Learning to discern information in the time of coronavirus

Every day, a mountain of information about coronavirus COVID-19 appears in the media: origin, symptoms, treatments, vaccines, protective measures… Some media outlets contradict information from others, leaving citizens and even healthcare staff lost amid so much unverified information. We have found a very interesting Twitter thread on the matter that can help us understand the reasons behind all this conflicting information and how to separate the wheat from the chaff.

The solidity of many works regarding the origin, symptoms, treatments, and vaccines of COVID-19 is non-existent, basically because the vast majority of results or theories we see in the media are being pulled from ‘preprints’ in repositories like @medrxivpreprint or @biorxivpreprint These are unreviewed studies not yet published in a scientific journal, and therefore by no means definitive. Every day, countless claims about coronavirus appear in national or international press based on these findings: whether reinfection is possible, whether the virus survives 3 hours on cardboard… And many of these findings have already been refuted or denied in other ‘preprints’. In many cases, results from these ‘preprints’ are passed off as 100% established science and spawn conspiracy theories, such as claims that coronavirus contained HIV fragments because it was manufactured. It was debunked and the author retracted it, but it was already making all the headlines. It has been like this from the beginning: claims that the virus started with someone eating bat soup in China, or that it was the pangolin’s fault. All of this originated from ‘preprints’—which in the worst cases are the scientific equivalent of a Twitter thread—and was never corroborated. But there it is.

Much the same thing happens with amateur epidemiological statisticians. Someone at home predicts 600 million deaths, uploads it to a ‘preprint’ repository, and… presto! Tomorrow it’s in the newspaper. The traditional scientific publishing system—peer-reviewing over months and setting an embargo date—has been blown to pieces during this crisis, with scientists publishing non-stop and media and public hungry for information. We are suffering an information overload—scientific included—without any filter.

Advice: when you read something that catches your attention, wait a couple of days to see if it holds up. The latest one circulating: people with blood type A are more likely to contract COVID-19. The study is based on 200 patients in China, and since the majority were group A and the minority group O—boom, instant conclusions! The exact same thing happened with ibuprofen. They asked COVID-19 patients what medication they were taking, most said ibuprofen—because most people take ibuprofen for everything—and voila: “People taking ibuprofen are more likely to get infected.” Guaranteed panic.

Journals like @NEJM, @JAMA_current or @TheLancet have a dedicated section for coronavirus publications and do fantastic work.

{Thread by: Antonio Villarreal}