Using Integrated Research-Practice Partnerships to Move Evidence-Based Principles into Practice

Abstract

The implementation of evidence-based physical activity interventions is improved when integrated research-practice partnerships are employed. These partnerships consider both research- and practice-based evidence that moves beyond only assessing program efficacy. Our novel hypothesis is that integrated research-practice partnerships may lead to interventions that are practical, effective, reach more participants, and are more likely to be sustained in practice.

Summary for Table of Contents

We describe a process model for integrating research and practice systems to co-produce intervention strategies that are practical, effective, and sustainable.

Introduction

The predominant translational paradigm, as it relates to behavioral physical activity interventions, is based on a pipeline model that begins with demonstrating the efficacy of a given intervention when tested under optimal conditions (1, 2). This systemic view of translational research embodies a number of deeply held beliefs, rules, and structures that provide a basis for when an intervention is ‘ready’ for testing in typical community or clinical contexts (3). Flay, in a seminal article on the stages of research needed to determine the utility of a health promotion intervention, recommended that research move from (i) development and pilot trials to (ii) efficacy trials to (iii) effectiveness trials that include a large and representative portion of the target population to (iv) implementation effectiveness trials that include typical members from the target population and the typical staff that would ultimately implement an intervention to (v) demonstration projects across large systems.

Thirty years since the publication of Flay’s work, translational research continues to focus on a hierarchy of evidence necessary before an intervention is ‘ready’ for translation. This type of research is described as knowledge translation in Canada(4), scale-up or scalability in Australia(5), or dissemination and implementation research in the United States(6). Indeed, across these and other countries, research-to-practice translation is predicated on the movement of an evidence-based product from science to practice. Yet, across a number of scientific fields, there remains uncertainty as to the level of evidence needed to denote that a particular product is evidence-based(5). Further, the predominant proposition from these perspectives is that providing an evidence-based product with ‘how to’ resources and support—and a focus on fidelity while allowing modest adaptation—will result in improved physical activity for those that receive the intervention(4–6).

When considering the public health potential of integrating physical activity interventions in a timely manner, there are a number of issues that arise with the current translational paradigm. First, practice professionals and decision makers often do not share the value of a hierarchy of evidence and value multiple types of information, some more than traditional research evidence(7). Second, some practitioners may actively criticize evidence-base as not relevant—in that the evidence was generated in settings that were not representative of the ‘real world’ nor were the study participants representative of the typical population served by practice organizations(8, 9). Third, there is limited data demonstrating efficacy is predictive of demonstrating effectiveness when an intervention is implemented with a more diverse population of participants and delivered by a more diverse set of settings or staff. Fourth, and pragmatically, the current translational paradigm suggests that practitioners would or should wait until the best possible evidence is available rather that recommending they act on the best available evidence. This is especially problematic when staff members in clinical, public health, education, or employment systems are faced with addressing current levels of physical inactivity across a broad spectrum of patients, residents, students, or employees.

The predominant focus on implementing evidence-based products with high fidelity as a solution to low levels of physical activity in patient or community populations can also put researchers in a difficult situation when presenting these ideas to practice organizations. Specifically, practice professionals and decision makers have unique knowledge, skills, and strategies that may not be considered with an evidence-based intervention’s adoption, implementation, or scale-up(10). This can put a researcher at a disadvantage when having to defend (i) why research evidence, and specific forms of research evidence, is necessary and superior to practitioner expertise and (ii) the need to assure fidelity while avoiding adaption to the local context that is likely necessary to increase the likelihood of broad reach and sustainability(11). Using policy case studies as an example, those that were successfully adopted and implemented were characterized by positive user-testimony, assets-based approaches that leveraged existing resources, short-term evaluation of costs or resources saved at the local level, and evidence of short-term effectiveness for service users, and a potential for high reach(7). Further, when interventions were adopted they were adapted based on the intended audience, practice professionals, and/or community organizations that housed the interventions (7).

