Adapt-Ed: co-designing adaptations to a whole school intervention to improve the uptake and impact of food provision in special schools – scoping research for a future trial

Patient and Public Involvement

Public involvement and engagement informed the preparation of the ADA proposal, and the involvement of children and young people with SEND, parents and carers was embedded throughout the project, underpinned by the UK Standards for Public Involvement in Research (HRA, 2020). Public involvement was the focus of Work Package (WP) 1, that was co-led by AF, as public co-applicant with relevant lived experience. Further details regarding when and how the public were involved in the project are discussed in the sections on WP1 below in Methods; the outcomes from, and impact of, this involvement are covered in the Results and Discussion.

The study took a mixed methods approach, consisting of four parallel Work Packages (WP):

WP1 Public and stakeholder involvement (to address OBJ1). Guidelines on adapting interventions to new contexts recommend collaboration with stakeholders from the earliest stages (Moore et al., 2021). HZ was devised and is delivered by the charity School Food Matters (SFM), a partner on the grant. In consultation with SFM we identified stakeholders most affected by this research and with power to influence school food in special schools as within two main groups: (1) ‘public contributors’ including: (a) children and young people with SEND and (b) their parents and carers; and (2) ‘professional stakeholders’ including those working in (a) schools (teachers, school nurses, governors/trustees, catering and support staff) and (b) education and health policy and strategy (local authority policymakers, commissioners, voluntary/community sector organisations, catering providers)³.

Recruitment to stakeholder groups was carried out via a range of methods including general articles and announcements online (e.g. an article on the NIHR ARC East of England website), notices in newsletters (e.g. SFM newsletter, local authority mailing lists), emails to network convenors and existing contacts, and personal email invitations to individuals. We recognised the need to access children and young people with SEND via gatekeepers. Whilst we did not aim for ‘representativeness’ (as our purpose was not to generalise) we did want to achieve diversity of experience and expertise. We therefore collected optional demographic data, and took an iterative approach, targeting public and professional members who contributors told us were currently ‘missing’. We aimed for around six people in each group, which is ideal for focus-group type workshops (Bryman, 2004). We aimed to ensure any access needs were met, and included time and space during workshops to get to know each other, build trust, learn and share ideas. However, we set out to prepare detailed plans in collaboration with those involved. In particular, we recognised that person-centred involvement with children and young people, is most effective when developed with young people, rather than expecting them to fit into a model pre-determined by adults (Cuevas-Parra, 2020).

The gatekeepers (services and parents and carers) and children and young people we spoke to early on made clear that our initial plans to organise a project-specific young people’s group would not work for them in terms of access, inclusion or availability, so plans for involvement were revised in collaboration with those we wished to involve, and organisations working with them. We aimed to make participation as flexible and inclusive as possible, with ‘pockets of participation’ depending on young people’s interests, availability, and personal circumstances (Franks, 2011). In practice this meant we contacted a variety of organisations working with children and young people with SEND with information about the project and, during an initial meeting, found out more about the group and how they could be involved in ways that worked for them.

1.   Public contributors

a.   Parents and carers (n=6)

We formed a group of six parents and carers of children with SEND, who met with us online three times, for around 90 minutes at each workshop, between July and December. The group included a mix of in terms of gender, ethnicity and child’s needs regarding food and eating. However, most were mothers, White British, with ASD/autism related eating difficulties most common among their children. The latter attended a variety of educational provision, including mainstream school, special schools, and Education Otherwise Than at School (EOTAS). In addition, we met with 17 parents in two special schools as part of WP4. Of those parents who shared demographic information, most identified as being on a low income (defined ‘subjectively’ as below what they needed (Gordon, 2006:51) and/or having a child eligible for FSM) (see Table 1 for a breakdown of meetings and attendance).

b.   Children and young people (n=42)

Between September-November, we met with 41 children and young people (aged 4–25 years old), through attending three youth groups for children and young people with SEND, a session with a School Council within a special school, and a 1:1 meeting, using a mix of modes (in-person and online), ranging from 30 to 90 minutes. Our approach was underpinned by the Lundy (2014) model of participation which is focussed around four core concepts (space, voice, audience and influence) all of which need to be in place to meet young people’s rights to be heard in both personal and public decision making. By ‘voice’ we mean communication, in whichever form works best for the child or young person. Broad topics and questions were discussed with group facilitators beforehand and sent ahead of the meeting. Due to the project timeline, and capacity and resources of each group, we were only able to meet with each group on one occasion during the funded period (Table 1).

2.   Professional Stakeholders

a.   People working in schools (n=7)

This group included three school Chefs/Cooks, a Head of Food Studies, a Deputy Head Teacher, a Teaching Assistant (all in special schools), and a Public Health Officer working with schools.

b.   People working in policy and with schools (n=17)

This group included representatives of Trades Unions, parent/carer networks, catering companies (including a Local Authority Trading Company, LATCo⁴), charities for disabled children and their families, local authority Public Health Officers, the Department for Education, and the Department of Health and Social Care.

We met with each professional stakeholder group separately, twice, between July-November, then together, with members of the parent/carer group, at a joint meeting in December. Each meeting lasted around 90 minutes (Table 1).

Wider stakeholder engagement (approx. n=41)

Alongside our ‘core’ stakeholder group meetings (above) we held ‘ad hoc’ 1:1/small group meetings with professionals and organisations identified via our scoping research and networking. This included nine individual meetings with approximately 41 people in total, including the Special Educational Consortium, The Royal College of Speech and Language Therapists, and people in Public Health and Education related roles in various local authorities.

In line with our ethical approval, online stakeholder engagement was automatically recorded, with contributors’ informed consent, and/or notes were taken. Automatic transcriptions of core stakeholder workshops were corrected against audio recordings by a professional transcriber. Discussions with children and young people, and most wider stakeholder engagement, was recorded in fieldnotes and written up immediately afterwards. At least one member of the research team completed a Patient and Public Involvement (PPI) log (PIHWE, 2023) after each meeting. For all core stakeholder workshops, we asked for members’ feedback about what worked well (or not) and sent summaries via email of what we thought we had learned from our discussions. A simplified ‘Framework’ approach (Gale et al., 2013) was employed to organise and make sense of information gathered (transcripts and notes of discussions), charting this against predefined topics/questions whilst leaving room for unanticipated discoveries.

