Estimated numbers of Covid19 antibodies based in random samples in Santa Clara County reveals approximately 75 times more people had COVID-19 than actual government reporting numbers. What this means; the numbers we have been using for symptom reporting and mortality may be all wrong. – Anthony
COVID-19 Antibody Seroprevalence in Santa Clara County, California
Eran Bendavid, Bianca Mulaney, Neeraj Sood, Soleil Shah, Emilia Ling, Rebecca Bromley-Dulfano, Cara Lai, Zoe Weissberg, Rodrigo Saavedra, James Tedrow, Dona Tversky, Andrew Bogan, Thomas Kupiec, Daniel Eichner, Ribhav Gupta, John Ioannidis, Jay Bhattacharya
doi: https://doi.org/10.1101/2020.04.14.20062463
Abstract
Background Addressing COVID-19 is a pressing health and social concern. To date, many epidemic projections and policies addressing COVID-19 have been designed without seroprevalence data to inform epidemic parameters. We measured the seroprevalence of antibodies to SARS-CoV-2 in Santa Clara County. Methods On 4/3-4/4, 2020, we tested county residents for antibodies to SARS-CoV-2 using a lateral flow immunoassay. Participants were recruited using Facebook ads targeting a representative sample of the county by demographic and geographic characteristics. We report the prevalence of antibodies to SARS-CoV-2 in a sample of 3,330 people, adjusting for zip code, sex, and race/ethnicity. We also adjust for test performance characteristics using 3 different estimates: (i) the test manufacturer’s data, (ii) a sample of 37 positive and 30 negative controls tested at Stanford, and (iii) a combination of both. Results The unadjusted prevalence of antibodies to SARS-CoV-2 in Santa Clara County was 1.5% (exact binomial 95CI 1.11-1.97%), and the population-weighted prevalence was 2.81% (95CI 2.24-3.37%). Under the three scenarios for test performance characteristics, the population prevalence of COVID-19 in Santa Clara ranged from 2.49% (95CI 1.80-3.17%) to 4.16% (2.58-5.70%). These prevalence estimates represent a range between 48,000 and 81,000 people infected in Santa Clara County by early April, 50-85-fold more than the number of confirmed cases.
Conclusions
The population prevalence of SARS-CoV-2 antibodies in Santa Clara County implies that the infection is much more widespread than indicated by the number of confirmed cases. Population prevalence estimates can now be used to calibrate epidemic and mortality projections.
Oh Christ
They, allowed the subjects to bring a child from the household
~900 of the 3500 subjects are from the same household
80% of case transmission is family to family
BZZZNT
overestimates the prevalence
Huh?
Plus motivation from people who got symptoms but couldn’t get a PCR test around that time might be higher to self-register for being tested.
That’s a problem with non-randomized subjects.
YUP
here is a peer review, says the same thing
https://medium.com/@balajis/peer-review-of-covid-19-antibody-seroprevalence-in-santa-clara-county-california-1f6382258c25
Thanks for the link, very informative!
I should pay more attention to the confidence intervals in the future. Always slips my mind especially if I already see non-statistical confounding problems with the study design or its interpretation.
alice: One of my kid tested positive in feb, can we get tests for the whole family?
Doctor: No severe symptoms no tests.
Alice: but we are in contact daily!
Doctor: sorry CDC rules, no severe symptoms,no test, fly to Korea if you want to be tested.
…..
jane: hey Alice, they are doing antibody testing, did you see it on facebook?
Alice: no let me check, oh wow, I can go get tested and bring one of my kids
jane; ya, you can find out if you had it when they refused to test you, call your friends
let them know, everyone from our coffee group should go and bring their kids.
Result? Non random sample .
