Children's Cooperation Denmark: a 3-year System Dynamics Trial

The Child-COOP trial was developed according to the United Kingdom Medical Research Council's framework for developing and evaluating complex interventions. Two feasibility studies were initiated in 2021 and still running - and Child-COOP was subsequently adapted and adjusted.

INTRODUCTION Lack of physical activity (PA) and sedentary lifestyle in Danish children is a major challenge. New strategies are needed to combat this development. Early awareness and promotion of healthy PA behaviour is important, as PA behaviour in childhood often is manifested across adolescence and into adulthood. Social inequity in health is also mirrored in PA behaviour among children. Children in families with few resources are known to be less physically active compared to peers in more affluent families. One way to promote healthy PA behaviour in children, which also considers inequity challenges, is to develop more health-enhancing active leisure communities. Unfortunately, dropout from active leisure communities is common in the transition from childhood to adolescence. These high dropout rates call for identification of initiatives that encourage children to engage in active leisure communities throughout childhood and youth. However, both sedentary behaviour and moderate-to-vigorous PA have been found to be independent predictors for health status, and they are not mutually related. Hence, focussing solely on reaching the recommended levels of daily PA is not sufficient. Instead, a comprehensive strategy is warranted to ensure that more time is spent on light, moderate and vigorous PA and less time on sedentary activity. Promoting (initiating and sustaining) a healthy PA behaviour is not an easy task because PA behaviour is shaped by a complex interplay between a wide range of factors at individual, family, community and society levels. Interventions building on a participatory system dynamics approach has been suggested as a feasible way to address such complex problems as they combine the current evidence base on prevention, best practice and local wisdom to achieve new knowledge and create solutions. A participatory system dynamics approach implies that a variety of engagement and solutions at different system levels may occur at both the interpersonal and organisational levels. This strategy has already proven successful in several international childhood obesity prevention interventions. The participatory system dynamics approach remains to be tested in a large-scale Danish context. During the past year, our research group has tested the participatory system dynamics approach in different settings in Denmark. We propose to conduct a participatory system dynamics trial across ten communities in five municipalities to create local solutions to promote healthy PA behaviour in Danish schoolchildren through Child-COOP.

AIM

The overall aim is to evaluate a participatory system dynamics approach to promote (increase and sustain) healthy PA behaviour in 6- to 12-year-old schoolchildren in local communities in Denmark through the following sub-aims:

• Measuring the effectiveness on the children's PA behaviour by assessing changes in time spent sedentary and time spent at light, moderate and vigorous intensity at three-year follow-up
• Estimating changes at system level such as assessing changes in community systems, community networks structure, policy and practice at three-year follow-up
• Identifying health-economic implications by assessing both a cost consequence analysis and an explorative analysis of the derived effects at three-year follow-up
• Conducting a process evaluation to map a final systems program theory on "what works for whom under what circumstances" to be used as a framework in future recommendations and up-scaling.

TRIAL COMPONENTS

The Child-COOP trial consists of three stages:

Stage 1: Preparation: selection of communities, training and monitoring (12 months) Selection of communities: The communities will be selected and defined by the municipalities based on their local challenges related to children's PA behaviour, socially deprived areas and motivation.

Training: Employees from all participating municipalities will receive training in how to run local community initiatives based on a participatory system dynamics approach. The target group for the training is Child-COOP municipal coordinators and other employees in the municipality taking part in the project. The training will consist of seminars and homework. The purpose of the training is to build capacity for municipal employees to understand and act within the framework of the participatory system dynamics approach and to solve complex health challenges regarding children's PA behaviour.

Monitoring: A local childhood health profile will be generated to outline daily PA behaviour, sleep patterns and leisure time activities at baseline (based on survey and objective data) and factors associated with PA behaviour in children, e.g. mental health, screen use and parental support. This profile will be generated at baseline based on survey and objective data (see details in "Individual level: data collection and measurements").

Stage 2: Recruitment, group model building (GMB) and systems mapping (9 months) Recruitment of key leaders and local stakeholders: A coordinator from each municipality will together with the research team identify and recruit key leaders (e.g. local politicians, department heads, municipality leaders) and local stakeholders (e.g. school board members, school principal, school nurse, sport club representatives). Key leaders and local stakeholders will be selected based on authority and capacity to initiate actions that are likely to influence the children's PA behaviour across sectors and organisations. During this stage, key leaders must commit to allocate resources to ensure subsequent implementation. The aim is to recruit 12-15 key leaders and local stakeholders from each community.

