1000 Hometown Heroes: Mobilising community social influencers for COVID-19 prevention, care and coping.

Coronavirus disease 2019 (COVID-19) created life-disrupting stressors that disproportionately burden disadvantaged communities with devastating impacts that extend far beyond the burden of the disease itself, including joblessness, housing and food insecurity, educational system upheavals, isolation due to disrupted relationships, worsened mental health and substance use and violence. Socially interconnected community members are resources whose efforts can be mobilised to improve COVID-19 coping within their social networks. This research examined the feasibility, acceptability, and reach of a social media-based peer influencer intervention for COVID-19 coping. Over a 9-month period in 2020, the project enrolled 1253 social influencers in Milwaukee-primarily ethnic and racial minorities-who regularly received and passed along messages to members of their social networks that provided advice about COVID-19 pandemic coping, economic survival, health protection, mental health, family needs, social justice and other impacts. Messages were shared by influencers with others over social media and also text messages, phone calls and conversations. Facebook social media tracking metrics objectively measured the community reach of social influencers' messages. Quantitative surveys and qualitative follow-up interviews with a subset of influencers also measured the feasibility and acceptability of the intervention. Social media monitoring metrics showed that, by the end of the project, influencers' messages reached an average of 7978 unique individuals per week and had an average of 13,894 total views per week, with more than 140,000 total cumulative organic impressions. More than half of social influencers indicated that-beyond Facebook message sharing-they shared COVID-19 prevention, care, and coping messages with social network members in conversations, phone calls and text messages. Social influencers reported that they valued having the opportunity to help community members to cope with pandemic stressors by conveying practical COVID-19 coping advice.

What is known about the topic:

The COVID‐19 pandemic has produced stressors that disproportionately burden disadvantaged communities.

Research on disasters has shown that social media can be harnessed to help community members prepare for and cope with severe threats.

Prior research in the HIV prevention field shows that socially‐connected opinion leaders can be enlisted to communicate health and support messages to others in their social networks.

What this paper adds:

It is possible to recruit a large cadre of predominantly minority social influencers who are highly interconnected with others in their community networks.

Messages that provide practical advice and support in COVID‐19 prevention, care, and coping were regularly provided to members of the social influencer cohort and were then passed along by influencers to others in their networks.

Social media tracking metrics, postintervention surveys and interviews with a subset of influencers established intervention reach, satisfaction and feasibility.

INTRODUCTION

Coronavirus disease 2019 (COVID‐19) has profoundly affected the well‐being of all Americans but has disproportionately burdened communities of colour (Abrams et al., 2022; Ettman et al., 2020; McKnight‐Eily et al., 2021; Morales et al., 2021). COVID‐19 hospitalizations and deaths are greater amongst African Americans and Hispanics than amongst non‐Hispanic whites (Mackey et al., 2021; Romano et al., 2021), including in Milwaukee, where this research was conducted (Milwaukee County COVID‐19 Epidemiology Intel Team, 2021). At the time of the study, African Americans in Wisconsin were 2.1 times more likely—and Hispanics 1.7 times more likely—to be hospitalised with COVID‐19 than non‐Hispanic whites (Wisconsin Department of Health Services, 2021). COVID‐19 disparities are attributable to employment in frontline service sector jobs where remote work is not possible, living in large households with extended family and where the isolation of ill persons is difficult, having to leave home frequently for work and daily shopping, high prevalence of underlying health conditions and limited access to primary healthcare (CDC, 2021).

The devastating impacts of COVID‐19 extend far beyond the burden of the disease itself. Early in the pandemic—and as large swaths of the economy shut down and as schools and businesses closed—joblessness soared, creating stressors that disrupted life across multiple domains including housing and food insecurity, educational system upheaval, disrupted social and family relationships, worsened mental health and substance use disorders, increased rates of violence and other stressors (Center on Budget and Policy Priorities, 2021; Javed et al., 2020; Nicola et al., 2020; Panchal et al., 2021; Park et al., 2021; Prati & Mancini, 2021).

