This is a very very good article about the sensitivity of the CORVID-19 testing and the predictive positive and predictive negative results and their impact.
It's quite pithy, but ends up with these conclusions:
"Either way, it is clear that low sensitivity diagnostic tests are a major problem in this COVID-19 outbreak."
The sensitivity of the current CORVID-19 test ( done on a thoat swab) is currently only about 40%.
Contrast that with the current rapid molecular testing for Influenza A or B done in most US hospitals which is about 90-95% sensitive.
This is a huge problem right now for epidemiologists. With a low-sensitivity test in a high-prevalence area, this paper explains you would have to test the same person 8 different times to get a statistically valid negative result. There is just no way any country can do this. Right now the best lab can only process 6,000 specimens a day. If the practice was to re-test negative results up to 8 times to assure they were negative, that would mean they would only test something like 1500 people a day, not 6000
Statistics, Decision Making, And Containing COVID-19 With Unreliable Diagnostic Tests
It's quite pithy, but ends up with these conclusions:
"Either way, it is clear that low sensitivity diagnostic tests are a major problem in this COVID-19 outbreak."
The sensitivity of the current CORVID-19 test ( done on a thoat swab) is currently only about 40%.
Contrast that with the current rapid molecular testing for Influenza A or B done in most US hospitals which is about 90-95% sensitive.
This is a huge problem right now for epidemiologists. With a low-sensitivity test in a high-prevalence area, this paper explains you would have to test the same person 8 different times to get a statistically valid negative result. There is just no way any country can do this. Right now the best lab can only process 6,000 specimens a day. If the practice was to re-test negative results up to 8 times to assure they were negative, that would mean they would only test something like 1500 people a day, not 6000
Statistics, Decision Making, And Containing COVID-19 With Unreliable Diagnostic Tests
Last edited: