The role of community leaders and other information intermediaries during the COVID-19 pandemic: insights from the multicultural sector in Australia

Community and religious leaders and other natural leaders from culturally and linguistically diverse (CaLD) backgrounds have been postulated as a gateway into communities. They act as information intermediaries that enable public health messages to reach individuals. However, there are currently limitations regarding our understanding of these information intermediaries’ regarding their capacity, role, and reach. In-depth interviews were undertaken to understand the perceptions of those working in Australia, including multicultural health, communication and other social support roles focused on CaLD communities, towards the role and impact of information intermediaries in promoting and supporting COVID-19 public health communication and engagement activities. Forty-six semi-structured telephone interviews were undertaken with key stakeholders who have an active role in delivering services and other social support to CaLD communities. Four key themes emerged related to the role of information intermediaries during the interviews. Ideas focused on their role in “bridging the gap” and supporting pandemic-related information delivery into communities. Participants felt that there had been a failure by Federal government agencies to recognise the role of these stakeholders early in the pandemic and a failure to provide sufficient resources and support. However, concerns were also raised that public health messages may be inappropriately interpreted or translated by the community information intermediaries or potentially blocked if the message does not align with the broker’s own beliefs. Finally, concerns were raised about the potential for burn-out among information intermediaries. In preparing and responding to pandemics and other disasters, community leaders and other information intermediaries recognise they have an important role to play and must be provided with resources to enhance and sustain their involvement.

Background P opulations at risk of COVID-19 infection are diverse and differ in COVID-19 literacy and social, behavioural, cultural and health practices (Greenaway et al., 2020). Data from several countries has highlighted that compared to the white population majority, there has been a much greater risk of infection and adverse outcomes from COVID-19 among Black, Asian, and Minority Ethnic [BAME] groups, Black Americans, Hispanics, Latinos, other people of colour, and Indigenous groups (Cheshmehzangi, 2022). While the exact reasons for this remain unclear, they are likely due to a complex interplay of factors rather than a single cause (Katikireddi et al., 2021;Saatci et al., 2021). Within Australia's culturally and linguistically diverse (CaLD) communities, a higher proportion of recently arrived immigrants may be working in public-facing occupations (e.g., retail, transportation, or service roles) or within health, aged or community and childcare, jobs that would prevent physical distancing. Other factors associated with an increased risk of COVID-19 among CaLD populations include housing arrangements and extended family groups living together (large, intergenerational households), collectivist approaches to childcare and the cultural expectation of family members providing care for each other when sick (Greenaway et al., 2020).
Concerns have been raised that members of CaLD communities have not been meaningfully included in public health planning since the outset of the COVID-19 pandemic (Guttman and Lev, 2021). Public health efforts must be adapted to the linguistic, cultural, and social circumstances of marginalised groups to reduce the unequal burden during a pandemic, as well as ensure equitable and timely uptake of an appropriate pandemic vaccine (Feinberg et al., 2021;Habersaat et al., 2020;Kumar et al., 2021;Vuong et al., 2022). This includes ensuring effective and equitable two-way crisis communication (Guttman and Lev, 2021). Communication barriers can severely hamper access to and use of healthcare services by people from CaLD backgrounds because of limited knowledge about services and little understanding of recommendations (Henderson and Kendall, 2011). Formalised responses to this issue by statutory and non-statutory agencies often include providing translated health information, producing uncontextualized text-heavy resources and training health staff in cultural competency (Abd et al., 2020;Wild et al., 2021). The challenge is that these strategies can be limited in scope in that: (1) not all people from CaLD backgrounds have the critical literacy and health literacy skills to read and understand the materials (even if written in their language) (Jessup, Osborne, Beauchamp, Bourne, and Buchbinder, 2017); and (2) having a translated resource does not necessarily mean that the messages are reaching the target audience nor are they meaningful for CaLD communities (Ogie and Perez, 2020).