“The test kit used in this study (Premier Biotech, Minneapolis, MN) was tested in a Stanford laboratory
prior to field deployment. Among 37 samples of known PCR-positive COVID-19 patients with positive
IgG or IgM detected on a locally-developed ELISA test, 25 were kit-positive. A sample of 30 pre-COVID
samples from hip surgery patients were also tested, and all 30 were negative. The manufacturer’s test
characteristics relied on samples from clinically confirmed COVID-19 patients as positive gold standard
and pre-COVID sera for negative gold standard. Among 75 samples of clinically confirmed COVID-19
patients with positive IgG, 75 were kit-positive, and among 85 samples with positive IgM, 78 were kitpositive. Among 371 pre-COVID samples, 369 were negative. Our estimates of sensitivity based on the
manufacturer’s and locally tested data were 91.8% (using the lower estimate based on IgM, 95 CI 83.8-
96.6%) and 67.6% (95 CI 50.2-82.0%), respectively. Similarly, our estimates of specificity are 99.5% (95
CI 98.1-99.9%) and 100% (95 CI 90.5-100%). A combination of both data sources provides us with a
combined sensitivity of 80.3% (95 CI 72.1-87.0%) and a specificity of 99.5% (95 CI 98.3-99.9%).
ouch,
WRT to allowing people to bring their children.
dumbass test design.
‘ After weighting our sample to match
Santa Clara County by zip, race, and sex, the prevalence was 2.81% (95% CI 2.24-3.37 without clustering
the standard errors for members of the same household, and 1.45-4.16 with clustering). ”
why? why make your analysis harder by including close contacts.?
same with the adjustments for race/sex.
why? The enrollment process would allow them to accept enrollees in such a way that
you got representative samples with no need to adjust the data.
Consistent with ZERO.
“These results represent the first large-scale community-based prevalence study in a major US county
completed during a rapidly changing pandemic, and with newly available test kits. We consider our
estimate to represent the best available current evidence, but recognize that new information, especially
about the test kit performance, could result in updated estimates. For example, if new estimates indicate
test specificity to be less than 97.9%, our SARS-CoV-2 prevalence estimate would change from 2.8% to
less than 1%, and the lower uncertainty bound of our estimate would include zero. On the other hand,
lower sensitivity, which has been raised as a concern with point-of-care test kits, would imply that the
population prevalence would be even higher. New information on test kit performance and population
should be incorporated as more testing is done and we plan to revise our estimates accordingly.
jesus what a mess.
Age distribution is the critical parameter to control for in the sampling.
Age distribution is the critical parameter to control for in death stats as well.
‘This study had several limitations. First, our sampling strategy selected for members of Santa Clara
County with access to Facebook and a car to attend drive-through testing sites. This resulted in an overrepresentation of white women between the ages of 19 and 64, and an under-representation of Hispanic
and Asian populations, relative to our community. Those imbalances were partly addressed by weighting
our sample population by zip code, race, and sex to match the county. We did not account for age
imbalance in our sample, and could not ascertain representativeness of SARS-CoV-2 antibodies in
homeless populations. Other biases, such as bias favoring individuals in good health capable of attending
our testing sites, or bias favoring those with prior COVID-like illnesses seeking antibody confirmation are
also possible. The overall effect of such biases is hard to ascertain.
The Premier Biotech serology test used in this study has not been approved by the FDA by the time of the
study, and validation studies for this assay are ongoing. We used existing test performance data to
establish a range of sensitivity and specificity, including reliable but small-size data sourced at Stanford.
Test sensitivity varied between the manufacturer’s data and the local data. It is possible that
asymptomatic or mildly symptomatic individuals may generate only low-titer antibodies, and that
sensitivity may be even lower if there are many such cases.23 Additional validation of the assays used
could improve our estimates and those of ongoing serosurveys. ”
1. the age bins reported are stupid. 0-4, 5-18, 19-64, 65+
why?
2. Positive rates are not reported Per Age bin.
Thanks Steven,
Being a resident of Santa Clara and an amateur number cruncher, this study’s results seemed contradictory to everything else virus-related I have been absorbing since Jan. 17. Your link to the review helped me understand why it may be almost worthless in describing the current local situation in regard to COVID-19. It is highly appreciated and will be shared with the many people I have been communicating with about this for three months. Innumerancy is a contagious disease itself, and the many subtle possibilities and clear explanations offered in this review help vaccinate me against the wild misunderstandings I see in so many comments.
There may be some things that “rescue” the results, but they need to release more data
I would like to see their recruitment criteria.
A) did they tell the subjects that they would NOT tell the subject whether they tested
positive or negative? this would ameliorate the selection bias
B) did they ask the subjects if they had
1. had any symptoms whatsoever
2. Know anyone who was tested
3. worked in a high human contact service job
4. Requested a PCR test in the prior 3 months and been denied
That might make me more accepting of the results, but they do not document their recruitment
protocal.
