIRT in ways that were identified as likely to promote model fidelity. These themes were: repeated learning, having the right tools, and a desire to uphold social work values. Repeated learning. Participants who experienced repetitive exposure to SBIRT material seemed to have a deeper understanding of the material, which in turn facilitated their implementation of the protocol with a greater likelihood of maintaining model fidelity. Among the faculty, this shared attitude is encapsulated in one of the participant’s comments that “the more frequently you do it, the easier it becomes and you can easily weave it in” and another who described being able “to teach it and help the students to learn it in just a more knowledgeable way” by the third year. Students were also aware of the importance of repeated learning and expressed the importance that material “was reiterated a bunch of times by my professor.” Another student observed that faculty used various pedagogical strategies to repeatedly convey content “because he wanted us to get a concrete foundation of this training, so when we go out in the field we can use it and implement it in the right way.” Fieldwork placements were another venue for repeated learning that provided opportunities for practice, which was appreciated by students. As one student stated, “The more you use it, the more you become familiar and be able to apply it fully.” In contrast to the theme of “being thrown off by client responses,” the repeated learning added confidence to implement the material and opportunities for students to have the experience to “actually practice it in real life” and, when clients responded, to discover, “Oh—it really works!” Students who saw field as an opportunity for repeated practice gained confidence in fully implementing the protocol. Having and using the right concrete tools. Faculty approaches that promoted implementation and model fidelity included using a wide variety of training tools. These included concrete tools such as videos, role-plays, and feedback on process recordings. Specific and pointed feedback on process recordings was particularly important to students: The professor did use the process recordings . . . to help us and tell us where – to give us pointers to where we could ask some questions and why, and . . . to just give us insight on how we could have did something better, or add to it. This pointed feedback was likely central to model fidelity as specific correctives could be made. Active teaching and learning also happened in interactions between faculty and fieldwork instructors: In my [agency] visits a couple of times I actually brought out that laminated [visual aid] and sort of walked [the field instructors] through it, which definitely seemed to be helpful. They felt as overwhelmed, the ones who had not been exposed to it before, felt as overwhelmed I thought as our most overwhelmed students did . . . So when I broke it down for them in the visit, also it was face-to-face in that initial ADVANCES IN SOCIAL WORK, Fall 2018, 18(4) 1231 visit, that seemed to help out. And a lot of times then they were like “oh, okay, so this is kind of what we do already.” Faculty members also saw that having concrete tools increased student confidence and considered that the concreteness of the tools and specificity of the practice might contribute to model fidelity and certainly to implementation: I felt that the students had a vocabulary, they had words to use to make the assessment, because I think that they want words to use for everything. They want to know what do I say in this situation, that situation, and every situation, so this was something that gave them some words to use which was helpful. And we talked about how you use the words and it’s not just the words themselves, but it’s how the engagement and the warm handoff and all that, but it was a process, it was very concrete steps and visual aids, a lot of things. For students, having and using concrete tools, such as the SBIRT manual and visual aids that were developed for this specific SBIRT project and the reliable and validated substance use disorder measures, promoted comfort and confidence and contributed to an affinity towards using a universal screening approach. One student said, “So every time I had to use it, I would actually go to the [SBIRT manual] and make sure I know where to go if we continue the process.” Approaching the practice in close consultation with the SBIRT manual allowed students to feel they had a “safety net” that increased a sense of “I know what I’m talking about” and “helped me feel more confident in speaking