9

Recently I saw the local media (Romania) presenting a case that was initially considered COVID-19 positive and two subsequent tests led to negative results. As far as I know no official bothered to offer an explanation and I assume that this is expected in some cases (e.g. initial test was not done correctly).

Like any other tests, I expect to have an accuracy < 100% and I am wondering what that accuracy is for current tests.

I have read here about the issues with US tests and this article shows how tests are done in China:

(..) the main way to test patients is using the Reverse Transcriptase Polymerase Chain Reaction (RT-PCR) technique.

This paper deals with RT-PCR accuracy, but I cannot understand its conclusions.

By accuracy I mean (total tests - false positives - false negatives) / total tests.

Is there any figure about the accuracy of COVID-19 tests?

4
  • Does this answer your question? How accurate are coronavirus tests? – cbeleites unhappy with SX Mar 20 '20 at 15:37
  • @cbeleitesunhappywithSX - yes, just that the question was asked 12 days after this one and it is strange that no one notice that it is a duplicate of this one. Thanks. – Alexei Mar 20 '20 at 15:42
  • I also noticed it only now because of the new answer. I'll also flag the other as duplicate and let the mods sort it out. – cbeleites unhappy with SX Mar 20 '20 at 15:46
  • @Alexei I've merged your question into the other one, which means the answers to this question are now merged into the duplicate. I went by which question had the most answers and comments, not by which was first. – Carey Gregory Mar 20 '20 at 19:02