Beware of Fake COVID ‘Studies’ – Lifehacker

Photo: (Shutterstock)

Seeking out information about COVID-19 online means wandering through a wild morass of truth, speculation and outright lies. It’s important to start with trusted sources and see what they’re saying, rather than assuming that just anything you come across is legit. Some of what’s out there is misinformation, pushed by conspiracy theorists or by people motivated by political agendas rather than by science. And sometimes they do a very good job of making a shoddy argument look like an authoritative study.

We saw something similar with Plandemic: a nonsensical collection of untruths dressed up with high production values and all the surface signifiers of a Serious Documentary. People who shared it probably didn’t bother to check whether it was legit, because it looked legit.

Something similar is going on now with other faux “studies,” including one on hydroxychloroquine that’s made the rounds on social media and even surfaced on Fox News. It’s, and it’s not a study, but it’s definitely packaged to look like one.

What it actually is, is an anonymous website with many of its sources credited to also-anonymous Twitter accounts. But it wears the garb of science: The charts mirror the style of other epidemiological graphs that have been floating around; the typography smacks of certain scientific journals’ style.

But what’s even sneakier is the way it co-opts the language of science and deliberately misuses words. The phrasing has changed after criticism, but at one point the site claimed to describe a “country-randomized trial” of literally billions of people.

Randomized trials are good, and the bigger the better, generally. If someone had actually conducted, as the site says, “a large trial with 2.0 billion people in the treatment group and 663 million in the control group,” that would be an amazing and incredibly expensive undertaking.

But of course, no one did.

So what is this study?

As many actual epidemiologists have pointed out, all these folks did was to look at COVID-19 death rates in a handful of cherry-picked countries and conclude that the countries that made a “strong decision” in favor of hydroxychloroquine usage had lower death rates.

That’s not a randomized trial of billions of people; it’s a retrospective look at a few dozen data points (each country in the analysis has a death rate and a yes/no on hydroxychloroquine).

A randomized trial, on the other hand, is where the people running the study assign subjects to one group or another, randomly. That is, without letting them choose and without allowing other factors, like demographics, skew the decision of which group they are in.

People in these countries could choose whether they wanted to take the drug or not. These countries’ governments could choose whether they wanted to set a policy or make a recommendation about the drug. Nothing whatsoever about this is randomized. The problems with the study go on from there, and you can read lengthy critiques from the likes of epidemiologist Gideon Meyerowitz-Katz and biologist Carl Bergstrom—the latter of whom literally wrote the book on calling bullshit. (This study is bullshit.)

Why put so much energy into faking a study on hydroxychloroquine?

The website makes every effort to masquerade as a scientific publication, except for a few kind of important things:

  • Publishing it in a place that scientists publish (even a preprint server, which exist to put forth early-stage research for discussion)
  • The authors’ names

The website just directs people to a Twitter account. The site claims the authors include scientists, but without names or other information, there’s no way to verify who’s behind it. Names aren’t necessary to judge the publication on its merits, but accountability is important. Who put this study up, and why?

Remember, hydroxychloroquine has become radically politicized. Hundreds of drugs have been tested on COVID-19 patients this year, and as anyone familiar with drug development could have predicted, most of them don’t work. There are a few exceptions: Remdesivir may slow the virus’s ability to replicate, and dexamethasone seems to boost survival rates among the very sick. That’s great news!

But ever since the President endorsed hydroxychloroquine, rumors of that drug’s effectiveness have taken on a life of their own. The doctor who was passionately preaching about hydroxychloroquine on the Supreme Court steps (yes, the demon sex one) was part of a group reportedly funded by “reopen” PACs. QAnon supporters have created whole story arcs about how the FDA is trying to hide the drug’s supposed effectiveness. One meme I saw in a conspiracy theory Facebook group says “I have an idea! The Trump supporters will get the hydroxychloroquine & the democrats will get the Bill Gates vaccine!” (They think Bill Gates is using the vaccine as a way to microchip the world’s population. Don’t ask.)

It wouldn’t be so weird to collect information about hydroxychloroquine if the drug didn’t keep failing the well-designed trials that are being done to see if it works. Is it possible that there’s a use of the drug that could wind up helping fight COVID-19? Sure, but that possibility is very remote, and getting smaller by the day. As Meyerowitz-Katz writes:

Hydroxychloroquine probably doesn’t work for coronavirus. It doesn’t help people who have severe illness. It doesn’t prevent illness in high-risk groups. There’s no evidence it benefits people with mild disease either, even in combination with azithromycin. There’s still a bit of a question mark over whether HCQ might reduce the risk of getting the disease for people who are low-risk, but the initial results are not looking good.

Now, you can take those results as you like, but it’s worth noting that these are really quite impressive trials. Running a properly controlled randomized study of over a thousand people in less than six months is an immensely impressive work, and the findings are both stark and uncompromising: HCQ probably doesn’t benefit COVID-19, and may be harmful.

The people behind the website—whoever they are—aren’t behaving like scientists with an analysis to discuss. They’re pestering people on Twitter in a bot-like manner, and have affixed a shiny red “under attack” ribbon to the top left corner of the homepage, as if it were some kind of award. Telling a story about being silenced is a classic conspiracy theorist’s ploy for attention; we saw it with Plandemic’s deplatforming, and we’ll see it again.

If you’re not a scientist, you don’t need to pore over a faux-scientific paper looking for logical and scientific inaccuracies. Instead, be suspicious of anonymous work, be suspicious of anyone arguing a little too hard for their pet theory and above all, double-check trusted sources to see what they’re saying rather than assuming that anything with neat typography and pretty charts is inherently scientific.