To address these issues, we developed a systems-based model as a generalizable process to improve the likelihood that a given intervention approach will achieve a local public health impact(12). This process model is based on the following propositions. Our first proposition is twofold in that integration of scientific and community/clinical systems to address questions that are scientifically innovative and have practical implications for stakeholders will (i) result in an increased likelihood of sustained local implementation and (ii) provide replicable evidence of the generalizability of the integration process across systems. Our second proposition is that the development of sustainable program, practice, or policy approaches is best achieved using a vertical and horizontal systems approach(13). Vertical components of our systems-based process model are typically operationalized to include both the staff that interact with the potential recipients of a given intervention and the organizational decision-makers who can approve intervention adoption, implementation, and institutionalization. Horizontal components of a systems-based approach include the engagement of units or sectors across an organization or community that can contribute to promoting physical activity. Our third proposition is that research synthesis that focuses on evidence-based principles (i.e., core mechanisms of change) rather than products (i.e., evidence-based programs) has a higher likelihood of achieving wide adoption and high-quality implementation(12, 14). Finally, our fourth proposition is that there is a higher likelihood of scale-up and sustainability when organizational or system governance, values, resources, strategies and structure are leveraged to ultimately design for dissemination(12).

In this manuscript, our purpose is to define the key elements of our integrated research-practice partnership process model and provide examples from our work in physical activity promotion across projects and partnerships. We used the RE-AIM (reach, effectiveness, adoption, implementation, maintenance) framework(15, 16) in project planning with a goal to support partnerships in developing strategies with broad and representative reach, which could effectively change and maintain participant physical activity and be adopted, implemented and sustained across partnership systems when target metrics were achieved. Similarly, within each partnership, we pragmatically identified RE-AIM dimensions within the evaluation and decision-making phases outlined in the model below.

The Integrated Research-Practice Partnership Process Model

Similar to other processes including community-based participatory research (CBPR) and educational research practice partnerships, our process model includes the need for mutual benefit for both the research and practice systems as well as transparency, trust, and sharing of resources to prioritize action across partners(17–19). Distinct from these approaches, our process model allows for varied duration in partnerships from acute, single issue specific partnering to the longer-term partnerships across multiple projects described in CBPR and educational research-practice partnerships(18, 19). In addition, our model is explicitly intended to focus on the application of scientific knowledge in practice settings, but is predicated on the use of evidence-based principles, rather than evidence-based products, and the inclusion of vertical and horizontal system representatives as members of a partnership(12). Therefore, in our integrated research-practice partnership process model, we assert that partnerships should be structured to promote a collaborative, multi-level systems approach to the development, testing, and sustained use of evidence-based principles within real-world settings(12).

What is the process?

The Figure provides an updated visual representation of our integrated research-practice partnership process model.(12) Central to the figure is the iterative process used to co-produce research-based and practice-relevant, evidence. This process was heavily influenced by the innovation decision-making process outlined in Rogers’ Diffusion of Innovations.(20) Specifically, co-production is operationalized by the partnership’s role as the decision-making unit with underlying principles described by Rogers’ five stages – knowledge (exposure to and understanding of and innovation), persuasion (favorable or unfavorable attitude toward innovation is formed), decision (activities that lead to innovation adoption or rejection), implementation (new idea is put to use), and confirmation (seeking reinforcement of decision).

In our process model, there is a continual, overarching emphasis on the collaboration among practice professionals, organizational decision-makers, and scientists. The notion of knowledge – decision-makers exposure to an innovation – occurs early in the integrated research-practice partnership process—typically during the transition from the problem prioritization phase to the strategy selection phase. In our process model, the Rogers’ concept of persuasion is addressed, initially, during the strategy selection stage. This is when the partnership members identify and select the most favorable option from existing evidence-based approaches or sets of evidence-based principles—that is, for ongoing partnership success, initial persuasion that the evidence-based principles may address the problem is needed to then determine potential adaptations. During the strategy adaptation stage, scientists and practice partners engage in a process to ensure fidelity to the underlying evidence-based principles of the intervention and that the decision-makers have and maintain a favorable view of the adapted intervention. As such, the integrated research-practice partnership model is intended to result in a product that leveraged the practice partners’ expertise and developed a sense of local ownership.

Implementation and confirmation components of the innovation decision making process(20) are operationalized as integration trials and the evaluation, decision-making, and translational solution steps, respectively. Determining if a decision is made to sustain and scale-up an intervention, sustain the intervention with further adaptation, or discontinue the intervention and pursue another strategy reflect the final steps within our integrated research-practice partnership approach. We propose that RE-AIM is a useful framework for these final steps and provides metrics that are easily communicated, measured, and addressed in practice settings(21, 22). These metrics can provide direction during the decision-making stage for potential intervention adaptations or for specific components to look for in a new intervention.