Forty years ago, the world was on the brink of a different pandemic as the first cases of another previously‐unknown disease—AIDS—were being diagnosed. Like COVID‐19, AIDS carried a stigma, instilled fear, and took its greatest toll on marginalised communities (Centers for Disease Control and Prevention, 2021). Indigenous, grassroots neighbour‐to‐neighbour campaigns such as ‘Stop AIDS’ were mobilised in the early 1980s by members of gay communities hard‐hit by AIDS. These and similar efforts brought together networks of gay men who sometimes met in one another's living rooms to plan personal and community protective actions (Wolfeiler, 1997). Applied behavioural research later established the effectiveness of community interventions that mobilised ‘popular opinion leaders’ and influence leaders within social networks of men who have sex with men (MSM) to talk with peers about ways to protect against HIV. This body of research showed that opinion leaders' endorsements of safer sex practices to their MSM peers produced reductions in high‐risk behaviours amongst network and community members (Amirkhanian et al., 2005, 2015; Kelly et al., 1997, 2006, 2020). In the HIV arena, grassroots community programs such as Stop AIDS and research‐based interventions both demonstrated that socially influential community members can be engaged to serve as resources for HIV prevention and coping within their day‐to‐day networks.

Research has also demonstrated that social media can be harnessed to mobilise communications and support coping amongst community members during times of natural disaster. Messages posted and shared within online social networks help community members respond to threats posed by disaster, convey important information to mitigate disaster effects, assist in accessing needed survival resources, help in coping with stress and uncertainty, regain a sense of resilience, counter misinformation and rumours, and maintain community cohesion in the face of natural calamities (Duffy, 2012; Houston et al., 2015; Keim & Noji, 2011; Kongthon et al., 2014; Lindsey, 2011; Mavrodieva & Shaw, 2021; Mills et al., 2009; Peary et al., 2012; Reuter & Kaufhold, 2017). Social media can potentially be harnessed in similar ways to facilitate community responses to COVID‐19 (Quinn, 2020).

The present research sought to determine the feasibility, acceptability, and community reach of an intervention delivered over social media platforms that recruited a large cohort of volunteer social influencers and encouraged them to pass along COVID‐19 prevention, care, and coping messages within their social networks. Identified by the name ‘1000 Hometown Heroes’ (1000 HH), the 9‐month rapid response project regularly posted social media messages to the cohort of social influencers. As in other types of network interventions (Valente, 2012), messages were meant to be shared by influencers within their social networks to provide practical information to assist others in coping with personal, social and economic sequelae of the pandemic. Because of the stark COVID‐19 disparities affecting communities of colour, 1000 HH especially sought to reach racially‐ and ethnically‐diverse community members. This paper reports on the methods used in project implementation, the intervention's focus and content, characteristics of the social influencer cohort, and quantitative and qualitative evaluations of intervention feasibility, acceptability and reach, all of which are constructs that underlie the successful implementation of an intervention (Gaglio et al., 2013; Glasgow et al., 1999).

METHODS

Project background

1000 HH was implemented in May, 2020, with data collected through a follow‐up point in January, 2021. Procedures followed a protocol approved by the Medical College of Wisconsin IRB. To ensure that the campaign reached those affected by COVID‐19 and responded to their needs, a community advisory panel (CAP) was convened to provide recommendations for project priorities, methods, content, and strategies for engaging with an initial cadre of community members widely connected with others on social media. The CAP included faith leaders, agency and neighbourhood volunteers and other local stakeholders. 1000 HH Facebook and Instagram pages and a project website were created, and the social media campaign was launched.

Identification and recruitment of social influencers

Recruiting and engaging with social influencers was ongoing for the full project period. Recruitment of influencers relied on outreach, referrals, links with community organisations, CAP recommendations, Facebook ads and announcements placed on other organisations' websites. The study staff identified local leaders, agencies and organisations that worked with populations in Milwaukee most affected by COVID‐19. Influencers were also identified by consulting with community leaders, faith leaders, activists and other groups. Advertisements were posted on Facebook and other sites inviting interested community members to follow 1000 HH. Online announcements about 1000 HH were featured on local news websites that reached Hispanic and African American community members. Interested volunteers were considered influencers if they followed the 1000 HH Social Media pages on Facebook or Instagram and received project posts and messages. Moreover, throughout the project, individuals who ‘liked’ a post but were not yet followers were invited to join 1000 HH. Most posts and messages to influencers were in the influencer's language of preference, English or Spanish.

