Implementation of a diabetes prevention program within two community sites: a qualitative assessment

Abstract

Background

Despite numerous translations of diabetes prevention programs, implementation evaluations are rarely conducted. The purpose of this study was to examine the implementation process and multilevel contextual factors as an evidence-based diabetes prevention program was implemented into two local community organization sites to inform future scale-up. To build the science of implementation, context and strategies must be identified and explored to understand their impact.

Methods

The program was a brief-counseling diet and exercise modification program for individuals at risk of developing type 2 diabetes. A 1-year collaborative planning process with a local not-for-profit community organization co-developed an implementation plan to translate the program. A pragmatic epistemology guided this research. Semi-structured interviews were conducted with staff who delivered the program (n = 8), and a focus group was completed with implementation support staff (n = 5) at both community sites. Interviews were transcribed verbatim and thematically analyzed using a template approach. The consolidated framework for implementation research (CFIR) is a well-researched multilevel implementation determinant framework and was used to guide the analysis of this study. Within the template approach, salient themes were first inductively identified, then identified themes were deductively linked to CFIR constructs.

Results

Implementation strategies used were appropriate, well-received, and promoted effective implementation. The implementation plan had an impact on multiple levels as several CFIR constructs were identified from all five domains of the framework: (a) process, (b) intervention characteristics, (c) outer setting, (d) inner setting, and (e) individual characteristics. Specifically, results revealed the collaborative 1-year planning process, program components and structure, level of support, and synergy between program and context were important factors in the implementation.

Conclusion

This study offers insights into the process of implementing a community-based diabetes prevention program in two local sites. Successful implementation benefited from a fully engaged, partnered approach to planning, and subsequently executing, an implementation effort. The CFIR was a useful and thorough framework to evaluate and identify multilevel contextual factors impacting implementation. Results can be used to inform future implementation and scale-up efforts.

Contributions to the literature

• Comprehensive reporting of implementation strategies, context, and processes that contribute to implementation are needed to advance the field and support future implementation and scale-up studies.

• Despite numerous translations of diabetes prevention programs, implementation evaluations describing the implementation process are rarely conducted and are needed to help advance the field.

• The Consolidated Framework for Implementation Research was a useful and thorough framework to guide data analysis

• As diabetes continues to rise, diabetes prevention programs are needed to meet the rising need. Fitness centers may be feasible community-based venues to deliver a sustainable individual, brief-counseling diabetes prevention program

Background

Worldwide, 837 million people are diagnosed with, or at risk for, type two diabetes (T2D), costing 10% of global health expenditure [1]. Individuals with T2D have increased risk for kidney disease, cardiovascular disease, eye disease, and foot and lower limb complications [1]. With the economic and health burdens of T2D, increased attention is needed to prevent diabetes. Diet and exercise modification programs have demonstrated success in reducing future risk of developing T2D by up to 80% [2, 3] and indicated greater long-term effectiveness than pharmaceutical interventions (5.7–9.4 years) [4]. However, these efficacy studies have been executed under intensive, tightly controlled experimental conditions and are not suitable to implement and sustain in real life settings [5]. To reach the large number of at-risk individuals, programs must be implemented in community settings in a sustainable and scalable way.

To date, there have been multiple translations of diabetes prevention programs (e.g., [6,7,8]) with demonstrated effectiveness (e.g., [9, 10]). Despite many countries developing large-scale translation efforts (e.g., the USA, India, Australia, see [11]), there has been a lack of reporting on the implementation process, strategies, and context. Further, there have been no large-scale translations within a Canadian context. Implementation, as defined by Greenhalg, is the “active and planned efforts to mainstream an innovation within an organization” ([12] p582). There is a dearth of research in this area. A meta-narrative review of 495 sources on diffusion of innovations in healthcare identified the most serious research gap as limited research examining the processes of implementing and maintaining innovation in health service delivery and organizations in specific contexts [12]. Understanding the process and strategies used across studies can improve the field of implementation science by identifying effective strategies to improve future implementation efforts.

