Objective: To determine the degree to which health systems and state health departments used social media to address COVID-related racial disparities in terms of written content and visual representation.

Data Sources: Twitter content (words and images) from May/June 2020 and May/June 2021 were collected from the two largest non-governmental hospitals and state health departments in the five states with the largest racial disparities in COVID-related mortality.

Study Design: The study was a retrospective analysis of Twitter content. The research team coded words in pandemic-related tweets to determine messaging themes. Survey respondents (n=60), with equal representation across five racial/ethnic groups, coded images from pandemic-related tweets to determine whether individuals from minority groups were represented in images.

Data Collection: All tweets and retweets (n=6790) were coded along several lines. For May/June 2020 and May/June 2021, posts were coded as pandemic-related or not. Pandemic-related content was further categorized with open-coding methods into content areas, including: COVID education, hospital or public policy, addressing misinformation, and providing resources. Survey respondents self-identified their race/ethnicity and coded pandemic-related images (n=198) as presenting individuals of a similar race/ethnicity or not.

Principal Findings: In May/June 2020, health departments posted more pandemic-related content than hospitals (µ=204 and 71 tweets respectively; p=.03), including more about health disparities (µ=14.3% of tweets and 2.11%; p=.03). Neither group included significant content addressing COVID misinformation. Between May/June 2020 and May/June 2021, content addressing health disparities decreased for both groups (47% decrease for health departments and 69% decrease for hospitals). Regarding visual representation, black respondents were more likely to feel represented in images from health departments than in those from hospitals (44.3% of images and 23.7%; p=.05). Both hospitals and health departments were far more likely to use images where white respondents felt represented (hospitals = 76.1% of images; health departments = 59.7%) compared to images where respondents from racial/ethnic minorities felt represented (hospitals = 19.3% of images; health departments = 21.4%) (p<.001 for hospitals; p=.004 for health departments).

Conclusions: The public requires information that addresses COVID-related health outcomes and vaccination rates, including visual representation of those most negatively affected.

Poorva Sheth and Chloe Thompson wrote the entirety of the manuscript with editing guidance from Dr. Rozier.

Dr. Rozier was a key part of the development of this manuscript. While Poorva and Chloe were working on writing the various parts of the introduction, methods, etc, Dr. Rozier was always free to give feedback and reassurances that the research being conducted was noteworthy and valuable. He also pushed Poorva to develop manuscript writing skills by empowering her to narrate the research process itself rather than simply making a table with all of the data points.