Thread by @gro_tsen

🧵 A comparison between hydroxychloroquine and lockdowns. ⤵️

Recently I compared lockdown proponents with crackpots who believe 5G causes covid: I admit I shouldn't have written this. But the comparison with HCQ proponents, on the other hand, works very well. Thread: •1/36

Ⓐ So, first, in both cases we have something which is supposed to work against covid for a simple and not completely idiotic reason (i.e., the idea is, at least, worth considering!). •2/36

In the case of hydroxychloroquine (“HCQ” henceafter), the theory is that HCQ could serve as a zinc ionophore, transporting Zn²⁺ into the cytoplasm where it serves to inhibit viral ARN-replicase. •3/36

In the case of lockdowns (“LD” henceafter), the theory is that LD serve to diminish the number of contacts between people and thus inhibit viral transmission on the population scale. •4/36

Ⓑ In both cases, this simple theory is borne out by simple models: for HCQ, we have in vitro evidence of its antiviral properties, for LD, we have SIR-like models that support their efficacy in reducing deaths. •5/36

Ⓒ In both cases, however, a great deal of complexity is being ignored: the difference between a simple model (cells in vitro, a mathematical model) and a complex system (a living body, a human society). •6/36

E.g., to defend LD on the basis of “reducing contacts” we are ignoring layer upon layer of complexity of human behavior: that making something forbidden may displace it elsewhere or elsewhen, for example, … •7/36

… or that reducing contacts for everyone uniformly may in fact increase mortality because we are preventing the less vulnerable from becoming immune. It's a bit like thinking “patient is loosing lots of blood, let's make them eat salt as this increases blood pressure!”. •8/36

Ⓓ We have a bigshot, multiply awarded, science guy standing behind these two approaches: Didier Raoult for HCQ, Neil Ferguson for LD. But in both cases, their reputation is put into doubt by erroneous predictions and past behavior, scientific and otherwise. •9/36

Didier Raoult has made claims so far about the covid pandemic that turned out to be very wrong, he has signed way too many publications to have even read them, he was even banned from an international journal. •10/36

In other matters, Raoult also expressed skepticism regarding climate change, and has been accused of turning a blind eye (or even favoring) various kinds of harassment in his lab. •11/36

Neil Ferguson has made some incredibly overpessimistic claims about both the present covid pandemic and several past pandemics (mad cow's disease, bird flu in 2005, swine flu in 2009…). •12/36

And he had to resign from the UK's “SAGE” group (Scientific Advisory Group for Emergency) for failing to abide by the rules of the LD he had himself advised and called for. •13/36

Ⓔ Independently of the bigshot guy's general reputation, the main paper used to justify the usefulness of HCQ resp. LD is borderline scientific fraud and should probably never have been published. •14/36

In the case of HCQ, Raoult's team's main paper on the subject is a non-blind non-randomized study in which they conveniently ignored differences between the two arms of the study. See forbetterscience.com/2020/03/26/chl… for more analysis. •15/36

In the case of LD, Ferguson's team's main paper on the subject conveniently postulates that only non-pharmaceutical interventions have an effect on epidemic dynamic, to arrive at the conclusion that lockdowns do. See necpluribusimpar.net/lockdowns-scie… for more analysis. •16/36

Ⓕ On a broader level, the “evidence” we have in favor of HCQ and LD are based on the same kind of intellectual deception: “see, for this patient / country, it worked” conveniently ignoring the fact that the disease might have ebbed for other reasons. •17/36

Either a vague argument that this set of patients/countries was comparable to this other set, or some kind of timing argument (hmmm… how are we supposed to know that LD caused the epidemic to recede and not the general panic which also caused the LD to be decreed?). •18/36

Ⓖ In fact, both in the case of HCQ and LD, we now have some evidence that, in fact, they do not work as advertised (for HCQ this takes the form of counter-studies; for LD, by evidence from countries or regions which didn't have a timely LD and still had a similar peak). •19/36

But in a sense, this is irrelevant: it's for the proponent of a measure to offer clear evidence that the measure works, not for its detractors to offer evidence that it does not. And this is again something common between HCQ and LD: … •20/36

… their proponents are subtly trying to shift the burden of proof: having offered crappy evidence, they say “if you don't like our evidence, please provide some counter-evidence”. But this is not how science is supposed to work: it is up to them to demonstrate efficacy. •21/36

Ⓗ Another common theme between HCQ and LD is the insistence on early treatment. Faced with evidence that HCQ provides no benefit, Raoult retreated by explaining that it only works if administered sufficiently early in the disease. •22/36

Similarly, LD fanatics insist that not only should countries be given strong doses of their treatment, but it should also be administered very early. Various countries have had debates not on whether lockdown was beneficial but whether it was started early enough! •23/36

Ⓘ Perhaps the most important point of all: proponents of both HCQ and LD choose to ignore, dismiss or deny all the negative side effects of their “cure”. •24/36

HCQ can cause long QT syndrome and ventricular arrhythmia, but this has been variously dismissed as well-known, irrelevant, exaggerated, or manageable by HCQ proponents. Dying patients have been excluded from studies! •25/36

Similarly, LD proponents are very keen to ignore all the negative consequences of lockdowns (be they social, psychological, medical, economical or otherwise), and retreat behind the injunction that they need to save lives, as if these consequences didn't also cost lives. •26/36

Basically, HCQ and LD proponents are ignoring one of the cardinal rules of medicine, “primum non nocere” (“first do no harm”) which commands that we fully investigate the negative consequences of any intervention before undertaking it. •27/36

Ⓙ And this breach of medical ethics probably stems, in both cases, from the overwhelming desire to do something in what looks like a hopeless situation. Basically, “we have do do something! (so let's do anything!)”. •28/36

It's really hard to admit that sometimes the best course of action is to do nothing — or rather, that is to say, to stick with basic things which are proven to do no harm even if they provide only limited benefit. •29/36

https://twitter.com/gro_tsen/status/1303726285093564418

So individuals are keen to place their hopes on whatever miracle drug they hear about, or governments on miracle measures, thinking that limited evidence (see Ⓐ&Ⓑ) is better than no evidence at all, and that “we have do do something!”. •30/36

https://twitter.com/gro_tsen/status/1303726286398017537

This is particularly true in a situation where all available remedies (medical: anti-inflammatory drugs and respirators; or epidemiological: hand-washing and masks) seem grossly insufficient. We don't like to feel hopeless, so we prefer the illusion of control. •31/36

This is the reason, I think, why so many people have jumped on the HCQ or LD bandwagon (but very rarely on both: once you place your hopes in one place, you don't really need a second). •32/36

Ⓚ But there's another reason, which is the political push they got, in opposite sides, from authoritarians. HCQ was pushed by Donald Trump, because it was a way to deflect guilt from his utterly damning failure to act under the “it's just a flu” slogan. •33/36

LD, on the other hand, were pushed by different authoritarian governments, who saw it as a convenient pretext to crack down on civil liberties and gain new powers under the pretext of saving lives and public health. •34/36

✱ Now I'm not saying, of course, that there aren't also significant differences between HCQ and LD. But in both cases we have quack remedies hat have been pushed aggressively, despite lack of serious evidence, and in abandonment of the “primum non nocere” principle. •35/36

And what disturbs me is how many people are able to see the light on one side and not on the other, despite the high level of parallel. (In fact, that people tend to believe in HCQ or LD but rarely both is circumstantial evidence of bad faith!) •36/36