We did not establish criteria for influencers, and we allowed interested persons to self‐nominate. 1000 HH was not meant to be an exclusionary ‘club’, and any community member who wanted to help was welcome to join. Influencers were provided with 1000 HH buttons and stickers to recognise their roles but were not financially compensated for sharing messages or bringing others into the cohort. Influencers received gift cards only for completing evaluation measures.

Development and content of social media posts

The heart of 1000 HH were posts that could be passed on by influencers to others in their social networks. A total of 394 Facebook and 95 Instagram posts were provided to social influencers over the project period, all developed to be concise, easily understood and include both a written message and an image. Posts were reviewed and edited by a diverse staff team to ensure that they were clear and culturally appropriate, including their bilingual forms. The posting schedule targeted days and times with the greatest number of online users on weekdays and weekend.

1000 HH posts included content recommended by the CAP, project workgroups, community‐engaged staff and surveys completed by social influencers that identified new and emerging areas of community concern. Topics evolved in response to community needs over the course of the pandemic. The project team also monitored other organisations' social media pages and shared relevant COVID‐19 announcements with the influencer cohort. Posts generally did not restate basic COVID‐19 facts but instead provided practical strategies and resources for coping with COVID‐19 stressors relevant to community members. Contents of posts included:

Practical help in meeting economic survival needs caused by the pandemic, including locations of food banks, ‘pop‐up’ food pantries and organisations offering free meals; sources of rent and utility assistance; sources of healthcare for the uninsured; and navigating procedures to apply for unemployment assistance;

Protecting oneself and family members from COVID‐19, including where to get free adult and children's face masks; ways to reduce virus exposure risk by people who had to be close to the public and co‐workers; steps to protect others in the same household when one family member has COVID‐19; and reducing the need for frequent trips for groceries or other essentials;

Coping with mental health stressors worsened by COVID‐19, such as handling the impacts of stress caused by income loss and isolation; sources of assistance for victims of domestic and intimate partner violence; identifying activities to replace those unavailable due to COVID‐19; strategies for remaining connected with friends and loved ones; and accessing virtual mental health services, support and faith support communities;

Helping children and older people cope with COVID‐19 fears and life disruptions, including tutoring children in virtual learning; helping young people cope with having less in‐person contact with schoolmates and peers; recognising and handling signs of depression including where to get professional assistance; helping young children, teenagers and elderly relatives stay connected with friends over extended periods of time; building resilience to uncertainty, especially about when the pandemic would end;

Linking with important social justice movements also impacting communities during the COVID‐19 pandemic. 2020 was not only the year when the COVID‐19 pandemic swept the United States, but also a year of widespread protests and mobilisation against racial injustice. Social influencers were encouraged to share with friends, family members and neighbours' practical advice on how persons who participate in Black Lives Matter protests and other demonstrations can maintain COVID‐19 precautions. We believed that the credibility of 1000 HH required that the project address important contextual social justice issues that co‐occurred with the COVID‐19 pandemic and affected the lives of people concurrently impacted by multiple syndemics and health disparities.

Feature stories that profiled community members who were playing roles to help others during the COVID‐19 pandemic. 1000 HH posted human interest stories about local community members who donated their time and effort to assist others during this time of crisis, such as by volunteering at a food pantry, helping seniors with shopping and playing other help‐giving roles. These features reinforced themes of supportive concern for the community and allowed the project to highlight the selflessness of people assisting others in the midst of the pandemic.

Study measures

The project's feasibility, acceptability, and reach were evaluated with five types of measures: (1) enrollment surveys completed when an individual first followed 1000 HH to characterise the social influencer cohort; (2) monitoring surveys from a subset of influencers to elicit feedback and identify new community COVID‐19 concerns and impacts; (3) social media platform tracking metrics (Facebook ‘insights’) that provided general demographic characteristics of 1000 HH followers and objective, real‐time data on the reach and user interactions with each week's posts; (4) final quantitative evaluation surveys completed by influencers at the end of the project; (5) qualitative review of comments, feedback and post‐participation telephone interviews with a subset of influencers to elicit additional feedback about the project participation experience. Participants were presented with an online informational consent letter prior to being asked to complete any study measures and clicked an agreement button to indicate their consent to participate. This online consent procedure was approved by the institution's IRB.

