One size doesn’t fit all: methodological reflections in conducting community-based behavioural science research to tailor COVID-19 vaccination initiatives for public health priority populations

With over 769 million confirmed cases and 6.9 million deaths, the coronavirus disease 2019 (COVID-19) continues to pose significant public health challenges across the globe [1]. Health systems, economies, and social structures have been widely impacted by COVID-19, leading to changes in public health policy in many countries [2‐4]. Epidemiological data shows that the infection can affect people of all ages, although older adults and those with underlying medical conditions are at increased risk of severe illness and death [5]. Certain racial and ethnic groups, as well as those living in densely populated areas, are also at increased risk [6‐9]. For example, individuals of Black and Asian ethnicity are at increased risk of COVID-19 infection compared to white individuals [10]. Several factors contribute to higher rates of infection in ethnic minority groups, including racism and structural discrimination, lower socioeconomic status, living in larger and multi-generational households with shared facilities, and being employed in essential jobs with frequent exposure to infection and proximity to others [10].

Addressing these inequities requires multifaceted interventions, from mitigating structural disparities to promoting the uptake of public health and social measures aimed at preventing the transmission and spread of COVID-19 [10, 11]. While a wide range of measures have been recommended by the World Health Organization [12], the challenge lies in their effective implementation, especially concerning vaccination. Globally, promoting COVID-19 vaccination uptake (and boosters) during the ongoing (and constantly evolving) pandemic remains a challenge [13‐15]. This is complicated by the politicization of COVID-19, the pervasive spread of misinformation within social media platforms, the mistrust of scientific authorities and institutions, cultural norms related to family, peers, religious leaders, and community leaders, and vaccine confidence, which have influenced behaviours towards these health measures and vaccination efforts, hindering their effectiveness [16‐19]. Health inequities for structurally marginalized Canadians were exacerbated by COVID-19, underscoring the need for community-specific and responsive solutions [20‐22]. These solutions should be based on an understanding of the unique barriers and facilitators to information dissemination and uptake within different communities to ensure their relevance and acceptability [20, 23, 24]. Factors such as language barriers, socioeconomic status, access to technology, and historical experiences with healthcare systems can significantly impact the relevance and acceptability of public health measures [20, 23, 24]. Therefore, addressing the health inequities exacerbated by the pandemic requires a comprehensive approach that prioritizes responsive solutions, ensuring that these are tailored to the assets, needs and challenges of each community.

In Ontario, Canada, Public Health Units (PHUs) have been at the forefront of the pandemic, battling not only the disease but also facing operational, ethical and communication challenges [25]. Ontario PHUs are the agencies responsible for delivery of local public health programs and services. One of the main challenges has been the scale of the pandemic, highlighting critical weaknesses in health systems. These weaknesses include limitations in the capacity and flexibility of service delivery with PHUs struggling to manage the large number of cases; inconsistencies in decision-making and coordination at different levels of the health system and government; and the significant strain on the health workforce revealing shortages and the need for additional training in epidemic management [26]. PHUs navigated complex ethical and political issues related to the pandemic response, such as balancing individual rights and public health priorities, as well as tensions between government and PHUs [27, 28]. In Canada, while the federal government provides overall guidance, PHUs, operating at the municipal or regional level, are primarily funded and governed by provincial or territorial governments [29]. This can lead to discrepancies in resource allocation, priorities, and decision-making processes, impacting public health initiatives. PHUs have also faced challenges in communicating with the public, especially dispelling misinformation and countering increasing mistrust [30]. There is a need for increased collaboration and coordination among public health authorities at the global, national and sub-national levels to share best practices, resources and expertise; PHUs need access to timely and reliable data to inform their decision-making and response efforts.

Applied behavioural science has the potential to play an important role in supporting PHUs’ programs and strategies during public health emergencies such as the COVID-19 pandemic [31, 32]. Behavioural science provides understanding of how key healthcare actors—including individuals, communities, healthcare professionals, and policy-makers—interact within health-related areas. It enables better knowledge about the attitudes and beliefs, and decision-making processes of these actors, and understanding of the underlying processes and mechanisms that drive specific health-related behaviours [33, 34]. Behavioural science also provides evidence about effective (and just as importantly – ineffective) ways to support behaviour change [35, 36]. Evidence-based solutions grounded in social and behavioural science have been shown to be more effective and sustainable in the long run [32, 37]. This is key for PHUs to design and implement more effective interventions to prevent the spread of COVID-19 [33, 34]. An area where behavioural science has provided important evidence is the factors that influence vaccination uptake [38]. By drawing on this evidence, PHUs could develop more effective campaigns and interventions that resonate with the public and encourage vaccination uptake [31]. PHUs have long-standing partnerships with specific communities, which could be leveraged to introduce behavioural science to develop effective solutions.

