It’s important to address two criticisms of our work. The first is that we exaggerated the FDA’s warning on ivermectin. The second is that Merck’s stance on ivermectin proved that even the company that developed ivermectin thought that it doesn’t work for Covid-19.
First, we didn’t exaggerate the FDA’s warning on ivermectin. Instead, the agency changed its website after our article was published, probably to reflect the points we made. Second, Merck had two incentives to downplay ivermectin’s usefulness against the novel coronavirus. We’ll explain both points more fully.
This is from Charles L. Hooper and David R. Henderson, “The FDA’s War Against the Truth on Ivermectin,” American Institute for Economic Research, October 18, 2021.
Read the whole thing.
READER COMMENTS
john hare
Oct 20 2021 at 4:45am
I find it interesting how the incentives line up in your article. I was unaware that companies could get hammered for recommending other uses for existing drugs even though reading about it in passing before.
Interesting how the generic drug rules incentivize downplaying vulnerable products. I would be interested in following the incentives to a way of using what is available instead of the latest/greatest/most expensive solutions. I was once prescribed a $1,100.00 drug that was changed for a $20.00 drug when I was cash customer and balked.
JFA
Oct 20 2021 at 7:08am
My understanding is that many of the more “high quality” trials (in humans) showing ivermectin was useful against Covid were retracted or shown to have many data irregularities.
Alan Goldhammer
Oct 20 2021 at 8:53am
You are correct, there is no convincing evidence that ivermectin works against Covid-19. The original hypothesis came from an Australian in vitro screening study. This prompted a lot of excitement until a close look at their data showed that the concentration needed for viral inhibition was TEN TIMES the serum level of the drug that can be achieved. Thus, even very high doses of ivermectin will not lead to a therapeutic dose based on the Australian data. Those researchers later acknowledged this (Dylan also writes about this problem in a subsequent post). Secondly, nobody has come up with a plausible mechanism of action for the drug. There were a large number of screening experiments done in the early days of the pandemic to see if there were possible anti-virals that could be used for treatment. Nothing came of this and a lot of people took drugs such as ivermectin and hydroxychloroquine that did not provide any treatment effect at all.
MarkW
Oct 20 2021 at 1:32pm
While that may be true, it is no excuse for (clearly politicized) lying on the part of the FDA.
Charles Hooper
Oct 20 2021 at 3:04pm
Alan,
Can you comment on the clinical studies I referenced?
Charles Hooper
Oct 20 2021 at 2:17pm
That has been the common understanding. Unfortunately, for many reasons, that idea is just plain wrong.
There are scientists who tally published results from clinical trials for therapeutics such as ivermectin. With ivermectin for COVID-19, there have been studies in 37,558 patients, even after removing all the studies that were retracted or had data irregularities. Even considering only randomized, controlled trials, the number of patients tested still hits 4,583. In addition, the peer-reviewed studies of ivermectin have included 17,126 patients.
Gilead Science’s Veklury (remdesivir), which has received an FDA approval for COVID-19, has been tested in 81,264 patients (once the bad studies were removed). That’s about twice as many patients as for ivermectin but consider the results. Veklury has shown an improvement in 13 percent of patients versus ivermectin’s 69 percent.
The WHO still doesn’t recognize ivermectin as a useful therapeutic for COVID-19 even though the WHO does recognize ivermectin for two other indications (other uses) that were tested in a total of 591 and 852 patients, respectively.
When I worked at Merck & Co. years ago, Vasotec (enalapril) was one of the company’s biggest drugs. It was tested in 2,987 patients before FDA approval. Mevacor (lovastatin), another of Merck’s big drugs at the time, was tested in 6,582 patients in the EXCEL Study. At the time, that was thought to be a massive trial.
Ivermectin has been more thoroughly tested than Vasotec and Mevacor and its results are better than for Veklury. What gives?