with clients and I was able to refer back to a secure resource.” Using the SBIRT manual in this way is consistent with model fidelity, as students stayed close to practice taught in the classroom. One student described, “In class when we got all the charts, it was much easier to bring that up . . . knowing that you have that information.” That student also appreciated the SBIRT manual’s “listing of drugs and possible outcomes of over-usage and things like that. So I’m not so familiar with those types of drugs but by having that, my knowledge just got extended and now, talking to someone who is using that substance, I’m more informed.” By contrast, not having the right tools can lead to lack of implementation and lack of model fidelity as illustrated by one student who “lost the paper” s/he was using as a cue during the SBIRT interviews and “after that . . . there was no more structure.” The student understood model fidelity was lost, even without using that term. Fieldwork instructors sometimes approached the tools, particularly the SBIRT manual and standardized assessments, as “a script” or a “formula” from which to practice and used it to help focus on the details of the protocol. Those who saw the benefit of paying attention to the details saw the connection between the details and the overall philosophy of the practice and its fit within social work values. Desire to uphold social work values. Among all participants there was a desire to uphold the core social work value of respecting the dignity and worth of the person. This value prevailed over concerns about model fidelity. However, many participants identified motivational interviewing, which underlies SBIRT, as being in line with this core social work value. For those participants, model fidelity was not just about adhering to concrete steps in practice but adhering to the spirit of practice. As one faculty member described: Vinjamuri et al./IDENTIFYING HOW SOCIAL WORKERS’ 1232 My approach really is about how do we not judge and not stigmatize, and I like . . . talking about how do we challenge our own assumptions about people. So that way it fit… my professional identity . . . that we’re a profession that’s about working with marginalized communities . . . The SBIRT approach is very much about how . . . you help a group that’s already feeling marginalized feel less so. Similarly, some students articulated that fit with social work values was central to their implementation of the protocol: “The opportunity for a client to make choices is very, very, very important. But not the social worker or the social work intern making a choice for the client. That is really a big, big difference.” Another student said: I think it gave me a better perspective as far as treating the person as a whole. You know, not just the mental health issues, not just the substance abuse issues, not just the environmental issues, just as, the person as a whole. And, you know, helping develop a complete plan, treatment plan. I think it’s helped me for that. Seeing a fit with professional values was thus in line with implementation and model fidelity. Discussion This article builds on the Implementation Model created by the authors (Ogden et al., 2016) and helps develop urgently needed implementation knowledge and theory by examining processes through which an evidence-based protocol reaches clients and whether it does so with model fidelity. Our findings confirmed, added, and developed several impeding and promoting factors already theorized in implementation literature. The Implementation Model provided a useful framework to find points of intervention to promote model fidelity during the implementation process. Findings of the present study were focused on SBIRT and implementation that moves a practice directly from an educational setting to the field; however, the implications are transferable to other evidence-based protocols. Approaches that impeded both implementation of SBIRT and model fidelity to the protocol were viewing the universal screening principle as optional and the ways in which participants encountered and responded to agency barriers. Specifically, practitioners were affected by agency-level buy-in, level of knowledge and training, and perceptions about the degree of fit between the EBP and agency clients and services. One striking feature of the focus group data was the lack of consciousness of model fidelity as an issue across all three groups. No participants explicitly discussed model fidelity; rather, problems with model fidelity were detected through critical data analysis. This may be a central problem with