The largest translation study to date, the U.S. National Diabetes Prevention Program, conducted an evaluation aimed to report from an implementation perspective [13]. Successful scale-up was demonstrated, and promising site-level and program-level features were documented to boost engagement [13]. However, this study did not provide in-depth information on how or why such features were implemented, lacked description of the scale-up process, did not include provider-level perspectives, and did not describe specific implementation strategies. Such information is essential for understanding mechanisms leading to successful implementation. The Kerala Diabetes Prevention Program is a translational study that was culturally adapted to Kerala, India [14], and a thorough implementation evaluation has been conducted [15]. The evaluation of the cultural adaptation reported several contextual details including the impact of community stakeholders as a success factor and the use of a local resource person. In addition, the implementation evaluation reported several important implementation markers (e.g., feasibility, acceptability) from multiple perspectives. However, the study did not report specific implementation strategies. The present research builds on previous research by conducting a process evaluation of implementing a diabetes prevention program within a community organization by specifying the implementation process, strategies, and context.

Greenhalg identified context and confounders as a paradox at the center of implementation efforts that must be studied to understand what leads to success or lack of success of an implementation effort [12]. For example, within the National Diabetes Prevention Program, only 25% of sites received full Centre for Disease Control recognition within the study period (implemented the program for 12 months and achieved program outcomes per the Standards and Operating Procedures, necessary to continue to receive funding) [13]. To increase the likelihood of future success, in-depth evaluations on implementation processes can help identify contextual features from sites where there was successful, or unsuccessful, implementation. Building on the work of Greenhalg, and 18 other theories, the Consolidated Framework for Implementation Research (CFIR) was developed as a multilevel framework to systematically assess the implementation context to understand, “what works, where, and why” ([16] p2). The CFIR framework established a list of factors likely to influence implementation through individual, program, and organizational levels resulting in 39 fully operationalized constructs organized into five domains (intervention, inner setting, outer setting, individuals, and process) [16]. The intervention domain includes eight constructs associated with characteristics of the intervention that may influence implementation (e.g., complexity). The inner setting domain includes 12 constructs associated with characteristics of the organization that is implementing the intervention (e.g., culture). The outer setting domain includes four constructs associated with characteristics of the broader environment in which the organization is situated (e.g., cosmopolitanism). The individual domain includes five constructs associated with characteristics of individuals involved with the implementation (e.g., self-efficacy). Finally, the process domain includes eight constructs associated with strategies used to implement the intervention within the organization (e.g., opinion leader). In the current study, CFIR was used to systematically identify relevant factors that contribute to the success (or lack of success) of implementation, at multiple levels of the implementation context.

The CFIR has been adopted and used widely. In 2015, a systematic review identified 429 unique articles referencing the framework; however, the authors, including the author of the CFIR framework, stated only 6% of studies used CFIR in a meaningful manner [17]. Meaningful was defined as “to guide data collection, measurement, coding, analysis, and/or reporting” ([17] p2). The review also provided recommendations to help researchers apply the CFIR. This current study applied two of the reviews’ recommendations through (1) justifying selection of CFIR constructs (see the “Methods” section) and (2) meaningfully using the CFIR across different phases of implementation. In this present study, meaningful use of the CFIR occurred by using a during-implementation evaluation to inform future program scale-up initiatives. To date, the CFIR has been used in a pre-implementation study for integrating community health workers in primary care for diabetes prevention [18], in a protocol evaluating the implementation of a Veterans Affairs diabetes prevention program [19] and in community-based diabetes self-management education classes [20]. To our knowledge, this research is the first to use the CFIR to examine a community-based diabetes prevention program.

The present study utilized a qualitative approach to evaluate the process of implementing a diabetes prevention program within two sites of a local community organization. The purpose of this research was to examine the multilevel factors that facilitated or impeded implementation to inform future scale-up. Understanding the process of implementation, as well as the key contextual factors as the program enters a new organizational context, will help explain the success (or lack of success) of implementation and provide useful information to inform future scale-up. CFIR represents an ideal fit to help elucidate important process factors as it is a comprehensive, multi-level implementation determinant framework with strong theoretical heritage and fully operationalized constructs. This study contributes to the field a thorough implementation evaluation, allowing for future studies to understand the implementation process, strategies and factors impacting implementation.

