Professor Philip Nolan:

I will answer the question the other way around. A lot of the things people are thinking about are serial testings in particular settings in order to pick up an outbreak early and then move in and manage that outbreak. It is an utterly different type of circumstance from the one-off test we are discussing. To give an example, if one gathers 60,000 people together, there is another problem with the one-off testing. If one gathers 60,000 people in Croke Park at a higher prevalence than we have - we expect that by September prevalence will be lower - so let us say 150 per 100,000, with the test sensitivity at 70%, which is generous to antigen testing, as opposed to 99.8%, one can expect about 90 infections in that crowd. Perhaps the test will pick up 40 of them, miss about 20 and would also generate about 120 false positives. If the specificity is worse, there would be more of those. The first thing to mention is that of those 90 people, perhaps 40 or 50 of them would be symptomatic. They should not even be considering going to see a match. They should not be going for antigen testing. They should be calling their GP and seeking a PCR test. We are now down to 45 asymptomatic people. In that 60,000 people that we are considering to rapid antigen test, there are 45 with asymptomatic infection. We might detect one in two of those with a rapid antigen test. We will find 22 or 23 cases and let 22 or 23 potentially infectious people into the event. They may not be as infectious as the symptomatic people, but they are potentially infectious people. There will also be 120 other people. We get five false positives for every case we are detecting. It is not a big thing-----