4 ms·
Pretty much what I thought then. And each result should then have a confidence interval attached to it that the result "patient infected" is correct. Also, dist
by hef19898 5y ago
Pretty much what I thought then. And each result should then have a confidence interval attached to it that the result "patient infected" is correct. Also, distinguishing between "viral DNA present" and "virus present" seems like a lot of nitpicking to me. You can't have the one without the other.
- nradov 5y agoYou're really missing the point. There is no way to calculate a confidence interval for the presence of an active infection based solely on a PCR test. That is a yes or no judgment call made by a clinician based on multiple objective and subjective factors. Recovered patients often continue to test positive even long after the infection has cleared and they are no longer infected. Thus a PCR test on it's own isn't a reliable way to detect re-infection. https://www.uptodate.com/contents/covid-19-diagnosis https://www.uptodate.com/contents/covid-19-diagnosis
- hef19898 5y agoSo now we are talking about re-infection, not about simple infections anymore. But ok, I'll bite. Every clinical testing procedure has to be approved. Every used machine has to be approved and regularly calibrated. So you don't need a confidence intervall for each test, just one for the used procedure / equipment combination. That combination is tested and approved. Are the reults perfect? Of course not, no test is. And no, how many cycles and how the test is conducted is not up to the lab tech. And for re-infection, that's were you need a doctor to properly diagnose, don't you? And that doctor would use PCR test results as part of his diagnosis.