It will probably end up being 0.3-0.5% when all said and done.

The Chinese have some serious explaining to do and they need to be held accountable monetarily for the cover up.

I'm not sure why you think that

unless you think a therapeutic is really going to kick this thing's ass

Generally, finger stick tests aren't going to be as accurate as a lab done Elisa.

It is known that SARS antibodies cross react with SARS2CoV spike protein, so you aren't actually going to get absolutely specific. On the other hand, very few people in the US have had SARS.

The human CoVs are on other branches of the family tree, so it may be possible to distinguish SARSCoV2 antibodies from HuCoV antibodies.

What does a positive test and cross reactivity mean, cross reactivity would imply your antibodies recognize/bind to the protein used in the test (meaning SARSCoV2). That may offer some protection, but doesn't imply you ever had it.

Finger sticks are qualitative, meaning do you have antibodies above the threshold detection limit. That doesn't tell you how strong an immune response you will have, so doesn't necessarily mean immunity.

False positive rates play havoc when there is a small % of the population that has actually been exposed. False positives of 5% could produce more false positives than true positives if ~5% of the population actually has antibodies. In the latest WH testing briefing from Friday, the last couple of pages discuss the need to confirm positive antibody tests with a second test. There are some good tables showing numbers for scenarios where the true positives range up to 30%.

IgM vs IgG is interesting and I wonder if they are using different binding targets for IgM and IgG. There are two good surface protein targets in spike protein and nucleocapsid protein. It is possible there is more cross reactivity to one than the other

B-1 83 said:

---

These tests are quite the double edged sword, it seems. On one hand, people have been screaming for more testing, on the other hand as we do it we're finding lots more people with exposure and antibodies........that death rate from it is plumeting.

---

I don't think as many people as you'd like to believe are cheering for negative stats from the antibody tests. A lot of people are simply cautious about whether they are accurate or not, which up to this point has been a valid concern.

If they are in fact accurate, and give us a picture of a much bigger infected rate and thus lower IFR %, then most people will see that as good news. I know that's hard to believe after all the TexAgs mob warfare, but it's true.

Just when I feel I'm pretty smart, I read some of these posts and only feel I grasp 50% of it. I'm just hoping for a big "false negative of my understanding %"

cone said:

---

plummeting to... a 1% IFR

the thing about this bug is that it's not nearly the hospitalization threat we thought it was, but it's still a stone cold killer

---

With the greatly expanded testing, it appears to be dropping WAY below 1%.......approaching that of .....of.....well, you know.

I'm sorry but the NYC study doesn't back up your math

cone said:

---

I'm sorry but the NYC study doesn't back up your math

---

I'm sorry, BUT NOT EVERY PLACE IS NEW YORK CITY. Why is that so difficult for people on this forum to understand? What are the NATIONWIDE numbers?

Player To Be Named Later said:

---

Just when I feel I'm pretty smart, I read some of these posts and only feel I grasp 50% of it. I'm just hoping for a big "false negative of my understanding %"

---

Agreed.

My feeling from reading more deeply into the antibody tests the past few weeks is that they aren't nearly reliable enough yet to give us any real information.

Hopefully we get better testing soon to give us a grasp on what % of our population has had the virus.

B-1 83 said:

---

cone said:

---

I'm sorry but the NYC study doesn't back up your math

---

I'm sorry, BUT NOT EVERY PLACE IS NEW YORK CITY. Why is that so difficult for people on this forum to understand? What are the NATIONWIDE numbers?

---

I understand what you are saying 100%.

However, why would the NY state death % be THAT much worse than the rest of the country when their hospitals were never over max capacity? I know they were bad in NYC but were they bad enough to cause additional, needless deaths because of lack of care?

One view is that NY state is the only place that we have gotten a true sense of how deadly this is (in the US) because they are the only state that has gotten enough of a % of their entire populace infected to give a meaningful read.

I know the argument for the other more cheery picture. I just don't know if I buy it yet. I hope your view ends up being the truth and with better viral remedies (at minimum) hopefully that is what happens.

B1,

No need to get angry and attack people on this board. The problem with the seroprevalence studies in areas that have not had large outbreaks is that it's hard to distinguish signal from noise. In Santa Clara county, the anti-body tests found that 2-3% of people tested positive using a test that has a 2-3% false positive rate. Effectively, every person who tested positive could have been a false positive. That's not the only issue with the Santa Clara study -- their sample recruited people on Facebook and by email, which people worry is sampling bias -- i.e. people who think they might have had COVID were more likely to go out and get the test. There were some other ethical issues, including the people leading the study promoting the results before they even did the tests and them promising people that the tests were FDA approved and would allow them to return to work if they were positive. Finally, the quality of the tests and it's cross-reactivity with other coronaviruses is not really known yet.

The nice thing about the NY is that since the % of people testing positive is higher, it makes the false positive rate smaller as a percentage of the whole. For example, in NYC if 20% test positive with a 2-3% false positive rate, you're error is a much smaller percentage of the whole.