Community engagement has been integral to the pandemic response [39]. A recent review revealed a disconnect between communities and government in disseminating public health information during the pandemic [40]. To bridge this gap, engaging with local communities and community organizations to address specific health information needs and provide tailored strategies to various groups and communities can help in re-establishing public trust, enhancing inclusivity and increasing uptake of health information from credible sources [40]. More specifically, community engagement can support counter-messaging to address circulating misinformation, designing impactful messaging appeals, and identifying trusted health messengers within the community [22, 41‐43]. Achieving a balance between public health priorities, which require urgent action, and building trust with communities, which requires time and resources, is challenging but necessary for effective communication and increased uptake of health information from credible sources [44]. There is a need to conduct community-based, behavioural science research with marginalized populations to identify community-specific and responsive solutions.

There is a knowledge gap in understanding how effective collaborations and relationships can be established between PHUs and communities to conduct behavioural science research. This paper presents the methodology we developed for community engagement with ethnically and socioeconomically diverse populations in Ontario, Canada to conduct behavioural science research to guide PHUs in their strategies to promote COVID-19 vaccination uptake. We share lessons learnt in applying the methodology, covering key considerations for fostering effective research partnerships with communities.

Our study yielded insightful findings on the complexities of aligning public health needs, community engagement and behavioural science research. We explored the nuances of working with diverse communities, and the critical role of trust-building in fostering effective partnerships. We outline here the lessons learnt in applying the methodology.

We developed and applied a dynamic community engagement approach involving close collaboration with PHUs, community leaders, and residents. This approach facilitated community-based behavioural science research with equity-denied groups to inform public health strategies to enhance vaccination uptake and curb the spread of COVID-19. After PHUs identified priority communities and vaccination targets, we collaborated closely with community members. This collaboration helped in the recruitment of their peers for theory-informed interviews and facilitated in-depth understanding of, and community-driven perspectives on, the factors influencing COVID-19 vaccination uptake, as well as interpretation of findings and co-design of timely recommendations tailored to each PHU in Ottawa, Peel and Toronto.

Our study shows why and how a ‘one size fits all’ approach to community engagement does not work when engaging with individuals from a wide range of sociocultural backgrounds, particularly around a highly politicized and polarized subject. The principle of ‘one size does not fit all’ starts at the study design and grant application stage; we must be nimble and prepared to adapt based on the community and their needs as they evolve throughout the project. While we aimed to adopt a rapid, agile approach to the current project given the pandemic context (including the use of a novel, rapid analysis approach for interviews with residents), it still took several months to recruit enough participants to achieve a desirable sample size in target populations, conduct the co-design process and produce recommendations to PHUs. It was imperative to tailor our approach to community engagement, carefully considering factors such as the specific partnerships researchers establish within each community, the cultural and linguistic nuances of the community (e.g., language skills and preferences, communication styles), the historical and social context surrounding the issue at hand, the accessibility of resources for community members, and the support systems and organizations within the community. Overall, this approach requires researchers to adopt a more flexible, empathetic, and community-centric perspective, moving away from a one-size-fits-all methodology and pre-established protocols. Although the urgency of a public health crisis can result in projects being conducted more quickly, it remains possible to build trusting relationships if meetings and feedback occur regularly and consistently while respecting the overall project timeline [41, 42, 55, 57]. Community members were understanding of the urgency surrounding the situation and agreed to shorter timelines when approached transparently. Trust can be established and maintained even in the face of time constraints by maintaining open lines of communication, and consistently involving community members in the decision-making process.

