Covid-19 - Emerging treatment news thread (no politics, no tin foil hats)

Status
Not open for further replies.
I would be careful of drawing too many conclusions from this article. They are using the number or people who tested positive in the antibody test to determine the fatality rate. I don't see where the article indicates anything about the sensitivity or the false positive rate of the antibody test they used. Both of these are important factors and can really skew results when the amount of infection found in the community is low. This is due to Bayes' theorem, which has a lengthy wikipedia article. I think an easier to understand explanation is the figures below. The sensitivity and false positive rates were taken from news reports a couple weeks ago about the antibody test that was being developed. I don't know if the same test was used in this study, so the sensitivity and false positive rates may be different. But what these figures do show is that depending on the level of true infection found in the community, you can get wildly different ratios of detected true infections to false positives. This becomes more important when the true level of infection is a small percentage of the population. For instance, in the left figure the observed rate of infection is 5% whereas the true rate of infection is actually 1%. For the right figure the observed rate of infection is 13% whereas the true rate is 10%. So the error is significantly higher at lower levels of infection, and does so in such a way that it overestimates the number of people who have become infected by this virus. And the LA Times article states that 4% of people tested positive for antibodies, which falls in the category of low levels of exposure where the observed positives is likely much higher than the true positives. Of course, the caveat is that we don't know how good the antibody test they used for the study is. If they used a better antibody test with a much, much lower false positive rate, this would be less of a problem.

View attachment 137390 View attachment 137391

Full disclosure, these are figures my friend who has a PhD in biochemistry made and then shared on Twitter for others to use. I am an organic chemist and we seldom if ever use statistics, so I was only useful as a clueless sounding board in the attempt to make easily understandable figures.

Does this take into account that random sample was obtained from an add/solicitation on Facebook?
 
Here is a Twitter thread from a UC-Berkeley statistics professor about the problems he and others noticed about the statistical methods employed by the team who conducted the Santa Clara study.


"However, I'm flabbergasted that yesterday afternoon, a group including several of the same authors circulated a press release describing new results for a similar study in LA County, without any accompanying technical report and before correcting the Santa Clara County preprint."

This incident really highlights the potential benefits and drawbacks for publishing your work as a preprint. Preprint journals are these journals that have popped up in the last few years that allow you to put your work out to the larger community without it having to go through a peer review process, with the idea that the paper will be published relatively soon in another journal that has a peer review process. This allows you to get your work out faster, which can be good for a variety of reasons in normal times and is really useful during this pandemic. But the obvious drawback is you don't have a peer review process that can catch these problems before the paper is shown to the world. And with the paper being accessible to the entire world, people will talk about the results as if they are peer reviewed research.
 
Ho boy, I didn't catch that on the first read. I guess desperate times call for desperate measures, but that seems like a great way to over inflate the number of actual infections in your study.
How about inflating the number of deaths. Does anyone take this into account?

Should “COVID-19” be reported on the death certificate only with a confirmed test? COVID-19 should be reported on the death certificate for all decedents where the disease caused or is assumed to have caused or contributed to death

 
Last edited:
Should “COVID-19” be reported on the death certificate only with a confirmed test? COVID-19 should be reported on the death certificate for all decedents where the disease caused or is assumed to have caused or contributed to death

If you look at the difference between the current mortality rate and the background mortality rate, and then compare that to the current death count for the virus, it becomes clear they are actually under counting the death toll in the official counts. But that under count is going to vary from region to region. Here is a good Twitter thread about this from data scientist and head of FiveThirtyEight Nate Silver:


Also, don't kill the thread.
 
If you look at the difference between the current mortality rate and the background mortality rate, and then compare that to the current death count for the virus, it becomes clear they are actually under counting the death toll in the official counts. But that under count is going to vary from region to region. Here is a good Twitter thread about this from data scientist and head of FiveThirtyEight Nate Silver:


Also, don't kill the thread.
Excess deaths aren't unexpected when the majority of the healthcare system is only focused on c-19. Couple that with the media scaring the hell out of most people and they aren't seeking medical attention for other very serious issues.


"What we've seen over the course of the last six to eight weeks is that there are decreasing numbers of heart attacks and strokes showing up at US hospitals.
 
The continuously changing rules on masks is dumbfounding.
Last week, my wife while taking care of a 600 gram translucent baby born at 24 weeks gestation, was FORBIDDEN from wearing a medical grade surgical mask.
Today, I am supposed to wear some POS, homemade, cloth ‘comfort’ mask to the grocery store.

Is anybody even trying to make sense?
 
Noted it is early on but another study for you all(you know who you are) to discount. This could definitely shift the big scary mortality numbers that everyone likes to throw around



KEY POINTS
  • An estimated 13.9% of the New Yorkers have likely had Covid-19, according to preliminary results of coronavirus antibody testing released by Gov. Andrew Cuomo on Thursday.
  • The state randomly tested 3,000 people at grocery stores and shopping locations across 19 counties in 40 localities to see if they had the antibodies to fight the coronavirus, indicating they have had the virus and recovered from it.
  • With more than 19.4 million people residents, the preliminary results indicate that at least 2.7 million New Yorkers have been infected with Covid-19.
 
That would indicate a fatality rate of around 0.7%
LA’s antibody testing (if accurate) indicate 0.1-0.3%.
Both are a big decline from 4%
Stockholm thinks they are about 2 weeks from reaching herd immunity, which they say happens at around 60% or the population being infected.

Whether they have flaws or not, anti body blood test sound like a more reliable test than the nasal swabbing
I hear from my sister in law, who just got home from another ICU shift that the nasal swabs are continuing to be unreliable. They have patients test positive, then negative, then positive and vice versa. They think that positive cases without much for head congestion are frequently testing negative.
 
Noted it is early on but another study for you all(you know who you are) to discount. This could definitely shift the big scary mortality numbers that everyone likes to throw around



KEY POINTS
  • An estimated 13.9% of the New Yorkers have likely had Covid-19, according to preliminary results of coronavirus antibody testing released by Gov. Andrew Cuomo on Thursday.
  • The state randomly tested 3,000 people at grocery stores and shopping locations across 19 counties in 40 localities to see if they had the antibodies to fight the coronavirus, indicating they have had the virus and recovered from it.
  • With more than 19.4 million people residents, the preliminary results indicate that at least 2.7 million New Yorkers have been infected with Covid-19.

While I agree that lowering the death rate makes the number less scary, was 2-3% of symptomatic cases, with an average age of death only about 2yrs below life expectancy and usually with multiple co-morbidities ever actually that scary? We knew that from Italy weeks ago.
 
Status
Not open for further replies.
Back
Top