Methods

Program description and study setting

Small Steps for Big Changes (SSBC) is an evidenced-based diabetes prevention program with demonstrated success on health-related outcomes (e.g., improved cardiorespiratory fitness, moderate to vigorous physical activity) [21,22,23,24,25]. The program uses a client-centered approach to support adults (age 18+) at risk of developing T2D make diet and exercise modifications. The 3-week program set in Kelowna, British Columbia, includes six one-on-one sessions of brief-counseling and supervised exercise. The program has been previously described (see [26] and supplementary file A for full program details). Working with the community partner, a local not-for-profit community organization (YMCA of Okanagan)¹ housed within a broader national organization, a 1-year planning process facilitated co-development of a plan to sustainably translate the program into local community sites, with a goal to scale-up in the future [27]. The planning process engaged multiple stakeholders in face-to-face meetings, including a planning task force (consisting of researchers and upper-level YMCA leadership), site managers, prospective staff from the YMCA to enroll in the project, program alumni, and an external facilitator who guided two planning meetings (see [27] for further details). Throughout this process, both partners emphasized the importance of program fidelity, sufficient training, ongoing support, and open communication. These key factors were then targeted through the development of implementation strategies. In response to multiple calls for consistent language in the implementation field [16, 28], the strategies have been identified and described using the comprehensive list and definitions by Powell and colleagues (see Table 1) [28]. In doing so, this study will contribute to building an implementation evidence-base to support future implementation studies. The implementation plan was then executed to examine if SSBC can be successfully implemented within, and delivered by, YMCA staff. This current study examined the multilevel factors that facilitated or impeded the implementation process.

Table 1 Identification and description of implementation strategies used in SSBC

During the planning process, three local YMCA sites were considered. The first site had hosted SSBC since 2017, with research staff implementing the program. Despite exposure and prior knowledge of SSBC, staff were not involved in the program until the start of this current study. The second site was selected as the prospective staff were committed to the project and the organizational structure was similar to the first site. Due to differences in organizational structure and site readiness, the third site was selected to be a future scale-up site. This decision helped keep the project manageable and provided a location to implement lessons learned.

Paradigmatic position

A pragmatic epistemology was used to guide the design of this research, which prioritizes using research findings to develop practical recommendations to answer the research question and tolerates multiple truths [29]. A qualitative descriptive methodology was used to guide interpretation of the qualitative data. This methodology compliments a pragmatic epistemology as qualitative description is useful when seeking practical answers through understanding a phenomenon from the perspective of those involved [30, 31].

Participants and data collection

Each site nominated one staff as site-lead. This study involved two sets of participants: (a) staff who delivered SSBC and (b) implementation support staff. YMCA staff at both the delivery and implementation support staff level were invited to participate in the 1-year planning process. After hearing about the project, interested staff volunteered to enroll. Training occurred in two rounds and covered the counseling style, motivational interviewing (MI), and program delivery (see [32] for detailed description). Staff unable to attend the first training, in addition to any new staff wanting to enroll, were offered to attend the second training. Implementation support staff included the manager and site lead for each site, and the vice-president of health, fitness, and aquatics for the region. Ethical approval was obtained from the researchers’ institution, and written informed consent was obtained from all participants.

Semi-structured interviews

All YMCA staff (n = 8) who received training and counseled a minimum of three clients through the program were invited, and agreed, to participate in a semi-structured interview between 4- and 6-months post training (M = 5 months). This procedure was chosen to ensure staff had sufficient experience implementing the program, prior to an interview. Within the study period, the eight YMCA staff facilitated 32 clients through the program. The purpose of the interview was to understand the implementation strategies (e.g., training, support, program resources), provide feedback on the program, and understand the experience participating from an individual level. An interview guide was developed from previous research [33] and piloted with two staff, resulting in minor modifications (e.g., question wording, question addition; see Table 2 for sample interview guide questions). Staff ranged from 24 to 51 years old (M = 31.8, SD = 9.0), had between >1 and 17 years of experience at the community organization (M = 7.8, SD = 5.9), identified as 78% women and 89% Caucasian. Interview options (in-person [n = 8], or telephone [n = 1]), were provided to prioritize participant convenience and reduce burden. Staff interviews lasted 50–68 min (M = 58 min).

Table 2 Sample interview guide questions

Focus group

An in-person semi-structured focus group was arranged 6 months into the project with implementation support staff. Each site lead and one site manager were trained to deliver the program and had direct experience in program facilitation. The purpose of the focus group was to understand the implementation strategies (e.g., training, support, program resources) and experiences participating from an organizational level. The focus group also discussed challenges identified in the project and solutions to promote sustainability. A semi-structured interview guide was created, guided by identified implementation determinants in the literature, including questions on feasibility, acceptability, culture, complexity, and sustainability [34]. Focus group participants ranged from 24 to 51 years old (M = 38.4, SD = 11.1), had between >1 and 17 years of experience at the community organization (M = 10.0, SD = 6.1), 80% identified as women, and 100% as Caucasian. The focus group lasted 118 min.