implementation of EBPs into social work practice: Social workers “take away the most important points” as they subjectively see them, rather than adapting new material with model fidelity in mind. As such, practitioners and social work faculty alike need training in identifying barriers to model fidelity that will also raise consciousness of this aspect of implementation. Additionally, when agencies implement new protocols and when practitioners are trained, the importance of model fidelity should be explicitly addressed with those charged with applying the protocols with the clients. Raising consciousness and awareness of how one’s practices may or may not promote model fidelity moves ADVANCES IN SOCIAL WORK, Fall 2018, 18(4) 1233 practitioners from merely deciding to use an EBP to critically evaluating how one is using it and client outcomes (Mitchell, 2011). Challenges experienced by students as they responded to the uncomfortable situation of trying a new practice, and sometimes to negative client responses, suggests a need for training to help students and other practitioners to use a new EBP practice as given, with particular attention paid to the discomfort that may arise. Helping practitioners understand and grow from discomfort that comes with practices can address practitioner-specific attitudinal barriers towards implementation, such as resistance to change (Gray & Schubert, 2012) and skepticism about a new practice’s clinical value (Bellamy, Bledsoe, & Traube, 2006; Mitchell, 2011). Given that consciousness of ethical issues is a central element of evidence-based practice in social work (Gambrill, 2007), finding fit between new practices and values, and simultaneously striving for model fidelity, is an integral part of providing ethical services to clients. Highlighting the relationship between scientific practice and ethical practice is key. Results of this study shed further light on how agencies and practitioners can move from “process” to “performance” levels of implementing new EBPs (Fixsen et al., 2005) by revealing elements central to the faithful implementation of a new practice. These include repeated exposure to new material and opportunities to engage with it, having specific and concrete tools that remind practitioners of a new practice and support them in its use, and validating and supporting learners to uphold social work values in their new practice. Thus, passive learning is not likely to lead to faithful implementation; conversely, active and repeated learning likely supports implementation with good model fidelity. These findings are consistent with what adult learning principles identify as key elements of integrating new knowledge: finding applicability and relevance, being co-authors in one’s learning, and engaging in active problem-solving (Knowles, 1980, 1984; Plack et al., 2007). Additional and ongoing training closely tying new practices to social work curricula and existing agency practices might help students feel more comfortable in fully and faithfully using a new practice, while honoring clients’ responses to the new practice. Consistent with previous recommendations (Fixsen et al., 2005), our findings suggest the central importance of skillful and timely supervision and coaching throughout the implementation process and add the need to focus particularly and explicitly on approaches to model fidelity. The current study, both in design and results, illustrates the importance of understanding stakeholders’ needs and perspectives and of providing open channels of communication to create and sustain successful implementation (Mildona & Shlonsky, 2011). Finally, this study highlights a larger workforce issue: Fieldwork instruction is a voluntary activity. Mandatory training is thus infeasible. Prior to the students’ expected use of the SBIRT protocol, all of the fieldwork instructors in this project received information about SBIRT in the form of written materials and conversations with faculty. However, not all fieldwork instructors attended the additional voluntary trainings that could have increased their knowledge, competence, and commitment to SBIRT. Invariably, this leads to inconsistencies in implementation of evidence-based practices and supervision of students. This aspect of our social work professional pedagogy, integral to how the Vinjamuri et al./IDENTIFYING HOW SOCIAL WORKERS’ 1234 profession trains future professionals, can stand in conflict with faithful implementation of an evidence-based protocol when using this type of design. Further examination on how to resolve this conflict is warranted. Limitations and