Lessons Learned by Community Stakeholders in the Massachusetts Childhood Obesity Research Demonstration (MA-CORD) Project, 2013–2014

Introduction Childhood obesity is a multifaceted disease that requires sustainable, multidimensional approaches that support change at the individual, community, and systems levels. The Massachusetts Childhood Obesity Research Demonstration project addressed this need by using clinical and public health evidence-based methods to prevent childhood obesity. To date, little information is known about successes and lessons learned from implementing such large-scale interventions. To address this gap, we examined perspectives of community stakeholders from various sectors on successes achieved and lessons learned during the implementation process. Methods We conducted 39 semistructured interviews with key stakeholders from 6 community sectors in 2 low-income communities from November 2013 through April 2014, during project implementation. Interviews were audio-recorded, transcribed, and analyzed by using the constant comparative method. Data were analyzed by using QSR NVivo 10. Results Successes included increased parental involvement in children’s health and education, increased connections within participating organizations and within the broader community, changes in organizational policies and environments to better support healthy living, and improvements in health behaviors in children, parents, and stakeholders. Lessons learned included the importance of obtaining administrative and leadership support, involving key stakeholders early in the program planning process, creating buffers that allow for unexpected changes, and establishing opportunities for regular communication within and across sectors. Conclusion Study findings indicate that multidisciplinary approaches support health behavior change and provide insight into key issues to consider in developing and implementing such approaches in low-income communities.


Introduction
In the United States, the prevalence of childhood obesity is high: 16.9% of children and adolescents aged 2 to 19 years were obese in 2011-2012 (1). Racial/ethnic and socioeconomic disparities between children of normal weight and obese children also persist (2)(3)(4). Obesity is a multifaceted disease, demanding sustainable, multidimensional approaches that support change at the individual, community, and systems levels (5)(6)(7). Multidisciplinary approaches are more successful in addressing childhood obesity than are single-site interventions (8,9). A 2016 review showed the promising results of multicomponent community-based interventions designed to prevent childhood obesity (10). In public health research, multidisciplinary interventions play an important role and should be emphasized (11)(12)(13)(14). Funded by the Centers for Disease Control and Prevention, the Childhood Obesity Research Demonstration (CORD) project addressed this demand by incorporating evidence-based approaches (15). CORD is a multisite program that was implemented from September 2012 through August 2014 in Massachusetts, California, and Texas. Obesity is most prevalent in families with low socioeconomic status (4); therefore, CORD targeted underserved children aged 2 to 12 years (15).
This study focused on the Massachusetts site of CORD (MA-CORD). Evidence-based interventions were implemented in 5 community sectors: health care; early care and education; the Special Supplemental Nutrition Program for Women, Infants, and Children (WIC); schools; and after-school programs (16,17). Interventions targeted 5 key behaviors: fruit and vegetable consumption, sugar-sweetened beverage consumption, physical activity, screen time, and sleep duration. These behaviors have strong associations with children's weight development (17). To date, little information is known about the successes and lessons learned from a stakeholder's perspective for implementing multidisciplinary interventions. A Cochrane review called for more qualitative research as part of intervention implementation (18). Although researchers can gain valuable insight from stakeholders' experiences with interventions such as MA-CORD, few studies provide a detailed qualitative account of the implementation process (9,18,19). This qualitative study addressed this gap by outlining successes and lessons learned from the perspective of community stakeholders directly engaged with MA-CORD, including stakeholders from after-school programs, elementary and middle schools, health care, WIC, the parks and recreation department, and coordinators from each community. Methods MA-CORD was implemented in 2 communities in Massachusetts (population, 40,545 and 94,958) from September 2012 through August 2014. Poverty rates in both communities are approximately twice as high as the state's average, with a mean income per capita between $12,600 and $14,500 lower than the state average (20,21). Both communities have large non-Hispanic white (~68%) and Hispanic (16%-22%) populations. Interventions were implemented in multiple community sectors (Figure). Details on the intervention components and evaluation design for MA-CORD are available elsewhere (16,17,22). Stakeholders from all sectors who were directly (eg, teachers, pediatricians) or indirectly (eg, school principals, program directors) engaged in implementing MA-CORD were invited by email from October 2013 through April 2014 to participate in an interview. We had no other inclusion or exclusion criteria. Up to 2 follow-up emails were sent; stakeholders who did not reply after the third email were counted as nonresponders. We contacted 183 stakeholders and 40 (22% response rate) completed an interview. The study was approved by the institutional review board at the Harvard T.H. Chan School of Public Health. Stakeholders received a $20 gift card as compensation.
A semistructured interview guide was developed to support standardization of interview procedure (Box). Two authors (A.A., C.G.) conducted all interviews by telephone from November 2013 through April 2014. One interview was conducted with 2 stakeholders, the previous and current coordinator from 1 community, resulting in 39 interviews with 40 participants. Demographic information was collected at the end of each interview. The average interview length was 34 minutes, with a range of 16 to 87 minutes. Is there anything I haven't asked about MA-CORD that you think is important for me to know?

