CV: Stanford Antibody study results are out ...
Figured this was worth a new thread - the first truly rigorous and large CV antibody test results from anywhere in America.
This is Santa Clara County, the SF South Bay. 3330 residents tested in a county of about 1.9 MM.
Key takeaway - estimated that 2.49%-4.16% of the population there has been infected by CV by early April (the study was done on April 3-4) already. That prevalence estimate represents a range of 48,000 to 81,000 people infected in Santa Clara County by early April, which is 50x-85x the number of actually confirmed cases.
Link to the paper:
https://www.medrxiv.org/content/10.1101/2020.04.14.20062463v1.full.pdf
Of course, that is just Santa Clara County, but it gives us another estimate of the ratio between actual and confirmed cases. I've always been pegging that number at 10x-30x. This ratio is obviously larger than that.
More and more antibody studies to come in the time ahead. I think a similar Stanford study conducted down in SoCal (as opposed to NorCal) is next, and should be out within days.
April 17th, 2020 at 12:28 PM ^
Maybe one of you smart people can help me understand- so if most people still haven't gotten it and the economy is re-opened in 1, 3, 6 months, isn't it still super high risk that it explodes again and we're almost back to square one? I don't get how it works if there's no medicine, no drugs, and most people still haven't had it
April 17th, 2020 at 12:39 PM ^
i am not a smart person, but I think that’s why mass testing is so critical. So you can snuff out any new outbreak before it has a chance to get out of control. I don’t know how feasible mass testing is, but that’s the underlying premise being pushed by smart people I read about.
I am also not a genius, but doesn't this mean that the Santa Clara health system had no problem with a huge number of infections and that risk of dying are actually really low.
Wouldn't that mean that as it spreads through the country, as it will in any event, we stand a much lower risk of overloading the healthcare system.
We are already much further up the curve than anticipated. And the virus is less severe.
That means opening up causes much less concern.
I think you are jumping to conclusions that support your opinion based on too limited a data set.
I don't know when it becomes more scientifically accepted (ie., sample size), but the numbers are un-mistakenly headed in that direction.
This study.
Boston homeless shelter.
German Study.
Spanish study.
NYC pregnant women.
Percent of asymptomatic people testing positive in China as they land on flights.
Etc.
When do we call it?
April 17th, 2020 at 12:46 PM ^
NittanyFan > Fauci
(this is a joke people, I like Fauci very much)
Actually wait a minute...
NittanyFan ... AnthonyFauci
i think i'm on to something
April 17th, 2020 at 12:51 PM ^
Good discovery, now his tinder hookup comment makes complete sense...
April 17th, 2020 at 12:58 PM ^
You're half right.
I'm actually both Dr. Fauci and Dr. Birx (the latter is a Penn State grad)!
April 17th, 2020 at 12:52 PM ^
We. All. Are. Infected.
April 17th, 2020 at 12:54 PM ^
This can't be true. I suggested this possibility the other day and poster LewisBullox told me "that was not accurate". Can someone please email Stanford and let me know that they should hold off on publishing the study until LewisBullox peer reviews it?
You mean when you said "I wouldn't be shocked if we find out months from now that a much larger portion of the population than we knew had already been infected and that the stay at home orders didn't move the needle as much as we thought."
Cause both of those statements are still incredibly false. Most experts have already been assuming a prevalence near this, and the stay at home orders part is just wrong.
No, most scientists have not been assuming a rate this high. The full social distancing impact will be better understood once we have more data.
the question with the Stay at Home Orders isn’t “do they help slow virus spread?” Duh, of course they do. The question is how much do/did they help within the context of the actual human behavior during the Stay at Home Orders.
Physiology professor. Most of us have been assuming this, although estimates have varied widely.
Everyone has been assuming undetected cases. The "vary widely" is the important part. When projections are done, what percent do modelers incorporate?
From what I have seen, most don't incorporate anything near this. Many project an exponential growth that would not be possible if we were this far up the curve already.
Do you have an example?
Remember when Ohio's health department said that the prevalence was likely 1% of the state's population because there was community spread? That was weeks ago, so seeing 1-3% in California in early April fits right in.
I have to run right now but there was an interesting article a bit ago where different experts suggested everything from x10 - x300. I'll see if I can find ti for you later.
I think some of the confusion may come from the difference in asymptomatic and undiagnosed. They are very different things.
Sure. The Ohio health secretary was criticized by modelers though and they said that was an overstatement.
I saw the Oxford study that half of England might already be infected. The imperial model did not take that into account.
The fact is that the main models we use IHME and CovidActNow, etc. take a much more conservative view of unreported cases.
The epidemiologists who projected the highest numbers of unreported cases were considered fringe and none of the policymakers are relying on those projections.
Our laws are being made by people anticipating we are much lower down on the infection curve and that we have a much longer way to go than this indicates.