This is not that hard to do.
Ok
here’s a professional, detailing the issues
https://twitter.com/nataliexdean/status/1251309217215942656
Be careful finding what you want to believe
Good advice to share with your warmista friends, too, Steven.
I share it with everyone.
People who find a station adjusted poorly
People who find a really hot day, or really cold day
People who look at just one country (USA) for their temperature data
People who look at models that run really hot, or cold
People who only look at UAH
people who look at photos of subs at the north pole.
people who look at one tree in Yamal
The tendency to find what you want to find is pervasive.
so when a skeptic finds a warmist who finds what he want to believe
the skeptic has found what he wants to believe
and vice versa
Gotta lurve your postmodern approach to the truth Mosh, ‘The truth is what people believe it is’
Until you find yourself gasping for breath in a hospital ward
Huh.
if I thought that I would not correct people when I think they are wrong.
maybe you didn’t see me on this site when the death count was 0 and the case count
was 68..
telling people that the USA was not testing enough.
If I thought the truth was simply made up, why would I make any argument.
One thing I learned is that it was naive of me to believe that antibody testing would end the uncertainty.
On the whole a takeaway from mosh’s comments would be that we should ignore this study and wait for a well-designed randomized representative sample.
HIDE THE DECLINE!
I ran the numbers for WA state and Florida, and new cases peaked on 4/3/2020 for both states. The new Trump guidelines suggest that two weeks of declining infections are a prerequisite for easing stay-at-home orders. You are going to be seeing talking heads trying to hide the fact that new cases are declining.
50 out of 3300 were positive
based on the stats of their own test, 33% of these could be false positives.
Is this a John P.A. Ioannidis study?
Yup.
So major problems.
1. Not a random sample, you are likely to recruit people who
A) could not get a PCR test because they failed to meet the severe criteria required for that.
2. Included Children from the families and did not report out those stats
3. Test specificity and Sensitivity.
Do it again.
and report out all the raw data FFS
Ok.
This is a test done by recruitment over facebook.
Do you want to be tested?
well if you live in a county where you could not be PCR tested unless you had severe symptoms
then you would be motivated to sign up.
Hey, I was feeling ill in January, or I had a little cough in Feb, but I could not get tested.
I’m really curious, did I have COVID?
or
hey My girl friend tested positive and they never tracked me down to test me like they would
in Korea. I wonder maybe did I catch it? And gosh, maybe I gave it to my kid.
There’s free testing, lets go
And
Look at how many people faked data to show up for the test
101, more than tested positive
42 had invalid Id
59 had ID that didn’t match their survey
I will say this.
my prediction of less than 100 positive was correct.
It’s is good to see every reader practice skepticism and not try to fool themselves.
Feynman would be proud of all of you who read the paper BEFORE commenting.
and he would also be proud of all of you who bent over backwards to find issues with the study
That’s the skeptical spirit.
so if you did that, great.
if not, well no comment.
Bottom line: it is good news but not great news, and the science is not settled.
Plus they need to learn to post raw data and code.
The response to this study is as interesting as the content. In some ways, maybe more.
This makes maybe one time that I agree with Brother Mosher. You put ads on Facebook offering a free test to see if you have or have had Covid-19, who is most likely to sign up? People who think they may have had it, that is who. And they are allowed to bring a child, who typically would live in the same home?
This is not random sampling…
yep,
Telluride is probably better since they aim at testing everyone.
Still, the study is a good START cause it shows you how difficult it will be to get good data
IF
1. you are in an area where getting PCR tests was tough
2. you allow people to select in.
3. if the test sensitivity and specificity are low
4. If the spread is low..
Then you are going to have to get a very large sample.
So, the test will help others in designing a better test.
Folks should not object to doing a better job.
The authors themselves suggest as much
Food for thought:
This summer will come the adenovirus season. We now have the ability to identify, name, test and track every virus that comes along. Good thing, right? SO we can identify each “new” virus, yes? “New” means they have not been sequenced before, therefore each virus will be “new”. (Not new—ancient, coming cyclically like sine waves.) We were just never introduced properly.