For example, a target metric could be identified based on the reach of the new intervention with a goal to engage 25 or more people per county as the metric necessary for the partner system to commit to long-term implementation(23). A metric could also be based on effectiveness in that if physical activity is increased significantly by the intervention when compared to a matched-contact control, then the organization will sustain delivery(24). These individual participant level metrics can be balanced by also using metrics related to the feasibility of adoption, implementation, and sustainability from the provider and participant perspective(25). The process concludes an initial test of an intervention with a decision related to if the intervention should be sustained and scaled-up, sustained—but adapted further, or discontinued and the pursuit of an alternate strategy is necessary. In each case, the alignment of the intervention characteristics and principles is considered concurrently with intervention fit within the organizational system by both researchers and practitioners.

Who is involved in an integrated research-practice partnership?

Concentric circles to the left and right of the figure represent the research and practice domains, respectively. The research domain includes a strong focus on the evidence-base, where, and with who, interventions have been tested, and the core elements that reflect the underlying evidence-based principles or mechanisms of intervention action.(22) The practice domain includes a hierarchical and horizontal systems-based perspective that focuses on where, how, and by whom new or existing program, policy, or practice are implemented. This also acknowledges the explicit and tacit knowledge related to infrastructure and resources, participant population needs, and organizational values and workflow that are critical to initiating new health promotion strategies.(12)

Partnership composition varies based on local context, but consistently uses a team-based approach to develop a collaborative agenda and includes, at a minimum, a practice professional who interacts with the patient or participant population, an organizational administrator with decision making authority relative to taking action in response to identified priorities, and a scientist with expertise in the potential priority areas(12). In addition, these partnerships frequently engage patient and community representatives from the population intended to benefit from a given intervention. Practice members of the partnership can often expand to include an inter-professional team of organizational administrators, clinic program managers, and front-line program delivery staff. Similarly, research membership on a partnership may expand to include an interdisciplinary team of investigators, data managers, and student research assistants. The number of members and overall composition of teams involved in a partnership depends on the scope of practice, complexity of research design, and available resources(12).

Summary of process and roles.

Rogers posited that individuals or organizations move through each stage in a linear fashion but acknowledged that each stage may not be overtly distinguishable by the decision-maker. The initial action within our integrated research-practice partnership process includes identifying a gap in system services that is a high priority for resolution. Once a priority area is identified, the partnership selects an appropriate evidence-based strategy and determines the need for adaptations to ensure that the strategy has relevance and will fit with organizational resources and potential participant preferences. Adapted strategies are then tested within the delivery system to determine if the intervention can be integrated into existing practices. The integration trial design is collaboratively developed with a focus on trial features that will improve scientific confidence in the findings while also providing information necessary for pragmatic decision making among the practice partners.

Group Dynamics Approaches to Physical Activity Promotion

Approximately 25 years ago, Carron and Spink,(26) published a paper examining group dynamics-based strategies to promote physical activity targeting the development of group cohesion. They demonstrated that by actively targeting group processes, structure, and environment within exercise classes that participants would have increased perceptions of group cohesion and be more likely to attend physical activity sessions over time. This work was replicated by a number of authors using Carron and Spink’s (and other) group dynamics-based approach(27–31). It also provided details on the underlying group dynamics-based principles that were hypothesized to be required to produce a change in physical activity participation. This is similar in concept to Chambers and Norton’s description of intervention core components or mechanisms of action.(32) Here we refer to these core components as evidence-based principles due to the large body of literature that demonstrates the use of these principles across populations, settings, and different implementation strategies consistently leads to increases in physical activity(31).

We have summarized the evidence-based, group dynamics-based principles for physical activity promotion elsewhere.(33) Briefly, the strategies within a group dynamics-based intervention are implemented to develop a sense of distinction for the group (e.g., team name; program t-shirts)(23), facilitate accountability and physical activity intentions using group goal setting and feedback(34), use or foster geographic (e.g., host programs where potential participants already aggregate) or emotional (e.g., recruit teams of friends, coworkers) proximity(24), and facilitate ongoing group interaction through regular communication, cooperative activities, and intergroup competition(35). These evidence-based principles were used consistently across the studies described below. In addition, the Table provides a description of each study including the key personnel, processes, and outcomes of applying group dynamics-based principles. In each case the partnership members adhere to the underlying evidence-based principles, but collaborate to adapt the intervention to fit the system or setting. in different contexts—adhering to the underlying evidence-based principles—but working through participatory approaches to develop strategies and translate the work to different settings and systems. Below we present the findings of studies completed in chronological order.

