Wolfenden 2017.

Study characteristics
Methods	Trial name: no trial name Study design: cluster‐RCT for implementation outcome Intervention duration: 12–14 months Length of follow‐up from baseline: to assess the primary trial outcome, data were collected at baseline (April to September 2013) and at the completion of the implementation period (November 2014 to April 2015). Differences in baseline characteristics: there were no significant differences in baseline characteristic among schools. Unit of allocation: schools Unit of analysis: schools
Participants	School type: primary schools Region: Hunter region of NSW, Australia Demographic/socioeconomic characteristics: the trial region contained a socioeconomically diverse group. Inclusion/exclusion criteria: schools from the study region were randomly selected and invited to participate. Inclusion ‐ Schools were required to have an operational canteen. ‐ ≥ 1 items on their canteen menu that was restricted for sale ('red' or 'banned') or < 50% of menu items classified as healthy ('green' items). Exclusion ‐ Non‐government schools ‐ Schools with both primary and secondary students ‐ Schools catering exclusively for children requiring specialist care Number of schools allocated: 70 Numbers by trial group n (controls baseline) = 35 n (controls follow‐up) = 30 n (interventions baseline) = 35 n (interventions follow‐up) = 27 Recruitment: recruitment continued until 70 schools provided consent for study participation. Recruitment rate: 88%
Interventions	Number of experimental conditions: 2 (1 intervention, 1 control) Policies, practices or programmes targeted by the intervention The NSW Government launched a Healthy School Canteen Strategy (also known as Fresh Tastes @ School) to help prevent childhood obesity. The strategy classified foods sold by schools as 'red', 'amber' or 'green' based on their nutritional content. The strategy was adopted as policy by the government education department, and all government schools were mandated to remove items classified as 'red' from regular sale. Furthermore, schools were encouraged to 'fill the menu' with items classified as 'green' and ensure items classified as 'amber' did not dominate the menu. The primary aim of this study was to assess the effectiveness of a multi‐strategic intervention to increase implementation of the state‐wide healthy canteen policy. Implementation strategies EPOC: audit and feedback ‐ Performance monitoring and feedback menu reviews were conducted quarterly (unless menus were unchanged), and the results were used to compile written feedback reports to the canteen manager and school principal. Verbal discussion of the reports occurred during academic detailing visits or via telephone support calls. EPOC: continuous quality improvement ‐ Policy implementation support schools were allocated a support officer with qualifications in nutrition and dietetics and experience in supporting schools to implement the policy. Support officers contacted canteen managers every 2 months (via email, telephone or in person) throughout the intervention and used a continuous quality improvement framework of repeated goal setting, action planning, self‐monitoring and problem‐solving with canteen managers. EPOC: external funding ‐ Schools were also offered a small reimbursement to cover the costs associated with canteen manager attendance at training. EPOC: education materials ‐ Tools and resources – printed instructional materials, sample policies/menus, planning templates, pricing guides, product lists of policy compliant menu items, supplier contacts and menu assessment feedback were provided to all school canteen managers during the workshop or mailed to non‐attenders of the workshop. EPOC: education meeting ‐ Canteen managers, canteen staff and parent representatives were invited to attend a training workshop (5 hours) with the aim of providing education and skill development in the policy, nutrition and food label reading, canteen stock and financial management, pricing and promotion, and change management. Training combined didactic and interactive components including opportunities for self‐assessment, role play and facilitator provided feedback. Training was facilitated by a support officer. EPOC: education outreach visits ‐ School canteen visits were conducted 1 and 3 months after canteen manager training to enable support officers to observe the operational canteen environment, provide feedback and assist with problem‐solving barriers to policy implementation EPOC: local consensus process ‐ Meetings between support officers and canteen staff were held to discuss and reach consensus regarding the policy, how best to implement it and to develop local canteen action plans to co‐ordinate implementation tasks. EPOC: local opinion leader ‐ Executive support school principals were asked to communicate support for policy implementation and maintenance to teachers, parents, students and canteen managers during staff meetings, in newsletters and assemblies. Support officers also sought meetings with the executive of parent representative groups to garner their support for and input on policy implementation. EPOC: tailored intervention ‐ Individualised goal setting, action planning with canteen managers at different schools EPOC: other ‐ Quarterly project newsletters communicated key messages, provided information and case studies of successful implementation approaches to common barriers. ‐ Recognition schools with a menu assessed as adhering to the policy (i.e. > 50% 'green' items and no 'red' or 'banned' items) received a congratulatory letter and telephone call from the research team and were publicly acknowledged via marketing strategies. ‐ Canteen managers also received kitchen equipment to the value of AUD100. Theoretical underpinning: the selection of intervention components was guided by the Theoretical Domains Framework. Description of control: no contact was made, and no resources provided to control schools during the intervention period by the research team.
Outcomes	Outcome relating to the implementation of school service policies, practices or programmes ‐ The proportion of schools with a canteen menu that did not contain foods or beverages restricted for sale ('red' and 'banned') under the policy. ‐ The proportion of schools where healthy canteen items ('green items') represented > 50% of listed menu items. Data collection method: copies of canteen menus were collected from all participating schools and audited by 2 dietitians independently. Validity of method: not reported yet objective. Outcome relating to cost: not reported Outcome relating to adverse consequences: as a measure of potential adverse effects yearly income and expenditure of canteen profitability were calculated. Data collection method: canteen's financial records. Validity: objective Outcome relating to child diet, PA or weight status: mean (95% confidence intervals) energy, total fat and sodium of student purchases. Data collection method: direct observation Validity of method: objective
Notes	Research funding: Australian Research Council who provided funding for the trial [LP130101008]. In‐kind support was provided by the Hunter New England Population Health and the Hunter Medical Research Institute. Conflicts of interest: authors reported grants from Australian Research Council Linkage Project Scheme (LP130101008) during the conduct of the study; no financial relationships with any organisation that might have had an interest in the submitted work in the previous 3 years; and no other relationships or activities that could appear to have influenced the submitted work.
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	RCT. Random sequence generated using a random number function in Microsoft Excel.
Allocation concealment (selection bias)	Unclear risk	No information provided about allocation concealment and, therefore, it was unclear if allocation was concealed.
Blinding of participants and personnel (performance bias) Implementation outcome	High risk	Outcome group: all outcomes (including adverse events) School staff (principals and canteen managers) were not blinded to group allocation. Cost data: collected retrospectively: low
Blinding of outcome assessment (detection bias) Implementation outcome	Low risk	Outcome group: data collectors were blinded to group allocation. Cost data: collected retrospectively: low
Incomplete outcome data (attrition bias) Implementation outcome	Low risk	Outcome group: all outcomes 13/70 (19%) schools did not provide their menu for assessment at follow‐up. There were no significant differences in baseline characteristics among schools that did and did not provide follow‐up data. Analyses of study outcomes were performed under an ITT framework. Cost data: no loss to follow‐up
Selective reporting (reporting bias)	Low risk	There were no unreported implementation outcomes according to those planned in the published protocol. Cost data: retrospective economic analysis: unclear
Other bias	Low risk	Appeared free from other bias.
Overall risk of bias assessment	Low risk	Most domains were at low risk of bias.