The Meta-Analytical Fixers, Part II: Public Servants or Middling Saboteurs?
The Chloroquine Wars Part LXXVII
I don't have time for this level of garbage, so I won't spend much on it, but something needs to be said. The propagandists vastly outnumber the people of good faith in The Chloroquine Wars, if that is now my general term for this medical authoritarian battle.
I've been forwarded this BBC article multiple times.
It's interesting to see how the BBC presents Gideon "Health Nerd" Meyerowitz-Katz, Kyle Sheldrick, Nick Brown, and James Heathers as something like a team of COVID Superfriends, taking down COVID disinformation, though without anything like a track record to talk about. For instance, when I asked Health Nerd and Sheldrick each about this error-laden meta-analysis claiming a lack of ivermectin (IVM) efficacy, neither had much to say (IIRC, they simply did not respond, though I blocked Sheldrick on Twitter for what I felt was harassment a short time later).
Dr Kyle Sheldrick, one of the group investigating the studies, said they had not found "a single clinical trial" claiming to show that ivermectin prevented Covid deaths that did not contain "either obvious signs of fabrication or errors so critical they invalidate the study".
But the team of COVID Superfriends does have one notch in their belt. Maybe. In a high profile case, some of the COVID Superfriends found extremely obvious flaws in the claimed raw data of a now-retracted paper. The author claims that the data set is not his, and that he will challenge the retraction. The whole story seems fishy to me to the extent that it strikes me as entirely bizarre and unique that such obviously faked lines of data would even exist in a sharable form, much less be shared given that researchers have no requirement to share their data. Also, it seems giggle-worthy that it would take more than one of them to perform that task, but hey, people need friends to push forward with their work sometimes. It would be hard for me to look into the story closer than that.
At this point, I'll note that Brown and Heathers are "famous" for inventing the GRIM test, which is something like an error-checking method you might teach middle school students, and that most any competent statistician with number sense does instinctively, but sadly sometimes applies to actually published research. But to be fair, much of the time it applies, it likely represents something simple like researchers adding or removing a data point [appropriately] while forgetting to update the computation on their spreadsheet because one person in the lab didn't realize the other person didn't automate the computation (I've personally caught such errors, reported them to researchers, and been thanked for it). This is, in fact, exactly the kind of error Health Nerd and Sheldrick raved over when I discussed the "horrid IVM research" with them online. They seemed emotionally unable to let go of any sort of notion that before making damning statements about researchers---working in nations with access to fewer trained biostatisticians---and their results, it might be polite and reasonable to just email them about the problem. Heck, sometimes we just fix that stuff in preprint. Yawn.
I will also note that Heathers stands out as less-than-objective in my mind after this apology piece over The Lancet's publication of the Surgisphere paper. If asked to defend my downgrading of Heathers' objectivity on this level, I would simply refer to my remarks about standards here. If database readouts are not going to ever be checked on the simplest level, including by people in the field who might be able to ferret out phony data or apply a basic sniff test (which Surgisphere clearly did not pass), what the hell is the point of peer review? Under such standards, scientific publication is no longer able to make any judgments, and the process becomes a money-driven battleground.
Now, where to begin with regards to the entirely demolished IVM research pool, as the COVID Disinformation Superfriends would have us see it?
It sounds like Team COVID Disinformation Superfriends state that they can point out flaws in two or four or six of the studies? I'll let Computer Scientist Alexandros Marinos explain.
Aside from all the many good points Alexandros makes, the central thesis of the article is entirely rhetorical. There is an incorrect implication that randomized control trials are the only studies worth examining, which is a claim few statisticians agree with (here and here), all of whom seem to work in or for the pharmaceutical industry (including the media-pharmaceutical complex). There is also the implication that "to show that ivermectin prevented Covid deaths" means "achieved a statistical significant threshold of p < .05" without any kind of further discussion of trial powering, grouping, or the fact that Ronald Fisher, the statistician who invented the methods of inferential statistics used to measure such results, openly and adamantly disagreed with such rhetorical use of statistics.
It's sad and disturbing how so many doctors don't understand the history behind the forms of pseudo-statistical brainwashing they're fed in order to capture them into reading medical research statistics like code to be injected into their brains, but that's another story for another day.
