Quantcast

Antibody test accuracy and immunity passports

SteveAWOL

Distinguished Member
Looks like even when the UK finally achieves 100,000 tests per day we’ll still have thousands of people being incorrectly diagnosed, as whatever the NHS eventually end up using will provide similar accuracy to the antibody tests recently approved by the FDA over in USA.

This Cellex test has a "sensitivity" of 93.8% so of those who have actually been infected 6.2% will receive a false-negative result and end up self isolating unnecessarily.

Whilst the "specificity" of 95.6% means a false-positive rate of 4.4% and these individuals could be considered eligible for the “immunity passports“ despite never having been infected with SARS-CoV-2.

So if the proportion of general population who are infected is actually 1% then Bayes Theorem shows that 82% of positive results will in fact be false, meaning thousands of people every day potentially being lead to believe they’re immune when they’ve yet to be infected :eek:


1586853469424.jpeg
 

kenshingintoki

Well-known Member
Yeah, the 70% sensitive swab we've been using here is really not going to help with immunity passports etc.

Work (NHS) phoned me up a few days ago asking if I'd like to be tested. They noted that the test they're using is really only effective at identifying people between days 1 to 4 of illness.

So although our testing capacity for what we have today might be increasing; we're still extremely far away from the holy grail.
 
Last edited:

richp007

Distinguished Member
This doesn't look that great does it?!

One area of all this where I've not really delved. But was hoping for something more clinical than this.
 

GadgetObsessed

Well-known Member
Looks like even when the UK finally achieves 100,000 tests per day we’ll still have thousands of people being incorrectly diagnosed, as whatever the NHS eventually end up using will provide similar accuracy to the antibody tests recently approved by the FDA over in USA.

This Cellex test has a "sensitivity" of 93.8% so of those who have actually been infected 6.2% will receive a false-negative result and end up self isolating unnecessarily.

Whilst the "specificity" of 95.6% means a false-positive rate of 4.4% and these individuals could be considered eligible for the “immunity passports“ despite never having been infected with SARS-CoV-2.

So if the proportion of general population who are infected is actually 1% then Bayes Theorem shows that 82% of positive results will in fact be false, meaning thousands of people every day potentially being lead to believe they’re immune when they’ve yet to be infected :eek:


View attachment 1281770
If those figures are correct then the test is surprisingly accurate. (Compared to many other medical diagnostic tests.)
 

simonblue

Distinguished Member
Looks like one has been approved :smashin:


 

Ste7en

Distinguished Member
It sounds pretty good as well. Also reporting that antibodies could protect for up to three years.

We could do with some good news.
 

fat jez

Well-known Member

SteveAWOL

Distinguished Member
Last edited:

usako

Member


I have not seen any immunity against common cold.
 

Latest News

CES 2021 to go ahead 'in person'
  • By Andy Bassett
  • Published
Philips 800 Series OLED TVs come to UK in July
  • By Andy Bassett
  • Published
AVForums Podcast: 31st May 2020
  • By Phil Hinton
  • Published
MQA expands global partnerships for high end experience
  • By Andy Bassett
  • Published
Samsung 2020 TV app lineup upgraded
  • By Andy Bassett
  • Published
Top Bottom