Previously, Mileti and Darlington found that people from CaLD backgrounds generally prioritise social networks and interpersonal communication when seeking information and prefer to receive information from people with similar attributes as themselves (Mileti and Darlington, 2014), which has been confirmed in a range of other settings (Tierney, 2019). This is especially true of cultural groups that are more collectivistic and cooperative and like to share their experiences and information with their community (Chen and Choi, 2011). In settings like Australia, people from CaLD backgrounds often form tight-knit communities with solid internal linkages, and these social networks (including relatives, friends, colleagues, neighbours) may act as sources of information regarding appropriate medical care (referred to as interpersonal communication) (Cohen et al., 2000). Traditionally, a large amount of interpersonal communication occurs within physically local community networks. However, this has shifted to include online networks via short messaging apps like WhatsApp and WeChat and online platforms that allow closed and semi-private groups, such as Facebook.
Within these social networks, individuals can be purported to play an enhanced role in transmitting information within their community. "Gatekeeper" is a term that has been previously used to describe people in the community who can facilitate other people's information and health seeking behaviours. Borrowing from attachment theory, the gatekeeper model posits that individuals may find comfort in sharing their feelings with acquaintances (Keith et al., 2010). The model is also influenced by the public health principle of mass saturation of awareness, whereby the likelihood of community members intervening in a crisis increases with the proportion of capable gatekeepers (Lipson et al., 2014). Originating from the US in the 1970s, the model focused on the training of nonprofessionals, including mail carriers, bartenders and hairdressers to identify older adults with health (mental health or suicide prevention) or social issues and to refer them to formal service providers (Keith et al., 2010). It is built on the notion that these informal social exchanges occur more frequently, and hence the gatekeeper can identify any issues (Florio and Raschko, 1998;Sarason et al., 1987). In this current study, we acknowledge that gatekeepers may be the CEO of community organisations, community or faith-based leaders, bilingual caseworkers, or "natural" leaders (example: a person who has completed medical training but does not practice in Australia). In this paper, we recognise gatekeepers as people who link others to information, so we feel that the more accurate term could be community networkers or information intermediaries.
Australia's COVID-19 Vaccination Programme Culturally and Linguistically Diverse Communities Implementation Plan (Australian Goverment, 2021), released in early 2021, sets out the fundamental principles to ensure that the rollout of the COVID-19 vaccination programme addresses the specific requirements of people from CaLD communities. Critical considerations in the plan include: (1) ensuring that easy to read information is distributed widely and in a variety of languages; (2) that channels are available for people from CaLD backgrounds to ask questions; (3) the need to ensure that the vaccination workforce has the capabilities to work with CaLD people, including access to professional interpreters and (4) provision of the free vaccine. The plan defines "community leaders and young people, and multicultural organisations as the gateway" to support information delivery. Given the acknowledged role of these community leaders in the government's strategy, this project aimed to understand the perceptions of those working in services and other social support roles focused on CaLD communities towards the role and impact of community, religious and other leaders in acting as information intermediaries to support COVID-19 public health communication and engagement activities.

Methods
Semi-structured, in-depth phone interviews with key stakeholders and informants of~30-40 min were undertaken between January 2021 and April 2021. The focus was on CaLD communities that include people born in English-speaking countries, and where English is not the primary language spoken at home. The Human Research Ethics Advisory Panel at the University of New South Wales reviewed and approved this study (HC200776). Informed verbal consent was collected from all participants.
Sampling. Key informants were defined as those who actively deliver services via migrant resource centres, refugee health services, settlement services, community-based organisations, translation services, and primary care settings. Stakeholders were those who play a role in multicultural health and diversity-related ARTICLE HUMANITIES AND SOCIAL SCIENCES COMMUNICATIONS | https://doi.org/10.1057/s41599-022-01196-3 activities through advocacy, policy/programme development, or research. This principally encompassed personnel such as those from government agencies or CaLD community peak bodies/ Councils and CEOs from community groups.