Data analysis

The interviews and focus group were conducted by the first author, audio-recorded, and transcribed verbatim. A note-taker was present during the focus group and documented non-verbal cues and other observations that helped inform the analysis. Transcripts were read multiple times to promote familiarization with the data. A template approach [35] was used to thematically analyze transcripts and NVivo12 software [36] facilitated data organization. Data was coded using an inductive-deductive approach. After data familiarization, the first author independently coded all transcripts inductively and then met to discuss codes with the second and third authors. Inductive codes were discussed and revised (e.g., consolidated, adjusted name) as necessary to ensure accurate interpretation. To avoid overlooking data, first transcripts were coded inductively to ensure salient data was captured before deductively linking to CFIR constructs.

The CFIR technical assistance website [37] provided resources to facilitate data analysis including definitions of each construct within each domain. Rather than selecting which CFIR domains to examine, each inductive code was compared to all the CFIR construct definitions. By doing so salient CFIR constructs from relevant CFIR domains emerged from the data and any unrepresented CFIR constructs were excluded. The first author created an initial template that aligned the inductive codes with deductive CFIR constructs. The template included a section of inductive codes linked to deductive CFIR constructs that identified suggested modifications for sustainability. This section of the analysis helped identify modifications to inform future scale-up efforts. The template was reviewed with the second and third authors. Throughout the process, codes were examined for fit and consolidated, or separated, to better align with CFIR constructs. Definitions of CFIR constructs were constantly compared to inductive codes to ensure appropriate interpretation during the linkage step. The initial template was reviewed and discussed in multiple rounds of coding with the second and third authors, with the first author returning to the transcripts to ensure appropriate representation of the data. Discrepancies were discussed between all authors until consensus was reached. Relevant quotations were selected to accurately represent the CFIR construct and the participants’ lived experiences.

Results

Several constructs from each CFIR domain were matched to inductive codes, and all five CFIR domains were represented within the data: (a) process, (b) intervention characteristics, (c) outer setting, (d) inner setting, and (e) individual characteristics. The CFIR’s process domain was difficult to link to the current qualitative findings as the interview guide focused more on participants’ reflections on the implementation plan execution and less on the planning process. Where possible, evidence supporting the process construct was provided in addition to a short descriptive overview of the planning process up to this project. Results are discussed according to each CFIR domain (heading) and CFIR construct (italicized). Quotes to support each construct are presented in Table 3.

Table 3 Overview of CFIR constructs linked to inductive themes

Process

As noted, a 1-year planning process with the community partner engaged relevant stakeholders in co-developing the implementation plan [27]. A champion from the community organization spearheaded the partnership and was a respected, credible YMCA employee with formal influence on YMCA leadership, sufficient engagement in the initiative, and capacity to participate in the planning process. An opinion leader was invited to contribute to the planning process, a respected YMCA employee with formal influence as Vice-President of health, fitness, and aquatics. A product from the planning process was the creation of a formally appointed internal implementation leader (site lead), a role designed to facilitate communication between the research team and the site. A second product was the creation of a formally appointed internal implementation team, designed to have the champion, the opinion leader, and one front-line staff attend monthly implementation team meetings during the project. To appease both partners’ interest in assessing fidelity, an evaluation was developed to examine whether the YMCA staff executed the program as intended (see [32]). Finally, reflecting and evaluating was targeted both through formal evaluation (e.g., data collection tools) and ongoing reflection at monthly meetings at both implementation team and community organization staff levels. One CFIR construct in the process domain was unrepresented: external change agents.

Intervention characteristics

Staff praised the design quality and packaging of the program, training, and associated documents (e.g., manual, checklists). Although there was specific program content to be delivered, each staff could deliver the session content in the order, manner, and depth that made sense to each client. This adaptability, encouraged by the client-centered nature of the program, was described as a facilitator. Staff were able to contextualize each appointment to their client. Although no staff described the program as difficult to deliver, staff did mention perceived complexities. A few staff discussed appointments being emotionally taxing from the effort needed to use and apply MI techniques. Even 3 months into program implementation, staff continued to prepare for each session. This, coupled with having some sessions run over-time, caused additional stress, especially when staff had appointments either right before or right after a scheduled SSBC appointment.

