3 ms·
The EUA-ness of the test has very little to do with re-positives. RT-PCR tests can only test for the presence of viral RNA in the sample collected. They cannot
by icegreentea2 7y ago
The EUA-ness of the test has very little to do with re-positives.
RT-PCR tests can only test for the presence of viral RNA in the sample collected. They cannot tell you the viability of the viral particles, they cannot tell you if you going to get sicker, or going to get better the next day. We don't even fully understand the relationship between vRNA detection and contagiousness. And all of those unknowns are outside of the scope of the test.
Even in normal times, these tests can only be used to determine if genetic material exists (and sometimes the quantity if the collection mechanism and processing are controlled enough). It must be left up to clinicians to interpret these results.
The current protocol is to wait for 2 consecutive negatives 24 hours apart to call someone virus free. That's a protocol created by clinicians in a time of extreme pressure and need based on empirical measurements and prior body of knowledge.
A 510k cleared COVID test would have most of the same issues because the problem isn't strictly that the tests can't reliability detect viral particles, the problem is that we don't know what actually constitutes "cured".
- _wldu 7y agoRight. The point is that due to testing issues (false negatives at 33%[1], not done properly, etc.) a positive test result after a negative result (when COVID symptoms are present) is not a sign of re-infection or relapse but an indicator of testing problems/issues. [1] https://www.wsj.com/articles/questions-about-accuracy-of-coronavirus-tests-sow-worry-11585836001?mod=e2tw https://www.wsj.com/articles/questions-about-accuracy-of-cor...
- icegreentea2 7y agoYeah, it's a huge problem. But I think its also very important to be precise about it. Careless phrasing about test kit performance can make people think that the problem is that our test kits "just need to be better". These test kits can have fantastic sensitivity when measured on a per sample basis, and absolutely awful clinical sensitivity at the same time. We're incredibly unlikely to get significantly better clinical sensitivity, especially for tests appropiate for population level screening over any reasonable time frame, so our strategies and plans are going to have to accommodate that.