Strengths of the Study The current study has several limitations. Given that not all fieldwork instructors participated in additional SBIRT training beyond the basic introduction, knowledge of SBIRT and supervision around SBIRT was inconsistent. Some fieldwork instructors could have had up to nine additional hours of training, while others may have merely reviewed the manual and had a single conversation about it. There is likely some self-selection bias in the results, given that those who volunteered for the focus groups may be those who had strong responses towards SBIRT, either positive or negative. Focus groups in general present other limitations: While providing detailed information elicited through group interactions and participant sharing, social desirability bias likely hinders comments that sway too far from any particular group’s norm (Hollander, 2004). Finally, given the scope of this study, perceptions of SBIRT were not elicited directly from clients, who are the fourth key stakeholder in the implementation process. One strength of this study is its use of triangulation in data collection and data analysis: Data were collected from three groups of stakeholders, and the three authors engaged in a rigorous multi-stage constant comparison coding and analysis process to determine the key themes, which supported the integrity of the analysis and transferability of the findings. Finally, the study incorporated both inductive and deductive processes of generating knowledge, using an existing model grounded in data about SBIRT implementation, which enhanced the richness and trustworthiness of the findings. While using an existing model developed by the researchers as an analytical lens to provide sensitizing concepts, the authors also generated themes that added depth of understanding for how barriers and facilitators interact to promote and/or impede effective implementation of an evidence- based practice. Implications for Future Research Clearly defined model fidelity measures may help further determine the existence, source, and extent of factors that impede or promote model fidelity. Awareness of model fidelity needs to be improved in order to help practitioners, especially students and fieldwork instructors, to think about model fidelity as they learn and apply new practices. Combining model fidelity training with training in an EBP would be useful for implementation and deserves further research attention. With these considerations in mind, we recommend the development of a diagnostic tool to assess for individual- and organizational-level barriers and facilitators to both implementation and model fidelity. References Aarons, G. A., Hurlburt, M., & Horwitz, S. M. (2011). Advancing a conceptual model of evidence-based practice implementation in public service sectors. Administration and Policy in Mental Health, 38(1), 4-23. https://doi.org/10.1007/s10488-010-0327-7 https://doi.org/10.1007/s10488-010-0327-7 ADVANCES IN SOCIAL WORK, Fall 2018, 18(4) 1235 Acri, M., Hamovitch, E., Mini, M., Garay, E., Connolly, C., & McKay, M. (2017). Testing the 4Rs and 2Ss Multiple Family Group intervention: Study protocol for a randomized controlled trial. Trials, 18(588), 1-11. https://doi.org/10.1186/s13063- 017-2331-7 Atkins, P., & Frederico, M. (2017). Supporting implementation of innovative social work practice: What factors really matter? British Journal of Social Work, 47(6), 1723- 1744. https://doi.org/10.1093/bjsw/bcx091 Bellamy, J. L., Bledsoe, S. E., & Traube, D. E. (2006). The current state of evidence- based practice in social work: A review of the literature and qualitative analysis of expert interviews. Journal of Evidence-Based Social Work, 3(1), 23-48. https://doi.org/10.1300/J394v03n01_02 Bhattacharyya, O., Reeves, S., & Zwarenstein, M. (2009). What is implementation research? Rationale, concepts, and practices. Research on Social Work Practice, 19 (5), 491-502. https://doi.org/10.1177/1049731509335528 Cabassa, L. J. (2016). Implementation science: Why it matters for the future of social work. Journal of Social Work Education, 52(sup.1), S38-S50. https://doi.org/10.1080/10437797.2016.1174648 Fereday, J., & Muir-Cochrane, E. (2006). Demonstrating rigor using thematic analysis: A hybrid approach of inductive and deductive coding and theme development. International Journal of Qualitative Methods, 5(1), 80-92. https://doi.org/10.1177/160940690600500107 Fixsen, D. L., Naoom, S.F., Blase, K. A., Friedman, R. M., & Wallace, F. (2005). Implementation research: A synthesis of the literature. Tampa, FL: University of South Florida, Louis de la Parte Florida