Trolling case study: HCQ

2020/08/09

Categories: GeekStuff Politics

There is a fascinating thing which I noticed recently in online discourse, and I have found what is, at least until they notice it, possibly a reliable way to distinguish between people who hold a perhaps surprising position, and people who are trying to spread misinformation.

The context of this particular thing was discussions of hydroxychloroquine, henceforth “HCQ”, an anti-malarial drug which some people thought might be an effective treatment of some sort related to COVID-19.

What’s HCQ, and who cares?

This all started with a paper by Didier Raoult, a French scientist who has some history of publishing unambiguously manipulated data. In this paper, it was claimed that treatment with HCQ plus Azithromycin (“HCQ+AZ”) resulted in improved viral clearance, which is to say, patients who had the virus and were being treated got better.

Unfortunately, there were methodological flaws in the paper, and other studies generally didn’t reproduce the results. However, many people claimed that it was effective only when combined with zinc, or with Vitamin C. Note that Zinc and Vitamin C were not in the original paper. However, a number of sources claim that “Vladimir Zelenko” proved “nearly 100% efficiency” for the combination of HCQ, Azithromycin, and also Zinc (“HCQ+AZ+Zn”).

Other studies generally did not seem to reproduce any of these results, but a number of people came forward with new theories, such as “it works but only if you start it very early”.

And then, just to spice things up, Donald Trump claimed to be taking HCQ, and when people asked why, since he hadn’t been reported to have been infected, he said it was preventive.

So, at this point, we have the following claims to evaluate:

  1. HCQ+AZ is an effective treatment for patients who are infected.
  2. HCQ+AZ is an effective treatment for patients who are infected, but only when applied early.
  3. HCQ+AZ+Zn is an effective treatment for patients who are infected.
  4. HCQ+AZ+Zn is an effective treatment for patients who are infected, but only when applied early.
  5. HCQ+AZ is an effective prophylactic treatment for patients not yet infected.
  6. HCQ+AZ+Zn is an effective prophylactic treatment for patients not yet infected.
  7. HCQ+AZ is an effective prophylactic treatment for patients not yet infected, but specifically is NOT effective in treating patients who are infected.
  8. HCQ+AZ+Zn is an effective prophylactic treatment for patients not yet infected, but specifically is NOT effective in treating patients who are infected.

Never read the comments

In comments on articles about studies, proponents of HCQ, unsurprisingly, advance only the claims not disproven by a particular study. For instance, if a study claims to have found no benefit from HCQ+AZ, the responses will contain people saying that, obviously not, because it’s only HCQ+AZ+Zn that is effective. If the study talks about relatively severe cases showing no benefit, the responses will contain people saying that, obviously not, because it’s effective only when given very early. And if the study actually talks about HCQ+AZ+Zn given early, well, then you get people explaining that of course you wouldn’t see any improvement once people are infected, it’s exclusively a prophylactic treatment.

But what about sites where there’s interactions between comments? Here, it gets weird. Say that one commenter makes a reference to a study which found no benefit from HCQ+AZ given early; they will get a response pointing out that, since it’s only believed to work with Zinc, this study is meaningless. And say another makes a reference to a study which used HCQ+AZ+Zn on severely ill patients; they will get a response pointing out that, since it’s only useful when given early, this study is meaningless. And if someone points out that obviously at least some patients have other sources of zinc in existing medications or prescriptions, you’d expect to see an effect anyway, someone will jump in and point out that it was never claimed that HCQ would “cure” COVID-19, only that it was usable as a prophylactic.

Okay, that makes sense. Obviously, you’re only going to get critical responses from the people who disagree with your claims, right?

But here’s the mystery: If you have three people actively scanning a comments section for things they disagree with to respond to, and one of them believes that HCQ is absolutely an effective cure for infected patients, but only when taken with zinc, and another asserts that HCQ is never a cure, only a prophylactic… They will never argue with each other. Even though their claims are mutually exclusive, they don’t disagree.

If a reasonably normal person held the belief that HCQ was an effective cure for COVID-19, and saw someone else saying there was no point at all in giving it to known-infected patients, they would probably consider this a major error to be argued with, as failure to treat people with an effective cure could result in easily-prevented deaths.

So what’s happening?

What’s happening is that none of them believe any of these things. What they actually believe is “you shouldn’t believe that HCQ isn’t useful”. They don’t care why you think it’s useful; the important part is that you are on their side, saying that HCQ is useful, and not on the side saying it’s not. So for any specific position anyone takes that argues that HCQ isn’t useful, they are ready to jump in with a counterargument, but as long as someone is saying there exists a use for HCQ, there’s nothing for them to argue against. In dozens of articles on multiple sites, I’ve seen a huge number of these commenters, and none of them responded to each other, even when their claims were directly contradictory.

Why? Heck if I know.

Two obvious theories come to mind. One is called to mind by the frequency with which the people insisting that HCQ has a use, no matter what any clinical trials say, use phrases like “orange man bad” to describe the contrary view. In their view, since Donald Trump said that HCQ might be effective, any belief to the contrary is motivated entirely by personal animosity towards him, and unscientific, and therefore necessarily wrong. As is typically the case in the current political climate, they have zero interest in such abstract questions as what external reality might be like; they view the question as entirely and exclusively political, and will argue their faction’s side of it no matter what evidence might say.

The slightly more sinister theory would be organized attempts to create confusion and divert resources away from effective research or treatments and towards this particular thing, on the grounds that time spent researching HCQ isn’t spent researching other things, and thus increases the economic harm done to the economies of countries that are vulnerable to this.

This would align curiously with the spectacularly huge study claiming to have looked at results across 85,000 patients, and finding no benefit to HCQ, but which had glaring methodology problems, such as there being no way to confirm that they had access to the data, and a lot of confirmation that they can’t possibly have actually had access to the data they claimed to have. This “study” didn’t actually do much to change the state of things from a research perspective – plenty of controlled studies had already come to roughly the same conclusions for all the different combinations. What it did do was create a really obvious citation for everyone to cite to, which also turned out to be fake, allowing people to claim thereafter that “oh, that study was debunked” whenever anyone suggested that a study had found no benefits to HCQ.

Some useful links for further reading:

There’s a huge number of additional arguments people sometimes advance on the HCQ question. There’s claims about relative death rates in countries based on whether it’s part of a standard-of-care, which tend to be hand-wavy. There’s claims about “lots” of papers showing benefits, but not usually concrete links to specific unretracted papers. And there’s a lot of research that seems to be pretty consistently showing no actual benefit. By contrast, for instance, research on dexamethasone for severe coronavirus cases doesn’t seem to be particularly controversial, and doesn’t attract large crowds of angry people. It will be interesting to see followup studies and reporting on that one.

It’s not just HCQ

This actually happens on a lot of issues. A while back, Donald Trump walked from the White House to a church, where he was photographed holding a Bible. In the process of him getting there, security forces used tear gas to discourage people from being in the way, and among those dispersed were the people present at the church, including the clergy whose church it was.

Among the things I’ve seen people say when rejecting a claim that this happened:

These show the same property: A person who is arguing that the crowds were not hit with tear gas isn’t going to argue with a person who says that they were, but it was okay because they were out after curfew, or vice versa. Neither of them actually cares about the claim they’re making, the only point is to create doubt about the claim they’re attacking.

Whether this is personal political loyalty or organized sabotage, the fundamental purpose remains the same: To create uncertainty and make it hard for people to be confident that external reality continues to exist and admit a distinction between true claims and false claims.