Enrollment surveys

Social influencers were invited to complete a REDCap enrollment survey that included questions about age, gender, race, ethnicity, language of preference (English or Spanish) and zip code of residence. A total of 776 individuals completed enrollment surveys, 61.9% of the 1253 individuals who followed 1000 HH on Facebook or Instagram. So as not to exclude anyone who wanted to participate in 1000 HH, individuals were not required to complete enrollment or evaluation surveys in order to participate.

Monitoring surveys

Weekly monitoring surveys were sent by email or text message to a randomised subset of influencers (n = 298) to gain ongoing information about their reactions and onward dissemination of 1000 HH messages and to identify new and emerging COVID‐19‐related community concerns.

Social media platform tracking metrics

Facebook—the primary platform used in the project—provides data (‘insights’) to quantify interactions with the 1000 HH page (e.g. reach, engagement, page views and page likes/follows), demographic characteristics of page followers (gender, age range and language), and post shares and reach. These constitute objective measures of 1000 HH message dissemination, and organic impressions that reflect the number of times 1000 HH posts entered users' screens.

Final evaluation surveys

Social influencers who completed enrollment surveys were invited to complete a final evaluation survey during the project's last 2 months. The REDCap survey assessed exposure to 1000 HH messages; self‐reported sharing of messages whether online or in other ways (e.g. by telephone call, text message, email or in face‐to‐face conversation); the perceived impact of 1000 HH messages on the influencer's social network members; and influencers' experience participating in 1000 HH.

Qualitative data collection

Throughout the project, the study team monitored Facebook comments made in response to 1000 HH posts and recorded verbatim comments that provided insights about how influencers felt about the project, how they were dealing with the pandemic, and concerns or challenges experienced by those around them. Both the monitoring survey and the final evaluation survey included open‐ended questions asking influencers to describe evolving community concerns and needs, as well as impressions about the project, the impact and usefulness of 1000 HH messages, and whether and how messages were shared within their social networks. These data were collected, organised and summarised by theme by the study team.

In addition, and after influencers completed final evaluation surveys, project staff contacted and then conducted semi‐structured telephone interviews with 16 members of the social influencer cohort, purposively selected to maximise racial, ethnic and age diversity. Open‐ended telephone interviews explored in more depth perceptions about the project, reactions of network members to the messages and quality of overall participation experience. Interviews lasted 20–30 min and were conducted in English or Spanish by experienced qualitative interviewers. Interviewees received modest incentive payments for their time. The study staff took notes during each interview and read back verbatim quotes to the interviewee to ensure their accuracy and completeness. The interviews were formally coded and analysed. A qualitative research team reviewed and summarised these data to identify key themes.

RESULTS

Characteristics of the social influencer cohort

As Table 1 shows, the 776 influencers who completed enrollment surveys were diverse in age and race/ethnicity; 26% were African American, 27% were White and 4% were of other races; 43% were of Hispanic ethnicity although many who identified as Hispanic did not separately indicate their race. One‐fourth (24%) of enrolled participants received materials in Spanish. Participants ranged in age from 18 to 81 years old (M = 39.8, SD = 12.03), and 87% self‐identified as female. The most frequently reported zip codes of residence were Milwaukee's predominantly Hispanic and predominantly African American neighbourhoods. Influencers reported substantial social media presence, with an average of 643 Facebook friends (median = 300, SD = 813.22) and 449 Instagram followers (median = 200, SD = 583.22) amongst those who used the platform. Certain demographic data were available from social media platforms for followers of 1000 HH regardless of whether they did or did not complete study enrollment surveys. These platform data showed similar characteristics to the demographic characteristics of enrolled influencers, although the proportion of male influencers was slightly higher (21%) and Spanish‐speaking influencers slightly lower (13%) in the whole cohort.