Growing Healthy Zones

Central to our approach to working with stakeholders was the co-production of an Adapt-ed HZ logic model. Logic models and theories of change are ways of identifying the inputs and activities needed for an intervention to achieve the desired results and are commonly used to plan and design evaluative research. Collaborative logic modelling is one way that public contributors and professional stakeholders can be involved with the design of research, policies and services (Harris et al., 2018). Logic models and theories of change can be represented through metaphors, as these are widespread in everyday language and allow us to understand abstract thoughts and feelings that cannot be directly seen, heard, touched, smelled, or tasted (Lakoff & Johnson (1980); Lakoff & Johnson (1999)). Commonly used metaphors for logic models in consultations with public contributors are ‘roadmaps’ and ‘trees’ (e.g. Breen, 2023; CLAPS, 2024). In this project, we used a sunflower to present the conditions, steps and outcomes of ‘growing Healthy Zones’ (Figure 1).

Figure 1. ‘Growing Adapt-Ed Healthy Zones’ sunflower metaphor used to co-develop the final collaborative logic model.

Collaborative logic modelling is one way that public contributors and professional stakeholders can be involved with the design of research, policies and services (Harris et al., 2018). Commonly used metaphors for logic models in consultations with public contributors are ‘roadmaps’ and ‘trees’ (e.g. Breen, 2023; CLAPS, 2024). In this project, we used a sunflower to present the conditions, steps and outcomes of ‘growing Adapt-Ed Healthy Zones’. Each part of the sunflower represented a specific element for the final collaborative logic model, as follows:

• 1. Centre of Sunflower: the overall goal or aim of the HZ programme.

• 2. Leaves: the changes that need to occur within schools to achieve the intended goal.

• 3. Water and fertilizer: the activities and/or support required to make these changes.

• 4. Pot and soil: the inputs and resources required to help schools make these changes.

• 5. Sunflower petals: the short-, medium- and long-term outcomes and/or impact of the programme (i.e. benefits arising from the activities).

The process of co-producing the sunflower logic model had a number of stages. A separate model was initially co-produced with each adult group, at the second in a series of three workshops with each. (We only met with each group of children once during this stage of the project, so we did not introduce the logic model to our discussions). These separate models were then combined, and at this stage input from discussions with children and young people groups was added, along with any additional input from our wider stakeholder engagement. A key was used to indicate the source of the input, and show where it was contributed by more than one set of core stakeholders. The combined model was shared with representatives of each (adult) group in a joint third meeting, at which it was revised and refined, and consensus was reached regarding its completeness.

WP 2 Scoping Reviews of relevant interventions, outcomes and dietary assessment methods (to address OBJ2). Our initial reviews of the literature identified little existing research on school meal provision or experiences for children with SEND and no comprehensive reviews of evidence on food interventions with SEND children, appropriate dietary assessment methods, or outcome measures for this group. We set out to carry out scoping reviews (Grant & Booth, 2009; Munn et al., 2018) since these are considered the most appropriate review method to understand the scope and nature of a body of literature that spans disciplines and designs and has not been comprehensively reviewed (Brock, 2023; Peters et al., 2015).

We conducted two scoping reviews, incorporating academic literature to inform the future evaluative research, specifically study design, including identification of interventions and appropriate instruments and outcomes related to food behaviours and dietary intake for children with SEND⁵. Protocols for both scoping reviews were designed with reference to the Joanna Briggs Institute (JBI) methodological guidance for scoping reviews and JBI Manual for Evidence Synthesis (Aromataris et al., 2024; Peters et al., 2020), reported in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses extension for Scoping Reviews (PRISMA-ScR; Tricco et al., 2018), and registered with the Open Science Framework (Denyer et al., 2024a; Denyer et al., 2024b). Searches were conducted on PubMed, Scopus, EMBASE, Web of Science and PsycINFO databases. Using the software, Rayyan, the references were screened by two researchers independently and any discrepancies resolved via a discussion with a third reviewer. Data were extracted into a table and accompanied by a narrative summary.

WP3 Typology of Special Schools in the East of England (to address OBJ3). In England, there is no current, comprehensive survey, or routinely collected data, on school characteristics, practices and outcomes of school food provision. This initial stage therefore explored what existing administrative data could tell us about the overall picture of school food provision, whether new survey data could fill in any gaps, and if together this information could help inform the future study design and sampling strategy. Specifically, we wanted to know whether we could identify and measure relevant school characteristics in order to stratify them in our future trial design⁶. We also wanted to look at the typicality (or not) of East of England (EoE)⁷ schools’ characteristics and outcomes relative to the rest of the country.

A main outcome of interest was take up of school meals. This is a pragmatic proxy for exploring the impact of school meals on pupil outcomes, since the association between take up of school meals and child health outcomes is well-established (Holford & Rabe, 2020; Holford & Rabe, 2022; Holford & Rabe, 2024). Although there are no publicly available data on the take up of all (paid and unpaid) school meals in England, take up of means-tested and Universal Infant FSM (among those eligible) is routinely collected and publicly available. However, an important limitation is that this indicator is based on data collected on one day of the year – ‘census day’ – for the purposes of allocating core funding, hence it may not be representative of take up generally. We therefore also sought to explore whether we could fill any gaps with an online survey of schools in the EoE region.

To describe special schools in EoE, by relevant characteristics, and explore associations with means-tested and Universal Infant FSM take up among those eligible (hereafter FSM take up), we:

(a) combined routinely collected public data for analysis to create a cross-sectional dataset of (currently open) state-funded special schools, Pupil Referral Units (PRU) with specialist SEN provision and mainstream schools with a SEN Unit, Resourced Provision or Special Classes across England. The combined public data include the January 2024 Department for Education ‘underlying data’: ‘Get Information About Schools’ (GIAS; DfE, 2024a) resource (accessed 14^th August 2024) and ‘Schools, Pupils and Characteristics’ dataset (SPC; DfE, 2024b). We conducted descriptive analysis of eligibility and take up rates of means-tested FSM (all ages) and of Universal Infant FSM (Reception, Year 1, 2); and Ordinary Least Squares regression to explore the school-level predictors (such as ethnic composition of school intake and urban/rural location) of take up of means-tested FSM among FSM-eligible children (all ages), and of Universal Infant FSM among all infant children. We compare differences between school types within EoE; and within school types between EoE and the rest of England. Since we are interested in the explanatory power of school-level variables, and potential for school-level changes in policy, we give each school equal weight in the regression rather than weighting observations in proportion to the number of (FSM-eligible or infant) children in each school.

(b) designed and conducted an online survey of state-funded special schools and mainstream schools with a SEN Unit, Resourced Provision or Special Classes in EoE (N=341 eligible) using JISC Online Surveys. WP1 stakeholders and SFM provided input when designing the survey. We aimed to i) cover aspects of school food provision, policy and practice, and support for pupils’ SEND, that are not collected in the routine public data; and ii) establish interest in the future Adapt-Ed intervention and willingness to be recontacted. We identified and contacted local authority and county council contacts with responsibility for SEND and/or schools with information about the survey (including the online link) and asked them to circulate amongst eligible schools. Relevant schools networks in the region were also contacted. Following this, survey invitations were sent directly to head teachers or relevant SENCo and admin staff (if contact details were publicly available) using the JISC automatic distribution tool.