Does any projection in NYC indicate that nearly half the population might already be infected and the city is near the top of the curve for that reason alone?
This paper suggests ~2% have had prior exposure to Covid-19 not "half of England".
You have a fair point about the models, but this data won't change them much. IHME doesn't make any assumptions about unreported cases.
Changing the CovidActNow model for slightly more recovered shouldn't change its predictions much (although this warrants investigations).
Why are we assuming half of NYC is infected based on this study?
April 17th, 2020 at 12:55 PM ^
https://www.livescience.com/coronavirus-in-pregnant-woman-high-nyc.html
This is also a kinda interesting article. Apparently between March 22nd and April 4th in NYC, of 215 pregnant women screened at Columbia University Irving Medical Center and New York-Presbyterian Allen Hospital, 33 of those women, or 15%, tested positive for COVID-19 and only 4 of those 33 women showed symptoms. Not sure how that translates to NYC at large, but 15% is a large number
How many viruses are there, that we know of, that have symptoms ranging from literally nothing at all, to death?
Pretty much every virus In existence
Yeah, no. Not to this degree. Common cold isn't deadly. Pox viruses are practically never asymptomatic. There are examples (typhoid, Legionnaires', although those are not viruses) but "pretty much every virus in existence" does not have a vast range of outcomes such that both death and zero symptoms are reasonably common.
Did I put them on the same scale? No.
but literally every virus has a range of outcomes from asymptomatic to death. Ymmv
I disagree.
Common cold isn't deadly.
Actually, other common colds can cause up to 8% fatality in elderly populations.
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2095096/
Also, a significant number of people are hospitalized from colds.
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3805243/
So, this may be pretty consistent with the range of outcomes from other viruses. Its just new.
My entirely unscientific estimate has been 10-100x the confirmed cases in the U.S. because of the colossal and ongoing testing failure. At least on the west coast, we're still in "if you are well enough to not need hospitalization you stay home and don't get tested."
Regardless of the moronic, nonsense daily assertions otherwise.
So yeah, 50-85x seems entirely plausible.
That policy is in place in pretty much the whole country. But keep in mind, an infectious disease takes time to work its way through the population. And most of the country has been social distancing for close to a month now. While there are surely far more unreported cases than reported, the total number of Americans who have been infected is probably still just a small fraction of the population.
Certainly, in California, it would be inexplicable that there would be massive infection and not even 1000 deaths for a state of 40 million people - given what this virus has done everywhere else in the world.
It doesn't seem very plausible to me. It would mean that anywhere from 33 to 57 million people in the US have been infected. And that would mean we shut the world down for the flu.
April 20th, 2020 at 12:55 PM ^
Ding! Ding! Ding!
We have a winner!
You just stumbled onto a fundamental truth of this whole mess. While nastier than the normal flu, the overall effects of this virus are only minimally worse than the flu. We have allowed ourselves to be panicked by a media desperate for a bleeding story. When we start counting the bodies that are going to result form the shutdown (Delayed medical care, mental health blowups, job loss impacts) we are going to find that we saved approximately 60K lives at the cost of 30K lives and a utterly destroyed economy. The trade offs never made sense in the first place, and people are just coming out of their collective stupor to realize it. If there is any justice people will be calling for Fauci's head on a pike.
If you take the 50x number and apply it to NYC, it means 70% of the population would have it, which is the number people have thrown out for herd immunity. New York state would be at 55%.
And obviously it can't be 85x because that would be >100%.
Otoh New York has run a lot more tests, so their multiplier would naturally be lower.
At any rate, even a 10x-20x number would be good news as far as total deaths and going back to work.
That is so much a simpler and more elegant way to make the comparison than I have been working on. I calculated a Santa Clara IFR with 67 deaths of between 0.085 and 0.14. I then calculated the minimum possible IFR under the clearly wrong assumption that 100% of the population was infected and got a number of 0.085 for New York as a whole and of 0.106 for NYC. Knowing those numbers have to be low because infected patients are still dying, with herd immunity 100% isn’t realistic and given that they are continuing to see new cases, they can’t be at herd immunity.
That all made me throw up my hands and say that either I’m calculating something really wrong; something weird is going on in NYC, I’m misunderstanding some key concept somewhere or there is something weird with the Stanford results. My guess is on my math or understanding of some concept, but would love to know if anyone else has tried to reconcile the numbers. I will say that if I go back and use 0.37% from the Gangelt Germany results I get about 30% infected in NYC which seems more realistic..
One unknown is whether the amount of initial viral load makes a difference in a person's outcome. If you are exposed to a larger amount of the virus during your infection, does that lead to a worse health outcome? In effect, getting more of the virus to enter your body could put your body further along the infection path than a person who just got a small amount, and it might make it harder for your immune system to fight it off.
