Renzatic on 17/4/2020 at 22:52
Quote Posted by heywood
The 1.5% positive rate they measured in the population sample could be all false positives. How can you measure a virus penetration rate of 1-2% with a test that's only 95% accurate? Of particular concern here is the test specificity, which they estimated based on just a few hundred samples.
Like everything, take it with a grain of salt. We're trying to very quickly discover the mechanics of a virus while in the midst of an outbreak, and due to such, there are bound to be errors. Until it's peer reviewed, and contrasted with a number of other similar studies, it's better to view this as something interesting, rather than outright concrete facts.
In other words, it gives more reason to exercise caution, just in case, but it's nothing worth freaking out over.
...like these people are.
zombe on 17/4/2020 at 23:06
Quote Posted by Renzatic
If you want to see something truly strange...
(
https://www.medrxiv.org/content/10.1101/2020.04.14.20062463v1) COVID-19 Antibody Seroprevalence in Santa Clara County, California
In digest:
Quote:
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.
I am skeptical.
Facebook advertisement is very self-selecting for this kind of study - especially (!) on thous kind of perilous times. I don't see how you could get anything out of it - how do you count for the bias. Asking for "underlying comorbidities, and prior clinical symptoms." might help a tiny unquantifyiable bit - sure (unquanifiably), but ... ex: who is more likely to answer the call (lets assume neither has had any symptoms and is not in any risk group (*))
1) someone who meets a lot of people or for some other reason might wonder if he had it
2) someone who is pretty much the opposite
Now ask which one of thous had a bigger chance of encountering the virus?
To suppress this kind of selection bias - the targets need to, sufficiently, not care about the outcome ... in the middle of a fucking pandemic ... where people have a hard time to get tested ... really?
Also:
Quote:
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.
Which, in my book, makes the X-fold result - dubious.
*) similar biases would happen for the excluded groups too of course - excluded just for clarity.
----------------------------------------
On other, quite old, news ... our unfortunate island case i mentioned before. I later found out that the mass testing planned there is indeed for antibodies (found that some time later - so, not much point to edit that in). While it will be opt-in - i expect nearly everyone who can will do so (if i were living on an island where more than 1.3% are confirmed cases - i would be very motivated to check it out). Now, that would be some interesting data to have ...
Not sure when/if that will actually happen (i think they are still testing specificity etc of the developed tests ... don't know how the budget talks went, but not anticipating any hiccups there).
Renzatic on 18/4/2020 at 00:26
Quote Posted by zombe
I am skeptical.
Facebook advertisement is very self-selecting for this kind of study - especially (!) on thous kind of perilous times.
Going by the study, it was meant to be. They wanted people from Santa Clara, and they separated them by zip code.
I'd say the study would be good for showing how the virus could spread in a localized, populous community, but applying the lessons garnered there to the entire country at large would be pretty fraught.
zombe on 18/4/2020 at 01:57
It is 4:50 here and i am a bit loopy ... but that still makes no sense to me. I don't see how they could infer what they say from probably bad data ... and they kind of admitted that the data could be bad. Filtering bad data by location is nice, but clearly insufficient to fix it (null hypothesis: the huge infection estimate comes from bad self selection and does not pertain to reality ... for example: how did they sufficiently discount/workaround my previous example? If they did not then the null hypothesis stands and the results are hence of questionable usefulness.)
... but i am too loopy for this - need to sleep ... like ... right now. Anyone care to take a closer look at the thing?
Renzatic on 18/4/2020 at 02:27
Well, right or wrong, I posted it. Peruse and discard at your leisure.
Gray on 18/4/2020 at 03:27
Quote Posted by SubJeff
If you both live alone... move in together?
Well, no.
Her son is twelve. And she has OCD issues. And oh, there's a lockdown.
If it was my choice, which it isn't, we'd clearly live happily ever after. As it stands now, we can meet up once a day for a walk. Not ideal, but hey, on the upside, neither of us is dead yet.
lowenz on 18/4/2020 at 05:16
Quote Posted by jkcerda
so I got a feeling
feeling?
SubJeff on 18/4/2020 at 06:13
I'm not trying to be a dick, but I've just read the last two pages and the science up in here is pretty terrible. It would take me too much time to go through it all and explain why every bit is flawed, and it's time I don't have. I suggest people go read up on immunology because that's the worst part of it. (
https://www.elsevier.com/books/immunology/male/978-0-323-08058-3) I suggest this book if you're really interested. . If you want to listen to a podcast that's talking about this issue I suggest Science Vs from Gimlet.
Have fun.
heywood on 18/4/2020 at 13:06
Quote Posted by zombe
On other, quite old, news ... our unfortunate island case i mentioned before. I later found out that the mass testing planned there is indeed for antibodies (found that some time later - so, not much point to edit that in). While it will be opt-in - i expect nearly everyone who can will do so (if i were living on an island where more than 1.3% are confirmed cases - i would be very motivated to check it out). Now, that would be some interesting data to have ...
Whether or not it's any good for getting people back to work depends on the specificity of the test. The false positive rate has to be significantly less than the actual percentage of the population who have antibodies, otherwise positive results give people a false sense of security.
Let's say the tests improve and achieve a specificity of 99%. Sounds pretty good, but if the true number of people with antibodies in the population is 1%, then there's likely to be as many false positives as true positives, and half the people you're green lighting to go back to work don't actually have antibodies. If the true percentage is 2%, then 1/3 of the people you're clearing for work don't have antibodies. And so on.