St. Mary’s Physical Activity Promotion for Older Adults

In 1999, a partnership was developed among a lead behavioral scientist with expertise in group dynamics and physical activity, a doctoral student, and a research assistant as well as a veritical system component with a nurse manager (i.e., decision-making authority) and physical activity class leader at a local older adult assistance center (St. Mary’s Home Support Services; Table).(36) Together, the partnership identified the need for supervised physical activity programs in rural areas and small towns, such as the community in which St. Mary’s was located. Using the evidence-based group dynamics principles, a 6-week intervention was offered to older adults at St. Mary’s. Through a small randomized controlled trial (n=33 participants), the research team was able to establish further evidence in support of applying group dynamics-based principles and strategies to improve physical activity behaviors within this setting and for these participants. Specifically, participants were randomly assigned to either a team-building/group-dynamics condition, a placebo condition, or a control condition. All programs used standard physical activities completed at a moderate intensity that included exercises to improve strength, cardiovascular fitness, and flexibility. The team-building condition was collaboratively developed to introduce strategies focusing on addressing the evidence-based, group-dynamics principles. The placebo condition included having a research assistant attend and co-facilitate standard classes and the control condition was led only by a class instructor from St. Mary’s. As documented in the Table, the participants in the team-building intervention attended more classes and were more likely to return to the program after a typical summer hiatus. In addition, St. Mary’s Home Support Services practice team were able to sustain a service that did not previously exist in their community.(36)

Walk Kansas

Walk Kansas began as a partnership between behavioral scientists with a focus on physical activity promotion, a local health educator who delivered a local fitness program through Kansas State Research and Extension (KSRE), and a statewide specialist from KSRE.(23, 37, 38) This project focused on the horizontal nature of KSRE with coverage across 105 counties and had the vertical nature of the system operationalized in the presence of the local health educator (ultimate implementer) and the statewide specialist (organizational decision maker). Together, the partnership identified the need for a state-wide physical activity promotion program to address the high prevalence of physical inactivity in adults across the state. The partnership applied evidence-based group dynamics-based principles to promote walking, but due to the lack of expertise of most health educators in delivering a physical activity class and the limitations with scalability, the program was distinct from previous iterations of the intervention. Specifically, health educators recruited self-identified teams of individuals and provided information for the teams to (i) select a team name, (ii) set a group goal to walk the distance across Kansas over 8 weeks, (iii) report weekly progress to a team captain, (iv) receive information on progress relative to other teams, and (v) plan for sustained physical activity once the program was complete using weekly newsletters. Group size was also strategically set to increase the likelihood that each team member would meet the recommended guidelines for moderate intensity physical activity. Team size was limited to 6 participants so that if each participant walked 2 miles (approximately 30 minutes), five days per week, the team would cover the distance across Kansas during the program period. Inactive and insufficiently active participants significantly increased and maintained physical activity during and 6 months after the program, respectively. The Table includes a complete description of the implementation process and outcomes for Walk Kansas across reach, effectiveness, adoption, and maintenance. Implementation was not assessed, though a follow-up study found that health educators understood the evidence-based principles that were used to create strategies for the intervention and were able to make adaptations to improve efficiency of implementation without violating the underlying principles(38).

Move More

Move More(24) could be considered a hybrid program that included some elements of the St. Mary’s program (e.g., in-person, though less frequent, classes) and of Walk Kansas (e.g., group goals for activities done individually, but reported back to a team). The program and study were collaboratively developed by investigators in the Kaiser Permanente Colorado (KPCO) Clinical Research Unit and providers from KPCO Preventive Medicine and Family Medicine reflecting the horizontal components necessary to identified potential participants and deliver the physical activity program in KPCO (24, 39). The partnership included the Chief and Director of Preventive Medicine, the Director of Health Education, a primary care physician, and health educators who would ultimately deliver the program as well as behavioral scientists, a project coordinator, and research assistant. The partnership identified the need for a centralized physical activity program that could be used as a referral outlet for primary care physicians in the region.

The problem prioritization regarded the degree to which a group dynamics-based intervention could be delivered in a system with limited resources and, if so, would the program be sustained in practice? Not only did the program improve physical activity behaviors at 3-monthss, but also 6-months after the completion of the intervention, Move More! participants maintained, and even increased, minutes of moderate physical activity per week when compared to the enhanced standard of care (print-based materials). This trial provides evidence that group dynamics-based interventions can work in clinical settings as well as the community system in which it was originally tested.