This study used a range of national-local-personal approaches to recruiting participants. Firstly, an online search of relevant websites was conducted to identify potential candidates matching the selection description. Each candidate was then followed up via email with an invitation letter. Secondly, interested candidates were asked to directly recommend any colleagues who may be willing to participate. Lastly, emails were sent directly to known relevant research team contacts working in the relevant sectors. An effort was made to recruit at least one participant from Australia's states and territories to capture a broad range of views on the country's issues. However, we were unable to recruit any participants from the Northern Territory. Participants were only included in the study when full verbal consent had been received. This study did not collect any identifiable personal information from the participants.
Data collection and analysis. An interview guide was developed and reviewed by the researcher's HS, BHS, AH, IK, and AM to identify critical areas of interest for the study. The questions related to the following topics: perspectives towards the current communication approach being used by the government, factors affecting communication and engagement with CaLD communities, the communication roles and influences of different agencies, and suggested options that could be adopted to enhance communication and engagement around the COVID-19 vaccine programme. Questions were asked in an open-ended manner to allow room for expansion (Richards, 2014). During the interviews, member checking was conducted to ensure that the ideas identified during the early analysis phase were appropriate. The data were analysed using thematic methods of building codes into themes using the constant comparison approach by HS and AC using NVivo software. Themes were compared within and across the sample and structured around the vital interview topics, allowing inductive and cross-cutting themes to emerge from the interview data. Any differences between the coders were resolved via discussions with the broader team to ensure vigorous analysis.

Results
Forty-six interviews were undertaken with key stakeholders and informants across Australia. The characteristics of the participants and the interviews are described in Supplementary Appendix 1 using the CORE-Q reporting format (Tong et al., 2007). The themes identified that apply to community leaders and information intermediaries are described in the text below.
An essential role in bridging the gap. Across all the interviews, the perceived role of community information intermediaries, including community leaders, religious leaders, and other "natural" leaders, was raised. The term "information intermediaries" included community leaders, faith-based leaders, natural leaders, bilingual workers/case or settlement workers, youth leaders, bilingual health professionals, local council members, and translators/interpreters. It was also suggested that community members who have health backgrounds (medicine, nursing, pharmacy etc.), but who are not registered to practice in Australia could be included, given their ability to interpret and paraphrase COVID-19 information. Participants proposed that different people play a role in providing information within communities and that a community member may "go to different people for different information".
The need to identify and work with local community information intermediaries was recognised as being instrumental in bridging the divide within the community and ensuring that information reaches all community members. It was felt that these people understand their communities and have good networks and links to the community.
One of the first outbreaks, I think, in our region was in the meatworks, and most of the employees at the meatworks are from Karen and Burmese heritage. Many people in those communities don't read or write in their language, so thanks to the local community leaders, we set up a testing site, and most people heard of it from word of mouth. (Interview 24) It was stressed that the role of these community leaders was not just about passing on COVID-19 information but in some settings, they also set up support networks, homework groups, and Zoom sessions focused on a range of community-nominated topics, including mental health, family violence, finances etc.
However, a few participants questioned the role of community information intermediaries and whether these individuals had the necessary understanding to have any impact. In many situations, community leadership is self-nominated, and these leaders may not have the required reach or relationships to support engagement. Some stressed that they did not know if the information was "filtering completely down".
One of the biggest challenges we have is communicating with those community leaders. Sometimes their educational levels are questionable. When the government says, "Well, we've approached the community members," do you genuinely know whether that person is fit for the purpose that you want? Are they going to deliver that information and are they delivering as you wished? (Interview 13) Participants also raised concerns that the community information intermediaries may interpret or translate information inappropriately. Some may even block information from reaching communities. For example, issues were raised that some of these community ambassadors may hold their concerns about COVID-19 vaccines and would not be willing to pass on information. To circumvent this issue, participants spoke about using multiple channels to disseminate information, including via the community leaders and public information sessions and other bicultural workers. This was particularly important for community groups presented by various dialects or languages.
