Read on Twitter
NEW on Rapid Tests!
Exceptional data that should help path to at-home highly accurate rapid tests
In over 860 samples tested w @AbbottNews BinaxNOW, only ONE was falsely positive! The Specificity was > 99.8%!
Why is this so important... and what else happened here
1/
Exceptional data that should help path to at-home highly accurate rapid tests
In over 860 samples tested w @AbbottNews BinaxNOW, only ONE was falsely positive! The Specificity was > 99.8%!
Why is this so important... and what else happened here
1/
First: here is the pre-print
https://www.medrxiv.org/content/10.1101/2020.11.02.20223891v1
Which @apoorva_nyc covered briefly recently prior to the study release
https://www.nytimes.com/2020/10/15/health/coronavirus-rapid-test-binaxnow.html?referringSource=articleShare
2/
https://www.medrxiv.org/content/10.1101/2020.11.02.20223891v1
Which @apoorva_nyc covered briefly recently prior to the study release
https://www.nytimes.com/2020/10/15/health/coronavirus-rapid-test-binaxnow.html?referringSource=articleShare
2/
Second, we are so frequently talking about sensitivity bc no one wants to miss a case (will come back to this...) but what is equally or even more important when we consider frequent rapid antigen tests to suppress outbreaks with millions of them daily is SPECIFICTY
3/
3/
That is, how many false positives
Although Abbott EUA suggest 1.5% false positive rate (too high if using millions a day for public health screening) numerous reports now from “field use” are showing MUCH better
Here instead of 1/75, they found only ~1/850 !! false pos
4/
Although Abbott EUA suggest 1.5% false positive rate (too high if using millions a day for public health screening) numerous reports now from “field use” are showing MUCH better
Here instead of 1/75, they found only ~1/850 !! false pos
4/
A number of other studies and non-researchers using in different institutions are finding similarly VERY good specificity - all above 99.5%!
5/
5/
Sensitivity here was also very good and precisely as expected for an antigen test.
Here, >93% of positives with moderate or high viral RNA loads (measured via PCR) were detected.
PCR+ Samples with LOW RNA loads were uniformly missed. 6/
Here, >93% of positives with moderate or high viral RNA loads (measured via PCR) were detected.
PCR+ Samples with LOW RNA loads were uniformly missed. 6/
Essentially all data at this point is demonstrating that samples beyond Ct values of 30 (depending on the lab/assay, but this is largely the case) are largely unable to be cultured and thus do not appear to usually have viable / likely transmissible virus.
7/
7/
However, PCR can stay positive for weeks after culturable virus disappears. So, while antigen test in this study and others miss those samples - when it comes to public health screening, where the goal is catching INFECTIOUS people, this is OK!
(See graph in this tweet here)
8/ https://twitter.com/michaelmina_lab/status/1324251571392286720
(See graph in this tweet here)
8/ https://twitter.com/michaelmina_lab/status/1324251571392286720
Although the BinaxNOW is currently only authorized for symptomatic use by a medical professional, this study in random people on the street shows just how powerful these tests can be!
9/
9/
Right now our surveillance testing plus contact tracing is unlikely to identify >5% of ppl in time to interrupt their onward transmission
Wide deployment of antigen tests like this one, for frequent use, 10M/ day could greatly enhance this to 40% detection in time to act!
10/
Wide deployment of antigen tests like this one, for frequent use, 10M/ day could greatly enhance this to 40% detection in time to act!
10/
This study adds to the mounting evidence that these tests can be used in a comprehensive program to beat back this virus. And to do so in a way that could be independent of major lockdowns and enable individual decision making by people, in privacy of their own homes.