TABLE 1

Demographic characteristics of enrolled social influencers in the 1000 Hometown Heroes cohort (n = 776)

Gender
Male	87% (n = 592)
Female	12% (n = 84)
Another gender	1% (n = 8)
Missing or preferred not to answer	n = 92
Mean age	29.8 years (n = 527)
Race/ethnicity
Hispanic	43% (n = 283)
White	27% (n = 181)
African American	26% (n = 173)
Other races	4% (n = 25)
Missing or preferred not to answer	n = 114
Language of preference
English	76% (n = 590)
Spanish	24% (n = 186)

Notes: Questions asked separately about influencer race (African American, White, Asian/Pacific Islander, Native American) and ethnicity (Hispanic/non‐Hispanic). Because a large number of respondents who indicated Hispanic ethnicity did not answer the question about their race (n = 171, 60% of Hispanic participants and 22% of the total sample), we grouped influencers in one of the four race/ethnicity categories. Some African American/Black participants (n = 38, 5% of the total sample) were of Hispanic ethnicity.

Reach of hometown heroes message posts and engagement with posts

Facebook metrics provided an objective assessment of exposure to and interactions with 1000 HH posts. As shown in Figure 1, the number of individuals who followed the campaign's Facebook page alone grew to 1063. Other influencers followed 1000 HH on Instagram. During the campaign, Hometown Heroes messages reached an average of 4470 unique individuals per week (SD = 3722, median = 3821) and had an average of 8295 total views per week (SD = 5803, median = 7829). The ‘viral reach’ of the page—the number of unique individuals who saw content based on sharing by their friends—averaged 1673 people per week (SD = 1268, median = 1267). People engaged with campaign messages (such as by liking, commenting on or sharing messages) an average of 326 times per week (SD = 289, median = 221). Figure 2 shows the project's cumulative total number of organic Facebook impressions (the number of times 1000 HH content entered someone's computer screen) over the life of the intervention. Total cumulative organic impressions reached more than 140,000 on Facebook. This underestimates actual message reach because it does not include other platforms or ways in which messages were shared outside of social media.

FIGURE 1

Growth by month in number of Facebook followers of 1000 Hometown Heroes.

FIGURE 2

Lifetime total organic impressions for 1000 Hometown Heroes on Facebook.

During the campaign's peak months following the rollout, the campaign reached a mean of 7978 unique individuals per week (SD = 3822, median = 7140) and engagement reached 574 people per week (SD = 334, median = 509). Viral reach during these months averaged 2260 people per week (SD = 1425, median = 1653). Weekly post views in this period averaged 13,894 (SD = 5353, median = 13,495).

Final evaluation surveys were completed at the end of the project by 317 social influencers or 41% of those who had enrolled. Influencers completing the final evaluation survey did not differ from enrolled influencers who did not complete the final evaluation in age, race/ethnicity, language of preference or reported social media following, although women were more likely than men to complete the final evaluation survey (47% vs. 29%), χ 2(2) = 10.06, p = 0.002.

Tables 2 and 3 summarise evaluation survey data. The great majority of influencers felt 1000 HH messages were helpful (86%) and relevant to their community (83%). Nearly two‐thirds (64%) said they talked with people about 1000 HH resources or posts, and 71% reported saving information from messages for future use or reference. Eighty‐four per cent of influencers said participating in 1000 HH was a positive experience and 78% felt they did a better job protecting themselves and the community from COVID‐19 because of their participation. Nearly half (48%) of influencers said they personally accessed community resources they had learned about through 1000 HH. The great majority were satisfied with the variety of post topics (85%), message clarity (85%), post appearance (82%) and the regularity of new posts (81%). Sixty‐eight per cent of influencers who saw Hometown Heroes messages reported sharing them.

TABLE 2

Reported influencer interactions with the 1000 Hometown Heroes campaign amongst final evaluation survey respondents (n = 317)

	Per cent (n)
Saw messages from 1000 HH on Facebook or Instagram
Yes	85% (271)
No	7% (23)
Not sure	7% (23)
Shared 1000 HH messages a
Yes	68% (184)
No	18% (49)
Not sure	14% (38)
Accessed resources learned about from 1000 HH a
Yes	48% (122)
No	37% (93)
Not sure	16% (40)
Information shared from 1000 HH has helped someone they know b
Yes	52% (92)
No	6% (11)
Not sure	42% (74)
Missing	n = 14

Amongst those who have seen messages (n = 271).