This might explain why some hotspots have had such a hard time - and also why some otherwise healthy doctors/nurses are succumbing to the virus after treating lots of patients. But we don't know yet if it's the case.
There's too many variables for everything to make sense at this stage. Climate, initial individual exposure level, initial regional exposure level, rural/suburban/urban, ethnicity, false positives/negatives, level of social distancing, public mask wearing, health care PPE availability, amount of testing, age, wealth, health, recovered percentage......and I'm sure there's more.
Hell even something like internet speed and an area's technological status could keep more people working at home, ordering food from home, spread apart more which could not only slow the virus but lower initial load possibly giving the immune system a head start.
NYC has a huge population, the lowest percentage of car ownership and highest percentage of mass transit usage. People are on top of each other and in close proximity to strangers on a daily basis. It’s no surprise that any virus would spread more easily in that type of environment. They’re also padding their death stats.
They’re also padding their death stats.
Counting people who displayed COVID symptoms and died before they could be tested is hardly "padding" the rate. This is essentially what the CDC does when it produces its final estimate of the number of deaths.
Because if someone was displaying any of the very generic Covid symptoms, which are also symptoms of many other things, it necessarily means they had the virus? I’m sure that some of those deaths were caused by the virus. I’m equally sure that not all of them were.
They do require a positive test for COVID before they include them in the stats. I believe the post-mortem ones are tallied once a week.
I understand why they have more people impacted, but the morbidity numbers from the Stanford number don’t seem to be possible in NYC given the population and number if people already dead.
Fascinating results
For anybody who cares, the comparison to SARS/MERS immunity is 2-3 years, depending on the person.
In other words, if you catch SARS, your body will develop an immunity that will last for a couple of years.
As COVID19 is a SARS strain, it's reasonable to presume AT LEAST an immediate re-exposure immunity. I've yet to see a disease that can immediately reinfect somebody once they've overcome it, with the exception of immunodeficient people who, for whatever reason, have trouble producing antibodies.
Part of the problem is that there were some conspiracy theorists starting the "reinfection" story without any real proof. A guy was doing this on Youtube, and when the article he cited was actually read, it was pretty obvious that the testing for COVID was faulty.
This is good news. It means that the real mortality rate (not the case fatality rate using only confirmed cases) is much lower than initially estimated. Using the mid point of the percentage of the population infected and applying to the entire USA, you get about 10 million people in the country already infected. Using the total deaths thus far, you get a rough mortality rate of 0.35%. This is almost exactly the same as a German study which looked at an entire region and came up with an estimated mortality rate of 0.37%. That's still much higher than the typical seasonal flu at around 0.1% but much better than initial estimates of 1%. With natural immunity and a vaccine by next year, these mortality rates may eventually be comparable.
There is still a lot of morbidity beyond fatalities with the number of people needing ventilators or supplemental oxygen. That's what we're seeing in the ER. And the virus seems to spread easily, causing a surge of very sick or dying patients. So it still places a significant stress on our health care system, especially without measures to slow the spread. And if allowed to spread throughout the country, the above mortality rate still comes out to about 1 million deaths if everybody gets infected. But the overall impact might be much lower than the worst estimates if we continue slowing the spread.
There are the caveats that the sampling wasn't random and doesn't necessarily apply to the whole country.
The shutdown is placing more than “significant stress” on every other system in the country.
Did you read the estimate of 1 million deaths if left unchecked? Several times that many would be seriously ill as well. Can you imagine what it would be like in hospitals? They would be overwhelmed and we would have had to just let people die. That's what they had to do in Italy. And that might include people with major trauma, heart attacks, etc. It's horrible now but it would have been horrific.
Yes I did, and the same people who made ridiculous estimate were also very wrong about how many deaths would occur even with all of these restrictions.
to sum it up we don't know anything except we are in the midst of a pandemic
So let’s just ignore every piece of information as it comes available and assume our initial best guess was correct
So far, there have been 69 deaths in Santa Clara County (where I work) that have been attributed to the virus. At the low range of the study’s estimated number of infection, that amounts to a fatality rate of .14%, about the same as the flu. The devastation caused by the lockdown is going to be much worse than the virus. This insanity has to end.
Good news and it's why we need to get our head of our asses and get back to normal.
Such a strange disease, show almost no symptoms in some people, yet kill others.
Which is a lot like every other respiratory illness, actually. As some of the comments in this thread have pointed out, depending upon underlying risk factors such as age and obesity, you can have a bad time with colds and flu. Both will also cause death, though certainly not at the same rates. Still, viruses are nothing to sneeze at, pun intended.
One thing that has been made abundantly clear while going through this virus is that we as a society have got to get healthier. The difference in the outcomes between those that are physically fit and those that are obese is staggering. Sadly, and a bit ironic is while healthier restaurants are closed the fast food drive- thrus continue to be pretty busy.