https://www.aei.org/carpe-diem/map-of-the-day-us-population-divided-by-one-thirds-of-covid-deaths/
Hopefully you can tap the brakes on your enthusiasm over antibody testing. Some of the tests have false positive rates as high as 16%, and CDC and AMA caution against using antibody test results to make decisions.
physicians and the general public should not use antibody testing to consider anyone immune to the diseaseādoing so may lead individuals to falsely assume they can stop physical distancing and further the spread of illness,ā
2 Likes
IMHO, this is mostly a map of counties where COVID was not present in non-trivial amounts prior to the effects of sheltering in place (in green) vs places where it had already made inroads (not it green).
The rate of transmission dropped precipitously once shelter in place started. Thus, places where it had not yet established a foothold were less much affected (which is a Good Thingā¢).
What remains to be seen is whether those trends will hold once people stop socially isolating, esp. if they fail to take even simple precautions.
(Lake of the Ozarks, Memorial Day 2020)
2 Likes
Next 2-3 weeks should be interesting.
If the hospitals are still functional in 3 weeks, we may be in good shape.
If theyāre swamped because of last weekend, we have problems.
1 Like
If you meant to reply to me then I think youāve misunderstood the data science. That is the topic here. I agree that the tests should not be used to decide āindividualā risk. For instance, we donāt want an ICU nurse, like my daughter, to throw caution to the wind just because an antibody test came back positive. That is the CDC and AMA warning. All tests have an error rate but meaningful stats can still be obtained so long as the errors are not systematic. However the data for the article I linked is using large scale statistics, where random errors have an averaging affect, to estimate the total number of positive patients. That is very different use than what you are implying by the articles you linked.
Some interesting information about the tests and the error rates for the tests in this video.
The rest of the series by the same channel is very good as well. ( Not a doctor myself. Useful explanations for meā¦ )
To heck with it with you experts. Whose āscienceā are we supposed to follow anyway?
They are all guessing and blowing it out of their anusiā¦
I work in healthcare fraud analytics. Before that I worked in medical research, including analyzing stats for clinical drug trials. That said, I can assure you I donāt understand what some of the ādata scientistsā are doing.
It makes more sense now, since apparently data science means taking test results with false-positive rates far too poopy (the technical term our principal investigator used) for peer-reviewed publication and aggregate lots of those and use it to make policy. That explains a lot about some of the āartificial intelligenceā platforms being peddled to analytics folks.
1 Like
An important post. So far Iāve seen somewhere between 10-15 serology tests, mostly from labcorp, all of which have been negative. What does this mean? Absolutely nothing, of course, but it is interesting none the less.
2 Likes
2 Likes
combine it with this tech, and you are good to go
4 Likes
1 Like
He stole the idea from Homer:
2 Likes
I see this
A genetic quirk of the RAS could be giving women extra protection against the disease.
And think of this
Lol
4 Likes