The efficacy of Hydroxychloroquine and z-pack.
The Coronavirus, not the beer
- Thread starter Runnin
- Start date
The efficacy of Hydroxychloroquine and z-pack.
Metaphysicist
Not Actually Brian Hunter
It's only a big deal if you are desperate for confirmation bias. Look how thethe immediately perked up, reinvigorated in his trolling.
Its not a big deal that we are confirming the first death was a month earlier than reported?
Is this a serious position you are taking right now? Talk about bias here
Metaphysicist
Not Actually Brian Hunter
Assuming you mean the one from the AP News article? Okay, I'll look at it.
Seems like a normal study where they don't find particularly significant effects either way. No evidence outcome were better with the drugs. Supposedly higher deaths with the drug, but the confidence intervals on that are pretty wide (~1 - 6 times the deaths). Looks like they adjusted for the more severe cases being more likely to be prescribed the drug, but the uncertainty in doing that probably adds to the wider confidence interval (in addition to just limits on sample size). No idea how they actually calculated things.
I don't think it is dispositive on whether it might help some people, but it seems like pretty good evidence it isn't some kind of wonder drug. Don't know what PSM means. Don't know what buried "lead" you are referencing because I don't read the whacko-sphere.
Metaphysicist
Not Actually Brian Hunter
The article you posted says a known death in the same county 11 days later. No, 11 days earlier doesn't blow my mind, since that is still weeks after we already know people had it.
nsacpi
Expects Yuge Games
that's your problem
I'm not going to pretend to be an expert (Its clear that I am not) but I can research enough to determine that the methodology they used to 'normalize' the fact that sicker people naturally sought out the use of the drugs introduces more bias into the study.
The lead was the drug actually increases fatality. Sure, the raw number showed that but its not even close to the conclusion that was made by the study.
This works on simpletons as shown in this thread (NOT REFERENCING YOU) but when you open the hood for just a second you understand that this study has almost as much value as used TP.
There is a morbid vested interest to shoot down this treatment and its disgusting.
This study is basically a throw away.
The first confirmed death was beginning of February. That implies infection in the beginning of January and unless you want to believe the FIRST person to be infected died then its very clear the infection was here in late 2019. How late is hte question.
nsacpi
Expects Yuge Games
I think you are overrating the importance of this.
In its early phases there is a lot of randomness to the course of an epidemic. Some infectious people don't infect anyone else. Those first few infections can die out.
But at some point a version of the law of large numbers kicks in. Once you have several thousand infections, the rate at which it spreads becomes predictable.
I agree 100%. THat is why the number of deaths in the makeshift model I made had less than 200 deaths before March even with a seed infection mid Q4 19.
But understanding the start helps you understand how many infected there are and compare it to confirmed and fatalities to derive a true fatality rate.
Metaphysicist
Not Actually Brian Hunter
I mean, they adjusted for this, which is a pretty standard thing to do. It's possible their "methodology" was skewed somehow when they did this, but the mere fact of adjustment doesn't make the study prima facie TP. For example, they did all sorts of similar adjustments in the antibody studies you've been hyping. I'm skeptical that you know nearly enough about this to determine a good methodology from a bad (and the paper didn't really even go into much depth on this).
Actually, the raw number was not the most interesting part. What stood out, to me at least, was that the death rate for people who ended up on ventilators was noticeably higher.
I disagree. The French study you jizzed your pants about and claimed had a "100% cured rate" used way fewer people and showed a lot less than this did.
Last edited:
The french study was only better because they actually administered the treatment as opposed to doing a retrospective analysis of people that got treatment and requested medication conceivably due to their own condition.
nsacpi
Expects Yuge Games
NY state has already seen slightly more than 1 COVID death per 1,000 population.
Double checking on my decimals, that works out to 0.1% of the population.
What is the % of the population that is infected? Remember we are talking state not city here.
The worst hit parts of the city have been returning 50% positive on the people being tested. Presumably this group has a reason to be tested. Those not being tested very likely have a lower positive rate even in the worst hit parts of the city.
For the state it will be even lower. I'd say around 20% of the state has been infected. So if .1% of the whole state has died of COVID, this would imply about .5% of the infected state population. Could be a little higher, could be a little lower. But for now this is our best estimate of how deadly it is.
Double checking on my decimals, that works out to 0.1% of the population.
What is the % of the population that is infected? Remember we are talking state not city here.
The worst hit parts of the city have been returning 50% positive on the people being tested. Presumably this group has a reason to be tested. Those not being tested very likely have a lower positive rate even in the worst hit parts of the city.
For the state it will be even lower. I'd say around 20% of the state has been infected. So if .1% of the whole state has died of COVID, this would imply about .5% of the infected state population. Could be a little higher, could be a little lower. But for now this is our best estimate of how deadly it is.
I'm actually dissapointed they are doing a NY State test and not NYC.
Metaphysicist
Not Actually Brian Hunter
Not sure where you are getting "this must take a month+" to kill someone (presumably an old?) who doesn't receive medical treatment (they died at home).
I mean, I will definitely grant you that this means the virus was spreading in the community to some degree by at least mid-Jan. Like, these earlier deaths make that undeniable, whereas before it was at least a sort of open question. But I'm not really sure what that changes, ultimately.
If they find a death from like Jan 6... well then, yeah, now we are into 2019. (though again, not even sure what that would ultimately change)
I appreciate your response here.
It is not conclusive to say it was spreading late 2019 yet until we can get a similar autopsy result so I am stretching to fit my bias.
In terms of what it would change...That gives more credence to the idea that we have much wider spread than thought which would lower fatatlity rate.
Metaphysicist
Not Actually Brian Hunter
If you mean that the French study didn't even have a control group to have to adjust against, they just gave it to 80 people, then sure "that's better."
Metaphysicist
Not Actually Brian Hunter
Not really. The only thing that can tell us anything meaningful about that is testing.
Of course the study wasn't perfect but they actually administered the treatment and knew of the conditions of their subjects prior to introducing the variable.
Thats a big ****ing deal if you ask me.
Seriously?
If the virus has a specific base R naught that we are comfortable with and it was introduced earlier that doesn't indicate more infection in the population?