Stakeholders recommended incentives to encourage participation. Those who completed the survey could choose to enter their school into a prize draw to win one of three prizes (a £500 Love2Shop voucher, a consultation with SFM about ‘sensory gardens’ or a £50 National Garden voucher). The relatively small number of responses (n=33/341; a 9.7% response rate), despite recommended use of incentives, was anticipated (Sturgis et al., 2006). The small number of schools responding limited the kinds of statistical models (using regression methods) that could be used. There is a large number of determinants related to take up of school meals, making any model with small numbers of respondents difficult to interpret.

(c) conducted an ‘audit’ of EoE special schools’ websites (N=114) to explore what other relevant information could be gleaned from school websites. Specifically, we examined whether the following were available online for each school: (i) a written school food policy; (ii) in-house or external catering provision details (and if external, which provider); and (iii) current lunch menus. These topics were informed by the emerging findings across the various WP, including indicative findings from the online school survey. For pragmatic reasons, only special schools in EoE were included in the audit. Descriptive analysis was then carried out.

WP 4 Ethnographic case studies of food provision in three schools (to address OBJ4). We purposively selected three EoE special schools (in Norfolk, Hertfordshire and Essex⁸) to provide a contrast of contexts in terms of area deprivation (according to the Index of Multiple Deprivation, IMD), proportion of pupils eligible for FSM (a pragmatic proxy for low-income, (Ilie et al., 2017; Taylor, 2018)) and ethnic mix. We also sought to include schools in different settings in terms of their urban/rural location and type of catering provider.

As shown in Table 2, the three schools selected are all state funded ‘through’ schools catering for children from infants (or in one case, a nursery) up to the age of 19 years. All are mixed in terms of gender, with a greater proportion of males than females (around 3:1 in each case, reflecting the national picture). All three schools provide for children with Autistic Spectrum Disorder (ASD), however, they vary in terms of their other specialisms, and in their location, the proportion of children eligible for FSM, the ethnic mix of the school population, and catering arrangements:

We visited Schools A and B for two days each and School C for (part of) one day¹⁰ to explore school policies, provision and practices within these different contexts and examine the views of staff and parents on a) opportunities and challenges for school food provision, b) what improvement would look like, and c) acceptability and feasibility of the future intervention/ evaluation. Two researchers visited each school, one of whom (ROC) visited all three schools, with a third researcher (LMB) joining the visit to School A to facilitate the involvement of the School Council in the WP1 (public involvement) discussion. We asked schools for their support for us to hold a parents and carers coffee morning, and to help us recruit parents for this, using a flyer we prepared. We also asked to interview key members of staff and observe mealtimes, snack times, and other food related activities. (Anonymised recruitment materials are provided in supplementary files).

As shown in Table 3, consistent with a broadly ethnographic approach, we used a range of methods (Hammersley & Atkinson, 2019; Spradley, 1979) to engage with different target groups (staff, parents and carers, and children and young people) in each school. We prepared and referred to observation and topic guides that directed our gaze and questions, but we were also led by interests and opportunities that arose in each setting. With staff, we conducted formal (semi-structured) interviews (three Headteachers, a Deputy Head, two School Cooks and an Area Catering Manager) and carried out informal (less structured and unrecorded) discussions with others (a Deputy Head, two Educational Psychologists, four Teaching Assistants (TAs), a School Nurse and a School Cook). A Family Support Worker (FSW) helped facilitate our coffee morning in School A and also took part in the discussion. Seventeen staff took part in formal interviews and informal discussions in total. We met with parents and carers in two schools during a bespoke coffee morning (School A, 10 parents and carers) and a regular coffee morning (School C, 6 parents and carers); whilst School B sent out our flier and invited parents to meet with us, no one took up the invitation, and instead we spoke with a mother on the phone. Seventeen parents and carers took part in these discussions in total. With children, given the brief nature of our visits, we relied mainly on observation (Nind, 2008), supplemented by informal discussions in all three schools and a formal consultation with the School Council in School A (the findings of which are reported under WP1). We had formal discussions with ten pupils in school A and informal discussions with around ten children, alongside our observations of classes and eating occasions, in the other two schools (Table 3).

Formal (semi-structured) interviews were recorded (with consent) and transcribed in full. Fieldnotes and flip-chart discussions were written up as soon as practically possible after each visit, and each fieldworker’s notes were combined into one account. In terms of documentary evidence, we gathered or were subsequently sent lunch menus from each school. The external catering provider for School A also emailed their team’s suggestions for ‘growing’ Healthy Zones in special schools (the title of the co-produced logic model). None of the three schools we visited had a formal food policy, so we were unable to collect these. Analysis of the school-level data (fieldnotes, notes from meetings, and transcripts) primarily took a comparative case approach (Yin, 2003) in which schools were compared across relevant dimensions of the whole school food environment (WSFE), combined with thematic and Framework (Gale et al., 2013; Ritchie & Lewis, 2003) approaches. (Transcripts of formal semi-structured qualitative interviews have been anonymised and archived).

Work Package 1

It has been suggested that collaborative logic modelling with stakeholders can be ‘a type of creative practice whose value lies not primarily in the product (model) that is produced, but in the process of iterative reflection that developing a model promotes’ (Harris et al., 2018, p. 1).

The simple logic model that we co-produced with our stakeholders is included below (Figure 2). It identifies the growing conditions, activities, outputs, outcomes and impacts identified by groups as needed to achieve the HZ programme goal of making ‘nutritious food accessible throughout the school day to every student’. As shown by the icons indicating the source of the input, there were many areas in which the stakeholders’ contributions complemented and confirmed each other, as well as areas and concerns that were unique to each group. The model also includes information about how stakeholders want to engage with schools and researchers.

Figure 2. Final collaborative logic model for ‘Adapt-Ed Healthy Zones’.

The process of co-producing the final collaborative logic model involved four stages, as follows:

• Stage One: The collaborative process initially took place during the second (of three) meetings with each (adult) stakeholder group separately ((1) parents/carers; (2) people working in schools; and (3) people working with schools). Prior to this meeting (stage two), each stakeholder group member was sent a copy of the sunflower model (Figure 1) to familiarise themselves and consult with colleagues or peers, as appropriate.

• Stage Two: During the second meetings, the model shared with each stakeholder group was pre-populated by the research team, drawing on discussions during the previous introductory stakeholder workshops to avoid repetition. The online programme, Mural, was then used to create a more detailed version of the sunflower model. At this stage, three distinct models were generated for each of the three adult stakeholder groups to preserve the specific contributions made by each of the groups. The exception to this was for the ‘who are the gardeners?’ section, that we consulted on and added to during each meeting. By the end of the second group meetings, each group had its own ‘bespoke’ sunflower.