Additional studies have been coming out of areas in Italy that were hit hard that also seem to be pointing things into the 0.5%-1.0% range.

But, you don't have to be an epidemiologist to conclude that this virus is worse than the flu. In one month more people died from Coronavirus than all of last flu season. Additionally, 0.12% of ALL people in New York and 0.8% of ALL people in New Jersey have died from Coronavirus, which is roughly the IFR of the flu. The virus has killed 0.14% of people in Lombardy, Italy and >0.2% in the hardest hit towns.

exactly

NYC has the prevalence to provide an accurate denominator via antibody test

and their hospitalizations were not overwhelming

so you're looking at real(er) data there

Thanks for the summary and the insight. It is a fascinating field.

What do you make of this

https://www.genalyte.com/wp-content/uploads/2020/04/SARS-CoV-2-Product-Sheet-4_20.pdf

They say they are testing for 5 CoV-2 antigens. I interpret that as 4 sites on the Spike protein and 1 on the
Nucleocapsid. The RBD subsection of the S1 subsection is one. One site where S1 and S2 "come together."
One other S1 site and one S2 site.

Am I in the ballpark?

PJYoung said:

---

Player To Be Named Later said:

---

Just when I feel I'm pretty smart, I read some of these posts and only feel I grasp 50% of it. I'm just hoping for a big "false negative of my understanding %"

---

Agreed.

My feeling from reading more deeply into the antibody tests the past few weeks is that they aren't nearly reliable enough yet to give us any real information.

Hopefully we get better testing soon to give us a grasp on what % of our population has had the virus.

---

Part of the issues is FDA, after giving EUA to a few tests basically gave everybody a green light. Then we get lots of bad tests out there, some because they didn't have proper testing to ensure they are good, and others because some people are just crooks. Good argument for continued FDA approval process, even if it slows things down a bit. We need faster than usual review, but not no review.

FDA has scaled that back, but damage is done. Crappy tests with crappy results floating around giving people the wrong info.

Also, there is the false positive rate. Even the good tests have some error and when you have vastly more people who do not have antibodies compared to those who do, the false positive creates a lot of noise/uncertainty in the numbers.

This is why the WH is saying confirm a positive antibody test with a second test. The chances of a false positive coming up positive both time on two different tests is pretty low.

plain_o_llama said:

---

Thanks for the summary and the insight. It is a fascinating field.

What do you make of this

https://www.genalyte.com/wp-content/uploads/2020/04/SARS-CoV-2-Product-Sheet-4_20.pdf

They say they are testing for 5 CoV-2 antigens. I interpret that as 4 sites on the Spike protein and 1 on the
Nucleocapsid. The RBD subsection of the S1 subsection is one. One site where S1 and S2 "come together."
One other S1 site and one S2 site.

Am I in the ballpark?

---

Close. You don't really want to just present a smallish part of the protein as the immune system will make antibodies to any surface feature it can (and the hopefully get rid of anything that cross reacts with you).

Spike is a complex of 2 different subunits (S1 and S2). (FYI, the virus makes them as one and then they are cut apart to become active). They are presenting S1 by itself, S2 by itself, and the combination of S1S2. Nucleocapsid is a single protein, so they present that by itself. And then add in the receptor binding domain of Spike S1. An antibody to the RBD would potentially block the infection of a new cell by getting in the way of the virus binding to the receptor, so that can be an important one.

I think they are using SARS nucleocapsid as a control. SARS and SARS2 are very similar in spike, so someone who has antibodies to SARS would likely test positive to SARS2 spike proteins. This should offer some protection, maybe enough, but they may not have has COVID19.

Bottom line, if you test positive to all those 5 parts of SARS2 and negative to SARS, there is a really good chance you were infected.

Also, this isn't a finger ***** stick test that can be performed at the drive through clinic. To do that you would need to read a pattern of nearly 15 different lines. It does tell you a lot about how complete the immune response would be, so a very good confirmation test if the finger ***** test comes back positive.

Thanks for the information. I knew my mental model was too simplistic. I wouldn't have understood the question and certainly your response in January. :-)

Two questions:

I assume at some point there will be efforts to inventory or characterize various antibodies specific to SARS-CoV2. What is that called?


Is this photonic ring resonance technology from Genalyte an established thing? Sounds kind of Theranos-like.

plain_o_llama said:

---

Thanks for the information. I knew my mental model was too simplistic. I wouldn't have understood the question and certainly your response in January. :-)

Two questions:

I assume at some point there will be efforts to inventory or characterize various antibodies specific to SARS-CoV2. What is that called?


Is this photonic ring resonance technology from Genalyte an established thing? Sounds kind of Theranos-like.

---

Yes, people will start looking for specific antibodies that bind in particularly useful ways. Not sure how medically useful that will be, but it will be useful for research.

FDA approved their device last October, which would have required a lot of proof it works. Here is on article about it. https://www.fastcompany.com/40574949/like-theranos-but-it-works-health-startup-genalyte-proves-its-worth

Thanks!