Design. We conducted a one-year repeated cross-sectional study to assess changes in self-reported adult and child COVID-19 vaccine uptake, intentions, and perceived norms, difficulty accessing COVID-19 information, and trust in COVID-19 information sources and government institutions to ensure vaccine safety among Brigada Digital Latino individuals from the Washington, DC Metropolitan (DMV) area who have been exposed to Brigada Digital de Salud social media content.
Intervention. Beginning in May of 2021, we developed and disseminated approximately two to three weekly social media posts in Spanish to educate audience members about: COVID-19 variants, risk, and prevention; testing; vaccine recommendations, safety, and efficacy; COVID-19 treatment options; and to promote resources for vaccination, testing, and prevention. Given the rapidly changing information landscape, Brigada Digital social media content also sought to provide regular news and scientific updates, explain changes in COVID-19 policies and vaccine eligibility, and correct COVID-19 misinformation. Brigada Digital content was developed for audiences with diverse levels of literacy and education levels, and included explanations of scientific concepts, visual illustrations, and audio narration of text. Content was delivered in varied formats, including carousels, videos, and tutorials (See Fig. 1).
From May 8, 2022 to April 5, 2023, we disseminated a total of 141 unique posts across each of the Brigada Digital Facebook, Instagram, X, and TikTok pages, which were then shared by a trained cadre of 10 community health workers (CHW) with their social media networks and Spanish language, Latino-oriented public DMV-based Facebook groups. Brigada Digital CHWs also conducted digital outreach and health promotion activities to engage audience members, answer questions, and connect community members with resources. A comprehensive discussion of Brigada Digital content development, topics, communication and community-based outreach strategies, and overall audience reach and engagement has been published elsewhere (39).
Instrument and Measures. The Common Survey 2.0 was developed by the National Institutes of Health’s Community Engagement Alliance Against COVID-19 Disparities (NIH CEAL), a national consortium of regional research collaboratives. The survey instrument included measures for sociodemographics, media consumption, adult and child COVID-19 vaccine uptake and intentions, adult booster uptake, difficulty accessing COVID-19 information, and levels of trust in COVID-19 information sources and government entities.
Sociodemographic variables included age, place of birth/origin, sex, education level, employment status, and household income. The survey instrument asked participants’ birth year, and a variable for respondent age was created by subtracting the year of survey administration (e.g. 2022) from the respondent’s reported year of birth (e.g. 2021 − 1975, yielding an age for the respondent of 46). Age categories were then created, including: 18–25, 26–35, 36–45, 46–55, 56–63, and 64+. The survey instrument included eight response options for household income, which was simplified by collapsing household income into four categories, including $15,000-$34,999, $35,000-$74,999, $75,000-$100,000>, and “declined to answer.” Likewise, the original variable for educational attainment was collapsed from eight categories to three, including “Less than high school/Some high school,” “High school graduate/GED/some college,” and “Associates, Bachelor’s or Postgraduate degree.”
Participants’ English language competency was assessed on a 5-point Likert scale from “Speaks English very well” to “Does not speak English at all,” with higher mean scores representing greater English competency. Participants were also asked whether they had ever been diagnosed with a chronic health condition and whether they had health insurance coverage. In addition, the survey included questions about time spent consuming media, social media platforms used, and sources from which they obtained COVID-19 information (i.e., healthcare provider, faith leader, news outlet, social media, federal government). Daily amount of time spent consuming media was captured using a 5-point Likert scale from “None” to “Six hours or more.” Participants were asked to indicate which social media platforms they used, for example, Facebook, WhatsApp, Twitter (now X), Instagram, TikTok, and Snapchat. Sources from where respondents obtained COVID-19 information (i.e., healthcare provider, local television channel, social media, friends or family in the U.S., state or local government, and federal government agencies) were assessed using a 4-point Likert scale with options ranging from “None of my information” to “All of my information” from each specific source. Higher means (ranging from 1–4) indicate more COVID-19 information was obtained from that particular source.