• Stage Three: The research team combined these three individual ‘bespoke’ models into one model, using a ‘key’ to indicate the source of the input. As expected, there was some overlap, as well as unique contributions made by each individual group. Relevant contributions from discussions with the children and young people stakeholder groups were recorded in a simplified Framework analysis table (Gale et al., 2013¹¹). These were then added to the model during stage two by the research team.

• Stage Four: For the third and final online stakeholder workshop, each of the three (adult) stakeholder groups were brought together into one group. The combined model was shared with the group using the programme, Mural. Feedback regarding its completeness and accuracy was overwhelmingly positive. Group discussions suggested that we had broadly reflected the contributions from each of the separate stakeholder groups and that members understood the model. A few correction and omissions were raised, which were noted and later added to the model by the research team, alongside a final check against the simple Framework analysis table

Below, we summarise discussions that iteratively informed and were informed by the co-production processes: (a) views of each group about school food for children with SEND and what improvement would look like and (b) how children and young people, parents and carers and professional stakeholders want to be included in decisions about school meals and in the research. In line with a Framework Approach (Gale et al., 2013) themes include those identified both deductively and inductively – i.e. related to workshop topics whilst leaving room for unanticipated findings.

a) Views about school food and what improvement looks like

Children and young people:

Listen to child voice: “Having choices is important.” (Young person, face-to-face workshop with SEND youth group, Hertfordshire).

Children and young people with SEND want to be involved in making decisions about school food, including flexibility or options for customising their food. The children and young people we spoke to expressed that they like having choices, and they would like to help to design the menu and give feedback to make ongoing improvements, such as via anonymous suggestion boxes or feedback sheets.

No surprises: “One bad experience with school food has a lasting impact on diet at home as well as at school” (Young person, online workshop with SEND youth group, Norfolk).

The children and young people told us that there are high stakes when it comes to food, with bad experiences leaving a lasting impact on the foods they eat and subsequently their overall diet. Some children and young people talked about liking variety, but many told us that they also like to know exactly what to expect.

Reduce anxiety: “Ordering in advance would be easier than queuing due to being nervous in the line.” (Young person, face-to-face workshop with SEND youth group, Hertfordshire).

There was an underlying theme of anxiety around food, including queuing up for food such as not knowing if preferred options would be available. They suggested that pre-ordering their meals in advance or the ability to see what is left on the menu on a live board whilst you’re in the lunch queue would reduce this anxiety at mealtimes.

Be sensitive to sensory needs: “Making sure there is always a quiet place to eat when it becomes too much for me” (Young person, face-to-face session with School Council in School A). This highlights the importance of the physical environment e.g., somewhere quiet, in the lunch hall with friends or eating in the classroom) and the ability to choose where they would like to eat each day, highlighting a key theme of flexibility and choice. The CYP made it clear that small differences to their lunchtime environment could change their entire experience and ability to eat and enjoy school food.

Use visual aids: “Have pictures on the menu of what the food actually looks like, so we know what to expect.” (Young person, face-to-face session with SEND Youth Group, Hertfordshire).

Visual aids were important to all the young people we spoke to. Suggestions included providing photos of the actual food and drink that will be available in their specific school and presenting food attractively. Another suggestion was to serve foods separately for self-assembly if possible (e.g., put sliced fruit on separate plates).

Parents and carers:

Take care with language: “Living with [a child with] ARFID... I would cut my leg off for my son to try something new. And I think as well, even like the wording ‘healthy’ ‘healthy food’... you all food is healthy...actually all food is good food. And I think actually to a child with ARFID to be told ‘this food’s healthy and this food’s not’... I think it’s the language around food that needs to change. You know it’s not ‘this is healthy’ and ‘this is bad for you’ - it’s food.” (Parent-carer, online stakeholder workshop).

The language around food and eating was particularly important to this group. They recommended avoiding binary terms such as ‘healthy’ and ‘unhealthy’ that can be a sensitive topic for children with eating and drinking difficulties, including Avoidant Restrictive Food Intake Disorder (ARFID), and taken literally by some children with SEND who may feel foods labelled ‘unhealthy’ shouldn’t be eaten at all. Parents preferred the term ‘nutritious food’ and felt that the availability of ‘safe’ foods (that do not trigger sensory sensitivities) is also important to CYP with SEND as it is fundamental to their mental health and emotional regulation.

Food is about more than nutrition: “If they can’t even engage because they’re hungry, or the food is making them feel upset or worried, then you know – what's the point?” (Parent-carer, online stakeholder workshop).

This quote illustrates the anxiety that can be caused by mealtimes, as well as the overlap between hunger and fullness and the ability to concentrate and impact on overall mental health.

Sensory play and exploration for positive relationships with food: “It should be based on more fun things. Kids like the colour, or the texture, the look – what does it feel like, is it crunchy? It could do something good to your tummy.” (Parent-carer, online stakeholder workshop).

Parents and carers in this group highlighted the importance of sensory play and removing the pressure around food. Sensory play such as blending Cheerios or counting carrot sticks was suggested to encourage a positive relationship with food.

The importance of communication between home and school: “If there was a way that they could communicate back and be like, you know ‘he was happy in the lunch hall today’ ‘he found this stressful’... that communication pathway without having to contact the school office or having to contact the teacher when you pick him up at the end of the day ... just a port of contact would make a huge difference.” (Parent-carer, online stakeholder workshop).

The group also suggested it was important to strengthen communication between home and school and liked the suggestion of using profiles for each child that could be updated from home with a child’s food preferences/dietary habits.

Embedding food into the curriculum: “Food needs to be seen as a vital part of the curriculum not an isolated part of the day.” (Parent-carer, feedback after online stakeholder workshop).

Parents and carers felt it was important to include food in the curriculum and that mealtimes should be included within this, as food is central to school life and the ability to learn and concentrate.

People working in schools:

Filling bellies vs nourishing bodies: “A lot of people don’t really understand what nutrition really means, and they think that filling a belly is more important. That’s their priority is to fill the belly rather than the quality of the food that’s going in.” (Food Technology teacher, online school stakeholder workshop).

There was an awareness among this group that some children may not have access to a hot meal at home, so it is important that there is a priority for schools to give them this at lunchtime. However within the group there were different views and competing priorities with respect to ensuring a child is fed (so it doesn’t matter what they eat, as long as they’re not hungry) versus providing food that is nutritious.

Sensory methods: “As simple as you know raw vegetables in, you know like a tomato, a cucumber. Let them touch it, let them feel it, and then explain what’s going to be done with that later on. So later on, that’s going to be made into a salad or something like that. I think if we try and introduce it and make it seem fun, you know that’s a way to get the children involved.” (School cook, online school stakeholder workshop).

