healthmedicinet

Setting

Partnership

A partnership consisting of The Danish Cancer Society, The Danish Foundation for Culture
and Sport Facilities, and the Danish foundation Realdania had the vision to increase
PA in primary schools in Denmark by redesigning and renovating schoolyards. Together
they launched the Activating Schoolyards Campaign. The campaign had a budget of approximately
8 million USD, including 2 million USD of local co-funding. The Danish Cancer Society
funded the development of study and the scientific assessment. The Partnership appointed
a campaign secretariat that was responsible for all practicalities involved in the
recruitment process.

Primary schools in Denmark

In Denmark school is mandatory for children between the age of 6 and 16 years. Public
schools are free of charge and students do not wear school uniforms. Schools are typically
divided into junior (0-3 grade, 6-9 years old), middle (4-6 grade, 10-12 years old)
and senior (7-10 grade, 13-16 years old) tiers 26]. Each class has a maximum of 28 gender-mixed students. Students participating in
this study attend school for 33 (grade 4-6) and 35 (grade 7-9) hours per week. Approximately
60 min are allocated to recess per day, being distributed over two to four recess
periods 26]. In general, the lunch break is the longest recess, lasting 25-30 min. Recess is
typically characterized by free play without any organized curriculum. Teachers on
yard duty are supervising the students handling conflicts and accidents. Some schools
organize ‘Play patrols’ with middle block students organizing games to activate junior
students. The junior students must often stay outdoors during recess. There is wide
variation in whether schools have an outdoor recess policy for middle-and senior tier
students. Seniors are allowed to leave school during recess at most schools.

Study design

The design is based on a quasi-experimental long-term follow-up study of students
attending selected primary schools (grade 4-8) in Denmark. To be able to accommodate
both an exploratory and an evaluating part of the study, a range of qualitative and
quantitative methods were employed to facilitate exploration and evaluation. The Activating
Schoolyards Study is divided into four main parts: 1) exploratory study; 2) effect
evaluation; 3) process evaluation; 4) post-intervention user-evaluation. The studies
were divided into two different phases: 1) the project development pre-study phase
and; 2) the study phase. The aim of the studies conducted in the pre-study phase was
to provide input and create inspiration for the interventions. The aim of the study
phase was to evaluate the Activating Schoolyards Study in terms of effect, process,
and user-perspective. The study design with its different sub-studies and phases is
illustrated in Fig. 1.

Figure 1. Illustration of study design, timeline and methods

Case selection

The project schools (cases) were selected by means of an open competition in order
to stimulate local engagement and participation in the development of the interventions
23]. In October 2012, all primary schools in Denmark (approximately 1800) were invited
to submit a vision proposal for improvement of their schoolyard. Out of the 106 submitted
proposals, 17 cases were selected for further development in April 2013 by an evaluation
panel appointed by the Partnership. Each of these 17 cases received approximately
USD 17,000 to further develop their vision in self-constituted case teams comprised
of external consultants chosen by the schools (architects, landscape architects, designers)
and stakeholders (students, teachers, parents, neighbors, and local organizations).
The 17 project proposals were submitted in December 2013, and in February 2014 the
evaluation panel selected seven cases for realization. The case selection process
is presented in Fig. 2.

Figure 2. Flow diagram of case selection

The evaluation panel selected both the vision proposal and the final project description
to favor the following selection criteria: innovative solutions promoting PA, inspiration
to other schools, focus on less active target groups, integration of the schoolyard
in the surrounding local area, organizational initiatives to support the intervention,
student and stakeholder involvement, and diversity of locations and target groups.
The selected projects had to document that they could provide at least 50 % of the
budgeted cost of the project. The total budget for each of the projects ranged from
120,000 to 900,000 USD. The seven cases represent a wide range of schools. As shown
in Table 1, the seven cases differed considerably in geographical area, school type (urban or
rural), number of students enrolled (middle and senior tiers), socioeconomic status
(based on parental income), share of students with a non-Danish ethnicity, square
meters of schoolyard per child, number of play facilities, recess duration, number
of playground duty teachers, recess rules, and organized play activities during recess.

