No this means the 11.4% received a false negative and the test...

First 108 Last

Share
Profile

Follow
LV50

8,443 Posts.

175

28/10/20

13:01

Post #: 48453580
Share

Originally posted by Lazarus65: ↑

Are you sure about that? If there is 88.4% sensitivity, it means 88.4% chance of correct positive diagnosis. The remaining 11.4% can only be false positive....not false negative as you suggest. No?

Surely false positive would be of little epidemiological concern as you think someone is contagious when in fact they are not. This (ie the 'false' part) is only determined/confirmed after a subsequent test (likely PCR-lab based) to the positive Ag/RDT. Meanwhile, 2nd test results pending, as a 'positive', they are isolated and potentially contact tracing commences. High false positives incurs a degree of wasted effort/cost here, but has no negative epidemiological implications. Conversely false negatives are of a great concern to epidemiologists as you have people who armed with the false confidence that they are not infected, may go on to become spreaders or super spreaders. 100% specificity implies zero false negatives.

If my understanding is correct, then very high specificity Ag/RDT's form a key part of the pandemic management arsenal. An arsenal that includes follow up tests plus isolation and contact tracing measures. Very high sensitivity tests determine (inversely) the cost of that overall arsenal.

Expand

No this means the 11.4% received a false negative and the test didn't pick up the virus in the infected person. Hence letting them roam around thinking they are negative when in fact they are positive.

So as you say "Conversely false negatives are of a great concern to epidemiologists as you have people who armed with the false confidence that they are not infected, may go on to become spreaders or super spreaders. 100% specificity implies zero false negatives."

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2636062/

Last edited by LV50: 28/10/20

AT1 Price at posting: 32.5¢ Sentiment: Hold Disclosure: Not Held
1 Upvote

0 Great analysis

Reply