Amongst those who have shared messages (n = 171).

TABLE 3

Satisfaction with the 1000 Hometown Heroes campaign amongst final evaluation survey respondents who reported having seen campaign messages (n = 271)

Campaign evaluations	% Agreeing
Overall, I found messages helpful	86
Participating in 1000 Hometown Heroes was a positive experience	84
Messages were relevant to my community	83
Messages were useful to me	82
Messages I shared were useful to others a	82
I am doing a better job protecting myself and my community from COVID‐19 because of 1000 Hometown Heroes	78
Others responded positively to the messages I shared a	72
I saved information from messages for future use or reference	71
I talked to people about resources or posts from 1000 Hometown Heroes	64

Amongst those who have shared messages (n = 171).

Final evaluation surveys provided information about the project reach beyond social media. Social influencers' sharing of messages to network members was not limited to social media. Whilst 1000 HH only posted messages to influencers on Facebook and Instagram, more than half of influencers (53%) said they shared messages in other places and ways. These included sharing messages on different social media platforms, and also by phone, text message, email and face‐to‐face conversation. Those sharing messages reported sharing a mean of 13 messages shared on social media (median = 5, SD = 15.54) and also a mean of 12 messages shared in other ways such as conversations and text messages (median = 5, SD = 15.93). Respondents expressed willingness to continue participating in 1000 HH in the future. Eighty‐six per cent said that they would share future COVID‐19 information from Hometown Heroes within their networks, and 82% said they would like to share information about COVID‐19 vaccination with their community.

Social influencers' comments, feedback and post‐participation open‐ended interviews

Influencers provided feedback on the campaign in response to open‐ended questions in the final evaluation survey, Facebook comments and the telephone interviews following project participation. Five main themes emerged: (1) how it felt to be a Hometown Hero; (2) general perceptions of the project; (3) project messaging; (4) dissemination and sharing of information; (5) benefits to influencers. Direct quotes from open‐ended survey questions, Facebook comments and telephone interviews are used to highlight the main themes, also noting the language of the interview.

How it felt to be a Hometown Hero

Most influencers indicated they participated in 1000 HH to help their communities and be a community leader and trusted source of information:

I feel very appreciative to be seen as a Hometown Hero because I love helping my community. I like that my community has support, and we can be counted on.—34‐year‐old Hispanic female, open‐ended interview (Spanish).

It felt good to pass along information to others, especially to those that are struggling a bit and do not want to admit it. I really liked the sharing of community events.—66‐year‐old White female, final survey (English).

Whilst influencers reported satisfaction in helping their community, others said it also gave them a sense of belonging:

[This] made me feel like I was a part of something. I kind of stepped out of my comfort zone, because I'm really not too big on joining different things like that, but I feel like I gained a lot of knowledge.—27‐year‐old Black female, open‐ended interview (English).

General perceptions of the project

Overall, influencers were very supportive of the 1000 HH campaign and liked the variety of messages and how they could easily share them:

I think it's a great way to connect with the community. And if anyone wanted more information, they could always reach out to me. When I see something, I try to forward it on, … get the information out there to my friends … and I thought … the best way is word of mouth. You get it out there to the people, and they will share the information. I think the strategy was very successful in my eyes.—43‐year‐old Hispanic female, open‐ended interview (English).

Influencers appreciated the opportunity to shape the direction of Hometown Heroes messages and resource posts:

The messages were good because you were being reminded of what was going on. The surveys on the page were good cause you could share your opinions about how you liked the page … The messages about food were good because I was out of work with low resources. I really am grateful, because that really helped us.—32‐year‐old Hispanic male, open‐ended interview (Spanish).

