Re:Japanese man tests positive for coronavirus again
by ohazi on Tuesday March 17, 2020 @03:36PM
(#22609310)
I think you have that backwards.
False negative seems extremely plausible. If you're sick, but your viral load is low, it should be possible to collect a sample without a virus particle. If the sample doesn't contain a viral particle, it will necessarily come back negative.
False positive seems unlikely, as you'd have to have something that survives several cloning cycles that ends up looking exactly like the markers you're matching against.
Re:Japanese man tests positive for coronavirus again
by snowmaker on Tuesday March 17, 2020 @03:34PM
(#22609287)
It's the opposite. The PCR tests are more specific (99%) than sensitive (perhaps 90%, we're not really sure).
Re:Japanese man tests positive for coronavirus again
by jennyyang on Tuesday March 17, 2020 @03:35PM
(#22609305)
30% chance of false negative which is why they do 2 days in a row, which bring it down to 9% chance of false negatives.
Re:Japanese man tests positive for coronavirus again
by PeterisP on Tuesday March 17, 2020 @04:47PM
(#22610224)
Sooo, if you release a thousand patients as cured based on these tests, you'll have 90 false negatives...
Re:Japanese man tests positive for coronavirus again
by andruby on Tuesday March 17, 2020 @03:40PM
(#22609362)
It’s the opposite. False positives are highly improbable. False negatives happen more regularly.
If there are very little viruses, and the swab/sample didn’t contain any you would get a false negative test.
EDIT: for those down voting. What about this paper then?
https://pubmed.ncbi.nlm.nih.gov/32133832/