A reflection on the co-design approach to the development of the Medway Can campaign; a whole systems approach to obesity prevention using COM-B

Abstract Obesity is one of the highest metabolic risk factors associated with morbidity and mortality in the UK, with two-thirds of adults in the UK classed as overweight or obese. Whole systems approaches can be effective in tackling this public health challenge through stakeholder and key partner engagement. This article describes the co-design process for a social marketing obesity prevention campaign taking a whole systems approach based on the COM-B model of behaviour change. Development of social marketing campaigns through Agencies is often hampered by rapid turnaround and short timescales; we highlight how drawing on existing knowledge and co-design with beneficiaries can support the design, delivery and implementation of a social marketing behaviour change campaign.


Introduction
Obesity is one of the highest metabolic risk factors associated with morbidity and mortality in the UK, with two-thirds of adults in the UK classed as overweight or obese (NHS, 2021).This is following an upward trajectory and in 2021 there were over one million hospital admissions where obesity was a factor: a 17% increase since 2019 (UK Government & Public Health England, 2021;NHS, 2021).Medway is a district in the Southeast of England, UK, with a population of 279,800 (Office of National Statistics, UK Government, 2021) Locally, in Medway, UK, there has been an increase in overweight and obese children of all ages and adults between 2008 and 2019 compared to the England average which appears to be trending towards a plateau (UK Government, 2021, Medway's Whole System Approach to Tackling Obesity, p. 2).
Associated behavioural factors contributing to obesity levels are physical inactivity and poor nutrition (UK Government and Public Health England, Health Profile for England, 2021).Only 66% of the UK's adult population report meeting recommended physical activity guidelines (150 minutes/week of moderate to vigorous physical activity; UK government Physical Activity Guidelines, 2019) and only 1% are meeting National Health Service nutritional guidelines (Steenson & Buttriss, 2021).Health inequalities are evident in these behaviours, with physical activity levels falling as low as 50% in Black, Asian and Minority Ethnic (BAME) populations, those living with disabilities, and those living in more deprived areas (Public Health England [PHE], 2021).Further, adults meeting recommended dietary requirements (e.g., "five-a-day" fruit and vegetables) in 2019-2020 was just 55%, and even lower in BAME populations (46%), people living with disabilities (52%), and people in more deprived areas (46%) (PHE, Health Profile for England, 2021).
In light of these figures, obesity prevention is a public health priority, particularly for at-risk groups experiencing greater health disparities.The multifactorial drivers of obesity have increased interest in whole systems approaches to tackle this challenge (Rutter, 2011).In a review commissioned by the National Institute for Health and Care Excellence (NICE), Garside et al. (2010) endeavoured to define whole systems approaches as heterogeneous elements that interact to elicit change that would not have been possible at an individual level.Whole systems approaches have the potential to address implementation lag from research to practice, particularly in public health campaigns, by engaging key partners and community members (Bagnall et al., 2019;Greenhalgh et al., 2004;Northridge & Metcalf, 2016;Westerlund et al., 2019).These agents play an important role in identifying potential intervention implementation barriers early in the planning process.
Despite the complex interaction between the systems that impact human behaviour, there is a compelling argument that systems thinking and individual approaches are not, in fact, dichotomous; individuals constitute the various interacting systems, whether they are policymakers, healthcare professionals, town planners, government or non-government organisations, or individual community members (Sniehotta et al., 2017).Thus, it is equally important to understand the drivers of individual behaviours, groups of individual's behaviours, and the systems in which they interact.The COM-B model (Michie et al., 2011(Michie et al., , 2014)), and its associated behaviour change wheel, is a unified theory that supports the planning of an intervention by identifying mechanisms of action that lead to behaviour change, such as improving self-efficacy or beliefs about capability.Through a period of systematic research, a taxonomy of theoretically derived behaviour change techniques have been identified which are linked to specific mechanisms of action (Michie et al., 2014).These, in turn, are linked to specific models of behaviour and form the basis of the behaviour change wheel (Michie et al., 2011; Figure 1).
The behaviour change wheel offers 7 policy categories (e.g., service provision, legislation, communication/marketing) which can be targeted via 9 intervention functions (e.g., education, persuasion, enablement, etc.) to address the sources of behaviours (an individual's sense of capability, opportunity, and motivation).One criticism of the COM-B model is that it rejects human variability (Ogden, 2016).However, proponents of the COM-B model stress that the purpose of a systematic model of behaviour change is to embed complex theoretical explanations of human behaviour into an accessible and transferable model that can be employed efficiently by intervention developers and policymakers across various settings, organisations and industries; this lends itself to the frequent application in public health interventions (Willmott et al., 2021).As such, the COM-B model supports implementation in realworld interventions which was a priority for the development of a healthy weight campaign in Medway.
The COM-B model describes three components that need to be met for behaviour to change: capability, in the form of both psychological and physical capability; opportunity, in the form of both physical and social opportunity; and motivation, in the form of both reflective and automatic motivation.The nature of the COM-B model, which includes social, environmental, psychological, and physical determinants, means that it is suitable for a whole systems approach.Finally, COM-B has been successfully used to explain and predict health behaviours such as physical activity (Ellis et al., 2019;Howlett et al., 2019) and dietary choices (Craveiro et al., 2021).
The necessity of codesign in the development of whole systems public health campaigns that aim to tackle complex problems such as obesity cannot be overstated.A shift from viewing intervention beneficiaries as passive recipients to empowering them to codesign and codevelop interventions is being driven by a need to ensure that health-related interventions are fit for purpose and appropriate for the target population, that research focuses on issues that are important to communities, and that funding is allocated where it is most needed (Sendra, 2023).The use of codesign in traditional marketing is commonplace and is a successful strategy to ensure end-users and beneficiaries contribute to the development of products and services that meet identified needs.This technique has filtered into the development of those products and services, including mass media campaigns, that aim to change behaviour for social good (Green et al., 2019;Lee & Kotler, 2019;Sampogna et al., 2017).Taking a systems approach to the development of a social marketing campaign means that actors at all levels of the system must be engaged in the codesign process.
The current article describes the employment of COM-B in the co-design sessions with stakeholders and the development of the Medway Can campaign, a social marketing obesity prevention campaign in Medway, England.It is anticipated that this case study will lend itself to informing good practice in the development of future communication efforts in the promotion of health behaviours.