A number of influencers commented that the campaign felt inclusive and celebrated diversity. In addition to posts sharing COVID‐19 information and resources, some posts highlighted minority‐owned businesses, provided information about racial and ethnic heritage months, and spotlighted individuals from diverse backgrounds:

My family and people in my circle thought it was pretty awesome that this even existed, especially online like on Facebook. My aunt kept telling me that it felt like they were talking directly to us because of how y'all made sure a lot of the information was to the Black community.—27‐year‐old Black female, open‐ended interview (English).

Inclusivity was also achieved through bilingual posting. Even though Milwaukee has a large Spanish‐speaking Hispanic community, information regarding COVID‐19 resources is not always available in Spanish. One influencer explained how individuals who primarily speak Spanish sometimes often do not learn about these resources:

Hispanics, people with few resources like myself, [accessed] resources posted on the page. Usually, information is in English, so us Latinos do not have access to it, so we do not comprehend it. Great idea that it was in English and Spanish.—42‐year‐old Hispanic female, open‐ended interview (Spanish).

Project messaging

For 1000 HH to be successful, followers and community members needed to trust the information being provided. It was evident in many comments that 1000 HH was viewed as trustworthy, especially important against a backdrop of a lot of misinformation, mistrust and ‘fake news’:

We need reliable information to counteract the huge wave of false information coming at us daily.—59‐year‐old White female, final survey (English).

The [social media posts] helped me to be aware of the testing procedures, and to be more clear of information about testing and social distancing, as well as the recommendations from the CDC and other medical authorities… It was often difficult to decide what was true and what was not true. So, I thought it was a reliable source of information, and I appreciated that.—24‐year‐old White female, open‐ended interview (English).

I liked seeing all the social media posts, and I like that a lot of that was translated into Spanish. It was really nice to have reliable information in such a reliable manner.—65‐year‐old White female, open‐ended interview (English).

Dissemination and sharing of information

To reach as many people as possible, it was crucial that influencers shared the posted messages. Many influencers stated they regularly did this:

I tell my daughters. To co‐workers, friends, and families, and shared it on my social media page. I took a picture and shared it to Snapchat… also to my Instagram.—52‐year‐old White female, open‐ended interview (English).

Yes, I shared with my people on my social media page. About the impact it had, if I share it, they share it, and that is helpful to me because it is reaching other people who I did not share it with, and I want other people to see it as well. People have in‐boxed me to say thank you for posting that flyer and what not, it made their day. I know they are utilising the resources ‘cause I always got thanks for posting things.—36‐year‐old Black female, open‐ended interview (English).

In addition, many followers indicated that they shared the messages with people not engaged on social media, such as elderly family and friends or their church communities:

Yes, I shared it with my family, friends, and church friends. Especially I shared with my parents and older uncle and aunts, since they are a bit older and needed the information. My sisters and other family and friends were able to benefit from the food post. The impact for my parents and older family was that they could know where to go to get tested for COVID‐19 because it was very important for them that they stay in good health.—44‐year‐old Hispanic female, open‐ended interview (Spanish).

I really liked it because it was very simple to understand and I could share it with others. And with others I know that do not have Facebook or access to the internet, I could share and tell them the information.—35‐year‐old Hispanic female, open‐ended interview (Spanish).

Benefits to influencers

Influencers identified personal benefits from the campaign. Whilst the benefits from COVID‐19‐related information and resources such as testing and free mask distribution sites were most frequently cited, influencers highlighted benefits from many other resources such as those related to coping with food insecurity:

The biggest [way the project helped me] was the Boys and Girls Club was offering for your student a week's worth of snacks and lunches; that was a real timesaver for me. I shared that on my page right away.—43‐year‐old Hispanic female, open‐ended interview (English).

My mom recently lost her job and suffered the loss of her husband at the same time. So she's having a hard time right now. I've been trying to help her with finding some community resources, so this post comes at the perfect timing. (In response to a post about resources for senior citizens)—Facebook comment, 8/2/2020 (English).

DISCUSSION

The 1000 HH campaign was undertaken as a rapid response project to address urgent community needs during the COVID‐19 pandemic. Because our aim was to very quickly scale up the intervention, the project drew upon two evidence‐based frameworks: (1) identification and engagement of social network or popular opinion leaders in the community to disseminate health protective and coping messages within their networks; (2) use of social media as a strategy to accelerate the scale and reach of messages to assist community members in coping with disaster. The project applied these frameworks to address COVID‐19, guided by initial input from community advisors and ongoing feedback about evolving community concerns from social influencers themselves.