A few weeks ago I tangled with Health Nerd on Twitter, not knowing anything about him except that he wrote an almost entirely incomprehensibly incorrect thread about statistics of ivermectin studies.
The most amazing part was when, backed into a corner of impossible defense, he claimed that there is no such thing as a study confounder local to a single study in a research pool. In other words, all study confounders are necessary global! And when offered the chance to convince me of that on camera, he dodged. Then he and Sheldrick went on an absurd campaign to state publicly as many times as possible that it was me who dodged, but I'm happy for as many people as would like to read the thread and decide. Yawn.
It will be interesting to keep an eye on his team of Superfriends to count the number of confounders they spot in one, but not all of the studies.
Remember the Fake Oklahoma Story of Ivermectin Patients Flooding Hospitals?
That story was quickly unraveled by numerous people.
Further demonstrating his clear good faith in checking the evidence, Health Nerd was among those completely drawn into the medical-partisan warfare over the strange tale of IVM overdoses filling ER rooms in Oklahoma. I kept wondering who would even believe such a story, but this answers that question.
If you're looking for some way to understand the BBC's level of credibility in all this, take a look at how the BBC handles the IVM overdose story in their "superfriends take down ivermectin studies" article. The article mentions, without noting the way Health Nerd helped push that false narrative, that increases in poison control calls over ivermectin never amounted to much. However, note their last sentence in this paragraph contains no citation, much less an attachment to ivermectin usage (italicized emphasis mine, only), which is to say they're ambiguously mentioning some terrible things people experience in the same paragraph without making it clear that they've researched the problem in a way that attaches these side effects to anyone's actual use of IVM.
Calls over suspected ivermectin poisonings in the US have increased a lot but from a very small base (435 to 1,143 this year) and most of these cases were not serious. Patients have had vomiting, diarrhoea, hallucinations, confusion, drowsiness and tremors.
The naked insinuation is disturbingly irresponsible, and beneath journalist ethics, or else plain stupidity. I'll let you decide. But it gets worse...
Ivermectin and Peru
Peru is one of the nations that thrust IVM into the spotlight. Sure, a handful of doctors and medical scientists were talking about it for months, including Dr. Antonio Cassio Prado, who stated last year in this video that none of the 3,000 Brazilians he gave IVM to got sick with COVID-19 in Sao Paulo---a city notably ravaged by contagion. But these stories don't get out in the media, easily. But around the middle of 2020, while suffering one of the world's worst COVID-19 outbreaks, Peru turned to IVM to bolster its treatment arsenal. As TrialSiteNews reported, Peru saw a steep drop in COVID and deaths from that point---a trend that reversed itself once Peru's new President, Francisco Sagasti, restricted the use of IVM.
Multiple research papers emerged from Peru during that time that seem to have never made it out of preprint.
Dr Patricia Garcia, a public health expert in Peru, said at one stage she estimated that 14 out of every 15 patients she saw in hospital had been taking ivermectin and by the time they came in they were "really, really sick".
The BBC article also goes on to claim a co-ordinated harassment campaign of doctors who do not prescribe IVM, but provide no evidence at all to back that claim up. The vacuously presented statement follows a discussion of pro-IVM Facebook groups, so the implication seems to be that these communities are frenzied mobs, making the lives of physicians more difficult by the day during these hard times.
Back to Peru: Let's remember that Peruvian medicine is a complex machine given the way U.S. medical and intelligence authorities have participated in forced sterilization programs there. The politics in Peru involve an extremely nasty underlying battle for the future, sometimes pitting Peruvians with European ethnic roots against the native population. This means that among medical personnel, you can find those conscripted in that battle willing to say anything, particularly when paid. And while I didn't mention that while writing about The Meta-analytical Fixers (the two lead authors on those papers immigrated from Peru), it was certainly on my mind.
Maybe They're Right
For all my critique of Health Nerd's COVID Disinformation Superfriends team, I'm not entirely certain they'll be wrong. I will admit openly that I have been concerned since I discovered how the Surgisphere team posted an IVM study to preprint early during the pandemic, then later retracting it, that there might be an organized operation to produce and then dismantle fake studies with the goal of casting doubt on their efficacy. It wouldn't be the first time.
That still wouldn't tell us anything new about ivermectin's efficacy in treating COVID-19 patients.