Sorry Tucker, ivermectin still doesn't work

Who knew that Tucker Carlson was a fan of science from Iran, Iraq, and Egypt? It turns out that much of the evidence supporting Tucker’s favorite Covid-19 treatment relies on deeply flawed studies from those countries.

But let me back up for a minute first.

I wasn’t planning to write about ivermectin again. Just two months ago, I explained in some detail that ivermectin simply doesn’t work. I also pointed out that prescriptions for ivermectin have exploded, largely due to relentless promotion of it by conservative media hosts including Tucker Carlson, Sean Hannity, and Laura Ingraham.

Official sources including the FDA and the CDC have been trying to explain that ivermectin is useless, but they have been far too timid, as I explained before, issuing press statements that, while clear enough to trained scientists, sound very wishy-washy to the general public.

So what’s new? Well, two new articles have appeared that explain a bit more about where all the confusion comes from. The first is a scientific paper, still in preprint form (not yet peer-reviewed), by the scientists who published one of the recent studies that initially reported positive results about ivermectin.

In their preprint, Andrew Hill, Manya Mirchandani, and Victoria Pilkington re-analyze the studies that went into their recent meta-analysis–which, I should add, they have already retracted. Let’s look at their new analysis.

In their original study–one that promoters of ivermectin often cite in support of their arguments–Hill and colleagues looked at data from 24 different studies of ivermectin, and concluded that, on average, it did seem to reduce mortality from Covid-19.

But then they found that some of these studies were fraudulent, and others were so heavily biased that they didn’t trust their own findings. They retracted their own study just a month after it was published. Kudos to them!

Now they’ve gone even further. They went back and re-graded those 24 studies based on how biased or otherwise flawed they were. Their new report explains that there’s a very strong correlation between the amount of bias and the apparent benefits of ivermectin. In other words, the more biased the study, the greater the benefit. When they excluded the biased studies, the benefits simply disappeared.

In other words, the data from the high-quality studies show that ivermectin doesn’t work.

The lowest-quality studies were conducted in Iran, Turkey, and Iraq, and the worst of all were two potentially fraudulent studies out of Egypt and Lebanon. (In the Egyptian study, it was revealed that “79 participants were duplicates, some deaths were recorded on dates before the trial had started and instances of plagiarism were identified in the text.”)

So apparently Tucker Carlson, Fox News’s biggest promoter of ivermectin, prefers studies from Iran, Iraq, Egypt, and Lebanon over much higher-quality studies out of, for example, Colombia (Lopez-Medina et al., published in JAMA), India (Ravikirti et al.) or Brazil (Bermejo Galan et al.).

But why the heck is anyone getting medical advice from Tucker Carlson?

The second new article appeared in The Atlantic last week, written by James Heathers, a data scientist who has uncovered flaws in multiple studies of ivermectin. Heathers’ summary of the problems is terrific, and I highly recommend that readers check it out, but I’ll share a few tidbits here.

Heathers describes the process of digging into the details of these studies, which is difficult and thankless. He’s not paid to do this; he’s just a good scientist who cares about the integrity of the literature, and it obviously bothers him when he sees shoddy science that has passed through peer review. “We do it because we feel it should be done,” he writes.

One of his main points is that bad studies get published all the time, and they simply remain out there, because who has time to go through all of them? Most of them don’t matter, and they are quickly forgotten.

In the case of ivermectin, Heathers found flaws in study after study. More tellingly, he observed the same phenomenon that Hill, Mirchandani, and Pilkington found: the biggest benefits of ivermectin appeared in the worst studies. As Heathers explains:

“the studies we are certain are unreliable happen to be the same ones that show ivermectin as most effective.”

Why did this happen with ivermectin? In retrospect, it’s not surprising: some doctors and scientists really wanted it to work, they let that bias creep into their studies. Maybe they ignored the patients who got sicker after ivermectin treatment, or maybe they made sure the sicker-looking patients got the placebo. (These would be big red flags if they happened.) Or maybe they had outside motivations–political pressure, for instance–to produce a positive study about ivermectin.

The peer-review process is supposed to catch these problems, and it’s slowly catching up now, as reflected in the new review from Hill et al. But as Heathers points out,

“Publishing science is slow; highly contagious diseases are fast.”

Meanwhile, the data pretty clearly show that ivermectin is useless against Covid-19. We desperately need effective treatments, and we should stop looking up this blind alley.

Before concluding, I’d be remiss if I didn’t point out (again) that ivermectin is a genuine drug, used to treat infections by two parasites: a nematode called Onchocerca volvulus that causes river blindness, and an intestinal nematode called Strongyloides stercoralis. It’s highly effective against these nematodes, which are sometimes called worms because they are worm-shaped, even though they are microscopic.

Unfortunately, people continue to take ivermectin, so let me say it again here: ivermectin does not work. It doesn’t treat Covid-19, it doesn’t prevent Covid-19, and it doesn’t reduce your risk of severe illness from Covid-19.

Oh, and one more thing: if you’re looking for medical advice, don’t get it from Tucker Carlson.