With PCR for COVID-19 in a population with prevalence of 5%, you'd expect to see 49 true positives, ~6-7 false positives, for a positive predictive value of about 90% Given that most positives are retested if they remain asymptomatic, the true PPV is close to 100%
-
-
Show this threadThanks. Twitter will use this to make your timeline better. UndoUndo
-
-
-
yes, but what about the pretty colors?
Thanks. Twitter will use this to make your timeline better. UndoUndo
-
-
-
Specificity must be higher than that. You’re in NSW too, our test data is incompatible with claims that ~1% of negative people tested will result in a false positive. Otherwise we’d have 100+, daily
-
This is exactly what I was thinking. If there really are 6-7 false positives for every 49 real positives in widely tested populations with very low Covid prevalence (like NYC), I don't see how their positivity could be well below 1%.
End of conversation
New conversation -
-
-
I always wondered: how exactly can PCR have false positive? If the sequence is not there how could it be found? Is there other reason for a false positive PCR test other than a contamination? And even if it's there there could be heaps of sequencing errors and false negatives...
Thanks. Twitter will use this to make your timeline better. UndoUndo
-
-
-
She is also talking about testing a population irrespective of symptoms. No test performs well under these circumstances. Test people with symptoms in the middle of a pandemic and the PPV is quite high for any decent test
-
The performance for asymptomatic patients is more important in this case and false negative is the greater concern from a public health perspective. The harm from “treating” a false positive (quarantine) is nit substantial.
- Show replies
New conversation -
-
-
I think the sensitivity is actually much lower , 70% because it’s dependent on a good sample. The specificity is obviously higher than quoted - I think the point with CT values is you don’t know if they are on the way up or the way down , and the sample is not standardised
-
Specificity at worst case is better than 99.96% - sensitivity is only moderate - in testing an enriched population ie: symptomatic/contacts the major issue is risk of FALSE NEGATIVESpic.twitter.com/BUBQDUzZ1i
This media may contain sensitive material. Learn more
- Show replies
New conversation -
-
-
I agree with your point. Also, she completely misses that the bigger risk is false negatives. Do you have a reference for the sensitivity and specificity of a PCR test?
Thanks. Twitter will use this to make your timeline better. UndoUndo
-
Loading seems to be taking a while.
Twitter may be over capacity or experiencing a momentary hiccup. Try again or visit Twitter Status for more information.