Fit Ex

Finally, a group of county-based health educators as well as Cooperative Extension leadership (vertical system representation) worked with researchers from the Human Nutrition, Foods, and Exercise Department of Virginia Tech to identify a program that would fit the needs to improve physical activity across the state of Virginia (25, 40). Core elements of Walk Kansas were translated to Virginia, but local adaptations were integrated. The Virginia partnership included emphasis on fruit/vegetable consumption to accompany the physical activity component of the intervention. The resultant program was called FitEx. One research interest in this partnership was the degree to which health educators of the system would be more likely to adopt a program developed specifically for Virginia Cooperative Extension when compared to a program developed in more research-intensive settings (i.e., Cooper Institute’s Active Living Every Day). Thirty-six health educators were randomly assigned to either FitEx (n = 18) or Active Living Every Day (n = 18). Only 11% (n= 2) of the health educators assigned to Activity Living Every Day delivered the program whereas 89% of the health educators assigned to FitEx delivered the program.(40) Qualitative inquiry indicated that health educators thought FitEx was easier to deliver and were attracted to the fact that they could adjust the program to fit their county needs (rather than following a set program manual) (25).

FitEx was designed for dissemination to multiple state professionals that could engage in a group dynamics-based approach to physical activity promotion. In addition, other Extension systems could adopt the program and develop competition across state lines to further enhance a sense of cohesion within the program (i.e., between individuals on a team, teams against other teams, counties versus other counties, and state versus other states). By 2018, two state systems (Wyoming and Virginia) were delivering FitEx. Future areas of research could apply principles of a participatory approach to unite state-level specialists in identifying program principles that work within their state systems, determining appropriate adaptations for their system, and collaboratively gathering the same effectiveness outcome (Harden SM, Ramalingam NS, Breig SA, Estabrooks PA. Unpublished Data, 2019).

Discussion

We began to use an integrated research-practice partnership approach as a potentially generalizable process to improve the likelihood that a given intervention approach will achieve a local public health impact through the translation of evidence-based physical activity interventions into practice settings (12). We are not the only research group to focus in this area, but the description of this type of approach and examining outcomes that include individual and organizational factors is rare (10, 41).

In this paper we provide examples of using this approach with different clinical and community practice partners who serve a broad range of target audiences (older adults, primary care patients). These examples provide preliminary evidence in support of our first proposition—the integration of scientific and community/clinical systems to address questions that are scientifically innovative and have practical implications for stakeholders will result in an increased likelihood of sustained local implementation while providing replicable evidence of the generalizability of the integration process across systems. All four of the interventions demonstrated positive changes in participant physical activity providing generalizable knowledge related to intervention effectiveness. Indeed, the magnitude of effect for each of the trials fell within the range of effect sizes found in a systematic review of other group-dynamics based physical activity interventions ((31, 23, 24, 36, 40) and Harden SM, Ramalingam NS, Breig SA, Estabrooks PA. Unpublished Data, 2019). and each was institutionalized and sustained within the local setting beyond the life of the respective research projects. Our work on Walk Kansas and FitEx also provided support for our third proposition that using evidence-based group dynamics principles (rather than a packaged evidence-based product) resulted in local interventions that were more likely to be broadly adopted in the local extension systems (25, 40).

These examples, and our program of research, have yet to provide explicit evidence of our second (the development of sustainable program, practice, or policy approaches is best achieved using a vertical and horizontal systems approach) (13) and fourth (when organizational or system governance, values, resources, strategies and structure are leveraged to design for dissemination there is a higher likelihood of scale-up and sustainability) (12) propositions—presenting opportunities for future research varying the composition of partnerships and measuring organizational context factors that could predict scale-up and sustainability. In particular, to date, we have not applied metrics to the approach to identify the underlying principles and process of these partnerships that will increase the likelihood of success. However, the development of metrics to assess process and outcomes in dissemination and implementation science is a current focus of the field (see Society for Implementation Science Collaboration Instrument Review Project https://societyforimplementationresearchcollaboration.org/sirc-instrument-project/). In our current work, we have adapted and developed a number of measures to assess the potential principles by which integrated research practice partnerships facilitate implementation of evidence-based principles into sustained practice.

Still, as seen in the Table, each program was adapted—using underlying group-dynamics principles rather than any specific evidence-based intervention—based on the needs of the setting in which the program was intended to be delivered, suggesting that alignment with organizational values and resources are important in the adoption decision-making process. This supports the third proposition—that evidence-based principles may have a higher likelihood of achieving wide adoption and high-quality implementation when compared to a packaged evidence-based program. These adaptations are often inevitable when moving evidence-based principles (rather than products) to new populations and settings (42) One area for future research is understanding what adaptations were made, by whom, and why (32). For this reason, we assert that this process model itself can be an implementation strategy (43) to improve the likelihood that the resultant intervention characteristics lead to adoption and implementation and potentially sustainability.