Come October-November, and into next winter, we will have seasonal viruses come as follows (from CDC):
PIV-2,3
Rhinovirus(es)
RSV(s)
MPV
Then influenza(s)
Many of these will be “new”. With new names (identifiers). Tracking across all countries. New “medicines”. Lots of people selling new snake oil.
AND, all viruses are lethal (to the compromised). How will we react then?
In Karolinska Hospiltal, Sweden, 320 women that were about to deliver a child (all of them) were tested for COV-19 virus, no exeptions.
Of those, 23 tested positive.
All relatively young, resonably good health, female of course, no mention of symptomes.
Reported on swedish public television news, 10april.
Have they all be going to the same birth classes and/or sharing a midwife for education and exercises? Or did they simple got it from the hospital staff?
That’s a too small non-randomized sample.
No previous contacts, just all incoming soon-to- be mothers. Well some are probably already mothers.
So, quite random. Small sampe, yes.
1/. Government policy is being driven by death rates alone. This is interesting but wont affect policy.
2/. However what this does mean is that lock down is probably reducing the severity of the disease in the vast majority of the population and should make arriving at ‘herd immunity’ less painful than e.g. the WHO predict.
The prevalence of serious disease in densely populated areas does seem to indicate that case severity is critically dependent on viral load
questions you want to ask to stratify your test subjects
1. Do you take mass transit or exclusively use your car
2. do you attend church
3. Do you work or stay at home
4. Do you eat out
5. How often do you grocery shop
ect.
rather than gender and race why dont they collect variables that might differentiate on
social interaction.
Probly need a sample size at least 10x, with 50 positive cases they learn nothing about associated risk
factors
Well two things are emerging in the UK
1/. It’s raging through the densely packed cities and care homes and te medical profession and leaving the countryside essentially untouched.
2/. Where it rages BAME* people are massively more likely to die of it.
3/. Of course so to are the elderly etc. But that’s no surprise, as they tend to die of respiratory infections whenever they have other issues.
Point 1/. Essentially justifies lockdown. It almost shouts that more exposure equals far greater risk of serious illness and death.
Point 2/. Is extremely interesting. If it turns out for instance that massive doses of vitamin D make a difference…
Point 3/. is unimportant. It is to be expected.
“Black, Asian, Minority Ethnic”
In Germany, church attend is prohibited.
Who are the ones who get an immune over reaction (cykotine storm) and end up on intensive care with ventilators?
Elderly people and people with underlying conditions? YES
But also people who get the invitations for flu shots….
If it was a stronger virus mutatioon/string also babies and young children should be affected, since they have a weak immune system. Just like the normal flu pattern.
Since 2012 there is a lot of research on corona virus vaccins. A few years ago animals were tested with the 4 most promosing corona vaccins. The test seemed to go right, until the animals were months later exposed to a wild corona virus. Then it went terribly wrong. Massive immune over reactions with binding anti bodies, instead of neutralizing anti bodies.
The proposed medicine mix is Hydroxychloroquine, zink, and z-pack
Hydroxychloroquine is an immune over reaction modulator. Zink (anti viral infection), and z-pack (anti bacterial) are for cleaning up what the body immune system over reacts on.
https://aidsinfo.nih.gov/drugs/564/hydroxychloroquine/0/professional
Breaking news: COVID does not lead to ARDS, and we’re treating the wrong disease.
Confirmation of this from several sources so far:
A physician treating patients in NYC: https://www.hippocraticpost.com/covid-19-update/does-covid-19-really-cause-ards/
An Intensive Care Specialist (but not a physician): https://www.hippocraticpost.com/covid-19-update/has-covid-19-had-us-all-fooled/?utm_source=website&utm_medium=webpush&utm_campaign=notifications
The Chief of Pulmonary and Critical Care Medicine at a hospital: https://www.evms.edu/media/evms_public/departments/marketing__communications/EVMS_Critical_Care_COVID_19_Protocol__4_2_2020-revised.pdf
Better random sampling approach
I posted these comments on FB about a similar test from Massachusetts where the prevalence of SARS-CoV-2 is much greater than in the Bay Area:
My sister-in-law found a Massachusetts variant of this article. I replied, with some edits:
https://www.foxnews.com/science/third-blood-samples-massachusetts-study-coronavirus
Sigh. That Massachusetts article is okay, very promising, actually, but:
“Participants …provided a drop of blood to researchers, who were able to produce a result in ten minutes with a rapid test.”