GMB process and systems mapping: During this stage, three workshops (WS1, WS2 and WS3) will be held at each participating community. WS1 and WS2 will engage the key leaders and stakeholders, and local childhood health data will be presented to provide the first critical engagement step. In these two workshops, the key leaders and stakeholders will map the system by building a causal loop diagram (CLD) to understand how the perceived local system affects the children's daily PA behaviour in their community. In WS3, all community members willing to engage in changing the local system will be invited to identify priority areas for action based on the developed CLD from WS1 and WS2.

Stage 3: Implementation of actions, support and monitoring (24-30 months) Actions and support: The output of WS3 is the formation of local working groups that will focus on implementing the chosen actions. The working groups will be supported and supervised by a backbone office consisting of the Child-COOP municipality coordinator(s) and the research team. A follow-up workshop (WS4) will be held with the key stakeholders six months after completion of WS3 to review the consolidated priority actions. To increase and maintain motivation and actions in the local community, subsequent follow-up meetings will be held with the working groups when needed during the 24-30 months after WS3.

Monitoring: Survey childhood health data will be collected from the study population on an annual basis and from the newly enrolled schoolchildren (first grade) every year (see details in "Individual level: data collection and measurements"). At an annual meeting in each community, these data on the children's health and well-being in the local community will be presented to actions groups, key leaders and stakeholders to sustain and promote engagement.

RESEARCH PLAN

Child-COOP will be evaluated through an assessment of outcomes at individual and systems level, process and economic implications. Combined, these three elements will give an essential understanding of what it would require to upscale Child-COOP to regional or national level. Few guidelines exist on how to evaluate public health programs in complex adaptive systems. Child-COOP is inspired by both the Medical Research Council recent evaluation guidelines for process evaluations of complex interventions and the newly developed ENCOMPASS framework aiming to generate a comprehensive, yet practically realistic evaluation. The evaluation will be structured along the realist research cycle, and a critical first step is to develop an initial program theory to inform how the participatory system dynamics intervention is expected to produce change. Given the participatory approach in Child-COOP, the exact outcomes and actions are unknown in advance. Consequently, the evaluation will be constructed to allow for systematic documentation of processes and outcomes. A final systems program theory on "what works for whom under what circumstances" based on the participatory systems dynamic approach will be made, including recommendations for large-scale implementation across different settings.

System level: data collection and measurements

Engagement and commitment will be measured by a survey to document changes at community and organizational level. The survey will be distributed to all participants attending workshops and will enrol new workings groups' members continuously at base line (defined as first attendance) after one year and at two-year follow-up.

Participatory mapping of communities will be assessed through online workshops with the participants attending the workshops. The online workshops will supplemented by survey data provide evaluation and formative insight into the local network of each community.

Actions implemented will be evaluated by using Ripple Effects Mapping method which asses the number of actions and where they have been initiated in the system. Action Scales Model is used to asses leverage point of each action implemented. Both methods will be used in workshops with working groups every 6 months for a two-year period.

Change in local systems will be understood through Causal loop diagrams (CLDs) built during the GMB workshops, which will be enriched with observation data from the workshops and survey data to quantify actions, collaborations and system changes over time and to assess the impact of actions and networks on variables in the system. To support the GMB method, the software program "Systems Thinking in Community Knowledge Exchange" (STICKE) will be used together with internationally developed, tested and standardised scripts.

Process evaluation: data collection and measurements

Quantitative data for the process evaluation will be collected with the same procedures as described at system level (to assess reach). Moreover, we will also use the results of "ripple effect mapping" (REM) to understand the solutions and actions developed in the local communities. Additional data will be collected through interviews, focus groups, structured observations and follow-up interviews before, during and after the GMB workshops to access the overall process. The process evaluation will also include structured observations/checklists collected during workshops (to assess dose). Interviews or focus group discussions with workshop participants and participants dropping out of the GMB processes will be used to assess participant responsiveness. These data will be used to understand the mechanisms in Child-COOP. Furthermore, adherence to the participatory system dynamics approach will be assessed by routinely (every 6 months) collected information regarding implementation, engagement and actions of key leaders and stakeholders, political will and advocacy. In addition adherence, dose, quality of delivery, participant responsiveness and reach will be assessed to gain knowledge on how the participatory system dynamics approach was delivered and implemented and to inform a final program theory.