At the core of our approach is practicality. This participatory approach moves away from attempting to ‘push’ evidence-based interventions into practice where proposed changes to clinical or community services are often met with resistance or hesitation due to the uncertainty of the change (44). Secondly, this process model highlights the non-linear fashion by which these efforts are usually achieved. Rogers’ also describes the decision to adopt an innovation as a process that includes a set of activities that lead to a choice to either initiate the use of the innovation, or not. Our work suggests that this set of activities can be operationalized as a process to characterize priority areas, the selection of evidence-based principles, and the adaptation of interventions—based on those principles—that provides the best fit within the community or clinical system. This is closely aligned with the construct of reinvention (20) considered as a potential pathway to uptake and sustainability. Our work in a number of settings across varied populations suggests that reinvention or adaptations are inevitable. Use of the integrated research-practice partnership model ensures that the scientific principles and the system level values are continuously considered, over time, to improve patient/participant physical activity behaviors. It is proposed as a distinct stage to highlight the importance of ensuring that the intervention is relevant to, and will fit within, the practice partners’ organizational resources, mission, and service-recipient preferences (20, 45).

There are also a number of unanswered questions and potential limitations with this approach. First, we presented partnerships that were primarily associated with single organizations, which simplifies processes when compared to approaches that may engage partners from a wide variety of organizations across a community necessitating a larger group of stakeholders with more varied values, resources, and motivation. In some of our work focusing on childhood obesity we have used this model and successfully engaged multiple community stakeholders with vertical representation across organizations (46). Second, this approach may or may not generalize to policy application for regional or state legislative approaches or create generalizable interventions based on evidence-based principles. This limitation is balanced with the long-term goal of action research—to concurrently move science forward while having an impact in the practice organization or community involved in the research (47). Third, this approach is time intensive for research and practice partners alike. This provides opportunities to examine the potential cost effectiveness of integrated research practice partnerships across RE-AIM dimensions when compared to the more tradition translation of evidence-based products into practice. Fourth, to date, we have primarily viewed sustainability as a binary outcome that occurs directly after the partnership project is completed—did it continue to be offered or did in not. There is clearly a need to more explicitly focus on sustainability in more depth to examine sustained fidelity to underlying principles. Still, we did demonstrate that the majority of these partnerships led to interventions that were sustained without further research intervention—showing some sustainability promise for the approach.

We conclude by returning to the second proposition regarding the need for vertical and horizontal systems approaches (13). This type of translational research – which leverages systems change – is difficult because of the dynamic characteristics at play. This type of research has traditionally failed due to a tendency towards reductionism that ignores the complexity of an innovation and the implementation system(s). Similar to systems thinking (12, 48) we present an approach that respects and leverages characteristics of the system to achieve meaningful change, efficiency, and sustainability. By using a vertical and horizontal systems approach, our process model accounts for the perspectives of individuals involved in program delivery, decision makers at the delivery site, and the organizational system governing the delivery site. Similarly, while simultaneously focusing on the whole of the healthcare or public health system, our process model acknowledges the boundaries present in these systems. The definition of boundaries can be especially helpful for researchers to understand who among their practice partners have interests that should be served by research, what information will be most helpful to meet their needs and to establish a timeline of when those needs will be met throughout the research process.

It is evident that there is value in leveraging expertise and experience from partners within the horizontal and vertical structures of a clinical or community system. We have contributed to the understanding of the generalizability of group-dynamics principles to promote physical activity while our practice partners have valued the development of relevant, effective, and sustainable evidence-based physical activity promotion practices. We have also found that this approach contributed to team science (49) by creating a culture of mutual respect across disciplines. This systems-based integrated research-practice partnership approach serves to direct the purpose of translational research in a manner that will improve efficiency, speed, and public health impact. By drawing on dissemination and implementation evidence and systems-thinking concepts, we have created a generalizable process model for the co-production of intervention strategies to produce evidence-informed, sustainable solutions to local public health issues.

Figure: Integrated Research Practice Partnership Process Model here

Table 1. Summary of key personnel, processes, and outcomes of applying group dynamics-based principles in different settings and systems.