YMCA staff described the relative advantage of SSBC compared to other YMCA programs. Specifically, three different advantages emerged. First, participants described the MI training as an asset that should be used in other YMCA programs. MI emerged as a valuable skill transferable to other YMCA roles (e.g., group fitness, one-on-one training, conflict management). Second, YMCA staff described a SSBC appointment as different compared to how YMCA staff typically work with clients. SSBC allotted more time with a client while they exercised, which enabled staff to help clients learn about the equipment and physical sensations experienced in exercise. In addition, typical YMCA appointments did not include counseling time, which enabled staff to connect with their client, and genuinely listen and support them. Many staff were surprised by the amount of emotional support given to clients, requiring a tissue box in the counseling office. In addition, the counseling component allowed staff to discuss topics beyond traditional YMCA programs focused on physical activity, such as providing program-specific dietary content. Third, the staff felt the program structure (brief 3-week program, with home days) coupled with the behavior change techniques (e.g., action planning), helped clients build self-management and self-regulatory skills to build a diet and exercise habit. Staff explained that meeting a client regularly for 3 weeks is atypical for YMCA programming and enabled them to build close connections to clients. For many staff, they appreciated and enjoyed these differences between SSBC sessions and their other YMCA programming.

While the SSBC provided many perceived advantages relative to other YMCA programming, the program also included some relative disadvantages. Due to the schedule of SSBC sessions, only full-time staff could reasonably take-on clients as part-time staff could not accommodate the schedule needed in the first week of the program. In addition, many staff could only take on one client per month, which limited client capacity. Finally, while many of the CFIR constructs in the intervention characteristics domain were present, four were not: intervention source, evidence strength and quality, trialability, and cost.

Outer setting

Staff discussed how SSBC helped with peer pressure by setting the YMCA apart from other fitness facilities. Providing SSBC further supported their niche as a community, health-focused fitness facility that provided accessible and effective programming to support the health of the community. In addition, the YMCA helps fill a gap in the local health authority, private, and public health through supporting a diabetes prevention program. The cosmopolitanism of the YMCA, through its partnership with a research group, supports this community, health-focused vision by offering an evidenced-based program. Two CFIR constructs were not represented in the outer setting domain: patient needs and resources and external policy and incentives.

Inner setting

During the project, evidence emerged that networks and communication within both community sites supported staff with implementation. Staff utilized both formal communication during SSBC meetings and informal communication through a community of practice with other staff to discuss the program, troubleshoot, and practice skills. There were distinctions between structural characteristics within the two sites. One site was more established and larger compared to the other site (opened 2001 versus 2017, respectively). These differences impacted the social architecture, whereby the older site had a more “tight-knit” feel amongst staff compared to the newer site. The newer site also had more staff turnover compared to the older site.

Overall, the community organization had a supportive implementation climate. Staff saw SSBC as compatible with the YMCA at three different levels: (a) organization, (b) staff, and (c) client levels. First, SSBC was a good fit in the YMCA with its focus on helping individuals become healthier, especially in relation to increasing physical activity levels. Second, staff were well-suited to deliver SSBC as they shared many values with the program, including being passionate about health and fitness and helping individuals improve their overall health. Third, the YMCA provides an ideal environment for the clients to continue to achieve their goals post-program. While the program may end after the final session, the YMCA environment remains available for clients to benefit from. Relatedly, the organization had a positive learning climate where all participants worked together, regardless of rank in the organization, to implement the program and support each other.

The project was ready for implementation through its leadership engagement (site lead). The site leads supported implementation by helping with logistics (e.g., scheduling), supporting coworkers (e.g., check in with staff) and liaising with the research team. The project had available resources such as a dedicated project coordinator to provide ongoing support and lead monthly meetings. In addition, staff had access to knowledge and information through provision of implementation support documents (e.g., program manual, checklists) and ongoing access to an online training platform with support videos. Finally, discussions in the focus group identified a shared belief amongst management that everyone at the YMCA should be a SSBC staff, providing insight into the presumed culture at the YMCA to support such a program. While many CFIR constructs were represented within the inner setting domain, four constructs were unrepresented: tension for change, relative priority, organizational incentives and rewards, and goals and feedback.

Individual characteristics

Staff’s knowledge and beliefs of the program grew during implementation. Staff discussed benefits of program tools, program structure, and counseling style (MI), and believed that the program positively impacted clients. Overtime, staff’s self-efficacy for implementing the program increased, particularly regarding use of MI. Evidence demonstrated individual identification with organization as staff displayed commitment to the success of the pilot project. Strong indications of commitment included staff completing preparation time (including preparing at home), reflecting on their skills, and practicing skills amongst each other. However, staff also mentioned being impacted by the program itself and described wanting to do well in their role to better help the clients. This code was placed within this domain (individual characteristic), as some staff demonstrated more commitment than others. However, the outer (cosmopolitanism of the research group with the YMCA of Okanagan), inner (learning climate and culture), intervention (relative advantage), and process (engagement in the planning process) likely contributed to staff’s strong commitment.

