Why health misinformation spreads faster than the correction

A correction almost never catches the claim it is chasing. The most-cited evidence for this is Vosoughi, Roy and Aral's 2018 Science study of rumour cascades on Twitter, which found false news travelled roughly six times faster than true news and was about 70% more likely to be retweeted (https://pubmed.ncbi.nlm.nih.gov/29590045/). The mechanism the authors identified was not bots. It was people: novel, surprising claims provoked stronger emotional reactions, and those reactions drove sharing.

That finding matters for health specifically, because health claims carry the two ingredients that make a post travel — personal stake and moral charge. A post about a treatment being hidden from patients is not simply a factual error. It is an accusation, and accusations recruit outrage.

This is why UTA Intelligence separates two things that are usually collapsed together. Public-health harm asks what happens if someone acts on the claim. Spread mechanism asks why the claim moves. A low-harm claim with a high spread mechanism still deserves attention, because it trains audiences to distrust the source rather than the specific claim.

The scale of exposure is not marginal. Ofcom's Adults' Media Use and Attitudes research reports that around four in ten UK adults encounter misinformation in any four-week period (https://www.ofcom.org.uk/media-use-and-attitudes/media-habits-adults/adults-media-use-and-attitudes). A WHO-linked systematic review found that up to 51% of vaccine-related social posts examined were misleading, and linked exposure to increased hesitancy and to promotion of unproven treatments (https://pmc.ncbi.nlm.nih.gov/articles/PMC9421549/).

The practical implication for anyone doing outreach is uncomfortable. Publishing a correction after the fact competes on the losing side of the speed gap. The interventions with better odds are the ones that arrive before or alongside the claim: prebunking the recurring narrative rather than the individual post, equipping trusted clinicians who already hold the audience's attention, and targeting the topics where outrage load is highest rather than the ones where the factual error is most obvious.

Every figure above links to its published source. Where a number is indicative or synthesised rather than directly measured, we label it as such on the dashboard rather than presenting it as a finding.