General and Tailored Covid-19 Health Messaging to Minorities in the United States

Fatima Cody Stanford
Sarah Eichmeyer
Paul Goldsmith-Pinkham
Lucy Ogbu-Nwobodo
Carlos Torres
Anirudh Sankar
Pierre-Luc Vautrey
and the Covid-19 Messaging Working Group
In the US, recent statistics show that Black and Latinx communities bear a disproportionate burden from Covid-19. Can improved public health messaging ameliorate this situation?

Project Summary

Black and Latinx communities have been disproportionately affected by Covid-19. The disparities in prevalence and health outcomes reflect a complex set of factors including systemic racism, inequality in access to health services, discrepancies in educational opportunities, and occupational differences. There are also knowledge gaps, perhaps reflecting the lack of effort to design Covid-19 public health messaging with minority communities in mind. White experts deliver most messages, and they rarely address these communities’ specific concerns.

Motivated by this issue, the research team worked with the Center for Diversity and Inclusion (CDI) of Massachusetts General Hospital (MGH), and over 40 physicians from MGH and the Lynn Community Health Center, to form a “Covid-19 Messaging Working Group” and develop Covid-19 video health messages that would address the specific concerns of these communities. To evaluate these messages’ effectiveness at increasing Covid-19 knowledge and spurring health information-seeking behaviors, the team conducted an online evaluation involving approximately 15,000 Black and Latinx individuals. Some participants randomly received these physician-delivered video messages on Covid-19. For those who received video messages, further randomization made it possible to assess whether race or ethnic concordance (i.e. where the physicians’ race matched that of the respondent), acknowledgments of injustice, or information on social perceptions would make the message more effective, as measured by a reduced knowledge gap and change in information-seeking behaviors.

The primary outcomes were Covid-19 related knowledge as well as health information-seeking behavior. To determine knowledge, participants were asked about symptoms, transmission, and best practices such as social distancing and mask-wearing. To determine health information seeking behaviors, participants were asked about their interest in a variety of weblinks about social distancing and hygiene, health resources, and homemade protective face coverings. The impact on health information-seeking is determined by the number of links for which participants expressed interest.

What have we learned?

Public health messaging, as transmitted through physicians, are effective for reducing knowledge gaps about Covid-19.

The results show that public health messaging is a strong lever for ameliorating knowledge gaps, and physicians are a great vehicle for these messages. Regardless of physician race, public health messages directly improved knowledge of symptoms, asymptomatic transmission, and preventative practices for both Black and Latinx respondents.

Race concordant messaging was particularly effective for improving health information-seeking behavior for Black respondents.

For Black respondents receiving video messages, messages delivered by Black physicians spurred an increased interest in health information-seeking, such as interest in watching tutorial videos on making homemade masks and seeking additional health resources about Covid-19. This suggests the importance of race concordance in spurring these behaviors.

Public health video messages

Respondents who received public health messaging videos saw a sequence of three videos, covering, respectively, (1) major symptoms and preventative practices, (2) advice about social distancing, and (3) information about mask wearing. Randomized variants of this standard sequence included acknowledgements of injustice and/or information about perceptions of mask wearing in the community.
The race of the physician was kept fixed throughout the video sequence, where respondents were randomized to see messages from a race concordant (Black or Latinx) or discordant (White) physician.
As an example, we show a standard sequence of messages by Black doctors. Such a sequence closed knowledge gaps relative to no videos, and furthermore encouraged health information-seeking behaviors relative to race discordant doctors. 

See all public health videos used in this project on YouTube 

Research Results

View the full results on Annals of Internal Medicine. 

Visual abstract by Annals of Internal Medicine
Source: Annuals of Internal Medicine. Copyright 2020 American College of Physicians. Used with permission.
Figure  1 Physician videos decreased Covid-19 knowledge gaps for all races
Figure showing that physician videos decreased Covid-19 knowledge gaps for all races

Intervention Details

A control group was guided to the outcomes about knowledge and intended behaviors without receiving any public health messaging. Within the treatment group, participants were either guided to a standard sequence of messages or variants on the standard sequence of messages that included acknowledgements of injustice or inequality, or social perceptions about mask wearing. For each such combination, the messages were either delivered by a race-concordant or discordant physician.

Some treatment respondents were furthermore shown Dr. Deborah Birx delivering the CDC social distancing message from the White House instead of a physician, to find possible contrast in the response to this messenger.

All treatments receive basic information+ Includes social perceptions of mask wearing No social perceptions of mask wearing
"Standard sequence" messages about symptoms and social distancing Concordant/Discordant Concordant/Discordant
Sequence with acknowledgement about economic inequality Concordant/Discordant Concordant/Discordant
Sequence with acknowledgement specific to race Concordant/Discordant Concordant/Discordant
Sequence with Dr. Birx delivered message about social distancing Concordant/Discordant Concordant/Discordant