Data analysis
Audio files were transcribed and transcripts were reviewed for accuracy by 1 interviewer (C.G.). Final transcripts were entered into QSR NVivo 10.0 (QSR International Pty Ltd). Data analyses were conducted by using the constant comparative method based in grounded theory. An inductive approach was used (11,23). One coder (C.G.) read 5 randomly selected transcripts representing different sectors to develop a coding framework that reflected successes and lessons learned. This framework was then discussed (A.A., C.G., K.K.D.), and 2 coders (A.A., C.G.) coded 5 additional, randomly chosen interviews. Coding was compared and discrepancies were resolved by the 2 coders (A.A., C.G.). Additional categories were also discussed and added as needed. Remaining transcripts were coded by 1 coder (C.G.). The framework was scrutinized for overlap and subcategory relevance, and a final framework ( Table 1) was developed by 3 authors (A.A., C.G., K.K.D.) To attain reliability within the coding process, each decision on changes to the codebook was discussed and documented. Additional coding was conducted if needed. During data collection, an audit trail was used to track interview participants and procedures (24). All 3 coders have a background in public health and experience in qualitative research.

Results
Of the 40 stakeholders, 20 were from schools, 8 from health care, 4 from after-school programs, 3 from WIC, 2 from parks and recreation, and 3 were coordinators from the communities ( Table 2). A summary of key successes and lessons learned follows, along with an illustrative quote. Additional quotes are provided in Table  3.

Success stories
Intervention acceptability. Most stakeholders (24 of 39, 62%) supported the program and made it a medium or high priority, and most (27 of 39, 69%) felt that MA-CORD fit into their organization, for example, by delivering similar messages. One stakeholder from WIC said, "I think [MA-CORD] should just be a normal part of everyone's curriculum and messaging." Increase in parent involvement. About half of stakeholders (20 of 39, 51%) reported an increase in parent involvement. They observed higher participation rates in activities at schools and afterschool programs, increased involvement during appointments at health care and WIC offices, and children bringing more healthful lunches to school. Stakeholders pointed to consistent messaging about 5 key behaviors throughout the community, an increase in community-wide strategies, and awareness of childhood obesity as reasons for these changes. A health care stakeholder noted, "The parents are asking questions. They're more engaged when they come in for the visit. . . . Parents are actually coming over to the table asking questions, asking for the brochures -never happened before." Increased linkages. Two-thirds of stakeholders (26 of 39; 67%) reported improved connections to community resources, such as food services, the Safe Routes to School program, Head Start, and several community parks. Nine (23%) stakeholders said that visible and consistent messaging about MA-CORD and events helped to create linkages between community agencies and foster greater collaboration within organizations. As a WIC stakeholder noted, "We counsel on these same messages, so it's great that they're hearing it out in the community, too, whether it be at Head Start, at the park, at different after school programs."

