Purpose: The feedback control system includes a flow of information through sensor, controller, and implementer processes. This study compares control system theory functions in two distinct stakeholder community groups aiming to improve community population physical activity through a locally-driven rapid improvement cycle, Investigate-Design-Practice-Reflect (IDPR).  

Methods: In Wave 1, two rural communities were randomized to one of two data feedback systems, the Wellscapes (WS) or Collective Impact (CI) condition. One community engaged in a locally-driven IDPR cycle within the WS model while the other community engaged in the IDPR cycle within the CI model. The IDPR cycle is an iterative, data-driven decision-making process allowing communities to rapidly process and feed back information into the system to achieve a population health outcome. We recorded and observed four stakeholder meetings in each of the two communities (n=8) and conducted and transcribed stakeholder interviews from each community (n=2). Three researchers analyzed the interviews using a deductive-inductive, framework analysis approach to examine data by three core elements of a control systems framework: sensors, controllers and implementer. One researcher qualitatively analyzed and coded stakeholder meeting observations data. All data was later compared and triangulated.

Results: Observations of stakeholder group meetings and interviews with meeting members from the two community groups revealed differences in sensors, controllers, and implementers in each community. We identified one overarching theme: WS stakeholders made decisions from a polycentric perspective, while the CI stakeholders made decisions from various monocentric perspectives as evidenced by flow of information (e.g., community data reports, video observations, stakeholder stories) through feedback loops to drive stakeholders’ decision making. Three sub themes emerged identifying differences between WS and CI sensors (summary data vs. data points), controllers (employed vs. volunteer), and implementers (collective vs. individual).

Conclusions: Manipulation of the control system resulted in differences in the flow of information, The WS stakeholders thought about the IDPR cycle from a holistic, big picture and systems change approach, while the CI stakeholders’ approach focused on program implementation to achieve a specific outcome. IDPR presents a way to communicate process control theory functions in communities with different architectural structures.

Purpose: Community monitoring and feedback systems provide stakeholders with timely and relevant population data to inform local decision-making around youth health behaviors. This presentation will describe the process of establishing and implementing data monitoring and feedback systems in two rural communities for investigating population-level youth physical activity (PA).

Methods: Two rural communities were randomized to the Wellscapes (WS) or Collective Impact (CI) condition. The research team collaborated with local health departments and school systems to establish data monitoring systems and data sharing agreements (DSA). Stakeholders completed training to learn how to implement the online youth PA surveillance instrument, Youth Activity Profile (YAP), and use community data. Communities adopted PA surveillance as standard educational practice, and all 3^rd–6^th graders (Fall 2018, n = 465; Fall 2019, n = 501) were eligible to complete the YAP. Students’ self-reported in-school, out-of-school, and weekend PA behaviors were used in a calibrated algorithm to estimate group-level PA. The research team provided community-specific aggregate data reports of YAP outcomes to stakeholders as part of the “Investigate” phase of the Investigate-Design-Practice-Reflect cycle. The WS community received detailed, time-segmented (i.e., in- and out-of-school, weekend) feedback about time spent in moderate-to-vigorous physical activity (MVPA). In contrast, the CI community received an overall estimate of time spent in MVPA. Mixed model ANOVAs examined MVPA by community and season (Fall 2018, Fall 2019). 

Results: YAP response rates ranged from 86.1% to 95.4%, depending on the community and season. Baseline overall MVPA for WS and CI communities was 81.7±1.1 min/day and 74.0±1.1 min/day, respectively. From Fall 2018 to Fall 2019, youth in the WS community reported a significantly greater change in in-school MVPA compared to the CI community (p = 0.007). In-school MVPA increased from 29.7±0.2 min/day to 31.0±0.2 min/day (p <0.05) in the WS community, but there was no difference in the CI community. No differences were observed in overall, out-of-school, and weekend MVPA outcomes in both communities.

Conclusions: Community data monitoring and feedback systems are essential elements of resilient communities. These systems provide locally relevant population-level data in a timely manner to inform decision-making around improving youth PA outcomes. 

Purpose: Children spend a large proportion of time in adult-organized group settings (e.g., classrooms, youth clubs), and leader practices implemented within these settings drive the social structure that produces physical activity (PA). Communities need setting-level data feedback to achieve community resilience and PA outcomes. This presentation will describe the setting-level data collection and feedback system intervention and outcomes of the Wellscapes (WS) Project.

Methods: During Fall 2018 and Fall 2019, we video-recorded organized group settings (n=44) and meetings (n=130) and simultaneously collected accelerometer data for attending children. Meetings were time-segmented into smaller units (i.e., sessions). Sessions were coded for purpose (e.g., academic, PA) and matched with children’s accelerometer data. The number and duration of sessions with a PA purpose and mean percentage of time (%time) in moderate-to-vigorous PA (MVPA) were assessed for each meeting. Mixed effects models examined changes in number and duration of implemented PA sessions and MVPA outcomes. Communities were randomized to a WS or Collective Impact (CI) feedback condition. All communities received quarterly data reports providing descriptive data on setting %time in MVPA. The WS condition also received specific group type (e.g., school grade, sport type) MVPA outcomes and setting implementation data on number of PA sessions.

Results: Across all community settings, the number of implemented PA sessions did not differ from Fall 2018 to Fall 2019. A significant community-by-season interaction (p=0.04) for the school setting showed the WS condition had a greater increase in duration (minutes) of implemented PA sessions (6.5±1.6 to 16.2±4.1) than the CI condition (20.7±5.2 to 18.0±4.5). WS had a greater increase (p=0.03) in school setting %time in MVPA (3.2±0.6% to 6.3±1.1%) than the CI condition (4.9±0.9% to 4.6±08%). The data reports provided to each condition were used as part of the Investigate-Design-Practice-Reflect cycle.

Conclusions: Setting-level data feedback within the WS condition may have contributed to greater implementation of PA sessions and effectiveness in improving school MVPA compared to PA outcome data alone in the CI condition. Setting-level data feedback on leader practices and PA outcomes may inform community data-driven decision-making for improving youth PA.