This group emphasised the importance of using sensory methods to introduce new foods and doing this gradually, outside of mealtimes, to remove anxiety and to prevent children refusing to eat altogether.

Rewards: “Rewards don’t have to be food-based, you know, some time playing a board game with your friends. It doesn’t have to be food to be a treat.” (Teaching Assistant, online school stakeholder workshop).

The group discussed the use of sweets or chocolates as a reward in school but that it’s important that this is avoided for tooth health as children with SEND may struggle to go to the dentist, but also for overall health.

Improving communication between home and school: “I think if we can try and encourage the parents to try and do things at home, and then we follow through with it at school... because I think that is confusing for children as well.” (School Cook, online school stakeholder workshop).

The group felt there could be improvements in communication between home and school so that the same methods are used across both settings for consistency, and that this would be better for the child.

Links between lessons and mealtimes: “It’s not particularly a holistic approach across schools. So one department might be doing this, another department will be doing something else... some sort of support or encouragement for everybody to sort of knit together.” (Food Technology teacher, online school stakeholder workshop).

Additionally, the group felt creating stronger links between lessons and eating occasions e.g., break and lunchtime, would be helpful, alongside increasing mealtime length.

People working in policy and with schools:

Resources: “I think a key issue is resourcing, making sure that there are the resources in the school, particularly staff resources, to implement this properly. I mean so many of the school meals workers we talk to, they’re really passionate about their job but they say they don’t have enough time, there aren’t enough staff in the kitchen.” (Union representative, online policy stakeholder group).

This group emphasised the broader context of the education and catering landscape. It highlighted the inadequacy of funding for school meals, with schools operating within tight budgets, schools having to use other funds to supplement school meals, and the importance of campaigning for more school meals funding.

Visual aids: “Visual resources are really helpful. So they have pictures kind of outlining the different ingredients in the things that are on the menu. So, I think personalised support to kind of develop you know visual resources like that... so not just the menu.” (Public Health Officer, online policy stakeholder group).

The group also suggested methods for improving children’s access to nutritious food at school, including engaging CYP with SEND in school food using visual resources and interactive methods, as well as co-designing menus with children and their families/carers. This, combined with treating all children as individuals, providing menu sampling days and the option to pre-order meals to reduce anxiety about queuing were suggested.

Involving the whole school community: “Support around developing the best feedback mechanism for pupils. So how do pupils relate to the school or the caterer their opinion on the changes, and their experience of the changes. So whether that’s using you know like interactive... I’ve seen things you know like tablets with interactive smiley face symbols on and things for different foods and whatnot. But that would likely take more time and thought in a special school than mainstream.” (Policy Officer, online policy stakeholder group).

This group also highlighted the importance of everyone working together for cohesion, including catering companies and their staff, who are often overlooked and not included in the wider school community. They also felt that pupils feedback was important to ensure that any mealtime changes were right for them.

Communication between school and home: “Having the individual conversations with the parents to find out what they already do at home... and I think that’s really important... So if a child is already getting a hot meal at home at dinner time, then their focus is not on getting them a big lunch. It’s about getting enough to sustain them through the day in order to get home to have the dinner.” (Parent-carer representative, online policy stakeholder group).

Similar to our other stakeholder groups, the group felt that it was important to strengthen the communication with home and school so that approaches are consistent and the priorities are the same, to improve the child’s eating and overall health.

b) How children and young people, parents/carers and professional stakeholders want to be involved in decisions about school meals and in the research.

Children and young people said they like to provide input in a variety of ways e.g., feedback sheets with ‘emojis’ (smiley or sad faces), using a menu core board or communication book, suggestion boxes, typing answers on a computer, speaking out loud, handwriting or using picture cards. The groups also felt having questions in advance and time to prepare answers can be important.

Parents and carers told us that using a range of methods to engage parents is important and to be creative in order to reach parents and carers who do not do the school drop off if their child gets transport to and from school (may also live far away).

People working in schools suggested engaging wider school staff (e.g., teaching assistants) in data collection because they work 1:1 or in small groups with the same children throughout the year.

People working in policy/with schools suggested that schools maintaining good relationships with parents is very important, finding out what parents do at home and the priorities for the children’s eating, e.g., if they need a hot meal at school because they won’t get one at home. In addition, the group indicated for us to be careful with the wording of the research and not to single out a particular group, to be inclusive of all children with SEND.

Work Package 2

Scoping review 1 - Healthy eating interventions for children and young people with Special Educational Needs and Disabilities: An initial review of the literature showed limited studies related specifically to school meal interventions, and so the review also considered approaches with SEND outside of school meals that focus on improving eating and feeding difficulties in SEND. Interventions included school-based interventions (changes at mealtimes or involved nutrition education lessons), after-school interventions, those in a feeding clinic and remote based interventions via Zoom or an app, behavioural/sensory interventions and Environmental changes. Changes pre and post intervention were typically measured using standardised eating behaviour questionnaires (e.g. Children’s Eating Behaviour Questionnaire), Nutrition (Healthy Food Index) and measured the increase in fruit and vegetables intake, amount of food consumed via number of bites and/or amount of food left on the plate. Interestingly, mealtime intervention studies conducted with children with SEND during lunchtime and which utilised the existing interventions used in mainstream, did not report improvements in healthy eating (Taylor et al., 2013). This may be due to reporting methods but may link to the fact that interventions should be specifically designed for children with SEND.

We found only one intervention co-adapted for special schools or SEND children (Hubbard et al., 2014). These authors undertook a community-engaged process to adapt a Smarter Lunchroom intervention for youth with intellectual/Development Disability aged 9-22 years attending a specialised private residential school in the United States. Focus groups and interviews with school staff elicited recommendations for adaptation strategies which focused around increasing choice along with the heterogeneity needed to address food related challenges. However, this was not developed with school staff or CYP.

The results of this review suggest several different approaches exist to improve healthy eating, feeding and diets of CYP with SEND. However, most interventions have focused on children with a diagnosis with autism and outcome measures for healthy eating and been heavily weighted to addressing changes in consumption of fruit and vegetables, and the interventions varied significantly in their delivery time.

Scoping review 2 - Measuring diet and eating behaviours in children and young people with special educational needs and disabilities: This review focused on the assessment methods, tools and outcome measures used to assess diet (food, food groups, energy and nutrients, dietary quality and dietary diversity) and eating behaviours in CYP with SEND. Over 25 different assessment tools were identified. Some of these assessment tools were used to assess specific aspects of eating or drinking difficulties related to certain conditions. Most instruments designed to measure eating behaviours relied on parental report, with few studies including children’s self-report measures. There has been a focus on autistic children feeding difficulties and diet, and comparatively less on children with motor difficulties and/or intellectual impairment. Furthermore, there is a scarcity of research assessing children’s eating behaviours and dietary intake specifically within a school environment, however, this is vital to understand as children typically consume around half their total daily energy intake at school.

