South Africa’s health department set up a technical working group during the COVID-19 pandemic to track what people said, asked and worried about, using social listening.
A recent paper examining that system found 964 instances of misinformation across 12 themes and 573 recommended actions across six themes in 91 reports produced between 2021 and 2023.
The research, led by Herkulaas MvE Combrink, whose work focuses on infodemics and the use of artificial intelligence and machine learning to understand how misinformation spreads, examined the Risk Communication and Community Engagement working group’s approach.
In a paper co-authored with other infodemic researchers, Combrink described a circular process: listen, understand the problem, recommend a response, and keep listening.
Public conversations and feedback were tracked across social media, online news, public WhatsApp streams, call centres and community channels.
Regular social listening reports brought this information together to identify concerns, misinformation, rumours and gaps in understanding, alerting health authorities and partners to emerging concerns and providing recommended actions.
What people were worried about
At least two researchers reviewed the information independently so that findings did not depend on one person’s judgment, grouping statements that said the same or very similar things.
More than half of the misinformation identified was directly about COVID, with the rest concerning vaccines and various health-related issues. The reports reflected recurring areas of uncertainty and misinformation around vaccine safety and side effects, trust, treatments and government communication.
The recommendations focused on improving how health information was communicated and on working more directly with communities.
This meant going beyond broadcasting the same national message to everyone, with recommendations including community information sessions using local media platforms and engagement activities involving people within communities who could place complex health information into locally relevant language and context.
Why did the concerns keep appearing?
The study did not track what happened after recommendations were made, and could not determine whether individual recommendations were implemented, how widely they reached communities, or whether they changed what people understood, believed or did.
Combrink noted that concerns may have recurred because information environments change, with new questions emerging, information voids developing as circumstances shift, and misinformation reappearing or taking different forms.
A concern appearing again later does not necessarily mean exactly the same information problem persisted, he wrote.
Combrink identified five principles that appeared repeatedly across the reports. The first is to understand the concern, using social listening to show what people are worried about rather than assuming what they need to know.
The second is to use that information to improve how health information is framed, since general information about COVID may not address a specific concern about vaccine side effects.
The third is to work with trusted people, including community participation and involvement from public figures and other credible voices.
The fourth is to use appropriate communication channels, recognising that something posted online does not necessarily require an online response and that communication can run through community radio, newspapers, faith-based and community organisations, outreach programmes and community health workers.
The fifth is to keep listening because concerns change.
Combrink argued that the purpose of social listening is not simply to build another dashboard counting posts, likes and mentions, but to connect what is being heard to decisions about what needs to be communicated, to whom, through whom and through which channels, and then to listen again to determine whether the information environment has changed.
He posed a series of questions for when a concern is heard: Who is saying this? Why are they worried?
Could this cause harm? Who do they trust?
What language do they speak? Where do they get their information from?
And, after a response, has anything changed?
Combrink wrote that the findings matter beyond COVID-19, as it will not be the last time countries face uncertainty, fear and rapidly changing information, with the next challenge potentially being another infectious disease, a vaccine, a new treatment or an entirely different public emergency.
The unanswered question, he wrote, is what happens between identifying a concern, recommending or implementing a response, and determining whether anything changes in what people understand, believe or do.
Combrink disclosed that he had received funding from the British Academy for 2025 to 2026 to investigate the general impact of social listening and public communication.
Source: The Conversation
Comments