Staff identified other personal attributes from participating in the SSBC, such as building transferable skills for personal and professional development. Common skills described were listening, communication, teamwork, and conflict management. It was evident during the focus group that managers felt through providing SSBC, they were providing additional learning opportunities for their staff to build their resumé. One construct in the individual characteristics domain was unrepresented: individual stage of change.

Modifications for increased sustainability

Ideas emerged during the focus group on how to modify the program for increased sustainability.

Intervention characteristics

To increase sustainability, the program has been restructured to reduce the time-burden on staff in the first week of the program, to a maximum of two appointments per client, per week. This modification eases staff perceived complexity by reducing the amount of time dedicated to SSBC per week. In addition, the restructuring reduces the identified relative disadvantage and allows part-time staff and volunteers to enroll in the training, resulting in an increased number of clients that the program can support each month.

Inner setting

To increase sustainability, SSBC team meetings have been incorporated into the larger YMCA site meetings to encourage all YMCA staff to become familiar with SSBC. This shift in the formal networks and communication within the organization aims to maintain all the benefits of the formal team meetings, but also encourages wider YMCA staff buy-in and eases scheduling as staff are already required to attend these meetings. This modification intends to further support the organizations’ culture identified in the focus group, to encourage more part-time and volunteers to sign-up for the training and make staff aware of the professional and personal development opportunities that SSBC provides to staff. The YMCA plans to continue to ensure leadership engagement through having managers and nominated site leads actively support staff who take the SSBC training through periodic check-ins. Some staff may need additional support and/or training to feel confident in implementing the program, which was indicated by the reliance on the research teams’ project coordinator (available resources). To encourage sustainability, the research team is working on additional training to transfer leadership from a research project coordinator to a coordinator within the organization. Training will support the individual assume a similar role to the project coordinator from the research team (e.g., lead SSBC meetings, provide ongoing support, check-in on staff).

Discussion

To reduce the growing burden of T2D, more diabetes prevention programs need to be made available to those at-risk. To meet the need, there has been a focus on translating effective and scalable diabetes prevention programs into sustainable community settings. While there have been numerous diabetes prevention programs translated to date, there is a lack of research documenting the process of such translation efforts, the contextual factors impacting implementation and identifying the implementation strategies used to promote program uptake. The research team partnered with a local not-for-profit community organization to offer Small Steps for Big Changes with a goal of scaling-up to the national level of the organization. A co-developed implementation plan was used to translate the evidence-based program to two local YMCA sites. With context at the heart of implementation, an implementation determinant framework was used to analyze the implementation process and elucidate key contextual factors. The CFIR was a useful tool to frame the results. Moving from inductive to deductive analysis helped identify salient themes from participant perspectives prior to deductively linking to CFIR domains. This process enabled deep interpretation of the data as linkages were made to the CFIR framework and prominent CFIR constructs were generated from the data rather than preselected. As the analysis progressed, findings demonstrated an interaction between all levels of the CFIR contributing to successful implementation, similar to previous studies [38, 39]. The current study contributes to the field an in-depth during-implementation evaluation reflecting on the process of implementing a program using the CFIR framework to guide analysis and identify modifications to inform future scale-up efforts. Such a study also provides an overview of implementation strategies used according to definitions described from the Expert Recommendations for Implementing Change (ERIC) project [28]. The analysis included multiple, in-depth perspectives from those involved in the project (front-line staff, managers, site-leads, vice-president of health, fitness, aquatics). All five CFIR domains were identified, with relevant constructs supported within each domain.

