Context
LHDs are NSW Government funded health services responsible for providing or supporting the provision of health promotion services to address the leading risk factors for chronic disease in their community. The NSW Ministry of Health provides funding to LHDs to support the implementation of state-wide health promotion programs.(22) All NSW LHDs have received funding to facilitate the implementation of healthy eating and physical activity policies and practices in NSW primary schools for over a decade as part of the NSW Healthy Children’s Initiative.(22) However, although healthy lunchboxes have historically been a focus for health promotion activities in some LHDs and non-government organisations (e.g. Cancer Council NSW), the funding provided by NSW Ministry of Health did not explicitly focus on a formal school-based program to support the packing of healthy lunchboxes. In addition, whilst a core component of health promotion practice, Health Promotion Unit capability to undertake research and evaluation of health promotion activity has been found to vary across LHDs.(23)
Ethics and trial registration
The research will be conducted and reported in accordance with the requirements of the Consolidated Standards of Reporting Trials (CONSORT) Statement.(24) Ethics approval has been obtained via the following Human Research Ethics Committees: Hunter New England (2019/ETH12353); University of Newcastle (09/07/26/4.04); NSW of Department of Education (2018247); and the Maitland-Newcastle, Sydney, Wollongong, Bathurst, Parramatta, Wagga Wagga and Canberra-Goulburn Catholic Dioceses. This trial is also registered prospectively with the Australian New Zealand Clinical Trials Registry (ACTRN12623000558628). The protocol is reported according to the Standard Protocol Items: Recommendations for Interventional Trials (SPIRIT) (Supplementary File 1).(25)
Study design and setting
Within a Master Protocol framework,(16) we will undertake a Collaborative Network Trial. Specifically, independent randomised controlled trials to test strategies to implement or improve health care occurring at different sites (LHDs) will be undertaken by the Health Promotion Units at each LHD. The key trial methods, measures and data collection processes will be harmonised with agreement across sites to provide individual school-level data for planned pooled analyses as part of a collaborative, following a prospective meta-analysis framework.(19) The design allows for heterogeneity or natural variation in the implementation strategies being tested and the contexts (i.e. sites) they are tested in.(16) The study builds on a pilot network trial to implement the scale-up of SWAP IT program in three LHDs.(13)
Sample and participants
The study will be conducted with primary and combined schools located across 10 LHDs in NSW, Australia. The state of NSW has approximately 2100 primary and combined schools, and is socioeconomically and geographically diverse.(26) Department of Education (DoE), Catholic Schools NSW and Association of Independent Schools of NSW primary and combined schools located within the LHDs of Murrumbidgee, Hunter New England, Sydney, Western Sydney, South Western Sydney, South Eastern Sydney, Northern Sydney, Western NSW, Nepean Blue Mountains, and Illawarra Shoalhaven will be included in the study.
Primary and combined schools located within the participating LHDs who cater for at least one primary school year and have not implemented the SWAP IT program will be eligible to participate. Only schools that do not use the Audiri parent communication app will be eligible, as these schools are participating in another trial being conducted concurrently by the research team (ACTRN12623000145606). Schools catering exclusively for children requiring specialist care, for example schools catering for students with severe disabilities, will be excluded. Schools with secondary students only will also be ineligible to participate.
All eligible schools will be included in the study as part of usual service delivery provided by LHD health promotion staff to support schools to implement a range of health promotion programs. Eligible schools will be invited to participate in the secondary data collection component of the study, specifically the follow-up survey conducted with school principals (described below). Schools will be recruited for the follow-up data collection via an invitation email containing a link to an online survey and a study information statement outlining the purpose of the research and their involvement. Schools that are yet to complete the survey will receive up to three reminder prompts via telephone or email by the research team to encourage completion.
Randomisation and blinding
Prior to the delivery of the first scale-up strategy, schools within each LHD will be randomly allocated to either the intervention or control condition using a computerised random number function in a 1:1 (intervention: control) ratio. Randomisation will be stratified by school size and social socio-economic location, as determined by Socio-Economic Indexes for Areas categorisation using school postcodes,(27) given the socio-economic association with implementation of school nutrition programs.(28) Randomisation will be completed by a statistician not otherwise involved in the trial. Due to the nature of the intervention, participants will not be blinded to group allocation. However, research staff assessing the outcomes at follow-up will be blinded.