This must be the serologic antibody test that’s been long awaited.
“He [the city manager] added: “Still, it’s kind of sobering that 30 percent of a random group of 200 people that are showing no symptoms are, in fact, infected.”
No, the purpose of the antibody test is to identify people who _had_ the disease. The are, in fact, recovered.
At any rate, 32% positive in one of the nations hotspots (the NH vs MA comparison is quite amazing).
That article links to a similar article from Santa Clara county in CA, https://www.foxnews.com/health/coronavirus-antibody-testing-finds-bay-area-infections-85-times-higher-reported-researchers
“Earlier this month, Stanford University-led researchers tested 3,330 adults and children in Santa Clara County, who were recruited using Facebook ads, for SARS-CoV-2 antibodies and found that the population prevalence of COVID-19 in Santa Clara ranged from 2.49 percent to 4.16 percent.” [Sigh. Covid-19 is the disease, so it found the prevalence of people who _recovered_ from Covid-19.]
I saw that story last night. It’s really important that so many cases are so minor. In CA’s case, they’re a long, long way to developing herd immunity. In Chelsea’s case, they’ve made a giant step there already, I imagine that NYC is even further along.
That suggests to me people in NYC will be astounded at how quickly Covid-19 fades from the scene, even without a cure that passes whatever muster people are demanding, or vaccine late this year or next.
The Massachusetts immunity research found 32% out of 200 participants testing positive while the Chelsea infection rate was only 2%. Immunity seems to be about 16 times the infection rate, which is the same ratio as in the Dutch research over 4000 people that I mentioned in the comment below.
So far 15-16 times the infection rate seems to be a more reasonable indication for the immunity rate.
https://wattsupwiththat.com/2020/04/17/covid-19-antibody-seroprevalence-in-santa-clara-county-california-coronavirus/#comment-2969259
How well do we trust the antibody test? Could it produce false positives, due to a person having antibodies to an earlier, different coronavirus infection?
Dave Burton: “How well do we trust the antibody test?”
WR: Not too much. We will have to wait for more results and for more reports about the tests themselves. For now the tests give only an indication. One of the questions is whether the test is specific enough for Covid-19.
Besides, the infection rate mentioned above also depends on the way of testing. In the Netherlands there is very restricted testing: not all cases are catched. In fact this makes the infection rate not comparable to the infection rate for another area where testing has been (more) complete.
I wonder just how reliable these results are – given the group tested were all volunteers.
Is it not possible that people who had mild or severe flu symptoms and suspected they had nCOV, are more likely to volunteer to be tested?
If so, it wouldn’t take much over-representation to skew the results.
Testing random groups of people would be more representative.
I was a scientist in animal and human medical trials in large and medium sized Pharma. I am retired, but still do evaluations for State and Federal grants to universities.
All of these tests we are talking about here don’t seem to account for Placebo Effect. I remember studying about how to minimize this powerful effect. Placebo Effect can account for 30-40% of results.
Do I trust the doctor? How much? This is a powerful actor in these studies.
Is there confirmation bias? The mind is a powerful thing.
These are why we need double blinding in order to find out, with confidence, that we are accurately seeing what nature is telling us.
Also, we are looking for a rare event (death end point). When looking for a confidence result of, say 95% confidence (P<0.05), for a rare event, I remember being shocked that we would require about 200 people in each group, both treated and non-treated, where both the subject and the studier are removed in a randomized fashion.
We had a textbook that we would use called, "Statistics of Rare Events". Surprisingly large numbers of replication had to be used. As the rare event incidence is lower and lower, the number of replications needed goes up geometrically.
It is confusing for people trying to make sense of this stuff. Common sense has its limitations, because nature most often is counter-intuitive.
Enjoy life and forget about mocktons and jo novas who live in horrible countries like Britain and Australia with no personal freedoms dictartorships police states and no nothing about cold viruses or knows nothing about climate or metereology Viva America https://onlineradiobox.com/br/bossanova/?cs=br.bossanova&played=1&lang=en