Primary Outcomes. Self-reported adult COVID-19 vaccine uptake, adult booster dose uptake and intentions, and COVID-19 vaccine uptake and intentions for their children under age 18 were assessed as primary outcomes. To assess adult COVID-19 vaccine uptake, participants were asked whether they had received at least one dose of the COVID-19 vaccine, with the response options including, “I received one dose of a two-dose series,” “I received both doses of a two-dose series,” “I received a one-dose vaccine,” and “I have not been vaccinated against COVID-19.” For participants indicating that they had received a one-dose vaccine or both doses of a two-dose series, they were also asked whether they had ever received a booster dose, with response options including, “I received a booster dose,” “I received more than one booster dose,” “I have not received a booster dose, but I plan to,” and “I have not received a booster dose, and I do not plan to.” Results related to vaccine intentions are reported separately from vaccine uptake results.
Among participants indicating that they were a parent/guardian of at least one child under age 18, child vaccine uptake and parent intentions to vaccinate their child(ren) was assessed by asking whether they were in favor of vaccinating their child/ren against COVID-19, with response options including, “Yes, child/ren is/are already vaccinated,” “Yes, I plan to vaccinate my child/ren,” “No, I do not plan on vaccinating my child/ren,” and “I’m unsure/undecided.” Participants were also asked about reasons for deciding to not vaccinate their child(ren) against COVID-19.
Secondary Outcomes. COVID-19 vaccination subjective norms, difficulty accessing COVID-19 information, trust in different sources of COVID-19 information, trust in the FDA to ensure COVID-19 vaccine safety, and trust in the federal government to ensure COVID-19 vaccine safety for children were assessed as secondary outcomes. COVID-19 vaccination subjective norms were assessed using a 4-item scale that asked participants how many people close to them thought they should get the COVID-19 vaccine, and how many of their friends, family, and community members had received the vaccine. A 4-point Likert scale ranging from “None” to “All” was used, with higher mean scores indicating a greater number of individuals.
Perceived levels of difficulty “finding needed COVID-19 information,” "finding COVID-19 information in my preferred language,” and “judging whether COVID-19 information in the media was reliable” were assessed using three survey questions. Items were assessed using a 4-point Likert scale, and response options included “Difficult,” “Somewhat difficult,” “Somewhat easy,” and “Easy,” with higher mean scores indicating greater difficulty.
Participants’ levels of trust in various sources of COVID-19 information (i.e., healthcare provider, faith leader, news outlet, social media, and federal government) were assessed with a 3-point Likert scale using response options of “Not at all,” “A little,” and “A great deal,” with higher mean scores indicating greater trust in that particular information source. Two items also assessed participants’ levels of trust in the U.S. Food and Drug Administration (FDA) and the federal government to ensure COVID-19 vaccine safety generally and specifically for child COVID-19 vaccines. Responses for those two items were assessed using a 3-point Likert scale of “Not at all,” “A little,” and “A great deal,” with higher mean scores indicating greater levels of trust.
Intervention Exposure and Reactions to Brigada Digital Content. To assess participants’ exposure to content, they were asked about the source from which they typically received Brigada Digital social media content (i.e., a social media network contact, a social media account for a group you belong to, a social media account for a community health center, someone you don’t know), the frequency with which they received posts (i.e., more than once per day, once per day, a few times per week, once per week, less than once per week), and their typical actions upon receiving posts (i.e., read the post, like/react to the post, comment on the post, share the post, go to a link in the post, follow the post’s advice, attend an advertised event, don’t read the post).
Additionally, to assess reactions to the content, respondents were provided with four statements, including “Brigada Digital posts are informative,” “I trust the information that I receive from Brigada Digital,” “Posts address my concerns about the COVID-19 vaccine,” and “The way the information was presented in posts kept my interest.” Participants were then asked to indicate the degree to which they agreed/disagreed with these statements using a 5-point Likert scale ranging from “Completely Agree” to “Completely Disagree,” with higher mean scores suggesting greater agreement with the statement.
