tration even mention it like it's even an issue. But in terms of sizing and risk it's probably 10 to 100 times the danger of just losing some you know data privacy. I mean data privacy is real important but nowhere nearly as important as the influence. It's not even in the same universe.

Well we've got some new interesting stuff on both hydroxychloroquine and ivermectin, which I hasten to note are not approved medications in this country, at least not for COVID. But they're approved for other stuff.

Now hydroxychloroquine got a new observational study published by medRxiv, whatever that is. And they've found that the anti-malarial drug hydroxychloroquine could increase the coronavirus survival rate by as much as 200% if distributed at higher doses to ventilated patients who are severely ill.

Now what do you make of that? Well if you look on the internet you see people dancing in the streets and saying I told you so, I told you so, I told you, I told you. But do you think that they should really be so happy about this?

Well first of all it's number one an observational study. How do you rate the credibility of an observational study which basically is looking backwards versus a controlled study where you're designing a study in advance and then you do the, then you get the data? So which is more credible? The one where you're just looking at what happened on its own or where you're setting up a controlled randomized study where you're looking at data in the future? Not even close, right? The observational study will not be in the same league with a controlled study.

Do you think that the FDA approves medicine based on observational studies? I don't think so. I'll take a fact check on that but I don't believe any drug has ever been approved in the United States, I probably shouldn't say any drug but I don't think that's the process. I don't believe the process says you can be approved with this kind of a study. So how excited should you be about a study that I believe the FDA would not even use?

Secondly what do you make of the fact that it was published in medRxiv? I think that's a preprint place isn't it? I don't think that means it's been peer reviewed. So this would be close to the lower quality of study, right? So does this tell you something you didn't know? Not really, because we had stuff like this before. So I would say this is more of the same, which is not saying that this drug works or doesn't work. I'm not telling you either one. I'm just saying that it's just more low quality data that you need to be pretty skeptical about. You should be pretty skeptical about it. Doesn't mean it's wrong. It's just not that credible.

But what about ivermectin? Well if you're following the Brett Weinstein, he's making an effort to sort of raise the issues that the public can see that although the studies on ivermectin are not the gold standard type that is nailing it down with some near absolutes, but there are lots of studies. And if you did a meta-analysis of them, most of them but not all of them suggest that it works now.

So keep in mind there are studies that show it doesn't work. Did you know that? So they don't all show it works but mostly they do. But you know a majority, solid majority say it works. Now all of the studies have a pretty wide degree of uncertainty. So wide that it could go from it works to it doesn't work. That's pretty, it's a pretty big wide uncertainty in a number of them.

And of course the idea of using a meta-analysis has been criticized. And let me give you some specific idea of why a meta-analysis, which is basically taking all the lower quality studies and looking at them as if they were one big study. And the idea is that any errors in one study would be sort of averaged out and cancelled by the other studies. So as long as they had different errors they might sort of cancel each other out.

But let me give you an idea of what kind of problems that runs into. Number one, Merck, who actually makes ivermectin, who you would normally expect would be very pro their own drug, right? But this is what Merck says in February. So it's just February, not that long ago. About their