3.1 Study Design

We used a qualitative approach to co-design the framework with young people and health communicators specialising in social media. The process included three workshops and subsequent analysis to develop, refine and finalise the framework iteratively.

We engaged in participatory action research methods [30], beyond the co-design workshops and involved young people as co-researchers to strengthen the research process and improve the effectiveness and relevance of outputs to young people as outlined in the WH&Y Youth Engagement in Health Research Framework [31].

Ethics approval was obtained from the University of Sydney Human Research Ethics Committee (2022/887).

3.2 Study Team

Three youth co-researchers (D. V., T. H., I. M.) were engaged throughout the study as part of the research team alongside health policy and health communication researchers with diverse experiences and seniority. Co-researchers were aged 21–24 years, identified as man or woman and had diverse cultural backgrounds (Hong Kong, Kazakhstan and UK). Two were trained research assistants working with senior members of research team on other projects and one was a consumer trained in the required research methods for this study. One co-researcher also had experience as a social media content creator (67,000 followers across Instagram, TikTok and YouTube). Co-researchers were organised by PhD candidate (M.T.; woman, aged 28, Iranian background) who served as a liaison to reduce power imbalance and potential social desirability bias between young co-researchers and senior researchers.

The youth co-researchers were involved throughout the research process from recruitment to data collection and analysis. Their contributions included developing and piloting workshop activities, designing recruitment flyers to ensure engagement from young people, co-facilitation of the workshops, collection of field notes and synthetises of ideas raised at the workshops. They also led the set-up of the in-person workshop venues to ensure they were comfortable for young participants, selecting catering, background music, lighting and setting up a charging station to ensure participants' phones were charged.

3.3 Participants

There were two stakeholder groups in the co-design process: professional health communicators and young people. Whilst the framework was intended primarily for professional stakeholder use, it was ultimately designed with the needs and preferences of young people at its core to achieve the greatest benefit to them.

The professional stakeholders were health communicators with experience working in the social media team of Australian government state/jurisdiction health departments. They were purposively sampled from a pool of previously identified stakeholders for a related interview study (to be reported separately) and were recruited via a study invitation email. All invited stakeholders were involved in the official government COVID-19 social media emergency response in a range of roles (i.e., manager or content creator).

The young people were aged 18–24 years and were recruited via flyers advertised on social media. Purposive sampling (maximum variation sampling) was used to recruit diverse participants with a range of ages, genders and social media usage to enhance the diversity of perspectives obtained and ensure broad relevance of the framework.

3.4 Co-Design Workshops

Three co-design workshops were conducted between October 2023 and March 2024 (Figure 1). Workshop 1 was conducted with young people and focussed on generating initial ideas for the framework. Workshop 2 was conducted with professional stakeholders and focussed on the feasibility and refinement of young people's ideas. Workshop 3 was conducted with young people and focussed on finalising the recommendations in the framework with illustrative examples.

The workshops with young people were conducted in-person over 2 h, and the workshop with the professional stakeholders was conducted online over 1 h using Zoom teleconferencing. Workshops were facilitated by M. T. and youth co-researchers, D. V., T. H. and I. M. All participants were offered a $100 gift card to acknowledge their contributions to the study.

3.5 Workshop Activities and Data Collection

The workshops included a range of data-generating activities like finding and assessing examples of official social media health messaging and live graphic illustrations capturing key ideas, as well as World Café style [32] and traditional focus group discussions. Activities and resultant artefacts (tangible and shareable outputs of the research and design process [33]) are described in Table 1. Physical and digital artefacts generated in the workshops included annotated tables and affinity diagrams (visual brainstorms that allowed participants to organise ideas according to their natural relationships [34]). Example artefacts in Supporting Information S1.

Field notes were taken by all facilitators and the workshops were audio-recorded and transcribed.

3.6 Data Analysis

Data analysis was informed by the participatory action research approach of collaborative analysis [30]. Initial data analysis occurred during workshop activities since generating artefacts required idea synthesis [30]. For example, participants and youth facilitators organised ideas into categories to generate affinity diagrams and the professional illustrator summarised ideas during graphic recording, amending the graphic immediately based on participant feedback during the workshop.