PREVENTING CHRONIC DISEASE
Opportunities to implement new activities. Most stakeholders (35 of 39; 90%) participating in MA-CORD were able to implement new activities to support increased physical activity and improved nutrition, such as regular walks to school, providing physical activity equipment, adding more healthful choices for breakfast and lunch in schools, offering more fruits and vegetables in schools and after-school programs, and changing menu options in public restaurants. One school stakeholder mentioned, "I've always done something with a walking program, but I really focused a lot on that. We have a walking club. I do it every morning early on. A lot of these things have started or have continued because of the program." Opportunities to change policies and/or organizational environment. About half of stakeholders (20 of 39; 51%) talked about changes in the policy or food environments, such as eliminating vending and soda machines, providing water instead of soda, and changing the staff handbook to discourage staff consumption of unhealthful snacks in front of the children. A school stakeholder noted the following: Stakeholders' behavior change, buy-in, and perceived responsibilities as role models. Sixteen (41%) stakeholders reported positive changes in staff and child behaviors. In schools, several stakeholders reported that school staff made more healthful choices to model behaviors and that children subsequently changed their eating behaviors. As one teacher said, "I used to bring in a salad every morning. . . . My students actually started doing the same. Instead of eating chips and cupcakes and cookies every day, I'd say probably at least one-third of my kids started bringing in salads in the morning and healthy snacks." Nine (23%) stakeholders indicated that awareness about childhood obesity and the 5 key behaviors increased. Stakeholders mentioned that they are more aware than before that children are watching too much television or eating too much sugar or that parents are sending requests for more information about the MA-CORD program.
All stakeholders said that they would participate in MA-CORD again, because they were aware of the childhood obesity problem and the impact it was having on their communities and because they believed in the program, as stated from a stakeholder from a parks and recreation department.

Lessons learned
Leadership and administrative support. Almost all stakeholders (35 of 39; 90%) reported that the presence of leadership and administrative support for the program reduced feelings of conflict between program implementation and other priorities among staff members. A school stakeholder mentioned, "We have very, very good support . . . with the principals in each building. They're extremely approachable about anything that we ask. If we say, 'Hey, you've got an assembly coming up. . . . Can one of those have a MA-CORD component?' They're like, 'Okay.'" Likewise, the challenges resulting from a lack of buy-in from leaders were described by a school stakeholder who experienced challenges with program implementation when administrative support waned: "They do not even mention it [MA-CORD] anymore. . . . Last year it was 'We want you to do this curriculum,' and this year it's not even mentioned by the administration." Preparation for unexpected changes. Most stakeholders (22 of 39; 56%) named several unforeseen events during planning and implementing MA-CORD. Turnover caused by retirements, job loss, and resignations was experienced at all levels of staff. A stakeholder from the health care sector said, "The school department, they're so understaffed right now. . . . Trying to get into the school department to try to spread the message or be involved is tough." Also, new staff were hired and became part of the implementation process. Another unpredictable event was inclement weather, which lead to cancellations of many trainings in the school and after-school sectors, causing delays in program implementation. Early involvement of stakeholders to assess existing resources. Twelve stakeholders mentioned the importance of assessing the processes and tools that organizations have in place before planning and implementing interventions. They mentioned that they already had access to resources (eg, a system to track height and weight in the clinical sector) and informational material on childhood obesity prevention before MA-CORD was implemented, and either did not understand why their systems should change, or did not find the changes helpful. A stakeholder from the healthcare sector noted, "A lot of the things that they're discussing now, we've already learned or done." Cross-sector communication was particularly important. Twelve (31%) stakeholders cited the benefits of exchanging information and ideas during cross-sector training sessions, which helped them to explore new ideas and to discuss their experiences with intervention components and events they had planned. Additionally, stakeholders addressed a communication tool, such as an online platform as opportunity to discuss what is and is not working. An afterschool stakeholder said, "The opportunity to share with the other teams and hear what they're doing, working with the administrators of the program and the specialists to get ideas has been good." Account for family life circumstances and other barriers. Although a range of strategies were used to accommodate the various needs of families to improve involvement in MA-CORD, 19 (49%) stakeholders named families' lack of financial support and transportation challenges as two of the most common reasons for low program attendance. One WIC stakeholder mentioned, "Our participants are coming in with a range of needs including housing, lack of food, other social issues. Sometimes nutrition is not what we talk about."