Implementation strategies
Implementation strategies were developed for each of the LHDs independently, based on their existing capacities and local contexts. Implementation strategies for each participating LHD (‘site’) were co-designed by LHD health promotion staff and other stakeholders, with support provided by National Centre of Implementation Science (NCOIS) implementation scientists and SWAP IT developers from the University of Newcastle. The development process included: i.) planning workshops facilitated by University staff that drew on tacit knowledge and experience of health promotion staff who had considerable experience working with schools; ii) evidence regarding barriers to school adoption and implementation of SWAP IT collected by the research team as part of previous SWAP IT trials, iii) data from systematic reviews and pilot trials regarding the effectiveness of strategies to facilitate adoption.(29) During the workshops, theoretical framework tools were used to facilitate the selection of strategies to address barriers that were aligned to individual LHD capacity and contexts.(30-32) Processes may have also been undertaken by LHDs to identify strategies to support access and engagement of priority populations within their region to ensure school adoption and implementation of SWAP IT does not further exacerbate health inequities. This may have included consultation and engagement processes with Aboriginal, or Culturally and Linguistically Diverse individuals, groups or stakeholders.
Across the 10 participating LHDs, six broad implementation strategies emerged and are described below. The combination of these six strategies employed by each LHD will differ and is described in Table 1. For all LHDs, these strategies will be executed over a period of six months. A timeline for the delivery of the implementation strategies is provided in Table 2.
1. Sector support and endorsement: Policy makers from Health will target principals to communicate, support and endorse the program and its outcomes, its alignment to sector policies and recommend its adoption. This endorsement will occur via a maximum of two targeted letters or emails developed by the research team, approved and endorsed by local and state-level Health partners. The letters or emails will also contain a link to resources and the enrolment website. As an additional strategy, some LHDs (outlined in Table 1) will use their existing connections to obtain endorsement for the program from local educational and wellbeing liaisons within the NSW Department of Education. This endorsement will be promoted to schools via an email distributed by the liaisons directly to schools receiving this strategy.
2. Local facilitation: Health promotion staff from LHDs have developed strong and trusted local relationships with schools for over a decade and represent credible sources of local nutrition expertise. LHD health promotion staff will use up to two of their existing planned school contacts, conducted via telephone call or face-to-face meeting, to assess interest in the SWAP IT program, address any school-specific barriers to adoption, and facilitate goal setting and action planning. Scripts developed by the research team to guide the local facilitation will incorporate motivational interviewing techniques to be employed by health promotion staff to address school barriers to program adoption.
3. Develop and distribute educational materials: Targeted at principals to address perceived barriers to adoption, the strategy will initially aim to create tension for change (e.g. via outlining parent and carer interest and expectations); and then communicate the attractive program attributes (e.g. simplicity, no-cost). This communication will consist of up to two contacts, including a printed information pack (consisting of a flyer, SWAP IT pen and example parent booklet) at the commencement of the intervention period followed by an email to promote the program. As an additional strategy, one LHD will offer printed parent booklets promoting the SWAP IT program to all parents and carers with children commencing the following school year within their school kindergarten orientation packs along with a flyer encouraging the school principal or wellbeing coordinator to adopt the program.
4. Local opinion leaders: Promotional materials, including one printed information pack (consisting of a flyer and example SWAP IT parent booklet) and one email, will be delivered to other leaders that may be influential in a schools decision to adopt health promotion programs, specifically the school administration manager and parent committee. The aim of these materials is to promote the SWAP IT program and encourage school adoption.
5. Audit and feedback: Data and feedback on school adoption of SWAP IT will be automatically captured through electronic registration records and be provided to schools via other implementation strategies, including educational materials, local facilitation and local opinion leaders. For example, educational materials provided to principals, school administration managers and parent committees will include information on the number of schools that have registered for SWAP IT, a link to view an online list of schools have already adopted the program (to create tension for change and social norms) and provide instruction on how the school can also register for the program.
6. Educational meeting: Health promotion staff from LHDs will conduct one webinar with schools within their LHD to assess interest in the SWAP IT program and address any barriers to adoption. Webinar content will be developed by the research team in collaboration with health promotion staff.
Control group and contamination
Registration for the SWAP IT program is publicly available and freely accessible for all schools, including schools allocated to the control group. The implementation strategies to be delivered to the control group across LHDs is described in Table 1. For most schools allocated to the control group, the comparison will be ‘no implementation support’ or a singular strategy. Execution of the implementation strategies will be monitored centrally by the research team in consultation with health promotion staff from each LHD to minimise risk of contamination. Nonetheless, school exposure to the implementation strategies will be assessed at follow-up via an online or telephone survey with school principals (described below).
Study outcomes and data collection
Trial outcomes were discussed and agreed upon by participating LHDs. Data collection for all trial outcomes were harmonised across all LHDs and will be collected centrally by the research team at the University of Newcastle. The centralisation of data collection represented an efficient means of collecting and managing data for all participating LHDs. All demographic, operational and trial outcome measures are harmonised (i.e. identical item, measure and data collection method) to facilitate comparability and analysis. Each participating LHD will retain access to their trial dataset.
