Developing capacity for learning community systems: Experiences from the 100 Million Healthier Lives SCALE Initiative

Abstract Introduction This paper explores the capabilities that contribute to community transformation and the common pathways followed by communities in the 100 Million Healthier Lives SCALE (Spreading Community Accelerators through Learning and Evaluation) initiative in their transformation journeys towards a “Culture of Health”. Methods Funded by the Robert Wood Johnson Foundation (RWJF), from 2016 to 2020, between 18 to 24 community coalitions nationwide participated in SCALE, the goal of which was to co‐design, implement, test, and scale up a model called the Community of Solutions (COS) Framework, that built community capacity around a set of skills and behaviors to advance culture change and create sustainable improvement in health, well‐being, and equity. We adapted and applied two qualitative research techniques, meta‐ethnography and participatory action synthesis, to evaluate SCALE initiative data. Results Eight concepts emerged that represent the knowledge, capabilities and practices commonly acquired and utilized across the communities. Overall, these concepts emphasize individual and team leadership, quality improvement skills, an intentional focus on equity, and partnerships for spread and sustainment. Concepts were linked into lines of arguments which were unique storylines explaining the transformation pathways. Three stories of the transformation process emerged from the data. Causal Loop Diagrams (CLDs) were created to represent non‐linear system relationships and visually capture some of the most important dynamics of the process of transformation. Even with vast heterogeneity among the SCALE communities and the diversity of activities that the communities undertook, our analysis showed there were a few basic principles that undergirded the process of building capability for transformation. Conclusions The knowledge from our findings should be useful to expand further research and practice in community learning systems.

The COVID-19 pandemic has merely reinforced what has been known for a long time. Health and well-being are intimately connected to community level contexts and require collaborative, transformational solutions that are built on communities' capacity to learn, in real-time, how to improve what matters most to them. 3 In this paper, we present the implementation, evaluation, results, and learning from a multiyear, multisite project called SCALE (Spreading Community Accelerators through Learning and Evaluation), part of the 100 Million Healthier Lives Initiative, that focused on building the capacity of community coalitions to develop relationships and partnerships and learn methods to transform health, well-being, and equity. Using a detailed description of the implementation process, we describe the most common pathways that communities followed in their transformation journey, the knowledge, capabilities, practices and relationships that were used, and the constraints and barriers that communities faced. The knowledge from our findings should be useful to expand further research and practice in community learning systems.

| Context: The 100 Million Healthier Lives (100MLives) SCALE Initiative
Convened by the Institute for Healthcare Improvement (IHI), 100MLives was a global movement and collaboration from 2014 to 2020 in support of Robert Wood Johnson Foundation's Culture of Health (COH) Strategy. 4 SCALE, the flagship initiative under 100MLives, was a learning collaboration that initially was cofacilitated by four organizations working on health and healthcare improvement-IHI, Communities Joined in Action, Community Solutions, and the Network for Regional Healthcare Improvement-together with community coalitions selected through a competitive application process. The goal of SCALE was to co-design, implement, test, and scale up a model called the Community of Solutions (COS) Framework that built community capacity around a set of skills and behaviors to advance culture change and create sustainable improvement in health, well-being, and equity. 5,6 The four organizations trained and supported the communities in their improvement efforts. SCALE communities applied the COS model to a variety of topics including healthy food access, infant mortality, refugee education and health, chronic disease, self-management, youth nutrition and physical activity, homelessness, access to public parks and green space, and youth development and education. Regardless of topic areas of focus, all SCALE communities agreed to apply an equity lens to the work. During the first phase of the initiative from 2016 to 2018, 24 community coalitions were part of SCALE. Eighteen stayed on in the second phase from 2018 to 2020.