For more, see: https://c19early.com
Dylan
Oct 21 2021 at 1:14pm
Charley,
This comment, taken on its own at least, doesn’t impress. Surely you know, the number of patients in a trial is just one factor in the trial power, the other one being effect size. Larger effect size means you can get by with a smaller trial size. I don’t know the details for the example trials you give, but the fact that you don’t even mention trial power or effect size in your argument doesn’t give me confidence.
As for the site you linked to, here’s a good examination of the many things that are wrong with relying on it.
https://arstechnica.com/science/2021/09/the-anonymous-meta-analysis-thats-convincing-people-to-use-ivermectin/
Charley Hooper
Oct 21 2021 at 8:35pm
Dylan,
I read the article in the link you referenced.
If you believe what was written in that article, then you should:
Never believe any clinical trial results
Entirely discount the field of econometrics
Why clinical trials? Because the patients in the active and control groups are never the same and they never get exactly the same treatment. If 50 are given ivermectin and 50 are given a placebo, how likely is it that the 50 in the one group are going to be identical to the 50 in the other group? Answer: epsilon. Plus, one person might have been treated slightly earlier than another or with a slightly higher dose. Maybe one person was taking vitamin D supplements during the trial and another wasn’t.
Why econometrics? If an economist does a study looking at the value of a college degree, what’s the chance that all those people were the same age, had the same work background, took the same major, had the same grades, were the same sex, worked in the same field, etc.? Answer: epsilon.
The field of econometrics works because it asks questions such as: Of all those in the last X years who have received an undergraduate degree, what is their annual income compared to those who didn’t?
What the c19early.com people are doing is the same thing. Of all those patients studied in a clinical trial for COVID-19 in which ivermectin was a treatment arm, what percentage of those who got ivermectin went on to need a ventilator compared to those who didn’t get ivermectin?
Are you saying this isn’t a legitimate analytical exercise?
The answer, by the way, is 27 or 1,131 ivermectin patients moved onto a ventilator while 49 of 977 in the control group did.
JFA
Oct 22 2021 at 7:04am
“Of all those patients studied in a clinical trial for COVID-19 in which ivermectin was a treatment arm, what percentage of those who got ivermectin went on to need a ventilator compared to those who didn’t get ivermectin?”
I think the point is that the exercise the c19 website does not take into account any confounds. In fact, the whole edifice of econometrics has been built in order to take into account the various problems that your question poses: omitted variables, selection bias, measurement error, etc. Presumably you’d at least want take into account whether the results come from randomized trials and separate out those results and also to present results based on whether ivermectin was the only drug used.
I appreciate the data visualization, but this is not really a good attempt at meta-analysis.
Dylan
Oct 22 2021 at 12:28pm
I agree, we should be skeptical of clinical trial results and econometrics! But, if we’ve learned anything from John Ioannidis, meta-analysis is hard to do right. As JFA mentions, you need to at least try to control for confounding variables and differences in methodology. From what I can tell, the people behind the site haven’t done any of this.
Dylan
Oct 20 2021 at 8:00am
All of these words, yet you don’t acknowledge the biggest criticism I saw after your last article, which is that the therapeutic concentration needed to get the result seen in vitro are impossible to reach in humans, and are far, far above the dose approved for anti-parasitic use?
Epigen
Oct 20 2021 at 8:45am
This has never struck me as a particularly damning point. Cell culture is…cell culture. We use highly abnormal cells (often cancer derived, usually with many mutations) adapted for continuous growth on plastic in non physiologic media because it easy and relatively cheap to use this approach to rapidly screen compounds for various activities. Sometimes this assay system picks up interesting compounds, but most of the time the in vivo data is wildly different than what was observed in cell culture. Inhibitory concentration in vito is one consideration taken into account when identifying good candidates for future study, but it’s not the only one. Ivermectin was well studied, safe and cheap enough that some people thought why not test it in people, and they did. The question isn’t whether or not the cell culture system required concentration rules out ivermectin usage but whether ivermectin works in human to treat covid. If it does work in humans, then it just means our drug and cell culture assay is behaving (shockingly, I know) differently than our drug and humans tests.