Table 1. Case characteristics regarding the study target group; middle and senior block students

Development of interventions

The interventions were developed using a participatory bottom-up approach inspired
by Community-Based Participatory Research ideas 27]. Building on existing capacities in the ‘case’ community, the interventions (e.g. target groups, areas and components) submitted in the project proposals reflected
local challenges and needs. The interventions contain both physical and organizational
changes. During the intervention development process, all case teams had access to
a campaign website that provided various materials for inspiration including a large
number of short thematic case descriptions of other schoolyard renovation projects,
as well as short videos with interviews with students, school principals and researchers.
The case teams were also obliged to attend two workshops. In May 2013, a start-up
workshop was conducted for the 17 case teams aiming to provide inspiration, stimulate
innovation and share knowledge from previous schoolyard interventions. Moreover, findings
from the exploratory study on the students’ perceived barriers for recess PA were
presented at this workshop to inspire the development of the organizational changes.
A second workshop was organized for the seven case teams in February 2014 focusing
on qualifying and anchoring the projects, and providing inspiration for organizational
initiatives. Furthermore, the process evaluation was designed to help the case teams
think through the decisions made during the intervention development.

It was left up to each case team to decide if and how the provided information and
feedback could be incorporated. The whole process led to highly tailored interventions
with considerable variation in intervention components between the seven cases. In
some of the cases the interventions took place in the existing schoolyard whereas
other cases expanded their outdoor area by including adjacent spaces (e.g. forests and streets). Even though the design and dimension of the intervention components
varied widely, some features were present in several cases, e.g. the introduction of climbing walls, balance-bars, amphitheater-stages, skating areas,
trampolines, and outdoor lunch eating areas. There were also similarities in the planned
organizational changes, e.g. implementation of a movement policy and changes in recess duration. An overview of
the intervention elements per case can be found in Table 2. All interventions will be implemented between summer 2014 and summer 2015.

Table 2. Planned intervention components

Data collection and measurements

As described above, the study consists of different parts and each part has its own
data collection method and measures, described in more detail below. Prior to the
Activating Schoolyards Study a pilot study was conducted to test objective and subjective
measurements of PA and classification of movement behavior patterns using accelerometers,
global positioning system (GPS), questionnaires, class-diaries and interviews. Based
on these findings small adjustments were made to improve the data collection procedure.

All parents of the students who participated in our study provided active informed
consents, and all participants could withdraw from the study at any time. Data were
collected in accordance with the Helsinki declaration. The study and its data-management
procedures have been approved by the Danish Data Protection Agency (2013-41-1900 and
2014-41-2801).

Exploratory study

The aim of the exploratory study was to get an understanding of the students’ PA patterns
and perceived barriers for PA during recess 26]. Non-systematic participant observations were conducted to gain insight in the students’
movement patterns, behavior and social interaction during recess 28] whereas interviews were carried out to gather in-depth data of the students perceived
barriers for PA during recess 29], 30]. To facilitate the conversation and evoke memories the interviews were carried out
in groups walking around in the schoolyard inspired by the go-along interview approach
31]. Data were collected during a one-day visit to the 17 cases selected for further
development between April and June 2013. A total of 460 min of recess were observed.
The observations were documented with field notes and photos 32]. A nominated teacher who knew the students was asked to identify three boys and three
girls from fourth grade classes (10-11 years), representing different levels of PA.
We recruited children representing different levels of PA to avoid stigmatizing of
the least active children and to make generalizations of this group more reliable
33]. Seventeen go-along group interviews (one in each case) were conducted. In total
111 students (53 boys and 58 girls, mean age 10.4 years) participated in the go-along
group interviews. The go-along group interviews lasted for approximately 60 min and
were conducted during school hours.

Effect evaluation

The aim of the effect evaluation is to examine the effect of the tailored interventions
in each of the cases on student’s PA and movement patterns during recess. Movement
pattern is defined as PA intensity levels at specific geographic locations. The primary
outcome is the difference in the objectively measured average activity level (in counts
per minute, CPM) during recess in the schoolyard, before and after the intervention.
The secondary outcomes are more exploratory examining intervention effects for the
least active students, and exploring the change in behavior in specific areas of the
schoolyard.

Baseline data were collected April to July 2014, and follow-up data will be collected
in the same period (April to July) in 2016. A combination of accelerometers, GPS and
geographic information system (GIS) was used to assess behavior changes in time and
space in each of the seven cases. Objective PA was recorded as an activity-count every
15 s using the ActiGraph accelerometer model GT3X. The ActiGraph accelerometer has
previously been recognized to provide acceptable validity and reliability for measuring
children’s activity levels and energy expenditure 34], 35]. The students’ locations were measured every 15 s using QStarz BT-Q1000xt GPS trackers.
The Qstarz GPS units have a median dynamic positional error of 2.9 m in real-world
conditions, within various urban environments and during different modes of transport
36]. The schoolyards were mapped in detail using ArcGIS 10.2 and the total schoolyard
area was calculated. During the week of measurements all participants completed an
electronic survey, inquiring about self-reported PA, neighborhood and school experiences,
and background characteristics.