| SCALE theory
The theory of change for SCALE is the COS model shown in Figure 1. These elements of the model were influenced by the following literature on community capacity building and adapted for SCALE based on practical experience of the implementing partners. The community problem solving and change framework 7,8 specifies the need to build operational and problem solving capacities of communities to enable them to engage in comprehensive change strategies. Zakocs and Edwards 9 and Zakocs and Greenberg 10 describe leadership style, member participation, group cohesion, participatory decision-making, and involvement of local government as key characteristics for successful coalition effectiveness and capacity. Flasphohler et al 11 emphasize the need to build different types of capacity (general vs innovation specific) at multiple levels (individual and community). Research from the field of implementation science posits that while training is necessary to develop skills, training alone is not enough to achieve outcomes and that training needs to be reinforced through technical assistance (eg, coaching and problem solving), tools, and iterative quality improvement processes. 12

| SCALE intervention components
The primary activities based on the COS model is the developmentco-designed with communities-of skills of community coalition members in five areas (personal growth and leadership; working with others; design, improvement and implementation science; centering equity; and long term planning) through intense training, coaching and support activities. The key activities are shown in Table 1

| SCALE evaluation
The primary intent of the evaluation was to advance knowledge about the processes of change in the complex environments within which the communities operated, rather than attempting to assess the extent to which these processes resulted in the achievement of specific outcomes. The evaluation was intentionally designed as a participatory partnership between the funder, the implementation partners, the communities, and a multidisciplinary evaluation team led by the Universities of North and South Carolina. Recognizing the heterogeneity in coalition structures, community contexts, and areas of improvement emphasis, the evaluation approach was designed to be flexible, adaptive, and process focused. The evaluation also needed to be sensitive to the fact that much of the knowledge about the implementation was generated by and existed within the communities, and that their interpretation of the data was an important part of the knowledge synthesis process. The approach was related to learning or developmental evaluation 14,15 that are iterative methods using rapid learning cycles to improve implementation processes. The evaluation was segmented into a formative and a summative evaluation component. Both have been described elsewhere and a summary of each is presented here.

| Formative evaluation: Learning for improvement
The formative evaluation approach is described in detail in Scott et al. 16 It was focused on the skills component of the theory of change Monthly network calls-to provide additional training and support, share progress and challenges, highlight the work of one SCALE community and co-design next steps for the SCALE initiative 3 Monthly coaching and peer support in regional networks (3- Training on planning and running an equity action lab-equity action labs provide a highly adaptable structure and strategy to bring together a diverse team to make meaningful progress on a complex goal in a short amount of time (generally 100 d). 13 9 A continually refined and streamlined system to continually track progress, assess general capability and use this data for improvement.
and was designed as a learning system for ongoing adaptation of the program. Three data collection methods: inquiry, observation, and reflection were used to obtain data about the implementation to capture as much contextual information as possible. Together, these methods provided a multifaceted view of the SCALE implementation process that was synthesized and periodically shared with the implementation team and the communities to stimulate improvements to the program design and its execution.

| Summative evaluation: Collaborative meaning making
The summative evaluation was designed in the latter half of the pro-  highlight the nonlinear nature of the transformation process. Table 3 summarizes the seven phases of meta-ethnography and how they were applied to the evaluation.

| RESULTS
We describe results from both the formative and summative evaluations, organized by the questions of interest presented earlier in this paper.

| Key contributors to achieving capability for community transformation
The formative evaluation revealed four key areas that were critical  These eight cross community concepts emerged in Phase 6 of the ToAST process as the output of multiple levels of synthesis shown in Table 3. The earlier phases of this process captured data on the specific knowledge, tools, and relationships at the community level.