Dylan
Oct 21 2021 at 1:22pm
I agree, by itself, the in vitro data isn’t a slam dunk against ivermectin. But, when people are using these studies to support the proposed mechanism of action…without mentioning the therapeutic concentrations required are impossible – and attempting to get close to that level is dangerous (note that we have at least one overdose death)
Combined with the fact that clinical trials have not shown a benefit either, and yeah, it seems pretty damning. I honestly don’t understand how which drug works or doesn’t work against Covid has gotten so politicized.
Alan Goldhammer
Oct 20 2021 at 8:53am
Quite right and even the Australian researchers who first published on ivermectin have acknowledged this.
Charles Hooper
Oct 20 2021 at 2:18pm
Please see my comment to JFA above.
AMT
Oct 20 2021 at 9:20pm
Which says absolutely nothing about the therapeutic concentrations required? Great response! I’m sure all of those studies in that c19 link are extraordinarily well done…
https://www.bbc.com/news/health-58170809
Unfortunately that BBC article doesn’t provide any links, but there are many sources saying mountains of these “studies” are basically garbage or outright fraudulent.
https://ebm.bmj.com/content/early/2021/05/26/bmjebm-2021-111678
https://www.nature.com/articles/s41591-021-01535-y
https://cosmosmagazine.com/health/covid/data-detectives-ivermectin-studies/
And yeah, of course these authors are completely impartial:
Lol
Charles Hooper
Oct 21 2021 at 3:52pm
You raise a good point, that we need to be careful to consider only good studies.
Please keep in mind that there is just as much bias against ivermectin as for it. For example,
Note that the López-Medina study still showed a large positive benefit while Together showed a smaller, but still positive benefit.
The BBC article you referenced states:
That statement about the Together Trial is incorrect for two reason. First, the Together Trial was of poor quality.
Second, the Together Trial found that while 22 or 678 patients in the control group died of COVID-19, only 18 of the 677 patients in the ivermectin group did. That looks like ivermectin is helping some patients (i.e., the patients on ivermectin were 18% less likely to die).
The BBC article you’ve linked to has been attacked by a number of researchers for inaccuracies, which I’d be happy to dive into detail if you’d like. Let’s just say that even if the studies the BBC complained about were removed from the ivermectin meta-analysis, the results don’t change much. In fact, most meta-analyses have already removed those studies.
Further, while we should exclude poor studies, the bad behavior of some does not implicate all. For you to make your point, you’ll need to show that ALL 63 of the studies are bad. That includes 63 studies of 47,461 patients by 625 authors and, if you want to focus only on peer-reviewed studies, 44 studies of 17,126 patients by 466 authors.
I look forward to seeing your report.
AMT
Oct 21 2021 at 11:57pm
You don’t seem to be grasping that the BBC article wasn’t just talking about a couple of the ones that are already retracted.
“more than a third of 26 major trials of the drug for use on Covid have serious errors or signs of potential fraud. None of the rest show convincing evidence of ivermectin’s effectiveness.”
You can’t be left with much after you take a huge fraction of the major trials away…especially since that’s just what’s been discovered to be untrustworthy so far.
I have no idea where you are getting your quotes from, so yes it would be nice for you to provide details.
Charles Hooper
Oct 22 2021 at 5:22pm
AMT,
I had to reply to myself because I couldn’t reply to you. I have two points about the BBC article. First, it is full of inaccuracies. It claims to be a fact-checking article but the fact-checkers have substantial complaints about it. Here’s one:
https://trialsitenews.com/how-ivermectin-became-a-target-for-the-fraud-detectives/
Second, even if the studies referenced in the BBC article are removed from the analysis, ivermectin still appears to be an effective drug. Based on 47 studies, ivermectin showed an 84% benefit in prophylaxis, 73% in early treatment, and 45% in late treatment. Those 47 studies include 37,558 patients.