I think the whole community leader thing, it's very convenient because you offload the responsibility to someone. Who are these community leaders? They follow their perceptions, their way of seeing and doing things. Yes, we need connections to the community, but also, we need to be very aware of how this is done. (Interview 42) There's no formal understanding of how well those people will then take notes and represent it." It's just an expectation that they are professionally qualified to deliver that. I'm almost certain that they are not professionally qualified to do that because they don't train to do that. (Interview 13) …some community leaders don't always act in their best interest because of power imbalances. They're worst, and thankfully this is not too common, but there are examples where it becomes a disservice because they try and block other channels of information for purposes of control and a whole bunch of complicated things. (Interview 44) Among those participants who were supportive of the role of community leaders was the sentiment that governments need to engage with them on an ongoing basis, including for emergencies: it was COVID, now it's vaccines, but it could just as easily have been bushfires, flood, or famine, may be not famine, but bushfires or floods, heat, extreme heat. Many emergencies continue to arise that we need to have good access to community members, through community leaders. (Interview 16) Failure to consult with community leaders. Participants were critical that the health departments did not recognise the importance of these information intermediaries until the later stages of the pandemic.
I asked the Health Minister how he engaged with the community, he said, "We're talking to community leaders of different cultural groups." I just thought, "That's interesting. That's a shift." We weren't hearing that earlier on in the pandemic. (Interview 12) Eventually, some health departments funded: "…community health centres to employ bicultural workers from their local communities for contact tracing, upright management, and general information". (Interview 16). In the early stages of the pandemic, participants mentioned they were concerned that minimal consultation was occurring. However, they did acknowledge that this improved during late 2020 and into 2021. However, participants continued to question whether governments actively involved community leaders and other information intermediaries or whether they were extracting information from them: "every time we need anything that has to do with CaLD, we go to the leaders, and then we get the information that we want, and then we forget about them. We don't even go back to tell them; this is the outcome of whatever we were doing is, and this is what we achieved." (Interview 6). One participant went as far as to suggest that the way governments interact with community leaders is not a new situation but a long-standing issue: …when we [the government] need you, we'll grab you and all the rest of the time, you don't matter, which is not right or fair (Interview 16) It was stressed that there must be structure in place so that governments can support and engage with community leaders, undertake more timely consultation, and check-in with communities before messages are distributed. One suggestion was to have a standing committee in place. However, concerns were raised that governments often find it easy to deal with one community member and assume "that the whole community thinks the same way", potentially disadvantaging some within the CaLD community.
Another issue raised was that community leaders are not necessarily allowed to contribute to the development of messages targeted at their communities but rather just given a script and put in front of a camera. In some settings, some community leaders may prefer this approach. Still, concerns were raised that this failure to involve the leader may contribute to audio and visual materials that do not necessarily resonate with the target communities.
Level of briefings and support. Concerns were raised across many of the interviews regarding the training and support provided to community leaders and other information intermediaries. One participant questioned the level of guidance available to support. In contrast, others raised issues around how they were being briefed ("randomly") and whether the briefings ("if they were occurring") were practical or successful in supporting the leaders to communicate correct information with their respective communities. It was suggested that it was "assumed" that the community leader was "up for it" (i.e., communicating about COVID-19) and that they may not have been prepared, which puts them in a compromised position.
The government tends to inform the professional public to further inform their constituents and customers… through media releases or government policy documents. They can consume that, understand it, and then regurgitate it. When I look at how the community leaders are being briefed, I can't see any guidance document to community members. (Interview 13) Participants raised additional concerns about the level of support given to community leaders (and other bilingual workers) in regional and remote areas. To fill this gap, one participant stated that they had created a community leaders forum, which included regular Zoom meetings with community leaders across a particular geographic area of Sydney to talk about the critical COVID-19 issues. External experts or government representatives were invited to attend the session as well. These sessions aimed to help ensure the consistency of the information being distributed. These sessions also allowed community leaders to raise issues about government requirements, advocate for changes/resources, etc.