Work Package 3

To explore the relationship between school characteristics and the take up of FSM, we conducted (a) analysis of routine publicly available administrative data; (b) an online school survey; and (c) an ‘audit’ of special schools’ websites.

a) Secondary analysis of routine publicly available data:

i) Descriptive analysis of free school meal eligibility and take up in the East of England: Our analysis of EoE schools found that special schools have around twice the proportion of children eligible for FSM as other types of school (Table 4). However, we found that take up of FSM in special schools lags behind other school types in EoE, with over one-in-four FSM-eligible pupils (27.4%) and nearly one-in-four infant pupils (24.5%) not taking up their entitlement (Table 5). The gap for infant FSM is especially stark when comparing special schools (75.5% take up), primary with SEN provision (84.7%) and mainstream primary (86.9%). Take up in Special and Primary schools tends to be slightly lower in EoE compared to the rest of England, and slightly higher for Secondary schools.

ii) Multi-variate analysis of FSM take up (England¹²): Special schools have similar take up of means-tested FSM compared to other primary schools with the same pupil demographic characteristics, location, and governance type (Table 6). Within non-special schools serving the same age groups, there is no significant difference in take up between those with a SEN Unit or Resourced Provision and mainstream-only schools. However, special schools have considerably lower Infant FSM take up than other schools (by 6-7 ppt). Aside from this, very little variation in FSM take up is explained by demographics, school type or location. To the extent that take up of FSM can proxy for a successful school food environment, this suggests that stratifying a future randomised trial only based on routinely collected would not improve precision, and that information about policy and practice within schools will be essential for this purpose.

b) Online school survey (East of England): 33 schools responded, including 16 special schools and 17 with a SEN Unit, Resourced Provision or Special Classes. Compared to non-responding schools, responding schools are somewhat positively self-selecting, with higher take up of means-tested FSM and UIFSM recorded in routine data (Table 7). Despite their apparent positive self-selection, only four responding special schools (25%) and one other school (6%) have a written school food policy; and 7 and 6 (44% and 35%) respectively have a named member of staff or governor responsible for school food (Table 8). Both school types had equal numbers with an external catering provider (i.e. local authority or private company) and in-house provision. With respect to other aspects of the whole school food environment (WSFE), all responding schools (n=33) include cooking, and most (n=24) include gardening, in food education activities. Other food education activities include community outings and using sensory approaches to encourage tasting or exploring of new foods. Schools also reported which aspects of the school dining environment may be important for school meal take up. Notably, pupils’ sensory issues were mentioned as a key factor, including: difficulties with noise and large groups of people; limited or specific food preferences (e.g. due to ARFID); how food is plated; and/or aversions to certain tastes, smells and textures. Other factors include portion size, cost and variety of meals; dining space/capacity; and the number of staff to support pupils with complex needs (e.g. gastrostomies).

The main challenges, as reported by schools, when providing food for or feeding children with SEND are not having enough staff, a lack of appropriate space or limited facilities, and limited funding. Other reported challenges included the type and extent of support that some children require, such as medical or physical mobility needs, swallowing, choking (dysphagia), allergens and providing specialist medical diets (e.g. via feeding tube or blended food). Some schools also reported that behavioural issues (e.g. oppositional behaviours), related to specific types of SEN, are more likely to occur due to sensory overwhelm, frustration from queuing, a lack of control and/or dislike of the food available, which can also affect the other pupils.

Whilst numbers are too small to draw firm conclusions, exploratory descriptive analysis suggests that the following factors (unobserved in admin data) may be associated with school meal take up: meal cost, time available for lunch and whether the school is part of an external food-related scheme or accreditation programme. There is no relationship between take up and catering provision (external vs. in-house) or whether schools have a food policy.

A majority of schools (n=27, 82%) said they were somewhat or very interested in receiving additional support to improve their school food. Thirty schools (91%) stated that they would or might be interested in taking part in a future study trialling an adapted Healthy Zones programme. Schools responding ‘it depends’ (n=2) expressed concerns about funding during the intervention and beyond the first year, staff time necessary to implement the intervention, and what the impact would be on pupils’ eating habits.

c) School website audit (East of England): An ‘audit’ of EoE special schools’ websites (N=114) explored what other relevant information could be gleaned from these. Only 11 schools (10%) had a school food policy available online. Seventy-nine schools (69%) had identifiable catering providers, of which 50 (63%) were ‘external’ and 29 (37%) were ‘in-house’. Lunch menus were available online for three-quarters of schools (n=87, 76%). Given the availability of menus online, they are a potential resource for a future study (e.g. analysis of nutrition and adherence to school food standards). However, as noted below (WP4: Table 9), menus may not accurately represent foods available.

Of the 114 special schools included in the website audit, 16 also completed the online school survey, which asked about catering provision (internal vs. external) and whether the school had a written school food policy. There was only one discrepancy in the type of catering provision recorded across these two datasets. However, of the four special schools that stated they had a written school food policy via the online school survey, only one of these schools had made the policy publicly available on their website. This suggests that conducting an audit of school websites is not sufficient to ascertain whether ADD 16 and 17 to bottom right two cells have a written school food policy.

Work Package 4

The ADAPT guidance (Moore et al., 2021: 4; Pfadenhauer et al., 2017) suggests that, in making decisions on the nature and extent of adaptation needed to achieve a good fit between an existing intervention and new context, the ‘use of an existing theory or framework (or elements of these) to structure thinking about context can be useful’.

In analysing the ethnographic data, we compared school food contexts across broad dimensions of the whole school food environment (WSFE) that, following Moore et al.’s (2023) framework, we summarised as: food policies and leadership; food provision; the physical setting; food education/activities: parental and child engagement in school food (see also Bryant et al., 2023). To these established WSFE dimensions, we added another aspect of school-level context: ‘educational philosophy and approach’, that emerged during our analysis as fundamental to understanding the provision of food for children in special schools, in that it underpinned schools’ approaches in the other areas. Table 9 summarises the comparison of schools’ approaches according to these dimensions, identifies aspects of context that were similar as well as those that seemed to make a difference to school food provision and practices, and provides brief selected quotes from interviews to illustrate key points. In comparing school contexts across the WFSE dimensions, as Table 9 shows, we found many similarities and some differences:

In terms of the educational philosophy and approach, in all three schools there was an emphasis on schools' role as providers of ‘care’ as well as ‘education’. Schools’ ‘ethic of care’ included recognition of the needs of individual children and the (often challenging financial) circumstances of many families - caring ‘for’ and caring ‘about’ (Tronto, 1998). This entailed an attention to ‘equity’ as opposed to ‘equality’, wherein what is fair or just was prioritised over what is the same. In terms of food this meant understanding that ‘one size does not fit all’ and that flexibility is needed in the application of rules. Priority was also placed on children eating and drinking something at school, so that they were able to regulate their behaviour, and because they might not eat at home.