The students were asked to wear the accelerometer and GPS in an adjustable elastic
belt around their waist for seven consecutive days. The equipment was not worn overnight.
Verbal and written instructions on wearing of the equipment were given to the students
by the research team. To increase compliance short reminder text messages were sent
out to the participants’ mobile phones twice a day. Two to three randomly selected
participants in each class were asked to fill out a short timetable diary containing
short questions about their school day and PA during class. Furthermore all schools
provided detailed class time tables for the data collection period. At baseline the
overall response rate was 52 % with 744 out of 1224 students in grade 4-8 participating.
The response rate differed between school and class with a maximum rate of 82 %.

Process evaluation

The aims of the process evaluation were to facilitate the intervention development
process and to identify barriers and facilitators in the implementation process. To
help facilitate intervention development, the process evaluation was designed based
on formative process evaluation principles 37], 38]. The process evaluation was carried out using an electronic survey to the 17 school
principals from the cases selected for development in June 2013 and focus group interviews
with the 7 final case teams were conducted in April 2014, and will be conducted in
spring 2016. The survey included questions about rules and policies regarding recess,
PA, outdoor teaching and activities outside school hours. Furthermore the schools
were asked about their initial plans and expectations towards the process.

The focus group interviews included between 5 and 10 members of the final seven case
teams and the interviews focused on the case teams’ experiences during the project
development process and their expectations for the coming implementation process.
The interviews took place at the intervention schools and lasted approximately 90 min.
The second focus group interviews with the case teams in spring 2016 will provide
insights to the organizational changes implemented in each of the cases, as well as
the intervention implementation process.

Post-intervention user-evaluation

The aim of the in-depth post-intervention user-evaluation is to explore how, and by
whom, the new elements in the schoolyard are used, within and outside of school hours.
The study will also explore how students perceive the organizational and physical
changes.

Data analysis

Exploratory study

Upon completion of the exploratory study, field notes, interview transcripts and photos
were ordered with the explicit purpose of identifying barriers influencing engagement
in recess PA across the cases 39]. The data was coded and arranged under headings derived from the social-ecological
model distinguishing natural, social, physical and organizational barriers 40].

Effect evaluation

The effect of the schoolyard interventions on PA will be assessed by calculating the
difference in the objectively measured average activity level (in counts per minute,
CPM) during recess in the schoolyard, before and after the intervention (? average
CPM during recess) using multilevel modelling to account for the nested structure
of the data (i.e. time points, students, class, school). The analyses will be adjusted for overall
activity levels, age, gender and parents’ socio-economic status. Furthermore, analyses
of changes in the proportion of time in sedentary, light and MVPA in the schoolyard
will be calculated to exemplify change in activity levels post the interventions.
To increase generalizability of the findings, the objectively measured average activity
level at the intervention schools will be compared to objectively measure average
activity levels of students during recess for approximately 40 other Danish schools.
This data is or will be available from other studies conducted by our university department.

The analysis of the secondary outcomes will be more exploratory requiring new methods
to clean and prepare useful variable based on combined accelerometer and GPS data.
Examples of secondary outcomes are: areas generating high level of activity (CPM or
MVPA) in the schoolyard, areas of the schoolyard most likely to encourage MVPA for
different groups of students (boys/girls, high/low activity groups, age-groups), exploring
routes of activity in the schoolyard.

Process evaluation

A descriptive analysis of data from the pre-intervention electronic survey was conducted
to identify the organizational structure at the cases regarding recess and schoolyards
policies, rules and practices prior to the intervention. The pre- and post-intervention
focus group data will be analyzed as a whole using a thematic analysis strategy 29], 30]. Relevant themes across cases related to how the process was experienced by the case
teams and school principals in the different phases will be extracted to identify
barriers and facilitators.

Post-intervention user-evaluation

Upon completion of the post-intervention user evaluation, field notes, interview transcripts
and photos will be analyzed using a thematic analysis strategy 29], 30]. Themes will be developed through a coding and re-coding process in order to identify
commonalities and divergences in how the students perceive and use their schoolyard
within and between cases 39].