Another participant highlighted that it was essential to distinguish between training and supporting people. In this situation, community leaders are not being actively sent out but instead invited to "have conversations" with their communities. In this setting, the participant spoke about holding forums that provide opportunities for open and honest discussions about the pandemic. They mentioned that at the end of the event, they emphasised to attendees that "if [they] feel comfortable, we'd love for [them] to have more conversations in the community". As these community leaders are volunteers, building their confidence to have conversations is essential.
We have to be careful because they (community leaders) shouldn't be responsible for answering complicated questions, but if they have a better understanding, they can at least point people in the right direction. (Interview 45) The danger of "burnout". A critical issue that was raised was the concern around community leaders burning out. Participants highlighted that they had heard from the community leaders that they felt "overworked and overstressed" and that a huge responsibility had been put on them to get information out, with pressure coming from "both sides". They are asked to translate repeatedly, understand, synthesise ("government-speak into community speak"), disseminate messages, and answer questions. This role can come with an emotional burden. Among those interviewed, issues about a lack of acknowledgement of the contributions from community leaders and lack of payment were voiced. Participants highlighted that these community leaders were volunteering their time on top of the other responsibilities (i.e., paid roles).

It's unpaid work. It's too much for them. (Interview 5)
There is a danger of the fatigue of these ambassadors… Then the lesson learned has to be capacity building for the next time it happens, and perhaps local governments having better connections with the community, knowing ARTICLE HUMANITIES AND SOCIAL SCIENCES COMMUNICATIONS | https://doi.org/10.1057/s41599-022-01196-3 who their different communities are, making sure that if we go into a lockdown situation (Interview 45) One idea was to pay community leaders for their time, acknowledge their contributions, and as an incentive to attend training. However, the capacity to provide this funding was questioned.
My recommendation is to engage them in a professional capacity to pay for their work, depending on the sessions. May be hourly rate, whatever. Just based on either per session or per hour or whatever it is to be able to deliver. (Interview 5) While funding may not be feasible, at the very least, governments need to do more to recognise publicly and show appreciation for the role that community leaders and other information intermediaries have had during this pandemic. Lastly, participants reflected on the lessons that need to be learnt from this experience, including capacity and relationship building, so that when it happens next time, governments have better connections with the community, know who their different communities are and have the relevant contact points.

Discussion
Information intermediaries can be differentiated into those occupying informal and formal (institutionally affiliated) roles, aligning to whether they are nominated for the part or have emerged from a social position. Research on information intermediaries suggests that they: (1) tend to have relatively higher education or language literacy skills; (2) are well known and frequently involved in their communities and (3) are multilingual and multiliterate (Bozkurt and Gürsakal, 2017;Lu, 2007). Research has also found that communities in almost all settings identify strategic actors for disseminating public health messages within groups (Agada, 1999;Buchanan et al., 2019). Currently, assumptions are made that information delivered by community leaders will circulate into their respective communities.
Previous studies on the role of information intermediaries in disseminating messages have suggested that their past experiences, knowledge, length of stay in a country and level of selfefficacy can impact the person's communication behaviours, especially around risk perception, information dissemination and filtering (Lu, 2007). By influencing the flow of information, information intermediaries can potentially shape and inform their community's reality and knowledge. For example, during the 2011 Brisbane (major city in Australia) floods, community leaders were found to have adapted information for CaLD communities by changing measurements from metric to imperial, making news "less scary", or filtering out information that was geographically irrelevant, or in some cases, not transmitting information at all (Shepherd and van Vuuren, 2014). If information intermediaries are selectively filtering information during a pandemic situation, communities may not be receiving timely or accurate information regarding public health strategies such as the vaccination programme. Reliability and trust are significant issues, as intermediaries can bridge gaps, but to be effective, it is critical that they themselves understand the information needs of the population they are serving and the implications of moderating or withholding information (Buchanan et al., 2019).
