Leveraging artificial intelligence and social innovation to reduce disparities in COVID-19 testing among African Americans

Background: Pandemic fatigue reduced motivation to follow protective behaviors and threatened ongoing COVID-19 control. This fatigue, combined with waning immunity, breakthrough infections, new variants, and persistent uncertainty, made it harder to sustain testing in underserved and medically or socially vulnerable populations, including African Americans. Since control strategies depended on improving communication approaches that increased perceived relevance of testing and reduced the influence of misinformation, this study aimed to identify how message design and presentation affect testing motivation among African Americans.

Materials/Methods: The study applied the Capability Opportunity Motivation-Behavior and Minority Health and Health Disparities Research Frameworks and used participatory research methods and artificial intelligence. A design-a-thon was hosted to create deep learning computer animations that delivered COVID-19 testing messages for African Americans in North Carolina. A 1:1 randomized experiment evaluated whether these animations increased testing uptake compared with a control condition.

Outcome/Impact: Study findings identified message approaches that improved motivation for COVID-19 testing among African Americans. Results provided evidence to support communication strategies with potential relevance for other key populations facing pandemic fatigue and misinformation.