Study design
The co-design sessions took the format of qualitative focus groups.The COM-B model (Michie et al., 2011) was used in the development of the focus group discussion guide and informed the analysis.

Participant sampling and recruitment
Participants were recruited through purposive methods via existing stakeholder networks.Stakeholder groups were identified as 'target audiences' for the campaign and were (i) primary school aged children (ii) local leaders in physical activity and health promotion (e.g., council members, sports club coaches, etc.) (iii) community members living within areas of Medway with higher deprivation scores.A local primary school acted as gatekeeper to support a co-design session within the school and pupils were recruited according to their availability during school time.Local leaders in physical activity and health promotion, and community members from areas of Medway with higher indices of multiple deprivation were recruited through existing stakeholder networks, via email invitations.The stakeholder group consisted of council members, local authority physical activity specialist/officers, whole systems obesity officers, religious leaders, schoolteachers, and community leaders who could support with reaching diverse community members.

Co-design session design
Discussion guides and activities were developed following the discover, define, develop, deliver model (UK Design Council, 2023) and COM-B model of behaviour change (Michie et al., 2011; Table 1) and adapted according to the stakeholder group.All groups were provided with supplementary materials such as mood boards containing past campaigns and activities to support the design of logos, straplines, and copy for the campaign.Primary school children were provided with drawing equipment and were supported by teachers and classroom assistants.Each group was first asked to define the audience and describe the barriers and drivers of health behaviour, design an intervention or campaign that would tackle the barriers from their perspective, and then review previously developed messaging and campaigns (Table 1).
Co-design sessions were facilitated by two members of the social marketing agency team and were assisted by school staff or members of the local authority, where appropriate.One facilitator took notes throughout the sessions.It was determined that sessions would not be recorded because data was not being collected for the purpose of research, rather for the purpose of co-design.Notes from sessions were collated and analysed using the COM-B model.The findings were used to inform the design of the campaign.