Overall, the implementation plan was successful, with positive views from all study participants. Results suggest the high level of engagement during the planning process with the community partner positively influenced the successful execution of the implementation plan. While a partnered approach takes time and effort, the process of building a partnership embeds commitment and critical thinking throughout, ensuring appropriate steps are taken and a quality product results [40, 41]. The current study supports literature that sufficient planning through early engagement, the inclusion of a wide range of stakeholders, and prioritizing planning for scale-up from the outset is essential for success [42, 43]. A successful implementation context likely formed through strong commitment to the partnership at the leadership level of the YMCA (opinion leader and champion) and engagement of prospective SSBC staff at the YMCA, both important factors for implementation planning [38, 44]. Involvement of prospective staff developing the implementation plan, an often-overlooked part of implementation, likely fostered buy-in and encouraged adoption [12]. Engagement with the community partner ensured implementation strategies (e.g., provision of a staff training manual, program documents, ongoing implementation support through monthly site-meetings) were deemed acceptable and sufficient to front-line staff prior to implementation. Overall, the highly engaged, multi-level planning approach supported multi-level buy-in.

The YMCA has been identified as a promising partner to disseminate diabetes prevention programs [45]. Synergies were evident at multiple levels. The program itself was well suited to the vision of the YMCA being a community-health focused fitness facility [41] and the YMCA staff were well suited to deliver the program as they already had knowledge, skills, and desire to help clients within the health and fitness industry [44]. This synergy may have led to highly engaged staff, a facilitating factor for successful implementation [16, 43]. Participation in the project provided staff with new professional and personal development skills. Specifically, MI emerged as a valuable skill and a key component in the program. Staff discussed many relative advantages of SSBC, largely due to MI coupled with other program components (e.g., counseling component, diet and exercise focus, and 1-month program) that are not typically included in YMCA programming. Having relative advantage is another facilitator identified in prior implementation studies [46, 47]. Specifically, in a program for individuals with overweight and obesity, health care workers who perceived relative advantage of an intervention had higher motivation to implement it [48]. To further increase the advantages and support sustainability in the future, feedback from the current results on complexity and the relative disadvantage have been modified to support future scale-up. A new program structure will reduce staff burden during the first week, an allow greater flexibility in scheduling, and allow part-time and volunteer staff to feasibly fit appointments into their schedules.

All participants commented on the level of support provided with the program. Support was provided through multiple communication channels, both formal (e.g., team meetings, implementation team meetings, one-on-one support through project coordinator, site lead, program manual, checklists) and informal (e.g., community of practice, research team visit to sites). Multi-directional communication channels have been identified as a facilitator in previous research [44]. Specifically, health care professionals that perceived more support from their colleagues, had greater motivation to implement an intervention [46]. Relatedly, differences between the two sites emerged in the inner setting, with one site discussing more informal communication (e.g., seeking support from co-workers) compared to the other site. This difference may be attributed to the structural characteristics of the two sites with the “tight knit” site being in existence longer, more staff working at one time (enabling overlap and communication opportunities), a site lead at a higher organizational leadership level, and less staff turnover. These differences may have impacted the level of informal support provided at the younger site, as one staff at the younger site struggled with the intervention at the early stages, despite the amount of formal support. A timely change in site-lead, enabled the staff member to receive additional peer-support to gain confidence in implementing the intervention. In prior research, specifically in relation to a study using MI, providing resources and communication between colleagues has been shown to enhance facilitators’ ability and experience of delivering MI by having constant access to feedback and support [49]. These observations demonstrate the importance of selecting an appropriate site lead and the impact of the inner setting on social knowledge; both have been identified in research using the CFIR [50].

More research is needed to understand important factors to cultivate an inner setting conducive to adopting an intervention and executing the intervention with success. The current findings support the strategy of having a site lead; however, selection of an appropriate site lead and provision of training will likely enhance the strategies’ success. In a study investigating the effects of champions as an implementation strategy, results determined the success of the strategy was dependent on the chosen champion and adequate organizational support [51]. Four characteristics of a successful champion were identified: engagement, influence, credibility, and capacity [51]. Building off this study, coupled with the characteristics identified for a successful champion, future research is needed to operationalize the site lead implementation strategy.

It is important to note the amount of researcher engagement within the current implementation effort. The first author acted as project coordinator and was engaged in the planning of the pilot, development of implementation tools (e.g., training manual, training plan), oversaw the execution of implementation activities (e.g., attended and co-led staff training, liaised with site leads, led monthly site meetings, attended monthly implementation team meetings, provided ongoing support to front-line staff), developed close connections to participants in the project, and acted as a knowledge broker to pass information between the implementation team, research team, and the front-line staff. Close connections, open communication, and provision of support were likely facilitated by the size of the organization and the ability to have face-to-face meetings—two factors that have been identified as facilitators to successful partnerships in community health promotion [30, 44]. Feedback from a 15-year scale-up project identified that a local project coordinator was fundamentally important to implementation success [52]. Within the Kerala Diabetes Prevention Program, researchers had close communication with peer leaders (e.g., provided ongoing support via telephone calls prior to and after each group session) and utilized a local resource person (a leader nominated from the local self-governing bodies) to facilitate implementation (e.g., organizes group sessions, advocate for the program and send reminders/follow-up with participants) [14]. The engagement of peer leaders and local resource persons were crucial to implementation [14]. In the current project, the project coordinator was the researcher (first author) who ensured implementation strategies were executed and feedback collected. As the project continues to be scaled-up, the role of the project coordinator will shift to the YMCA site lead role. Careful selection and support to the site lead will be of utmost importance for successful implementation.