Primary outcome
Adoption of the SWAP IT program, defined as the number of schools who register for the lunchbox nutrition program (SWAP IT), will be assessed within schools allocated to the intervention and control group via electronic registration records captured automatically following school registration to SWAP IT. As part of the registration process, schools provide consent for the de-identified registration data to be used for research and evaluation purposes. This outcome will be assessed at baseline and approximately 9 months after baseline data collection.
Secondary outcomes
Acceptability of implementation strategies, defined as the perception amongst principals that the implementation strategies are agreeable, palatable or satisfactory,(33) will be assessed in a telephone or online survey with school principals at 9-month follow-up. School principals will be asked if they recall receiving each of the implementation strategies during the intervention period. For strategies the participants recall receiving, they will be asked to rate how acceptable they found the strategy on a 5-point Likert scale (1=not acceptable; 5=very acceptable). Principals from 243 Catholic and Independent primary schools located across five LHDs (LHD 1; LHD 5; LHD 7; LHD 8; LHD 9) will be invited to participate in the survey. These LHDs have been selected as they are employing diverse combinations of the implementation strategies (Table 1). Including schools from these LHDs in the survey will ensure the acceptability of all employed strategies (across the 10 LHDs) will be assessed without surveying all participating schools.
Implementation of the SWAP IT program, defined as the extent to which the SWAP IT program components were delivered by the school to parents, will be assessed in the telephone or online survey with school principals at 9-months follow-up. Schools will be asked to report if they implemented the SWAP IT program at their school, and what program components were implemented (i.e. parent messages; school lunchbox guidelines; curriculum resources; parent and carer resources).
Sustainability of the SWAP IT program, defined as continued school use of the lunchbox nutrition program (SWAP IT) at 18 months after baseline data collection, will be assessed via electronic registration records captured automatically following school registration to SWAP IT.
School characteristics, including postcode, total student enrolments, geographic location (urban, regional, rural and remote), proportion of Aboriginal student enrolments, and proportion of students that speak a language other than English at home, were obtained from a publically accessible Australian Curriculum, Assessment and Reporting Authority (ACARA) database.(34)
Sample size and data analysis
We are anticipating a sample of at least 30 schools per group (and an average of 60 per group) in trials of each of the 10 participating LHDs. Descriptive statistics, including proportions, means and standard deviations, will be used to describe school characteristics, adoption, implementation and sustainability of SWAP IT, as well as the acceptability of the implementation strategies.
Analyses of trial outcomes will be undertaken under an intention to treat framework separately for each trial. For assessment of school level program adoption, the primary trial outcome, between-group differences, will be assessed using logistic regression. The model will include a term for treatment group (intervention vs control) and pre-specified covariates prognostic of the outcome. Little, if any, missing primary outcome data is anticipated at follow-up, as program adoption is recorded automatically for all participating schools. Nonetheless, we will employ multiple imputation for any missing data in the event that schools withdraw from the study and request that their data are not used. All statistical tests will be 2 tailed with alpha of 0.05. Assuming adoption of the program by 10% in the comparison group, a sample size of approximately 30 schools per group will be sufficient to detect an absolute difference between groups of 30%, with 80% power and an alpha of 0.05.
We will employ component IPD component network meta-analysis to compare and rank the effects from all the tested strategies on the primary trial outcome.(35) For this analysis we will also include the three randomised controlled trials from the pilot,(21) expanding the network and providing pooled individual level data from 13 randomised controlled trials. We will explore combining ‘educational meetings and educational materials’ into a single component for analysis given their shared underlying behavioural targets. We will adjust for prognostic factors and exploration of strategy—covariate interactions to identify if and to what extent effects vary by participant, population or other contextual factors (effect modifier).(35) We will also employ component network meta-analyses to model additive main effects (separate effects for each element or component of an implementation strategy); two way interactions (synergistic/antagonistic effects of components), and full interactions (different effects from each combination of components). The analyses will be performed under a Bayesian framework. There are no established methods for sample size calculations for component network meta-analysis.
Trial governance
The trial will be overseen by a Steering Group, comprised of representatives from each LHD, including: Aboriginal Health Promotion Managers; program developers, implementation scientists, trialists and research dietitians from the University of Newcastle. Roles and responsibilities will be documented in a Terms of Reference for the Group. LHDs will be responsible for the selection of implementation strategies for their jurisdiction, and execution of some of the strategies to schools. The University of Newcastle will be responsible for facilitating trial workshops, ethics, data collection, monitoring and quality assurance, data management and analysis. A Community of Practice, established in the pilot,(21) will also be employed to support the interpretation of trial results and pooled analyses, exchange tacit knowledge and experience and identify opportunities for improvement.