Results
A total N ¼ 55 participants took part in the co-design sessions.N ¼ 30 were children in the volunteer primary school, N ¼ 16 were stakeholders, and N ¼ 9 were members of the local community from lower socioeconomic areas of Medway.

Capability
One of the most salient findings relating to the Capability domain of COM-B was that all groups appeared to have sufficient psychological capability to support health related behaviours.There is an apparent sense of 'exhaustion' from all the information that is given about diet and exercise, and it was reported that intentions to change do not necessarily lead to real changes in behaviour.Stakeholders, children, and community members alike were able to cite the health benefits of physical activity and the balanced nutrition with little or no prompting.Some psychological capability needs were identified by participants; many felt that they would benefit from the seeing culturally and ethnically diverse healthy recipe options within a campaign.It was suggested that many healthy weight campaigns did not accommodate the needs of a diverse audience.
Similarly, many respondents-particularly within the stakeholder groupfelt that there was sufficient information in the public domain relating to physical activity, but there was limited information relating to the risks associated with sedentarism.This also lead to conversations about reframing physical activity, in that participants would rather be prompted to 'stop doing nothing' than to 'start doing something'.Despite the apparent high levels of psychological capability, there appeared to be a disconnect between having the knowledge and intention to act, and the motivation to drive that action.
The physical capability domain was more complex.Most participants in the children's co-creation session expressed a strong sense of self-efficacy for being active, but not for nutrition.As might be expected, they did not have autonomy over nutrition decisions in their homes.Their views aligned with other stakeholders in that there appeared to be a primary decision maker relating to food in the home, and this often took the form of the mother, wife or female partner.
In the adult co-creation sessions, it was expressed that existing overweight or obesity acted as a physical capability barrier to exercise-being overweight made it difficult to initiate physical activity as it was often accompanied by pain and discomfort.This also negatively impacted motivation for exercise and self-efficacy for engaging with healthy eating behaviours.

Opportunity
Participants in all groups identified physical opportunities for exercise.However, there were several barriers identified within the physical opportunity domain.While it was felt that there were sufficient services available to encourage physical activity (gyms, community sport centres, online classes, outdoor spaces, etc.), these were often perceived to be hostile environments.Many adult participants expressed a need for more welcoming spaces for people who already experienced overweight or obesity.This appeared particularly apparent in female participants who expressed that they would be much more likely to attend classes or engage with services that were facilitated or run by female staff.
Affordability was expressed as a physical opportunity barrier to both exercise and nutrition in that there was a perception that healthy eating is expensive and membership to exercise facilities is an expense that many families cannot justify.Participants suggested that there was a need for free or affordable exercise sessions and that they would like to see information about how to eat well on a budget.It was also suggested that physical activity should be reframed to include both changes in sedentary behaviour (e.g., sitting less) and non-tradition forms of exercise (e.g., playing with grandchildren).The physical opportunity domain highlighted some of the systemic barriers to healthy weight management through a sense of poor autonomy in participants.They felt that they had little or no influence over the opportunities that were provided to them through schools, local authorities, healthcare and the wider community.Lacking time was cited as a barrier to health behaviours (no time to exercise or prepare healthy meals) but, upon further probing, many adults in the groups identified that if they prioritised their health they would be likely to find the time to exercise and eat healthily by replacing other, sedentary, behaviours.
Similarly, social opportunity acted as both a barrier and a facilitator.Comments about creative concepts suggested that families and individuals who appeared in campaign imagery were considered "too perfect" and were not appealing to the audience.Despite the seriousness of this public health concern, messages that participants created organically were often focused on humour and light-heartedness and they felt positively about humour being used in past campaigns.Light-hearted straplines appealed to all groups.Equally, when participants could identify with the messengers (such as case-studies) within campaigns, they were more likely to attend to the call to action.A defensive response was elicited when imagery and campaign messaging seemed unattainable.
Participants identified potential driving 'actors' for health behaviours such as the decision maker in the house (again, often drawn or portrayed as the female), healthcare practitioners, and community leaders (e.g., religious leaders and teachers).Each group mapped out the decision-making process of specific behaviours; for example, children identified that parents often had intentions to be active, but that long days at work and early dark nights in winter often created barriers to acting on such intentions.It was interesting to note that the social opportunity relationship that existed between child and adult participants was reciprocal; school children cited their parents as social drivers of healthy weight behaviours.Equally, adults felt that children were a driver of healthy weight behaviours and would make impactful messengers within a campaign.