Quantitative and qualitative data—including objective tracking metric data documenting message reach and exposure provided by social media platforms, surveys completed by social influencers, and survey comments and open‐ended interviews with a diverse subset of influencers—provided support for the intervention's feasibility, acceptability and reach. Consistent with our goal to especially engaging with members of minority communities that experience COVID‐19 disparities, the project successfully enrolled a cohort that was predominantly comprised of African American and Hispanic community members, and one in four enrolled influencers received project materials in Spanish. A majority of racial and ethnic minorities of all ages in the United States have smartphones, use social media, and use Facebook in particular (Lohse, 2013; Pew Research, 2019; Thornton et al., 2016), and the pandemic forced many people to adopt remote forms of communication. This guided our decision to communicate with influencers over social media. Although Hometown Heroes communications with the influencer cohort were through social media, evaluation data showed that messages were shared in approximately equal numbers through social media and non‐social media channels, including telephone calls, text messages and in‐person conversations. This suggests that the project reached a much larger number of community members than social media platform tracking records measured and that COVID‐19 messages could reach community members not connected to social media or technology.

At the same time, the present study has a number of limitations. Whilst our findings support the feasibility, acceptability, and community reach of the intervention, the project did not assess the impact of the intervention on the community members with whom influencers shared messages. Future research will be needed to systematically establish benefits to community members beyond those who were in the influencer cohort. In addition, decisions were made at all stages of the project to balance rigour in evaluation with our aim to carry out a rapid rollout intervention to meet urgent community COVID‐19 needs. For example, the project reached out to recruit community members with substantial social media presence but also welcomed others who heard about—and wanted to be a part—of the effort. In order to maximise the reach of 1000 HH, we viewed the designation of ‘social influencer’ broadly. Consistent with the use of the term in the marketing field, influencers included persons with large social media followings. However, persons who did not have established social media followings could volunteer to engage in the effort and were included. Although designating community volunteers and 1000 HH followers as influencers differ from the usual meaning of ‘social media influencers’ in the marketing field, we believe this inclusivity was a strength.

Other limitations of the research also merit comment. Because this was an applied community program that aimed to engage as many community members as possible, the project requested—but did not require—participants to formally enrol or complete evaluation surveys. The majority did not complete evaluation surveys. Although Facebook provides certain demographic information about targeted audiences, the high number of survey nonresponders limits our knowledge of sample composition and creates gaps in the findings. Moreover, the majority of influencers were females. We are uncertain of the reasons for this gender imbalance, which could have included recruitment messages reaching or appealing more to women than men, more participation and interest of women in health‐related communication, different message dissemination in networks of men and women or other factors.

Public health campaigns typically rely on ‘top down’ communication from political leaders and health authorities to community members. However, community members often look to personally‐known members of their own social networks for trusted advice, guidance and support, especially in a time of crisis such as that posed by the COVID‐19 pandemic. Our findings indicate that community members are willing—and derive satisfaction—in taking on roles to help friends, family members, neighbours, co‐workers and others cope during a public health emergency. This demonstrates the feasibility of using social media to reach a community population diverse in age, ethnicity and race. This approach carries promise for addressing other issues, including the potential benefits of engaging social influencers in efforts to optimise COVID‐19 vaccination.

AUTHOR CONTRIBUTIONS

Jeffrey A. Kelly was the Principal Investigator of the overall project and oversaw paper writing. Jennifer L. Walsh and Yuri A. Amirkhanian headed the assessment workgroup, Laura R. Glasman and Erika Christenson headed the message development workgroup, Carol L. Galletly and Broderick Pearson headed the community concerns workgroup, and Katherine G. Quinn and Erika Christenson performed qualitative analyses. All authors participated in drafting sections of the manuscript.

CONFLICT OF INTEREST

The authors declare no conflict of interest. This research followed a protocol that was approved by the Medical College of Wisconsin IRB, and participants clicked agreement to complete study measures after reading an online informational letter.