Upon completion of the first workshop, M. T., D. V., T. H. and I. M. digitised and further collaboratively analysed the data by reviewing the contents of the artefacts, identifying and categorising concepts into an initial concept map using the web application Miro. Field notes and workshop transcripts (transcribed by external company) were also reviewed to ensure key concepts were not missed in analysis. The broader study team including J. A., K. M. and C. B. updated the concept map with subsequent iterations following each workshop based on participant feedback, further refining in weekly meetings until the final recommendations were developed. Youth co-researchers and three professional stakeholders who did not participate in the workshops but were interviewed for a related study reviewed and approved the final recommendations.

4.1 Recommendation 1: Involve Young People

4.2 Recommendation 2: Pitch at Right Level

Stakeholders considered this recommendation could be feasible during emergencies but that sense-checking and research opportunities are lost when messages need to be disseminated immediately. As above, they reported that if young people are already embedded in the system prior to the emergency, more tailored and age-appropriate messaging could always be enabled.

4.3 Recommendation 3: Capture Attention Fast

The stakeholders considered this recommendation as feasible during emergencies, yet less relevant due to heightened anxiety and interest around health issues.

4.4 Recommendation 4: Use Current Social Media Marketing

They further recommended that a registry of vetted influencers could be developed for proactive campaigns and budgets to be carved out specially for these collaborations. Young people in the subsequent workshop also suggested some ‘safe’ trends for health agencies including timeless references (e.g., classic film references or content featuring cute animals).

Preparation would be required such as maintaining relationships with previous collaborators or developing a registry of vetted influencers. Again, the need to engage in social media marketing was considered less relevant during an emergency due to heightened anxiety and interest around health issues.

4.5 Recommendation 5: Engage More With the Public

Professional stakeholders reported this recommendation was less feasible during emergencies due to time constraints and shift in focus to on one-way message dissemination. However, they also noted the increasing importance of community engagement during these contexts since the public are generally more anxious and have more questions for health agencies.

5.1 Principal Results

This study co-designed five key recommendations for public health agencies to use when communicating health messages to young people on social media; (1) involve young people, (2) pitch at the right level, (3) capture attention fast, (4) use current social media marketing and (5) engage more with the public. In general, the recommendations were considered feasible during emergencies if embedded in business-as-usual processes prior to the emergency. Some recommendations (capture attention fast and use social media marketing) were viewed as less essential during emergencies due to heightened public anxiety and increased interest in health messages. However, the recommendation of engaging with the public would become more important during times of uncertainty.

Like other co-design studies working with young people and professional stakeholder groups, the framework had to strike the right balance between the preferences of young people and the constraints of the health agencies [36, 37]. There were several ideas young people raised that were simply not possible for the professional stakeholders (e.g., sharing click-bait or ‘shutting down misinformation trolls’ in the comments). However, many of these ideas could be revised to reach a middle ground idea that was acceptable to both stakeholders (e.g., alternative hooks or constructive comment responses to misinformation).

5.2 Comparison With Prior Work

This study supports principles recommended in previous research on social media health communication to youth such as collaborating with young people as messengers, using age-appropriate language, and considering young people's interests in health messaging, outlined by Sobowale and colleagues [21]. This framework also shares elements with social media health communication frameworks designed for general populations. For example, de Vere Hunt and Linos [19] outlined the importance of involving members of target population in message development, tailoring of social media messaging to specific populations and leveraging shareability with social marketing strategies. Gatewood and colleagues [20] further recommended understanding and using appropriate platforms and consistently creating content that engaged audience using high visual images. A scoping review identified similar guidelines for public health agencies' use of social media during emergency contexts, including fostering public interaction and engagement, leveraging platform features, considering needs of message recipient and recognising members of the public as partners in co-creating and disseminating messaging [1]. Other principles are also echoed in official emergency communication guidance with WHO, CDC, ECDC and Canadian Public Health Association recommending involving consumers, simple and empathetic messaging, including calls to action and building trust with the public in their health emergency response frameworks [22-24, 38].

The proposed framework is unique in that it applies such principles to the specific needs of young people in the current social media context (i.e., considering newer platforms young people now use like TikTok). The framework also provides practical examples to guide health agencies in applying the recommendations in emergency and non-emergency phases. Social media is crowded with competing voices and information sources, which means health agencies cannot just assume any of their messages will automatically capture and hold the attention of its users, especially considering competitors are often very skilled at creating engaging content [39]. This is especially pertinent in the context of young audiences, who have unique social media cultures and norms [40]. As seen in our findings, young people have different needs when it comes to social media health communication. This includes both health literacy-related considerations, as young people need simple messaging with limited jargon, but also other age-related considerations including youth cultures and youth social media cultures [8, 21, 40]. These issues can be addressed by involving young people in the communications to ensure the messaging is appropriate for young people every time.