Do we truly understand community-driven communication mechanisms, whether the information being transmitted is accurate, or whether the process will support the adoption of recommended behaviours? These are critical concerns raised by our study participants about disseminating public health information about COVID-19 and promoting the COVID-19 vaccination programme. For communication to be effective, there need to be high levels of (1) trust; (2) information availability, accessibility, and readability, and (3) correct and positive subjective evaluation of the information. Focusing on the final point, information processing is influenced by cultural and strategic perspectives and by individual characteristics (Rhinesmith, 1992;Vuong and Napier, 2015). People can arrive at interpretations that capture realities, complexity, and ambiguity if they have a global mindset (i.e., the ability to recognise and adapt to cultural signals) (Barr et al., 1992). To aid interpretation, at an individual level, processes need to be put into place to enhance the knowledge and skills of the information intermediaries.
Participants acknowledged the need to move beyond relying on already-recognised community and religious leaders and identify other possible community contacts. One key reason for this was to reduce the issue of "burnout" among the community leaders, especially given the amount of work that some members put in during 2020. One possible suggestion was to enlist people from CaLD communities who have a medical background as vaccine champions. These include bilingual health staff and community members who may have trained as doctors, nurses, or other health areas but are not licensed to practice in Australia. The idea is that these individuals are still valued within communities as having a medical education and understanding can help disseminate information within communities in the community language.
The need to provide support and training to community information intermediaries has been well recognised in other public health and clinical sectors, including gatekeepers, as a strategy for identification and early intervention in the prevention of suicide (Isaac et al., 2009;Yonemoto et al., 2019). The training is aimed at specific groups of people to assist them in developing the knowledge, attitudes, and skills to identify people at high risk of suicide and refer them for treatment. Key components of the training focus on (1) preparing (setting the tone/expectations)' (2) connecting (reflecting on one's attitudes); (3) understanding (developing knowledge and skills to assist); (4) assisting (relevant strategies that can be used); and (5) networking (understanding local community resources and how to network) (Isaac et al., 2009). Similar programmes have been rolled out to support gatekeepers with mental healthcare in adolescent and adult populations (Lipson et al., 2014). In supporting information intermediaries to have an influential role in supporting public health action during a pandemic, training that adapts these five elements may be helpful to enhance the understanding, skills, and confidence of community information intermediaries to talk or address questions about testing or vaccination. It must support their skills and confidence to critically analyse online information and around building community skills to recognise misinformation. Training must also support skill development around interpreting information, the consequences of misinterpretation, and the tools and services available can help the leader. For example, a COVID-19 vaccination glossary was developed to provide plain-language meanings to complex immunisation and vaccine development words and terms (Seale, 2021). This glossary was developed to help community organisations, translators and interpreters, bilingual workers, and community leaders better understand and communicate about vaccine development and implementation. It was developed in collaboration with experts working in vaccine development, immunisation programmes and policy roles, and those whose work directly supports CaLD communities.
In early 2020, the WHO released a framework to support the response towards managing the COVID-19 infodemic (World Health Organisation, 2020). The framework included several pragmatic recommendations, including the need to identify sources (such as community leaders) credible to different audiences and share accurate public health messages through them. However, while the role of stakeholders was acknowledged as essential to enhance the management of the infodemic, there was no mention of how they could be supported in this role. Training has been provided for healthcare providers to help their capacity to respond. However, there seems to be a gap in the provision to other stakeholders, including community and faith-based leaders (Barua et al., 2020).
