Community Notes effectively targets medical misinformation on X

Social media has proven a breeding ground for medical misinformation over the past decade, particularly during the onslaught of the COVID-19 pandemic, but new research shows some platforms’ crowdsourced efforts to fight inaccuracies are paying off.

In a new JAMA article, researchers from the University of California San Diego’s Qualcomm Institute found that Community Notes, a feature on X (formerly Twitter), has been effective at combatting medical misinformation on the platform.

Community Notes is one of X’s efforts to address false misinformation posted to the platform. Users can attach a Community Note to a post to point out misinformation and provide correct information, including potentially adding sources. The post will then produce a flag with the fact-check.

The effect is to allow users can research, draw their own conclusions, and improve online information literacy via Community Notes.

“Rather than censoring misleading content, Community Notes fosters a learning environment where users can glean insights from corrections to misinformation to prevent similar misunderstandings in the future,” Eric Leas, PhD, a study co-author, assistant professor at the UC San Diego Herbert Wertheim School of Public Health, and a Human Longevity Science and Qualcomm Institute affiliate, explained in a press release.

“By providing context and credible sources alongside contentious posts, the platform empowers users to discern fact from fiction, a skill they will find useful as they navigate all claims.”

Typically, social media platforms are opaque in how they guard for false information posted to their sites. But the researchers said Community Notes offers a unique view into how social media websites are monitoring and flagging medical misinformation because it is entirely open source. This allowed them to study how effective it was at flagging medical misinformation.

The team looked at all notes that contained terms related to vaccines or COVID-19, as well as their corresponding posts, during the first year during which the Community Notes program was up and running (December 2022 through December 2023). That encompassed 45,783 notes, 657 of which addressed COVID-19 vaccination.

The researchers then collaborated with an infectious disease physician and virologist to evaluate the subject, accuracy, and source credibility of a random selection of Community Notes. Around half of the notes in the sample addressed adverse events linked to the COVID-19 vaccine, while 37 percent addressed conspiracy theories, 7 percent were vaccine recommendations, and 5 percent were about vaccine effectiveness.

The group found that Community Notes was fairly effective at flagging and correcting medical misinformation.

Nearly all (97.5 percent) of the Community Notes included in the sample were accurate, and 2 percent were partially accurate, meaning “addressed scientifically debated conclusions.” Only 0.5 percent were inaccurate, the researchers said.

What’s more, Community Notes used good information to back up their corrections.

Around half (49 percent) of the notes used highly credible sources, including primary data sources and peer-reviewed studies. Another 44 percent used moderately credible sources, like major news outlets and fact-checkers, and only 7 percent used low-credibility sources, such as blogs or tabloids.

Nimit Desai, a study co-author and a UC San Diego medical student, said the Community Notes research provided a glimpse into how public health can take action on social media’s medical misinformation problem.

“Notes typically addressed obvious misinformation, offering corrections from credible sources,” Desai said in a public statement. "It's remarkable to witness the online community's adeptness in steering conversations towards accurate and high-quality evidence when provided with the right tools."

Although the study did not entirely broach this subject, study co-author Mark Dredze, PhD, the John C. Malone Professor of Computer Science at Johns Hopkins University, said there could be significant ripple effects for accurate Community Notes.

“Although we couldn’t examine how these notes directly influenced people's beliefs or actions, the characteristics we analyzed have consistently been shown to predict a message's effectiveness,” he said in the press release.

Indeed, the notes included in the study were linked to X posts that average more than a million views, yielding anywhere from 500 million to 1 billion views for all posts related to the COVID-19 vaccine.

This could be a key tool in public health’s toolkit as experts continue to determine how they can address the growing amount of medical misinformation circulating on social media platforms.

"One viable avenue for the public health community to combat this threat is to actively engage in social media-based interventions, such as Community Notes," physician-scientist and study co-author Davey Smith, MD, chief of the Division of Infectious Disease and Global Public Health and professor in the UC San Diego School of Medicine, said in a statement.

"While only a fraction of vaccine misinformation posts are currently addressed, the ample room for expansion suggests significant opportunities to amplify the impact of Community Notes,” added Smith, who is also co-director of the Altman Clinical and Translational Research Institute at UC San Diego, and immunologist at UC San Diego Health.