Mental Health Institute, The National Implementation Research Network (FMHI Publication #231). Retrieved form https://nirn.fpg.unc.edu/sites/nirn.fpg.unc.edu/files/resources/NIRN- MonographFull-01-2005.pdf Gambrill, E. (2007). Views of evidence-based practice: Social workers’ code of ethics and accreditation standards as guides for choice. Journal of Social Work Education, 43(3), 447-462. https://doi.org/10.5175/JSWE.2007.200600639 Gray, M., & Schubert, L. (2012). Sustainable social work: Modeling knowledge production, transfer, and evidence-based practice. International Journal of Social Welfare, 21(2), 203-214. https://doi.org/10.1111/j.1468-2397.2011.00802.x Hollander, J. A. (2004). The social contexts of focus groups. Journal of Contemporary Ethnography, 33(5), 602-637. https://doi.org/10.1177/0891241604266988 Kerner, J. F., & Hall, K. L. (2009). Research dissemination and diffusion: Translation within science and society. Research on Social Work Practice, 19(5), 519-530. https://doi.org/10.1177/1049731509335585 Knowles, M. (1980). The modern practice of adult education: Andragogy versus pedagogy. Englewood Cliffs, NJ: Cambridge Adult Education. https://doi.org/10.1186/s13063-017-2331-7 https://doi.org/10.1186/s13063-017-2331-7 https://doi.org/10.1093/bjsw/bcx091 https://doi.org/10.1300/J394v03n01_02 https://doi.org/10.1177/1049731509335528 https://doi.org/10.1080/10437797.2016.1174648 https://doi.org/10.1177/160940690600500107 https://nirn.fpg.unc.edu/sites/nirn.fpg.unc.edu/files/resources/NIRN-MonographFull-01-2005.pdf https://nirn.fpg.unc.edu/sites/nirn.fpg.unc.edu/files/resources/NIRN-MonographFull-01-2005.pdf https://doi.org/10.5175/JSWE.2007.200600639 https://doi.org/10.1111/j.1468-2397.2011.00802.x https://doi.org/10.1177/0891241604266988 https://doi.org/10.1177/1049731509335585 Vinjamuri et al./IDENTIFYING HOW SOCIAL WORKERS’ 1236 Knowles, M. (1984). Andragogy in action: Applying modern principles of adult learning. San Francisco, CA: Jossey-Bass. Mildona, R., & Shlonsky, A. (2011). Bridge over troubled water: Using implementation science to facilitate effective services in child welfare. Child Abuse & Neglect, 35, 753-756. https://doi.org/10.1016/j.chiabu.2011.07.001 Mitchell, P. M. (2011). Evidence-based practice in real-world services for young people with complex needs: New opportunities suggested by recent implementation science. Children and Youth Services Review, 33, 207-216. https://doi.org/10.1016/j.childyouth.2010.10.003 Mullen, E. J., Shlonsky, A., Bledsoe, S. E., & Bellamy, J. L. (2005). From concept to implementation: Challenges facing evidence-based social work. Evidence & Policy, 1(1), 61-84. https://doi.org/10.1332/1744264052703159 Ogden, L., Vinjamuri, M. K., & Kahn, J. (2016). A model for implementing an evidence- based practice in student fieldwork placements: Barriers and facilitators to the use of “SBIRT.” Journal of Social Service Research, 42(4), 425-441. doi: https://doi.org/10.1080/01488376.2016.1182097 Otto, H., Polutta, A., & Ziegler, H. (2009). Reflexive professionalism as a second generation of evidence-based practice: Some considerations on the special issue, “What works? Modernizing the knowledge-base of social work.” Research on Social Work Practice, 19(4), 472-278. https://doi.org/10.1177/1049731509333200 Palinkas, L. A., He, A. S., Choy-Brown, M., & Hertel, A. L. (2017). Operationalizing social work science through research–practice partnerships: Lessons from implementation science. Research on Social Work Practice, 27(2), 181-188. https://doi.org/10.1177/1049731516666329 Plack, M. M., Driscoll, M., Marquez, M., Cuppernull, L., Maring, J., & Greenberg, L. (2007). Assessing reflective writing on a pediatric clerkship by using a modified Bloom’s taxonomy. Ambulatory Pediatrics, 7(4), 285-291. https://doi.org/10.1016/j.ambp.2007.04.006 Substance Abuse and Mental Health Services Administration. (2011). White paper on Screening, Brief Intervention, and Referral to Treatment (SBIRT) in behavioral healthcare. Retrieved from www.samhsa.gov/sites/default/files/sbirtwhitepaper_0.pdf Author note: Mohan Vinjamuri, LMSW, PhD, Department of Social Work, Lehman College - City University of New York, 250 Bedford Park Blvd. West, Bronx, NY 10468. mohan.vinjamuri@lehman.cuny.edu https://doi.org/10.1016/j.chiabu.2011.07.001 https://doi.org/10.1016/j.childyouth.2010.10.003 https://doi.org/10.1332/1744264052703159 https://doi.org/10.1080/01488376.2016.1182097 https://doi.org/10.1177/1049731509333200 https://doi.org/10.1177/1049731516666329 https://doi.org/10.1016/j.ambp.2007.04.006 http://www.samhsa.gov/sites/default/files/sbirtwhitepaper_0.pdf