3

I was surprised to hear Sanjay Gupta (CNN) say this morning (4/28/20) that there is significant disparity in the accuracy of the COVID-19 Diagnostic testing. He specifically mentioned that the inaccuracy is as high as 15% in reporting false-negatives.

Has the inaccuracy of the various tests for COVID-19 been publicly reported?

NPR and Others have reported, but what I am looking for is clinical peer-reviewed reports or papers that compare the various COVID-19 diagnostic tests for accuracy, particularly false-negatives.

BobE
  • 680
  • 3
  • 10
  • This really needs more context. A certain level of false-negative testing results is expected, when you test someone who's been infected but doesn't yet have a viral load high enough to detect. This is why concluding that someone doesn't have the virus requires repeated negative results over a period of time. – Mark Apr 28 '20 at 20:00
  • 1
    I think we need an exact quote of what he said. – Graham Chiu Apr 28 '20 at 22:30
  • If so many are saying this, then you should have no problem finding a link to include in your question. – Graham Chiu Apr 28 '20 at 23:12
  • 2
    Need a link to what test they're referring to – Graham Chiu Apr 28 '20 at 23:49
  • 2
    And I think you're supposed to show some element of research in your question. – Graham Chiu Apr 28 '20 at 23:52
  • I think I found your CNN article which you didn't link. It is talking about anti-bodies tests in re those 15%. Antibodies tests in the US have been criticized before. – the gods from engineering Apr 29 '20 at 06:28
  • @Fizz, actually the article you linked never mentions Gupta, was originally broadcast six or seven hours after Gupta's interview in the morning. While I cannot find a video clip of Gupta's interview, to the best of my recollection he was addressing the inaccuracy of the COVID19 diagnostic tests rather than the antibody or antigen tests. Regardless, I mentioned Gupta as it was his interview that brought to my attention. The other citations I provided (NPR and others) were addressing the diagnostic tests rather than the antigen tests (separate but similar problems there). – BobE Apr 29 '20 at 14:34
  • 1
    @Fizz, but I take your point and will edit the question to be specific to the diagnostic tests (do you currently have COVID19 versus did you develop antigens) – BobE Apr 29 '20 at 14:38
  • @BobE Just a quick correction on terminology: you mean antibodies. PCR tests look for viral genetic material (RNA), antibody tests look for host-produced antibodies (using lab-produced antigens), antigen tests look for viral-produced antigens (proteins) using lab-produced antibodies. – Bryan Krause Apr 29 '20 at 14:55
  • @BryanKrause thanks – BobE Apr 29 '20 at 15:19
  • 1
    No paper neither, but some weeks ago, Kekule (a German virologist and epidemiologist who is very prominently in the media right now) mentioned around 20 - 30 % false negative PCR tests due to difficulties with sampling (i.e. the swab not picking up the proper material) and that preferrably two swabs should be taken. – cbeleites unhappy with SX Apr 30 '20 at 15:21

0 Answers0