How well is the Innova test working in Liverpool?
This is a speculative calculation informed by data available from Liverpool. The source is the Liverpool mayor - reported 23,170 tests done, with 0.7% positive (so about 162) here.
liverpoolecho.co.uk/news/liverpool…
This is a speculative calculation informed by data available from Liverpool. The source is the Liverpool mayor - reported 23,170 tests done, with 0.7% positive (so about 162) here.
liverpoolecho.co.uk/news/liverpool…
Presume specificity 99.6%
No sensitivity data for this test - @UKPHE has data but not public
I estimate sensitivity of 75% based on independent evaluations done for WHO of other similar LFIA tests (range from 50% to 90%) - but in symptomatic patients
finddx.org/covid-19/sarsc…
No sensitivity data for this test - @UKPHE has data but not public
I estimate sensitivity of 75% based on independent evaluations done for WHO of other similar LFIA tests (range from 50% to 90%) - but in symptomatic patients
finddx.org/covid-19/sarsc…
Using a prevalence of 400 per 100,000 you get the following 2x2 table: 
11 of every 20 people told to isolate will not have Covid.
25 of every 100 people with Covid-19 will have negative results **** (I really hope they don’t now go and visit their granny)
One Covid case will have been found for every 330 tested.
25 of every 100 people with Covid-19 will have negative results **** (I really hope they don’t now go and visit their granny)
One Covid case will have been found for every 330 tested.
**** Remember the sensitivity is a big guess – I’ve heard @PHE_uk have done studies in asymptomatic people- it is a scandal that they are not shared.
Real concern that Boris says we can use this for test-and-release – the -ve likelihood ratio is 0.25 - has Boris met Bayes?
Real concern that Boris says we can use this for test-and-release – the -ve likelihood ratio is 0.25 - has Boris met Bayes?
• • •
Missing some Tweet in this thread? You can try to
force a refresh