Motivation
Motivation was the frequently cited in all groups.Motivation in children seemed to be driven by a sense of autonomy, identity, and fun.When asked to design a slogan or icon that they might find appealing, many of the children included the word "Medway" in their designs.They suggested "Motivating Medway," "Move Medway," "Moving Medway" and "Medway Can."They liked a play on words and seemed to find alliteration appealing, suggesting slogans such as, "Healthy Happy Habits".When prompted by facilitators to think about imagery that made them think about moving forward, progression, or actual movement, they drew arrows, footprints, triangles and circles.Similarly, community members recognised images of steps and ladders as symbolising moving forward.These images were embedded into the campaign design.
Stakeholders and members of the public were asked to comment on language use and the design of creative concepts and past campaigns in relation to motivation.There was a consensus that the word "challenge" associated with a healthy weight campaign is immediately objectionable for some audiences.Rather, it was agreed that the focus should be on personal strengths, the promotion of a sense of wellbeing and personal achievement.Participants in stakeholder and public sessions agreed that a sense of community should be established, rather than a sense of competition within the community.Stakeholders and community members were keen to see that the campaign delivered messages promoting behaviour change to increase energy levels and better overall health.
Further, it was expressed that a healthy weight campaign has the potential to invoke feelings of shame and embarrassment in the target audience, thus reducing motivation to change behaviour.Participants felt that humour had the potential mitigate this.Combined with messages about what can be gained by improving weight-related behaviours, rather than risk avoidance, there is less likelihood of audiences becoming defensive.Participants highlighted that humorous messaging is also likely to gain attention and lead to the retention of information and act as a buffer for the potential negative emotions that otherwise might be elicited.Some of the humorous messages that were suggested were "Get off the sofa, these foods need you!" and "End carrot stigma today!"When participants were shown images of humans that had been used in past campaigns, they enjoyed bright colours and realistic personas.They preferred images of people that were relatable in their weight, ethnicity, gender, economic status (depicted as "workers"), and age.Participants, particularly children, commented on the smiling, happy faces and the use of a positive message rather than messages about risk-avoidance.Motivation was increased when there was a sense of relatedness in campaign imagery.
Further, participants felt that the use of environmental cues for increasing health-related behaviours, such as signs encouraging the use of stairs and placement of healthy produce, might support behaviour change.Lack of motivation, both automatic and reflective, appeared to be a key barrier to behaviour change in all groups.Despite having awareness of the problem and many opportunities to change behaviour, motivation for change was lacking.

Delivery
Based on the findings from the codesign sessions, a strategy was developed for the implementation of the Medway Can campaign.An overarching brand was developed in co-design sessions, that aimed to be fundamentally motivating.The designed brand included a star figure, with five colours to represent the five boroughs in Medway.The brand had the ability to adapt for different audiences, organisations, and health behaviours (Figure 2).Specific behaviour change techniques were taken from the behaviour change wheel (Michie et al., 2011) to address the barriers and drivers of health behaviours that were evidenced in the co-design sessions.Table 2 details the behaviour change techniques employed, the targeted constructs (the mechanisms of change), the associated campaign strategy, and a rationale for the use of the BCT.
It was determined that the delivery of the campaign would include an 'always on' approach throughout the year to reach stakeholders and audiences through an owned social media presence, supported by three major campaign bursts targeting different audiences and behaviours (Figure 3).
Using the insight gained from the co-design sessions, the Medway Can website was created, enabling stakeholders and community members to engage with the Medway Can agenda.To encourage audiences to engage  with the campaign and increase motivation, the website included software whereby users could log the amount of physical activity they had completed, which was converted into miles.Audiences were encouraged to help Medway 'travel around the world' (24,902 miles) to give the community a shared goal.As audiences logged their activity, the Medway Can brand logo moved along a barometer to indicate distance travelled and countries reached while travelling to the overall target.To engage the diverse population of Medway, physical activity was reframed so that non-tradition forms of activity are included.To follow a whole systems approach, existing infrastructure was capitalised on through co-branding and aligning with the local authority events calendar.The website includes stakeholder toolkits and support, opportunities to host information about community group incentives, and information about ways to get involved (Figure 4).There was also a mechanism for community groups to bid for funding to deliver the campaign locally and to their community, to help reach diverse populations.
The campaign continues to be codesigned with community members and stakeholders.A full independent evaluation will take place following the completion of the first burst of the campaign.The campaign has had wide engagement across all audiences and continues to grow.As of January 2023, the residents of Medway had logged sufficient levels of activity to carry them 24,901 miles around the world (Figure 5).