Approaches such as iKT and CBPR have enormous value in informing public health authorities, especially when working with historically excluded groups [42, 43, 45]. Building on the expertise of community members, gained through lived experience, is essential for designing public health strategies and policies that reflect the realities of those communities. Effective partnerships are founded on mutual respect for each other's expertise, with researchers and community members working collaboratively [41, 45]. Before embarking on any research involving equity-denied groups, researchers need to be aware of the barriers to participation based on past history, potential harms, pre-conceived notions about research and uneasiness with the roles of researchers within society (past and present) [64]. We have to change the widely held (and often justified) belief that the researcher will parachute in, take what they need and walk away from the community without concrete measures to give back to the community [64]. Research involving community engagement has immense value for addressing issues that are relevant to communities and for proposing solutions informed by the expertise of the end-users in these communities. “Nothing about us, without us” is at the heart of community engagement. The voices of diverse members of the community are especially important in non-disease specific topics like public health, where people are coming to the table to share lived experience as members of a community, rather than with a specific disease or condition. This project demonstrated that it is not simply a matter of translating the ‘tried and tested’ patient engagement strategies to public health community engagement. In the public health sphere, there remains a need for agile and tailored community engagement strategies, and the current study fills a key knowledge gap in linking behavioural science approaches to iKT and CBPR [65‐67].

Presenting complex findings from behavioural science was pivotal in engaging CAGs in effective discussions in the OPTimise Platform. The use of ‘personas’ crafted as fictional individuals, each embodying the diverse themes and perspectives unearthed from the qualitative interviews, was deemed effective by CAG members. They not only recognized these personas as reflective of their community's diverse viewpoints but also found them relatable, akin to familiar neighbours and peers. This familiarity was crucial—it transformed our findings from abstract concepts into tangible, relatable narratives, fostering a deeper understanding among the CAGs. The personas served as a bridge, linking the theoretical aspects of our research with the practical, lived experiences of the community members. This, in turn, sparked rich discussions, enabling the CAGs to provide insightful observations and suggestions on potential strategies that could effectively resonate with the personas represented. Translating data into narratives that CAGs can connect with, we not only enhance their ability to engage with the findings but also empower them to contribute meaningfully to the discussion.

This study has some limitations that should be considered when interpreting the findings and considering the applicability of our approach to other contexts. First, our methodology was relatively resource-intensive, requiring significant time (around a year from the prioritization of key behaviours and populations by PHUs to the production of tailored recommendations) and effort from the research team and community partners. For many PHUs, this level of resource commitment may not be feasible, especially those with limited funding or staffing. However, we hope this approach also demonstrates the complementary–and at times, bridging–role that research teams can have in partnering with PHUs and communities, whereby research teams can provide complementary resources and expertise to PHUs. Furthermore, leveraging the methods and lessons from this process might improve efficiency in future applications. For future studies in low-resource settings, the use of existing and adaptable interview guides from this project could expedite data collection, as could partnering with research teams. Second, our recruitment strategy, which relied heavily on the existing networks of our community partners, may have introduced biases in the sample, study activities and outputs. While we made efforts to include diverse community members, there is a possibility that we missed individuals who are not connected to these networks or who face specific barriers related to trust, access to healthcare, and information. This could limit the generalizability of our findings to the broader community. Lastly, while we employed a variety of data sources to understand community-level perspectives including CAGs, we relied primarily on individual interviews and this can result in gaps in our understanding of the broader community context and concerns. Overall, while our approach provided valuable insights into community engagement and attitudes towards COVID-19 vaccination, readers should consider these limitations when adapting similar approaches to their own contexts.

When community engagement results in a positive, rewarding experience, community members are more willing and capable of advocating for inclusion in research that concerns them, as well as communicating to researchers which factors contribute to effective participation in the research process. At various points throughout the study, community members shared positive feedback on their engagement experiences. Many community partners from all three geographical regions elected to be notified of future opportunities to collaborate with research teams. The findings from a formal third-party evaluation of the project’s citizen and knowledge user engagement will be published.

As citizen engagement in health research becomes more common, best practices need to reflect the diversity of communities and how the engagement approach must not be conceived as ‘one size fits all.’ The COVID-19 pandemic has put a spotlight on public health and the role of PHUs as key actors in keeping our communities safe. Moreover, increased public interest in science and public health since the outbreak of the COVID-19 pandemic, including both positive and negative perceptions, should be responded to by researchers with active engagement efforts. As researchers, we have an opportunity to highlight the value of community engagement and bring community members together to tackle complex public health issues such as climate change, environmental destruction and food insecurity. Combining community engagement and behavioural science approaches can result in public health policies and recommendations that are truly relevant and meaningful to diverse communities.