The bottomless incompetence of Public Health England

The UK currently requires travellers entering the UK to go into quarantine for 14 days, except for those coming from countries in the 'travel corridor'. 

Some other countries simply test people for SARS-Cov-2 when they land, so why don't we?

The answer is that Public Health England - for it is them again - reckon that testing upon arrival would only identify seven per cent of cases. 

This is based on a PHE model which has a rather substantial flaw, as a report out today shows.

 

This 7% assumes that all infected travellers who are symptomatic or detectable with a test on departure do not board flights to the UK and therefore only travellers who become detectable during the course of their flight are included in the 7%.

 

You read that right. PHE simply assumes that a traveller who is symptomatic or detectable ‘prior to boarding their flight [...] does not make it onto their flight’ because of ‘exit screening or the traveller being too ill to fly’. This shows a faith in the integrity and intelligence of the public that experience suggests is unwarranted. It also assumes perfect testing prior on departure.

 

While the PHE paper assumes that all individuals intending to fly are infected (see section 4.5 below), only a proportion actually board the plane and travel to the UK. This is because the PHE paper makes the assumption that 100% of the symptomatic as well as the asymptomatic passengers who are detectable at the time of boarding do not fly. As these potential travellers are identified before exiting their country of origin, they are not counted towards the percentage of passengers detected (the 7% figure). However, this part of the infected population would also be detected if they flew to the UK and therefore need to be accounted for in considering the effectiveness of a testing regime. Hence, the paper bases its results on a false premise by assuming, but not accounting for, departure testing

This is not a complex technical issue. It is a basic failure of logic and common sense. The assumptions in the model make it impossible for testing on arrival to appear effective.

 

Therefore [in the PHE model], only passengers who become symptomatic/detectable during the flight can be detected on arrival. Clearly, this is a very small proportion of all passengers. Detection rates on arrival are therefore low based on the way the model is constructed.

 

When the data were reanalysed, the number of detectable cases rose from 7% to between 33% and 63%. Not perfect, but since only about a fifth of people who asked to quarantine for 14 days actually do so, it is a very significant improvement. 

 

There are other problems with the model which can you read about in the report if you're interested, but the take-home point is that...

 

The PHE paper is based on a theoretical model and is not calibrated to real-world data.

A fitting epitaph for this useless organisation.