Process Actions	St. Mary’s Older Adult PA Classes	Walk Kansas	Move More	FitEx	FitEx 2015–2017
Integrated Research Practice Partnership Structure	Research Behavioral scientist with expertise in group dynamics and physical activity Doctoral Student Research Assistant	Research Behavioral scientist with expertise in group dynamics and physical activity Master of Science Student	Research Behavioral scientist with expertise in group dynamics and physical activity Project coordinator Research Assistant	Research Behavioral scientist with expertise in group dynamics and physical activity Research & Practice Assistant Extension agent enrolled as PhD student	Research Behavioral scientist with expertise in group dynamics and physical activity Graduate research assistant (20% time)
	Practice Nurse manage. Physical activity class instructor	Practice K-State Research and Extension Health Specialist A local county extension agent with experience promoting physical activity (health educator who would ultimately deliver the program) Support staff	Practice Chief of Preventive Medicine Director of Preventive Medicine Director of Health Education Family Medicine Clinic Physician in Charge Family Medicine Clinic Medical Office Manager Health educators	Practice Physical activity advisory board Extension leadership	Practice Physical activity advisory board Food Nutrition and Health Program Team (administration, specialists, and health educators develop and monitor action items for Extension professionals)
Problem Prioritization Target Population	Older adults benefit from physical activity, but in rural areas and small towns access for supervised physical activity were limited	Need to improve physical activity participation for county residents 18+ years of age and for the intervention to be sustainable after research funding ceases	Improve and sustain physical activity for insufficiently or inactive adults (≥ 18 years of age)	Increase the number of health educators who adopt the group-dynamics based physical activity program	Need for a healthy lifestyle intervention with low direct/ in-person program delivery.
Research Question (s)	Determine if a 6-week group dynamics-enhanced class would lead to higher attendance and return rates following a summer hiatus when compared to standard physical activity classes based on the Canadian Center for Activity on Aging curriculum.	Determine program reach (number, percent and representativeness) Determine improvements in physical activity over 8-week program Determine yearly adoption rates at the health educator and county-level Long-term behavior change maintenance	Determine effectiveness and maintenance of a practical group dynamics-based physical activity intervention Determine the mechanisms of physical activity change	Determine the adoption rate Calculate the proportion and representativeness of county participants Calculated implementation costs (i.e., time) Determine improvements in physical activity over 8-week program Intent for system-level maintenance	Determine the ongoing reach (the number and proportion of eligible community members who participated in FitEx). Monitor ongoing effectiveness Determine ongoing adoption (the number and proportion of eligible county health educators who delivered in FitEx). Weekly implementation evaluation data Setting-level maintenance was operationalized by health educator retention over the course of the three years
Strategy Selection	Team-building intervention developed for adult physical activity promotion Decision based on evidence of previous research and practice perceptions that older adults would benefit from the increased social interactions generated through a group-dynamics based approach. Contextual resources Friendship Community Center with space and staff to deliver PA programming.	Group dynamics principles with strategies that encouraged social support Decision based on likely fit with previous programs delivered through KSRE Desire to make a program attractive to families, friends, and employee groups.	Decided to adopt group-based approach to physical activity promotion and integrate an evidence-based 5 A’s approach for individual counseling focused on maintenance of behavior change. Decision based on evidence documenting the efficacy these approaches in physical activity and other behavioral interventions. Decision also based on context specific resources requirements to leverage available in-person resources, but in a way that would allow for multiple concurrent sessions.	Selection of two evidence-based programs Randomization to determine rate of adoption	Health educators predominantly responsible for data collection Research team responsible for development of impact statements
Strategy Adaptation	Specific strategies were developed using a process developed to engage fitness leaders in generating teambuilding strategies after a training on the topic.	Group dynamics principles adapted for broad scalability and alignment to other existing Extension delivered physical activity programs.	Adapted existing resources from St. Mary’s and Walk Kansas Programs for two, 2-hour small group sessions spaced 1 month apart Single supportive telephone counseling session 2 months later that assessed participant behaviors, advised on maintenance goals, collaborative agreed on new goals, assisted with barrier resolution, and arranged for follow-up with family and providers.	Modeled after “Walk Kansas” with local adaptations such as the distance across the state and Virginia-based materials 1-walking mile = 15 minutes of moderate intensity physical activity	Enhance FitEx with web, smartwatch, and smartphone access/options