Discussion
This article describes the co-design process used within an agency in the succesful development of a social marketing behaviour change campaign in England, UK.Campaign development within agencies is often carried out within tight timescales with fast turn-around expected.This article highlights how existing knowledge can be rapidly translated into action through the use of Implementation Science frameworks such as COM-B without duplicating effort but by capitalising on beneficiary input.This is in keeping with previous research where intervention development has been successfully led by the application of COM-B (Coupe et al., 2022;Krusche et al., 2022).In addition to demonstrating the usefulness of theoretical frameworks in planning strategies for campaign development and launch, the article has also demonstrated the need for co-design in the application of such frameworks.Too often pragmatic or commonsense approach is applied to public health campaigning without due attention being given to the context in which the target audience enacts health behaviours (Kelly & Barker, 2016).Co-design processes have the potential to lead to impactful campaigns that can be transferred to different settings, populations, and behaviours.This is supported by previous research which has highlighted the benefits of co-design in intervention development in health-related behaviours (Bevan Jones et al., 2020;Brown et al., 2022;McGill et al., 2022).Co-design is an ongoing process and, as the campaign develops through stages, continued engagement with the local community is paramount to ensuring it success (Hawkins et al., 2017).
The complexity of taking a whole system approach has also been evidenced here; while integrated behaviour change models like COM-B lend themselves well to whole systems approaches, it is ultimately engagement with diverse stakeholder groups that leads to successful whole system approaches.To engage the different levels of a system, each component must understand the role it plays in influencing behaviour (Lee et al., 2017).This was demonstrated through co-design with stakeholders where learning was reciprocal; the session was an opportunity for stakeholders to realise their individual roles relating to the same agenda.For example, leaders of religious groups began to understand their role in conveying the health message to their followers who are not reached through traditional media, school head teachers understood their role as promoters of health literacy in their young cohorts, and local restaurant owners discussed their responsibility to promote healthy choices on their menus.By engaging these diverse stakeholders in one conversation, they were able to position themselves effectively within the system.
The strength of the article is in its ability to evidence the application of COM-B in different levels of the system.It's flexibility and ease of use lent itself well to translation for the different target audiences and different levels of the system which were engaged.There were, however, some challenges identified in the use of COM-B.Translation of behaviour change techniques into practical application within the campaign depended heavily on the individual's conceptualisation of the technique.Equally, categorising the determinants of weight-related behaviours into the COM-B constructs was occasionally challenging and open to individual interpretation.However, it should be noted that COM-B is designed for practitioners to enable the adoption of behavioural science into practice.Thus, it is likely open to some interpretation by intervention developers.
Finally, the article has demonstrated the practical ways in which the input of beneficiaries can support the development of a public health campaign; Medway Can was the name chosen by local residents, the Medway Can logo was designed with local children, and the look and feel and activities are developed by community members and stakeholders alike.Through co-production, cross-branding, and collaboration, Medway Can is supporting local people, employers, health and care professionals, food retailers, sports groups and education providers to make a positive change.

Figure 2 .
Figure 2. The Medway Can brand, utilising the colour palette and shapes symbolising movement according to co-creation sessions.

Figure 5 .
Figure 5.A Celebratory post on the Medway Can website congratulates residents on their 'around the world' achievement.

Table 1 .
Format of codesign sessions.