Thread by @ClaireBerlinski

It may well turn out to be exactly the same film. So far the evidence for ivermectin is poor quality and limited, but there are genuinely some hints it may be of value. Definitely a good enough reason to study it properly. And if good quality evidence shows it has real efficacy--

https://twitter.com/autkud/status/1411913882893914112

I'll become the biggest ivermectin booster on the planet, and all of it's advocates will say, "See! We told you so." But they told us so based on bad reasoning. They may hold a *true* belief, but it isn't a *justified* true belief,

and thus it is not knowledge, but a hunch--be it a lucky or unlucky one. Some powerful things are working against the hypothesis--among them, that at doses required to have the effect ivermectin does on SARS-CoV-2 cells in vitro, it would cause an overdose in humans;

and we seem to have a number of pretty good studies showing no efficacy at all, which isn't what you'd expect t see if it works. Usually, if something works in the physical world, it works every time (ceteris paribus); medicine isn't quantum mechanics;

if you're getting wildly varied results from giving people a drug, your first thought should be, "What's different about these groups of people, or the way I'm giving them the drug," not, "The studies that don't favor my hypothesis are the product of an elaborate conspiracy."

But none of this means that ivermectin doesn't work and is fated to be the next HCQ: It's possible that in some groups, at some stage of the disease, or administered in certain combinations, it does indeed show a therapeutic effect. If so: GLORIOUS news.

No exaggeration, no sarcasm. The world is desperate. People are dying. There aren't enough vaccines. We need all the help we can get. Do the research, do it well, and *hurry.*

But so far, the evidence for this hypothesis is a) mixed; and b) of poor quality; what's more;

even the most encouraging studies we have do not show anything like the *effect size* of vaccination. Even assuming only the most encouraging studies are correct--a bad assumption, but run with it--

they still don't show that you'd be safer taking ivermectin than you would be getting vaccinated. They show, to the contrary, that you would be three times safer getting vaccinated. So you'd be *insane* if you were one of the world's luckiest people, and had access to a vaccine,

to decline it. All of this is assuming there's a Covid19 pandemic, of course. (And there is.) If I prove that Drug A is safer than Drug B inherently, that's nice, but absent the context of the risk to be mitigated, meaningless: If Drug B works and Drug A doesn't, so what?

So ivermectin could also be safer than mRNA vaccines in some absolute sense, and given we have long-term data on ivermectin and none on mRNA vaccines, I am willing to concede that we have more long-term data on ivermectin.

But this is not the only thing we look at when assessing if a drug is safe. When I buy a box of eggs, I don't give one to someone else and then wait twenty years to see what happens before I eat the rest. I have a lot of experience with eggs and egg-like food;

I know my egg biochemistry and I know the telltale signs of a rotten egg, so I figure, "Probably fine." Not a 100 percent reliable heuristic, but not inherently bad reasoning, either. We have no special reason to think our vaccines unsafe, long-term;

and many reasons to think them safe.

I fully concede ivermectin has an excellent safety profile (except for the whole testicular dysfunction problem, but who cares about healthy testicles if the alternative is Covid19?) The thing is:

The safety profile of a drug is only meaningful in the context of its efficacy. The efficacy is only meaningful in the context of a particular symptom or disease. And this is only meaningful to *you* if you're at risk of experiencing that symptom or contracting that disease

Drug A is only safe in a *relevant* sense if by taking it, you reduce your risk of morbidity and mortality. If Drug A is safe but doesn't reduce your risk of dying of disease X--and if, arguendo, you have a 100 percent risk of dying of Disease X unless you take a drug--

Drug A's safety is irrelevant. If you take the safe but useless Drug A instead of the slightly-less-safe but highly effective Drug B, congratulations: You have successfully, *raised* your risk of getting sick and dying.

"its"

I don't know why I make that mistake so often given that I fully understand the difference and loathe seeing the error. Something about the way I see draft tweets doesn't lend itself to my eye picking it out as it usually would.