| Common pathways that communities follow in their transformation journey
The transformation pathways shown in Figure 2 were developed in Phase 7 of the ToAST process by linking six of the eight concepts in community] was focused on fixing the lake and not fixing the fish, which involves more long-term changes.
The "create a shared language" story was followed by community organizations that began with using community partners with deep connections within the community to develop a shared language that facilitated mutual communication and dialog. Initial conversations to strengthen the ability to communicate was seen as especially important by these communities as was the need to create safe spaces for open and honest discussions and sharing experiences. These efforts accelerated trust and relationships in these communities. A coalition member stated: We had a discussion on racism at every coalition and staff meeting. We created a safe and open space to discuss personal &/or patient experiences with racism or current events dealing with racism.
While all coalitions had a community champion from the outset, the "integrate people with lived experience (PLE) early" story emphasized developing community leaders as the first component of build- ing the core team's ability to engage with the wider community. They also used the power of the local social networks to build relationships to build relationships quickly to advance equity work.
As one core team member remarked: A focus on authentic engagement of partners, building trust and relationships was foundational to the ability to scale up and sustain work. SCALE helped us offer more leadership development opportunities for community champions and staff members. This was valuable for building local capacity to improve authentic engagement and peer mentorship opportunities to lead civic engagement and service leadership.

| Mechanisms through which communities brought about change
While the lines of argument are a convenient visual representation of how change happens, their sequential structure does not capture the complexity of the change process. CLDs allow us to better represent nonlinear system relationships. Using the data from Figure 2, the evaluation team developed the CLD in Figure 3 to visually capture some of the most important dynamics of the process of transformation.
In Figure 3, the blue arrows represent system characteristics that support transformation, while the red arrows represent constraints to transformation. The positive signs in the loops indicate that the variables move in the same direction: for example, the greater the relationships, the greater the collaborations. The input into the system is the available external funding that influences the extent to which SCALE activities can be supported. The output from the system is the SCALE goal, that is, the capability for community transformation. As   Figure 3 into opportunities. This is illustrated in a hypothetical future system shown in Figure 4. In this figure, the red constraint links in Figure 3 have become green opportunity links. As local expertise increases, the routine need for external technical assistance diminishes, leaving resources available to build specific targeted expertise as needed. This reduces the drain on program funds. At the same time, as communities increase their own ability to acquire funds, the overall availability of funds continues to increase. This enables communities to move from scarcity to abundance, which is a key component of the SCALE theory of change.

| LIMITATIONS
Our evaluation approach was designed to make the best use of existing project data, which was recorded in a variety of systems with varying degrees of completion. Consistent availability of high-quality data to meaningfully document progress and learning was an ongoing challenge.
The SCALE implementation teams introduced multiple measurement plans and mechanisms for data collection over the duration of the initiative, but in the end, there were still gaps in data quality and completeness. The ToAST process used the collective knowledge of the various stakeholder groups and triangulated data from various sources to develop the concepts and transformation pathways, but there were gaps in data that might affect the final results and conclusions. Also, our results have demonstrated progress in skills and behaviors, which are the first two components of theory of change. Given the range and complexity of the projects undertaken by the SCALE communities, there was not enough time for communities to demonstrate how this progress resulted in the achievement of the COH outcomes related to health, well-being and equity before the funding ran out. This is not unusual, and our inability to evaluate whether the capabilities acquired by the communities resulted in improved health and well-being is a serious limitation. The second CLD we have presented describes the scenario where communities have sustained financing to continue their transformation efforts. If this scenario were realized, it would be possible to follow communities over time to evaluate the impact of their work. Financial sustainability should be emphasized as a required component for future grants to facilitate this scenario.

| CONCLUSION
The United States has invested trillions of dollars in health, well-being and equity with little to show for these efforts. Simply investing more money and resources will not likely result in measurable improvement.
Complexity-informed approaches that approach well-being through the lens of learning rather than through the lens of technical program implementation are needed. SCALE is an example of such an approach. Communities learned how answers to complex questions about change are not apparent at the start of the work, and that practices of learning and testing with meaningful measures to guide progress will help uncover solutions. Learning in a context of complexity requires a willingness to "fail forward," in an environment where failures are opportunities for learning and growth rather than for judgment. Our results validate the relevance of the elements of the SCALE COS Theory of Change as the key components of skills and behaviors needed to bring about change.
The extent to which the acquisition of these skills and behaviors drive sustainable health, well-being and equity outcomes they aim to achieve should be a priority for future research.