Exploring Influences of COVID-19 Messaging on Behavior in Virginia: Summary of the Plenary Presentation for the 10th Annual Appalachian Translational Research Network Health Summit

This project was motivated by the increase in cases of the novel virus SARS-CoV-2 (COVID-19) in the United States in early 2020 and more specifically, in Virginia in March 2020. Nationally, the prevalence of COVID-19 in the U.S. increased from a single case reported on January 21, 2020 to a total of 1,023,659 cumulative cases, and resulted in  60,057 deaths as of April 29, 2020, according to the CDC. The Virginia Department of Health (VDH) first started reporting COVID-19 cases on March 17, 2020. By May 1st, VDH reported 14,961 cumulative cases and 522 deaths due to COVID-19 as of April 30th, statewide.

The researchers drew parallels between the COVID-19 pandemic and the 1918 Spanish Flu pandemics, as both resulted in tremendous differences in the number of outbreaks and deaths by geographic location. In Virginia and nationwide, COVID-19 was impacting minoritized populations in a disproportionate manner. CDC data reported in August 2020 showed 5.3 times higher rates of hospitalization in American Indian or Alaska Native, non-Hispanic persons, 4.7 times higher rates of hospitalization in Black or African American, Non-Hispanic persons, and 4.6 times higher rates of hospitalization in Hispanic or Latino persons than in White, Non-Hispanic persons. 

Individual public behaviors including wearing masks over one’s nose and mouth and social/physical distancing, have been essential to controlling the spread of COVID-19 as therapeutics and vaccines are being developed. The researchers wanted to better understand the public behavioral response to COVID-19 messaging in order to inform future communication efforts to more effectively “flatten the curve,” limit transmission, prevent illness, and save lives in Virginia and throughout the country. 

In order to understand the “whole story” about Virginians’ trust of and response to COVID-19 related public health messaging, the team chose to employ a mixed methods design. Collecting both qualitative and quantitative data allowed them to gain deep insight from diverse perspectives as well as triangulate between quantitative and qualitative data. The team used a mixed methods research design, including a cross-sectional online Qualtrics survey to collect quantitative (numerical) data and semi-structured virtual focus groups conducted via Zoom to collect qualitative (non-numerical) data. Participants were recruited via social media (including geographically targeted advertisements), listservs, and flyers with a QR code, to initiate snow-ball sampling. 

In addition to demographic information, survey participants provided information about their COVID-19 information sources; what COVID-19 related messaging they heard, believed, and which impacted their behaviors; how they changed their behaviors due to the pandemic; how seriously they took the pandemic at the time of the survey; and what influenced their perceptions of severity. 

Due to the conference focus on Appalachia, for this ATRN plenary presentation, results were limited to those of participants reporting a zip code within a county defined as part of Appalachia by the Appalachian Regional Commission (ARC). Preliminary quantitative data analysis from the online survey includes the following:

• Between May 19 and July 19, 2020, 3,694 Virginia residents completed the online survey
  • Of these, 1,041 (31%) had a zip code in an Appalachian county
  • Blacksburg zip codes were overrepresented due to the study being initiated through Virginia Tech, with 39% of the Appalachian participants reporting a Blacksburg zip code 
• Of the Appalachian participants, 76% were female, 90% were non-Hispanic White, and 53% had a master’s or doctoral degree
  • 45% identified as a Democrat, 15% as a Republican, and 40% as Independent, other, or no-preference in terms of political affiliation
• 72% of males, 82% of females, 57% of Republican, 94% of Democrats, and 73% of Independents, reported believing that COVID-19 was very serious
• Over half of the respondents reported the following impacting their perceived seriousness of the pandemic: 
  • Hearing about COVID-19 in other states or countries
  • Public schools closing
  • The governor declaring a state of emergency and mandated stay at home order
  • The CDC recommending face-masks
  • Restaurant dining rooms closing
  • Cancelled or postponed sporting events 
• 95% reported practicing social/physical distancing
• 89% said they washed their hands more often
• 89% said they shopped for essentials like groceries less often
• 86% reported wearing a mask when in public
  • 80% of men and 88% of women reported wearing a mask while 97% of men and 95% of women reported practicing physical distancing
  • Both of these reported behaviors increased with increasing age with 77% of 18-24 year olds vs. over 90% of those 60 years old and older reporting wearing a mask in public, though over 90% of those across all age groups reported practicing distancing
  • 70% and 86% of republicans and 96% and 99% of democrats reported mask wearing and distancing, respectively 
  • These behaviors also increased in reported frequency with increasing level of education
• Over 70% of respondents reported believing in and being affected by conventional messages such as “practice social distancing,” “the coronavirus is highly contagious,” “stay home, stay safe,” “stay home, save lives, slow the spread,” “flatten the curve,” “wash your hands for at least 20 seconds,” “your mask protects me, my mask protects you,” and/or “don’t touch your face” 
• 12% of respondents believed one or more alternative messages about COVID-19 including that it: “was developed as a bioweapon,” “was developed to lower social security payments to seniors,” “is a sign of the apocalypse/end times,” “is a hoax,” “can be treated with natural remedies,” “was developed for population control,” and/or “was developed to increase sales of cleaning supplies” 
  • A similar proportion of men (11%) and women (12%) believed one or more of these alternative messages
  • 30% of republicans, 4% of democrats, and 13% of others believed one or more of these messages
  • 27% of those age 18-24 believed one or more of these messages compared to between 3% and 14% of those in older age groups
  • Belief in one or more of these messages decreased with increasing level of education 36% of those with a high school diploma, 18% of those who completed trade school, 19% of those with some college, 14% of those with a bachelor's degree, 8% of those with a master’s degree, and 5% of those with a doctoral degree
• Most participants received trusted health information from National Health and Science Organizations like the NIH or CDC, including 86% of those who did not believe an alternative message compared to 66% of those who believed one or more alternative messages. 
  • Additionally, of those who did not vs. did believe one or more alternative messages, respectively:
    • 78% vs. 60% received trusted information from state or local health departments
    • 69% vs. 65% from health care professionals
    • 59% vs. 46% from online news
    • 47% vs. 49% from national TV news
    • 43% vs. 36% from local government leaders
    • 32% vs. 40% from local TV news
    • 28 vs. 38% from family and friends
    • 22 vs. 16% from radio
    • 22% vs. 36% from Federal Government leaders
    • 21% vs. 30% from social media
    • 18% vs. 12% from local newspaper
    • 4% vs. 11% from a faith leader

Study limitations included utilizing convenience and snow-ball sampling via online recruitment resulting in a non-representative sample of Virginia residents. Despite this limitation, we were able to recruit participants across Virginia within an 8 week period.

Quantitative data analysis from the online survey results will continue. Additional next phases of this mixed methods study involve qualitative data collection via virtual focus groups, conducted via Zoom. Focus groups have begun and have been completed thus far with a sample of participants who took the online survey, as well as with certain demographic groups of interest, including  Black/African-American participants and young adult participants. The researchers are still planning to conduct focus groups with Latinx Virginians, since that population has also been disproportionately affected by COVID-19. As qualitative and quantitative data continue to be analyzed, the researchers compare quantitative results and qualitative findings to triangulate and gain deeper understanding of how Virginians heard, trusted, and responded to public health messages about COVID-19 to help inform ongoing and future messaging efforts.

Research team: 
Natalie Cook, PhD - Virginia Tech
Mary Ann Friesen, PhD, RN, CPHQ - INOVA Health
Rachel Silverman, PhD - Virginia Tech
Emily Pomfrey Wells - University of Virginia at Wise
Sophie Wenzel, MPH, DrPH - Virginia Tech