Among our participants, there was a tendency to talk about community and religious leaders as potential information intermediaries in the community; however, previously, it has been suggested that young people can also fulfil this bridging role (Marlowe and Bogen, 2015). Given their linguistic capital and digital literacies, young people have the potential to be cultural brokers and links between decision-makers and the community. Their role could extend to community settings and within their multigenerational household. The challenge is that, despite the resources that young people bring, experts often miss them (Ingamells and Westoby, 2008). This issue appeared to play out during the first year of the pandemic when the focus seemed to be on the use of mass media and traditional channels of communication. However, there was a shift and a greater recognition of the role of younger influencers during the second year of the COVID-19 pandemic by the Australian Government, with an increase in the use of social media and other platforms, including TikTok. Recognising and supporting young people to be gatekeepers in their communities is critical, especially in situations where parents and family members do not speak the country's language. In these situations, the young people from CaLD backgrounds may represent the host society's primary or only linguistic link. For instance, Mitchell et al. (2008) presented two case scenarios that highlighted young people's role in being the mediator between their respective communities and external forms of support (Tom et al., 2008). Firstly, the case of young El Salvadorians who supported the response to multiple natural hazards (earthquakes, hurricanes, and landslides) and Vietnamese youth reacting to Hurricane Katrina. Mitchell highlights that young people not only play an essential role in providing a bridge to their families and communities but also have the potential to be informants within informal and formal risk communication networks (Lori, 2008;Tom et al., 2008). Parker and Handmer have even gone as far as to suggest that unofficial or information communication networks can be significantly improved by tapping into young people (Parker and Handmer, 1998). However, this has yet to be adequately explored.
As part of the COVID-19 vaccine programme roll-out, some local government areas in Australia introduced vaccine ambassadors or champions to support vaccine confidence and increase uptake. These ambassadors are volunteers or paid employees who provide education about vaccine efficacy by sharing personal reasons for confidence in the vaccine based on real-life experience. In some settings, the role was aimed at those with lived experiences, including taking the vaccine, being homeless, or being part of a racial minority group, to promote two-way communication and opportunities for concerns to be raised and addressed. Given the nature of the role, job descriptions were put out that highlighted that the ambassador would attend training, provide outreach via personal interactions and social media, facilitate two-way communication, and attend vaccine events and mobile vaccine units to offer peer support and liaison between the public health officials and the community. In one job description, it was suggested that the salary range should be "commensurate with the importance of the position in the community and the skills required to be effective". While studies looking at the effectiveness of these programmes on communities have yet to emerge, there are distinct advantages of these formal programmes, which need to be acknowledged. Firstly, in some instances, these ambassadors receive training on how to build confidence around COVID-19 vaccines, and they are supported with the resources to facilitate discussions about the vaccines. Secondly (and perhaps most importantly), their time is appropriately acknowledged and paid.
There is ample evidence of CaLD community leaders' motivation and passion for contributing to the health and wellbeing of their communities. However, there are still critical gaps in our understanding regarding the reach, role, and influence these information intermediaries have on citizens' intentions or decision making of citizens within their networks or communities. There are also gaps in our understanding of how best to support and enhance the role of information intermediaries to improve community members' understanding, motivation, and acceptance of health and safety recommendations during a crisis and non-crisis periods. Future studies should focus on understanding the costs and benefits of enhancing the role of information intermediaries versus supporting more traditional communication pathways.
Limitations. The study team acknowledges that CaLD communities are distinct yet heterogeneous groups with unique health delivery needs (Komaric et al., 2012). Efforts were therefore made to ensure stakeholders were recruited across a range of different CaLD communities. However, it should be acknowledged that we could not include participants from all the other migrant groups in Australia. The following are noted as limitations for this work: (1) interviews were only undertaken with a select group of participants, so the possibility of other important themes emerging cannot be ruled out; (2) the use of snowball recruitment may have also reduced the range of opinions amassed from participants; and (3) specific details regarding the participants' role was also not collected.

Conclusion
In preparing and responding to pandemics and other disasters, it is critical that there is recognition of the role of community leaders and other gatekeepers and that resources, training, and opportunities for financial remuneration are identified to enhance and sustain their involvement. Future pandemic plans must include communication strategies that acknowledge the role of different channels and actors, as this may be a cost-efficient approach to enhancing communication. Researchers and policymakers must work with gatekeepers to improve access to vital information about the community and to develop an effective policy for the communities they serve.

Data availability
The datasets generated during and/or analysed during the current study are not publicly available due to the sensitive nature of the topic raised during the interviews, but are available from the corresponding author upon reasonable request.