Limitations and future directions

The first author participated in all front-line staff meetings, implementation team meetings and was involved in the 1-year planning process. This involvement enabled the researcher to observe and simultaneously support the implementation strategies throughout the project. Although this dual role may introduce bias, it has been used in similar implementation studies for an in-depth perspective of the implementation process [43, 53]. All participants regularly interacted with/collaborated with the first author. While these professional relationships likely resulted from the collaborative partnership approach, it can also introduce social desirability bias. However, the close working relationship also enabled mutual respect enabling participants to provide honest feedback to the research team. The first staff to volunteer for the project may have been more motivated compared to staff trained in subsequent rounds. The study was interrupted by COVID-19 and not all staff had facilitated three clients through the program prior to facility shut down/interruption, which reduced the study’s sample size. The sample was primarily Caucasian women, which may limit generalizability to other studies or new sites with a more diverse staff. Finally, the study design itself (e.g., interviews and focus group) were effective strategies to reflect on the process of implementation. Useful feedback was obtained to modify the program for increased sustainability with the community organization. Future research should include qualitative evaluation during implementation to understand the process and learn from it, to be able to make improvements in the future.

Two of four recommendations identified in a 2016 systematic review on the application of the framework [17] were applied in this study (meaningfully use the CFIR to inform future scale-up initiatives and to describe how CFIR constructs were selected). Future research should work to include the other recommendations: associating CFIR constructs with study outcomes and integrating the CFIR throughout the research process. While the CFIR was integrated throughout much of the research process, future work should consider aligning data collection materials (e.g., interview guides) with resources available on the CFIR technical assistance website [37]. In this present research, had the interview guide been developed to align with CFIR constructs, unrepresented CFIR constructs might have had representation. For example, all participants were aware that the intervention was developed externally (i.e., intervention source), had demonstrated effectiveness (i.e., evidence strength and quality) and was being piloted (i.e., trialability). However, by not aligning the interview guide with CFIR constructs, salient themes emerged inductively rather than asking participants about all constructs that may or may not be deemed relevant by the participants spontaneously. While the research team was not able to accommodate all recommendations from the systematic review, results contribute to the CFIR’s research base by identifying relevant constructs within this project’s successful implementation effort.

The implementation strategies used in the project support their use in future scale-up efforts, which are underway. The current study was not able to assess which specific implementation strategies were most effective, rather the results presented are a result of the package of strategies used. Future experimental research should be used to tease apart the effects of specific implementation strategies and/or future systematic reviews and meta-analysis can be used to identify promising implementation strategies. In the current project, co-developed implementation strategies were executed with researcher support. In the next phase of program scale-up, implementation strategies will be executed by community organization staff. This tiered approach has been successful in prior research [53], after building off an initial study with strong involvement of the researcher [43, 52, 53]. The two sites where the project was conducted are well-suited to take on such leadership responsibilities having closely collaborated and shaped the current form of the program and will be invited to share lessons learned with new scale-up sites. When setting up a new site, sufficient support and engagement with prospective staff is warranted. While engagement requires additional resources and time, it has favorable outcomes when compared to more hands-off approaches (e.g., provision of a manual) that may not be sufficient to support implementation [38, 52]. Future research is needed to optimally operationalize provision of support in a sustainable way.

Conclusions

This study provides an in-depth analysis into the process of executing a co-developed implementation plan to have staff facilitate an evidence-based diabetes prevention program within a community organization. Overall, the implementation strategies used were acceptable and staff were satisfied with the implementation process. Project success was likely a result of the process in which they were developed, planned, executed, and reflected, which provides valuable feedback to prepare for the future scale-up of the project. This study demonstrates an effective partnered approach to the process of implementation and demonstrated important constructs from all five levels of the CFIR.