Discussion
Overall, we found a high level of stakeholder and community buyin to MA-CORD with all stakeholders reporting they would implement MA-CORD again. Stakeholders said that the program was a priority for their organization because it was consistent with their organization's goals and provided opportunities to implement new and old activities and policies and support existing ones. Other studies show that changing existing policies or using new policies can ensure program sustainability (19). A novel finding of this study is that stakeholders served as positive role models for families and were motivated to change their own behaviors. These successes may be due to the fact that MA-CORD was implemented by community organizations rather than by researchers. This type of experiential learning can be a motivational tool for behavior change when working with community stakeholders.
Half of all stakeholders described increases in parent participation in activities. Parent involvement is necessary for successful implementation of child health interventions (13,25,26). MA-CORD used diverse strategies for approaching and involving parents; these strategies ranged from in-person counseling at WIC and health care visits, school events that included a MA-CORD media competition (27), and materials promoting the 5 target behaviors that were distributed across sectors. Stakeholders also observed that families faced many challenges beyond nutrition; these are described elsewhere (28). In future interventions, parent involvement could be further enhanced through a more holistic approach that moves beyond a focus on children's diet and physical activity.
Although levels of community and stakeholder buy-in were high in both communities, levels of administrative and leadership support were sometimes low. During these periods, other events, such as an anti-bullying program, were given higher priority. A strong communication strategy directed toward administrators and leaders can help gain their necessary support. Regular staff turnover, particularly in schools, created challenges, because training new staff was logistically problematic. Developing a comprehensive training manual and using a train-the-trainer model may have alleviated some of these challenges. Unforeseen events can be addressed effectively if the project anticipates these possibilities from the beginning. Training sessions were often difficult to reschedule given the number of people involved. In the future, it may be advisable to prepare web-based trainings as alternative. Finally, stakeholders were enthusiastic about cross-sector interactions and communication. However, few of these opportunities were provided in MA-CORD. Future programs would benefit from creating multiple opportunities for cross-sector training and learning collaborations to permit the sharing of resources and lessons learned.
Qualitative studies add to existing epidemiological and behavioral evidence because they may suggest ideas for adapting interventions to community and individual needs (29). This study has several limitations. First, a low response rate could indicate a selection effect in which only the stakeholders most committed to MA- www.cdc.gov/pcd/issues/2017/16_0273.htm • Centers for Disease Control and Prevention CORD chose to participate. Another limitation was the use of convenience sampling. Aside from stakeholders' existing involvement with MA-CORD, no other exclusion criteria were defined. As a result, our sample over-represents stakeholders from the school sector. Because we invited all eligible stakeholders to participate, chances were high that a higher portion of school participants would be interested in participating. Finally, because MA-CORD was implemented only in 2 low-income communities in the northeastern United States, findings may not be generalizable to all communities; however, providing a detailed description about the study sample and the 2 intervention communities may still help other researchers to apply our results to their studies (17).

PREVENTING CHRONIC DISEASE
This study contributes to implementation research by identifying important successes and lessons learned in the context of a multisite and multisector program to prevent and control childhood obesity. The insight gained through this process will benefit future interventions by streamlining the implementation processes and anticipating challenges before they occur (18). Tables   Table 1.  Increase in parent involvement Increase of parent participation and interest in activities related to childhood obesity (eg, participation in school programs, greater interest at physician appointments). Includes information about parents behavior change since MA-CORD.
Increased linkages Increase of collaboration, communication, and connections, either within the community or within the organization.
Opportunities to implement new activities Opportunities to implement or maintain new activities (eg, nutrition, physical activity, policies) with the help of MA-CORD.
Opportunities to change policies, organizational environment, or both Stakeholders talking about the opportunity to change policies, the organizational environment, or both to prevent and control childhood obesity with help of MA-CORD.
Stakeholders' behavior change, buy-in, and perceived responsibilities as role models Change in stakeholders' behaviors and how that might have influenced children's behaviors.
Stakeholders' future intention to participate in MA-CORD The answer to the interview question "If you were given the choice to be part of MA-CORD again, would you chose to?" was coded here.
Lessons learned Leadership and administrative support Information given about the importance of support needed to implement MA-CORD (eg, leadership, staff, administration).  "When I first changed the policies for the staff handbook, there was no negative feedback. They completely understood, and they understood that they have to be the positive role models." (After-school program) "We can't just preach it to the kids, we have to model it." (After-school program) Stakeholders' future intention to participate in MA-CORD "'Cause I think it's so important. I think that we need to focus on these things. WIC is a perfect partner to help with that because of the number of kids that we see, the number of families that we interact with and have a positive effect on them. Absolutely, I would hate to see us not participate." (WIC) "I think that the concept and the structure of it is a really good model for other communities to follow. I feel like policy system and environmental change really provide the biggest impact at the community level, versus working with individual-level behavior change. Then, I feel like the model, in terms of all the sectors, with the consistent messaging, is also [a] best practice that other communities should be looking into. Everyone is on the same page with a common vision." (Parks and recreation) "I would. I think it's a good program." (School)