Conclusion: the BBC article is error-ridden and, ultimately, irrelevant to the final conclusions.
Dylan
Oct 20 2021 at 10:27am
I thought today’s XKCD on clinical trials was timely
https://xkcd.com/2530/
Henri Hein
Oct 21 2021 at 1:24pm
The likelihood whether Ivermectin is effective in treating Covid is not the main issue. The real issue is misguided top-down control imposed by FDA. It is the attitude that people should not be allowed to experiment, even when faced with a serious disease for which no other treatments are available. Of course, we should not experiment on unwilling subjects, and we should not lead people away from more-promising treatments to less-promising treatments. Those were not issues that FDA were addressing. They were pointlessly telling people to refrain from trying a drug we know to be safe. When ~2000 people are dying every day, and we don’t know how to stop it, let the people experiment. It’s the right thing to do, since individuals have inherent sovereignty over their bodies. Not only that, but the ones that are willing to experiment gives us data on which of the off-label treatments, if any, actually are effective.
Dylan
Oct 21 2021 at 2:34pm
Where is the FDA not allowing experimentation? There have been numerous clinical trials run and doctors remain free to prescribe ivermectin off-label if they think it would help. All the FDA did, as far as I can tell, was warn that there is not strong evidence that it works and tell people that it can be dangerous, particularly at the doses that some have been taking.
There is a difference though between clinical trials and individuals experimenting on themselves. In the former, we have the hope of generating new knowledge, I think the FDA is well within its purview to warn you of the dangers of the latter.
Henri Hein
Oct 22 2021 at 2:04pm
In the original warning, FDA stated outright: “You should not use ivermectin to treat covid-19 symptoms.” (Paraphrasing from memory, but it was something close to that.) I assume that would make some patients refrain from considering ivermectin, and many doctors refrain from recommending it.
Dylan
Oct 23 2021 at 8:00am
Well yes, and that is good advice. But I was using experimentation in the context of actual scientific experiments where knowledge can be generated.
You can certainly take the position that individuals should be free to imbibe any substance they want, and I’m certainly sympathetic to that view…but it’s not experimentation. And, in this case, people are more free than normal to do that (and have been!) The FDA recommending against it, pointing out the fact that there is not good evidence for efficacy and real chance of harm (at the doses people are taking) feels pretty on point for their mandate and way less heavy handed than a lot of other things they do.
Juan Manuel Perez Porrua Perez
Oct 23 2021 at 3:26pm
The American Institute for Economic Research? Really? Unironically?
That think tank has pretty much squandered what little credibility it had left after the Oregon Petition.
David Henderson
Oct 23 2021 at 4:14pm
You wrote:
Yes, yes, and yes.
You wrote:
Please explain.
David Henderson
Oct 23 2021 at 4:23pm
Further thought. Do you recognize that one can judge an argument independent of where it was published? If not, what criteria do you use other than the cogency of the argument?
Juan Manuel Perez Porrua Perez
Oct 25 2021 at 4:20pm
Appearances matter, too. I’m just saying even token effort at objectivity is much appreciated.
There are dozens of think tanks with the same interests and agenda as the AIER whose credibility hasn’t been (so much) damaged by fiascos like the Great Barrington Declaration or Oregon Petition against the Kyoto Protocol (a document promoted by AIER which also had problems with signatories with dubious credientials and false identities)
David Henderson
Oct 25 2021 at 5:59pm
You wrote:
I interpreted you to be saying way more than that. I took you to be saying that we can’t trust an argument because of where it was published. Are you now saying that I interpreted you wrong?
Juan Manuel Perez Porrua Perez
Oct 26 2021 at 3:23am
I’m saying that by publishing your arguments through an institution such as the AIER, which has proven not to be an intellectually honest actor, you taint them with their lack of credibility. And anyone would be more than justified in dismissing them outright. So yeah, you interpreted me correctly.
Comments are closed.