Participant Recruitment and Sample. To assess changes in COVID-19 vaccine–related outcomes of Brigada Digital audience members, we administered the CEAL Common Survey 2.0 in Spanish with participants in two waves: May 2022 (n = 480) and April 2023 (n = 348). Eligible participants were Latino adults ages 18 or older who resided in DC or Maryland and who spoke Spanish. Participants were recruited from among parishioners of 3 Maryland-based church partners and from among Maryland- and DC-based social media network members of 10 Brigada Digital CHWs. A convenience sample of participants were recruited by sharing a digital flier and making announcements to church congregations, and by Brigada Digital CHWs sharing the digital flier in posts to their social media networks. Participants contacted the study team to complete a survey by using a phone number included in the flier.
Only study participants indicating in the survey that they had ever seen a Brigada Digital post on a social media platform were included in the present analysis (n = 192 in wave 1; n = 123 in wave 2). Among those exposed to Brigada Digital content, 60 individuals in wave 1 and 24 individuals in wave 2 were recruited from among partnering church parishioners and 129 individuals in wave 1 and 98 individuals in wave 2 were recruited from social media network contacts.
Data Collection. Following informed consent, surveys were administered in Spanish using an interview format by trained, Latino data collectors in-person with individuals recruited from churches and by phone with individuals recruited from social media networks. Participant responses were entered directly into REDcap by data collectors using a tablet or laptop computer. The survey took approximately 35 minutes to complete, and participants received a $25 gift card incentive. All instruments and protocols were approved by the GW Institutional Review Board.
Data Analysis. Study outcomes were assessed among participants who self-reported having been exposed to Brigada Digital social media content at waves 1 and 2. To determine the comparability of wave 1 and 2 sample characteristics, we conducted descriptive analyses for socio-demographic, language competency, and health status variables. Chi-squares tests were used for categorical variables and t-tests were used for continuous variables. Means and standard deviations or frequencies and percentages were reported, respectively. In all analyses, the primary independent variable was the wave at which the survey was administered. Subsequent analyses discerned variations in dependent variables as a function of data collection wave.
To assess the primary outcomes of adult vaccine and booster dose uptake, booster intentions, and child vaccine uptake and parent intentions, odds ratios were estimated using logistic regression, while controlling for age, sex, income, language competency, and health insurance status. We controlled for age given its direct correlation with COVID-19 risk levels and the potential influence of age-based vaccine recommendations outcomes for vaccine uptake. Beyond these specific COVID-related reasons, age is generally an important factor to adjust for in health research due to its multifaceted implications on health behavior and outcomes. We also controlled for sex given that women tend to be higher users of healthcare services, including preventative care measures such as vaccination. Additionally, we controlled for income, language competency, and health insurance status since waves 1 and 2 exhibited statistically significant differences for these variables, all of which can influence healthcare access and behaviors.
The survey instrument included items that originally assessed COVID-19 vaccine/booster uptake and intentions concurrently in the same item for adults and children. Therefore, for analytical clarity, these original items were dissected into discrete dummy variables that were generated using response options corresponding to each distinct outcome. This approach permitted a precise portrayal of each distinct outcome; for example, future vaccination intentions could be assessed only among individuals who had yet to be vaccinated with the primary series.
For the secondary outcomes of COVID-19 vaccination subjective norms, difficulty accessing COVID-19 information, trust in COVID-19 information sources and the government, and reactions to Brigada Digital social media content, since all variables were assessed using Likert-type response formats, they were treated as continuous variables. Paired t-tests were executed for each variable to discern any differences in these outcomes between waves 1 and 2. For these paired t-tests, a difference was deemed statistically significant if the means differed at a significance threshold of P < .05. Means and standard deviations are reported for these variables.
To determine whether there were differences in responses between waves 1 and 2 for participant reasons for not vaccinating children and self-reported exposure to Brigada Digital social media content, chi-square tests were performed, with the level of significance demarcated at P < .05. Frequencies and percentages are reported for these variables. All analysis was conducted using STATA 17.