Integration Trials Design	Randomized controlled trial comparing two 6-week programs, one using a group dynamics approach and graduated exercise intensity over time, the other using the available PA promotion curriculum from CAA. Attendance and return rate after a summer hiatus were used as the indicators of effectiveness and potential maintenance.	Quasi-experimental, pre-post design with primary focus on reach, effectiveness, and maintenance at the individual level and adoption and maintenance at the organizational level; randomly selected subsample included in maintenance phase (6-month physical activity assessment post-program)	Randomized controlled trial comparing a 3-month group-based physical activity program and the standard of care PA promotion booklet. Self-reported PA and proposed mediators were assessed at baseline, three and nine months.	Pragmatic randomized controlled trial; assigning health educators to either the IRPP-developed program (FitEx) or to a efficacy-effectiveness-dissemination pipeline model program (Active Living Every Day)	Pre-post design for effectiveness, by year. Ongoing longitudinal data collection for adoption and maintenance.
Evaluation Processes Outcomes	Intervened to improve: Effectiveness: Participants in the group-dynamics program attended 90% of the physical activity sessions compared to 70% and 65% attendance for participants in a placebo and control condition, respectively. Maintenance: 91% of group-dynamics condition participants returned to the program after the summer hiatus compared to 73% and 40% returning from the placebo and control conditions, respectively.	Intervened to improve: Effectiveness: Significant increases in both moderate and vigorous physical activity. Maintenance Did not demonstrate a significant decrease in moderate or vigorous activity between program completion and 6-month follow-up. Reach: 4609 total participants; Participants more likely to be female, more active, and older than the adult population within the counties where they resided. Adoption The number of counties adopting the program increased across years from 48 to 97/105 over 5 years. Counties that did not adopt did not have an assigned health educator. Organizational Maintenance: All counties that adopted the program continued delivery in subsequent years over the 5-year study period.	Intervened to improve: Effectiveness: Participants from both conditions increased their PA significantly during the initial three months. Maintenance: Participants in the group dynamics condition sustained or increased physical activity further at nine months when compared to the standard of care.	Intervened to improve: Adoption: 14/18 HE adopted FitEx while two HE adopted ALED (X2= 21.8; p<.05). Maintenance (Staff) FitEx HE greater intentions for continued program delivery (p<.05)	Intervened to improve: Ongoing reach (so that participants received intervention components virtually). However, retention varied over the three-year period: 60% , 24%, 47% for 2015, 2016, and 2017, respectively
	Described: Organizational Maintenance: Group-dynamics strategies were integrated into all PA classes for older adults offered by St. Mary’s Home Support Services and programs were sustained in the community.		Described: Reach: 115 of the 181 patients referred to the study (63%) agreed to participate	Described: Reach: FitEx HE recruited n=1097 and ALED n=27 Effectiveness: Both programs improved physical activity Mincrease 7.98±25.46 minutes/week, p>0.05). Implementation: FitEx HEs took less time to deliver their program when compared to ALED (p <0.05)	Described: Reach: Over the years, recruitment numbers decreased 766 to 320 to 290, though participants were representative of the region they were recruited from. Identifying a specific denominator for rate was difficult given the varied recruitment strategies used. Adoption: Total of 18 agents delivered the program. Two health educators were retained across all three years and five health educators delivered the program for at least two of the years. Implementation: Compliance with process evaluation was limited
	Not reported: Adoption Implementation	Not reported: Implementation	Not reported: Adoption Implementation	Not reported: N/A	Not reported: Individual-level maintenance
Decision-Making New Strategy or Adaptation Needed? Ready for System Integration?	Instructors encouraged to carry on developing new strategies based on group dynamics principles. Ready for integration.	Walk Kansas was identified as an annual offering for KSRE with annual updates to program newsletters and a more formal inclusion of social cognitive theory-based strategies to promote fruit and vegetable promotion Automated and online support for health educators, team captains, and participants were developed to increase ease of implementation and participation.	Demonstrate potential for adoption of efficacious group dynamics principles into clinical model Ready for integration	Updates to online interface to match innovations in computer science Testing for online facilitation of group dynamics Family-friendly version developed	Integration of smartwatches for reach and retention, and objective measure of effectiveness. District-based (i.e., local) and online training for the program available to all agents.
Translational Solution Institutionalization Sustained Implementation	Institutionalized Sustained implementation not evaluated.	Institutionalized Sustained implementation	Institutionalized Sustained implementation not evaluated.	Institutionalized 2009–2012 Program re-launch 2015 (with addition of specialist support)	Continued program delivery (as of Spring 2018) Partnerships with other state Extension systems formed to apply for national funding

Key Points.

• Efficacious physical activity interventions exist, but do not reach a high proportion of those who could benefit.

• Through our program of research, we propose an integrated research-practice partnership process to co-produce intervention strategies that is informed by evidence-based principles and practice values/resources/infrastructure.

• Our proposed model may lead to interventions that are practical, effective, reach more participants, and are sustained in practice.