5.3 Practical Implications

These findings provide a framework with practical examples to guide public health agencies in using social media to reach young people. This framework can also be considered in circumstances when government agencies outsource communications to commercial firms. While participants raised some feasibility issues during emergencies, the framework is flexible, and the recommendations can be used in isolation or combination as appropriate to the context.

As seen in our findings and previous research, it is imperative to leverage two-way communication on social media to build trust and engagement with health messages [5]. Audiences are not passive information receivers on social media, but active participants who can interact with campaign messages, amplify reach and influence the wider online communities' perspectives contributing to message acceptance, uptake and trust [41-43]. Platforms also easily allow for monitoring of audience sentiment via ‘social listening’, as well as immediate responses to audience questions [44]. This is incredibly important during major health crises as fostering public engagement and trust with health agencies becomes more important whilst also providing them a platform for combatting misinformation effectively [1, 25, 45]. For example, research has found that rebuttal of COVID-19 vaccine misinformation by public health agencies on social media reduced the public's belief in the misinformation and increased their willingness to get a COVID-19 vaccine [43]. This is an appealing feature to young people, who are less interested in one-way didactic messaging, and prefer (and even expect) the two-way communication as seen in our findings. Young people may also observe these public interactions ‘silently’ without publicly engaging, but these interactions can still influence their attitudes towards health messages [46].

Using social marketing was another unique affordance to be leveraged for health messaging on social media, as highlighted in our study and previous research [21, 47-50]. For example, partnering with influencers, as recommended in our framework, has been successful in influencing social norms and health behaviours related to COVID-19 and influenza vaccination [47, 50]. Organisations like WHO have already implemented this, creating a registry of vetted influencers via the “Fides” programme, which health agencies could use as a model to develop their own local equivalents [51]. As seen in our findings, social media collaborations must be authentic and allow the influencer's own interpretation of the campaign to better speak to the audience for optimum effect [50]. However, as explained by the stakeholders in our study, such collaborations may be challenging to execute during emergency situations and may be risky for health agencies. Public health-related posts may contrast with usual influencer content and undermine content coherence, a key factor for social media influencer effectiveness [52]. This issue warrants further research on how trusted partnerships between health agencies and influencers could be built over time, in preparation for future emergencies. Different approaches may be needed on different social media platforms, such as TikTok versus Twitch [53, 54].

5.4 Strengths and Limitations

The key strength of the framework lies in its collaborative development. Bringing together the perspectives of young people and health communicators through co-design improves its acceptability and feasibility for all stakeholders, at multiple levels. The participatory approach involving youth co-researchers also strengthened the study. Involvement of co-researchers improved participant experience, and thus quality of the data generated, as well as introducing new ideas to study team about youth-centred research practice (i.e., importance of workshop ambiance and phone charging stations for young people).

A limitation due to convenience sampling meant there were only a small number of professional stakeholder participants, and they were all affiliated with the same Australian health department. Stakeholders from other organisations may operate under different processes which may have impacted final recommendations. However, all study stakeholders had diverse roles within their organisation and had experience working for other government organisations. The final framework was also checked by a larger and more diverse group of professional stakeholders from other health departments who were involved in a related study [18] to ensure the recommendations are appropriate for a range of organisations. Another limitation is that most participants were university students, whose health messaging needs and preferences may differ from non-students. Finally, the recommendations reflect the social media preferences that young people expressed during the workshops but may not translate to their actual engagement on social media. Future research could consider methods involving citizen science (i.e., data donation) and tracking a diverse group of young people to explore their actual engagement with social media content from public health agencies [55].

5.5 Future Directions

We recommend this co-designed framework is implemented for future health communication to young people in advance of emergency conditions as part of preparedness efforts. Whilst the recommendations were specifically developed with professional health communicators at Australian government health departments, the findings will be relevant to any organisation communicating health messages to young people via social media. Successful implementation of the framework ultimately requires investment, resourcing and buy-in from senior management [21]. Furthermore, social media campaigns should be evaluated to understand the real-world outcomes [19, 20]. Though social media is constantly evolving, the underpinning principles of this framework of youth engagement and leveraging new technologies for health communication will continue to be important as the landscape develops. Whilst this work aims to reduce the health literacy demands of the online information environment for young people, it must be paired with health literacy education and appropriate referral to health professionals to prevent issues like misinterpretation and self-diagnosis.