FALSE POSITIVE TEST RESULTS is a major issue & doesn't seem to be fully appreciated. An unknown false +ve rate & low prevalence is a recipe for trouble. With virus prevalence low & declining the data will be noiser. The false +ve rate of the test will mean that the results 1/n
2/n never reach zero +ves & will likely add more noise to the results (since not every test & batch will have an identical rate). But all we see are the overall results, without knowing how much is virus & how much is false +ves. One sign that prevalence has declined to near ...
3/n or below the false +ve rate would be that +ve test count figures stop falling. The R rate, if calculated from changes in +ve test counts, would then rise from below 1 to around 1 & remain there (with some noise) indefinitely. I fear how that might be interpreted.
End.
End.
@threadreaderapp unroll