The three schools were also similar with respect to policies and leadership; none had a school food policy and in each there was some resistance to the idea, with emphasised placed on children’s individual needs, as articulated in their EHC Plans, and the inappropriateness of ‘blanket’ rules. An exception to this was food safety: all schools had clear rules around this, for example, the refrigeration of food from home or not reheating previously cooked food. Unsurprisingly, then, none of the staff had food leadership in their roles, however some health was in the remit of some senior leaders.

In terms of food provision, in all three schools there was recognition from teaching, catering, support staff and parents and carers that food is about much more than nutrition, particularly for children with sensory sensitivities, and very important for children’s sense of control and wellbeing. However catering arrangements made a difference to how flexible catering staff could be. In all three schools, Cooks were able to and did amend presentation of food, serving foods in ways they hoped were more appealing for some children e.g. raw or separated. However, in terms of food provision, in the two schools in which the cook was employed by the school (Schools B and C), pupils were able to choose meals, and cooks provided completely or partially bespoke meals, but this did not happen in School A in which the catering was provided by a LATCo. Packed lunches and snacks brought by pupils were in all three schools observed and/or said to be high in pre-packaged and processed foods.

In terms of the physical setting, the importance of the sensory environment was acknowledged to shape children’s eating and in all three schools some or all the children ate lunch in classrooms rather than the dining hall. However, the school size and layout seemed to impact on the quality of the food, because food travelling in hot boxes to classrooms some distance away could spoil (e.g. chips could get cold and soggy), which was particularly challenging for children with sensory sensitivities. Hence, whilst the need for a calmer eating environment was acknowledged by all, there were challenges and trade-offs, particularly for the rural school with its larger footprint.

As ‘outsiders’ or ‘strangers’ (Schütz, 1964) to special education settings, we were struck by the centrality of food and food education / activities in the curricula of the special schools we visited. In all three schools, reflective of the purposes of special education, ‘food skills’ were recognised as fundamental to independence and wellbeing. Pupils in all three schools took part in regular cooking activities in ‘enabling spaces’ in which they were invited to explore and/or eat food. In all three schools, children took part in food shopping and other ‘community’ activities (e.g. making shopping lists, doing food shopping and making trips to restaurants). Whilst there was provision for growing food in all three schools this was less well supported, with a neglected balcony in one school and unused polytunnel in another due to staff led turnover. Food was also used as a medium for learning in all the schools, ranging from sensory play to the (more controversial) use of (less healthy) food as a reward, as food was seen as very ‘comforting’ or ‘motivating’ for some children.

Finally, with respect to parental/child engagement in school food, the embeddedness of children’s food and eating within family routines and relations (Albon, 2007; O’Connell & Brannen, 2016) was widely recognised and emphasised by staff. This meant schools saw the need for and value of involving parents in decisions about children’s eating, with most engagement 1:1 through a phone call or note in the home/school book or children’s bags. Because of wide catchments and use of transport, schools found it difficult to encourage parents to take part in group activities. However, two of them (Schools A and C) held regular coffee mornings whilst the other school (B) in a more affluent area found this did not work, as a large proportion of parents were in paid work. With respect to children and young people’s involvement in school food, it was quite striking, given the emphasis on children’s individual needs, that they (and parents) had only been formally consulted about school food in one of the schools (School C). This was prompted by the school’s participation in the Healthy Schools scheme. When asked about their views of school food as part of this process, the children had shared lots of ideas about what they would like to see on the menu, including fruit smoothies. The parents we talked to in Schools A and C, and the children in the School Council (School A) had lots of ideas about how to improve food at school, and suggestions for how childen might be supported to share their views. These are covered above (WP1) and in the next section on ‘school staff and parents and carers’ views, as well as in the draft sunflower logic model (Figure 2).

School staff, parents and carers’ views on acceptability, impact and feasibility

In addition to comparing school contexts in terms of their WSFE, we spoke with parents, carers and staff about their views on opportunities, challenges and acceptability of the intervention; what improvement would look like and the impacts we would see; and the feasibility and practicalities of delivery and collecting data. These are summarised below.

Opportunities, challenges and acceptability. The members of staff we spoke with in all three schools believed food and eating played an important role in children’s health and wellbeing, and Headteachers in all three schools expressed interest in taking part in a future intervention.

Schools recognised they had a responsibility and opportunity to support healthy food and increased dietary diversity. Headteachers also mentioned that eating at home could be difficult and parents’ anxiety that children ate could transfer onto children. However, their experience was that the school environment could relieve some of this stress:

However, staff recognised that pressuring children with highly restricted diets to try or eat foods risked them stopping eating altogether. The need to take a case-by-case approach was mentioned, and a diversity of methods for introducing new foods was observed. This included using ‘taster plates’ at lunchtime, to allow ‘room for growth’, even where children had packed lunches, or were mostly fed by gastronomy. However, there were different views; some suggested it was better to avoid the introduction of foods at mealtimes at all, and stick to other times, whilst others mentioned being confused about what was best and needing access to specialist advice. The need to consult with and involve parents was mentioned by staff in all three schools, as were the challenges of this, given what one Head called ‘fixed mindsets’, as well as the practical constraints of wide catchment areas and some parents not having cars. Parents highlighted the stress of mealtimes at home that at times triggered violence, and having a lack of time and energy, and other priorities.

What improvement would look like look like and the impacts we would see. Staff in two schools mentioned improvements they would like to see to the WSFE, specifically a salad bar that children could select food from (only one school had this). More than one parent mentioned the importance of children being able to see food, so they knew what to expect – visiting the kitchen or seeing photographs of actual meals. Some staff seemed to want to see an improvement in terms of the nutritional quality of the foods children brought to school to eat, in some cases appearing dismayed at pre-packed highly processed contents of many packed lunches and snacks, but also keen to emphasise they ‘understood’ the challenges and were ‘not judging’.

