Retractions for bad science claims, influencing public health

What are publishers’ responsibilities?

As everyone knows, the anti-vaccine movement has been increasingly demonstrating that bad public health claims have real consequences. Diseases that should be well-managed have erupted in ways that threaten never- and under-vaccinated kids, and those vulnerable from cancer treatments or other immune-compromised conditions.

Some of the problems stemmed from atrocious work done by Andrew Wakefield and published in The Lancet. Although it took far too long for Wakefield’s paper, science publishers do have some mechanisms to remove and shun poor quality and unethical literature: retraction. Some retractions are based on other issues — errors in experimental procedures or reagents or reproducibility. In either case, responsible publishers and researchers attempt to set the record straight. Another strategy is the “expression of concern” to flag questionable and disputed claims.

But not all of the anti-vaccine claims stem from Wakefield and the science literature channel. In fact, since his claims were related to the MMR vaccine which did not contain the preservative thimerosal, the fear around mercury-containing compounds pre-dated his efforts to generate fear. Publications outside the science journal channels has had real impact, like Loe Fisher’s book “A Shot in the Dark” and subsequent TV coverage. But internet-available misinformation has amplified unfounded claims and has unfortunate persistence.

Of course, bad claims will never be truly rooted-out of the echo chambers of misinformation. They fester and mutate at stunning rates. But beyond the true-believers, there are people who can be reached with information about corrections. That means we should at least try.

Misinformation on GMOs is distorting public perceptions in precisely the same way as the vaccine dramas played out. As with vaccines, a variety of nonsense peddlers and flawed research has dominated the debate. Everyone has heard the Pew data showing that scientists agree on the safety of GMOs at levels like that of those who respect the evidence of climate change: 88% of AAAS scientists say genetically modified foods are safe to eat; only 37% of the public agrees.


Just as with climate, researchers in the field are most confident: “ Similarly, a 2014 Pew Research Center survey found 88% of members of the American Association for the Advancement of Science (AAAS) and 92% of working Ph.D. biomedical scientists said it is safe to eat genetically modified foods.”

The misperceptions about GMOs have consequences for farming, food security, and food affordability. As with vaccines, making the wrong choice based on the misinformation can end up harming people who think they are making the right call for themselves. But an additional consequence is downstream: preventing access to solutions that improve sustainability ripples out to policy decisions that hamstring researchers and farmers around the world.

Years back, Robert F. Kennedy Jr. authored a piece about mercury in vaccines that was not only irresponsible, but wrong. “Deadly Immunity”’s claims from a respected source with a large megaphone ignited vaccine fearmongers, but caused outrage among scientists and public health practitioners. Eventually. Rolling Stone deleted the article, Salon retracted it.

Similarly, The Atlantic published an item by Ari LeVaux that ignited GMO fearmongers, titled originally “The Very Real Danger of Genetically Modified Foods” (see the URL that has original wording). Ironically, the original scientific journal article spawning this piece had absolutely nothing to do with GMOs, making LeVaux’s claims even more bogus. Quickly, qualified scientists stepped in to address the wild claims that he had made. Emily Willingham asked “Why did The Atlantic publish this piece trying to link miRNAs and GMOs?” and Christie Wilcox wrote about “The Very Real Scaremongering of Ari Levaux”. Researchers are always running into these hair-afire misinformed pieces like first responders into burning buildings, but damage can be done in the meantime.

Since this time, more and more evidence has accumulated that demonstrates claims from the original scientific publication has not been reproducible, and new data only confirms this.

Although the hair-afire title was changed by The Atlantic, and some text to distance themselves from the wrongness was added to the piece by the editors, it remains on The Atlantic website to come up in searches on this topic.

Worse, damage can even metastasize. In a conversation with Ari Levaux last year, we learned this: scientific textbooks have reprinted his piece.

Ari LeVaux claims his article has been reprinted in science textbooks.

What version did the textbooks use? The hair-doused version, with the caveats (which is still wrong)? Or are they propagating the misinformation wholesale? Think about the children getting the wrong information.

What is the responsibility of the author, and of the publisher, in this case? In a recent discussion about his responsibility, Ari LeVaux had this to say to me:

I have one cat. Facts may not be his strong point, explaining several things.

So I think I’m unlikely to see any accountability from this journalist. What can we do to chase down the the publishers? Does The Atlantic have responsibility? The textbook companies? How can we find them?

Retracting “Deadly Immunity” was the right thing to do. The Atlantic should follow suit and pull this piece. They should help us to find out where else the damage was done, as they bear responsibilty for its reach and spread.

If anyone has ideas on how to find the textbook publishers, please let me know. Check your kid’s textbooks for this. I have reached out to the National Center for Science Education that monitors creationism and climate claims in science textbooks, so they may know how to do it. But more strategies are needed to stop this from propagating. Think of the children.

One clap, two clap, three clap, forty?

By clapping more or less, you can signal to us which stories really stand out.