Bremer 2018.

Study characteristics
Methods	Trial name: daily physical activity (DPA) programme Study design: non‐RCT Intervention duration: 20 consecutive weeks Length of follow‐up from baseline: baseline, mid‐intervention (10 weeks) and postintervention (within 2 weeks postintervention) Differences in baseline characteristics: significant difference in number of white people between experimental and control group (Chi2 (1) = 3.31, P = 0.07) Unit of allocation: class Unit of analysis: ‐ Implementation outcomes: class ‐ Behavioural/health outcomes: child
Participants	School type: Catholic schools, grades 4–8 (ages 9–14 years) Region: Ontario, Canada Demographics/socioeconomic characteristics: not reported Inclusion/exclusion criteria: not reported Number of schools allocated Schools:7 Classes: 30 Students:362 Numbers by trial group n (controls baseline) = 11 control classes n (controls follow‐up) = 11 control classes n (interventions baseline) = 19 experimental classes n (interventions follow‐up) = 19 experimental classes Recruitment Schools: 7 schools from a local Catholic school board were chosen by the school board to participate. Teachers were asked to participate by their school administrators. 1–7 classes participated in each school, total of 30 participating classes. Students:all 783 students in the participating classes were approached to participate in the study. Recruitment rate Schools:not reported Students:46.2%
Interventions	Number of experimental conditions: 2 (1 intervention and 1 control) Policies, practices or programmes targeted by the intervention: DPA policy implementation Implementation strategies EPOC: educational meetings ‐ School teachers and student leaders attended a 1‐day workshop on how to deliver the programme as part of regular school activities and were provided with instructional materials to take back to their school for programme delivery. EPOC: educational materials ‐ Provided with instructional materials EPOC: other ‐ Participation in a community 5 km at end of intervention Theoretical underpinning ‐ Of the evidence‐based intervention/policy/practice or programme: not reported ‐ Of the implementation strategy: not reported Description of control: the remaining teachers who chose not to participate in the programme were expected to provide DPA to their students, as per the Ontario education curriculum (i.e. usual practice).
Outcomes	Outcome relating to the implementation of school policies, practices or programmes Outcomes included adherence to the programme, student behaviour, PA opportunities and student engagement in school‐based PA (i.e. number and length of PE classes and recesses per week). Data collection method: 21‐item questionnaire was developed for this study, and completed by the homeroom teacher at the last measurement point. Validity of measures used: not reported Outcome relating to cost: not reported Outcome relating to adverse consequences: not reported Outcome relating to child diet, PA or weight status: PA was assessed using the PAQ‐C Data collection method: PA was assessed with the PAQ‐C. The PAQ‐C was designed for children in grades 4–8 who had recess as a regular part of their school week. Validity of measures used: the PAQ‐C is a 9‐item 7‐day recall instrument that has good psychometric properties.
Notes	Research funding: not reported Conflicts of interest: authors declared that they had no competing interests.
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	High risk	Not a randomised trial, as teachers selected whether or not their class would participate in the programme or to act as a control class.
Allocation concealment (selection bias)	High risk	No allocation concealment, participants (teachers) chose intervention or control.
Blinding of participants and personnel (performance bias) Implementation outcome	High risk	Outcome group: all The nature of the trial precluded blinding, participants (teachers) were not blind to allocation, and students would have acknowledged a change from usual practice.
Blinding of outcome assessment (detection bias) Implementation outcome	High risk	Outcome group: participants (teachers) High: participants were not blind to allocation. Also, self‐report would have contributed to further risk. Outcome group: students Unclear: students would have been able to tell group allocation though no details around this.
Incomplete outcome data (attrition bias) Implementation outcome	Unclear risk	Outcome group: teachers Unclear: unclear if there was any loss to follow‐up with the teachers. Outcome group: students Low: attrition by the study end was 16% in the intervention group and 20% among the controls (Chi2 = 0.57, df = 1, P = 0.45)
Selective reporting (reporting bias)	Low risk	Protocol available on US National Library of Medicine (ClinicalTrials.gov). All outcomes appeared to be accounted for.
Other bias	Unclear risk	Unclear if at risk of contamination.
Recruitment to cluster	Unclear risk	Unclear at what time recruitment to cluster occurred.
Baseline imbalance	High risk	Fewer control than intervention participants. Table 3 indicated significant differences between groups. Unclear if these are taken into account in analyses.
Loss of cluster	Unclear risk	Unclear if losses to follow‐up resulted in loss of cluster (especially in the control where there were limited numbers).
Incorrect analysis	Unclear risk	Unsure if clustering was taken into account in analyses.
Compatibility with individually randomised RCTs	Unclear risk	Unsure, no statement on this. Unable to determine if herd effect existed.
Potential confounding	Unclear risk	Unclear if baseline differences in intervention and control schools adjusted for in analysis.
Overall risk of bias assessment	Unclear risk	Most domains were at unclear risk of bias.