Parents also said they wanted to see an increase in the variety of foods their children ate, especially nutritious foods like fruit and vegetables, and to have less stress around food and eating at home. However, some were sceptical about children’s capacity for change, whilst staff emphasised the need for parents to be ‘on board’, and both groups talked about the slow nature of progress, ‘small steps’ and not expecting ‘miracles’. The Head in School B said,

In terms of where impacts of improved school food could or would be seen, being able to access a school meal that they wanted to eat at lunchtime (in two of the schools) was seen to have an ‘immediate’ impact on children’s happiness. School B had recently introduced repeated popular meals daily, with one changing option, and children ordering at morning registration. The Head said that as a result there had been fewer behavioural incidents at lunchtime and ‘it makes playtime happier because you haven’t got children still indoors in crisis because they didn’t get what they wanted for dinner’. And in School B, that had a ‘smiley’ new Chef who encouraged children to enjoy school meals, the Executive Head said the children ‘have been much happier after their school lunches’. Staff also noted the impacts of sugar and additives on children’s behaviour, ‘especially’ for children with attention hyperactivity disorder (ADHD), implying there could be potential for improved behaviour with a reduction in processed and sweetened foods and drinks.

The Head of School A also said impacts of giving children what they liked to eat for lunch were that it ‘saves money, saves time’, whilst the Cook said there was much less waste. The Chef in School C mentioned the results he would see from an improved WSFE would be an increase in numbers having school meals and consequently a reduction in ‘cost per meal’. (It is important to note with respect to this potential indicator that figures collected from the schools on take up of paid and FSM for the weeks we visited diverged from the administrative FSM census day data; this suggests it could be a mistake to rely on administrative data to examine impacts on uptake at individual school-level).

Potential longer-term impacts on the diversity of children’s diets and eating behaviours at school and home were also mentioned. The Head in School A described using ‘tasting plates’ with a boy who’d had a very restricted diet, and as a result of gradually introducing new foods at school, he could eventually eat school meals. This was said to have had impacts on parents too:

Feasibility and practicalities of delivery and collecting data. According to a teacher in School B, feasibility depends on the intervention ‘slotting in’ to what the school is already doing. The Head of School B suggested using data already collected by teaching staff on children’s ‘engagement’ in learning: ‘that’s a big part of our lives, looking at engagement.’ She also suggested monitoring changes documented on children’s Education, Health and Care Plans (EHC Plans) as ‘if there’s a problem with eating then most of them would have something on their education health and care plan about it’. The Head in School A also suggested we could draw on data staff already collected on emotional regulation and behaviour, alongside researchers monitoring food through ‘food diaries’. However, she noted the importance of contextualising data about what children eat, as ‘who they sit with’ , for example, may influence what they eat. The Executive Head (School C) noted it would be ‘difficult’ to monitor progress as children have very different starting points. One Teaching Assistant (TA) told us TAs would want to support the study by collecting data, since they did the job because they cared about children’s wellbeing. Asked about their views, parents and carers suggested using a mix of approaches such as home-school communication books, Zoom and telephone. Whilst some staff mentioned challenges in engaging parents and carers, a Family Support Worker (School A) suggested they would be keen to have their voices heard, as they can often feel ignored.

Developing networks and partnerships

In addition to our core stakeholder workshops and scoping research we have taken part in extensive two-way engagement, giving invited talks and discussing the project with the Disabled Children’s Council’s ‘Special Educational Consortium’ (SEC), the School Food All-Party Parliamentary Group, and a webinar for the Nutrition Society Academy. We have also met with the Royal College of Speech and Language Therapists (SLTs), local authority (LA) leads, a school meals consultancy, special school Headteachers, a Head of School Food Curriculum in a large special school, an SLT specialising in dysphagia and complex needs, and an NHS Dental Nurse. Adapt-Ed has contributed to responses (by the Food Education Network and SEC) to the Curriculum and Assessment Review. The Executive Head of a Special Needs Trust of five special schools expressed an interest participating in the future study, 30 EoE schools said in our survey that they were potentially interested in taking part, and we have been contacted by LA leads in areas outside of EoE who want to know more. With additional funding from University of Hertfordshire’s (UH) ESRC Impact Acceleration Account, we are collaborating with public and professional stakeholders to co-produce outputs for different audiences, including a logo competition for children and young people with SEND, an inclusive face to face event and podcast supported by the Herts Young People’s Advisory Group. We are preparing a summary of evidence-informed resources requested by school and LA practitioners in our network and have prepared a Policy Brief that includes recommendations and an invitation to join our future Advisory Board. Crucially, we have strengthened our working relationship with School Food Matters (SFM), who will deliver the intervention and who have identified funding for the future Adapt-Ed HZ intervention.

We have also developed the networks and team required for the future research. We have the support of UH’s Clinical Trials Support Network, who will assist with methodological and statistical expertise in developing the future evaluative study. We have established links with the EoE Regional Research Delivery Network (RDN) that is building a schools network and, with a longer recruiting timeframe that will allow for portfolio adoption, will be able to support recruitment of schools in the future study. We have also added additional members to our research team with subject and methodological expertise in health economics, public health, nutrition and dietetics, speech and language therapy, complex interventions and process evaluation. We are in contact with an advisor at the Specialist Centre for Public Health (University of Leeds) to help support the upcoming funding application.

Collaborative logic model

A simple logic model using the metaphor of a sunflower (Figure 2) was developed collaboratively with our contributors and stakeholders to illustrate how and why the Adapt-Ed HZ intervention may lead to health and wellbeing impacts on children, families and schools. This was based on extensive discussions with our public contributors about their experiences of food in school, as children and young people and as parents and carers of children and young people with SEND. It is also based on extensive discussions in groups and individual meetings and interviews with professional stakeholders working in schools and with schools. For example, children expressed the importance of having a say about food at school and suggested different ways they could do this. Anxiety about food and its immediate impacts on children’s mental health and family relations was raised by parents, carers and school staff as a key issue, and they stressed the importance of strengthening communication with parents to support change. These are some examples of emerging conditions that would need to be met in a future intervention to enhance its chances of success. This logic model will be refined with core stakeholders and, in the next steps of this research programme, (feasibility study) implemented in a handful of schools to provide more information to support a full-scale definitive trial.

Outline research plan

From the scoping research completed in this ADA, an Adapt-ed HZ intervention and outline research plan have been developed. These have been co-designed based on our discussions with our Working Group, public co-applicant, public and professional stakeholders and engagement with SFM.

The plan is for a future trial that will adopt a Stepped-Wedge design, as the Adapt-Ed HZ intervention focuses on institutional change. Subject to a feasibility study, it is envisaged that the future trial will consist of three overlapping Work Packages (WP). WP1 will recruit schools to the Stepped-Wedge trial that will implement and evaluate the effectiveness of Adapt-Ed HZ on the health and wellbeing of children with SEND and other outcomes. WP2, a process and outcome evaluation, will use a mix of qualitative and quantitative methods to test whether and how the Adapt-Ed HZ intervention is working. WP3 will develop knowledge mobilisation and impact plans, and public and professional stakeholder involvement activities, which will be woven throughout.