Covid: the science and social context of testing

Submitted by AWL on 16 February, 2021 - 6:45 Author: Martin Thomas
Testing centre

Testing, especially rapid testing, was the subject of the third of the “Covid: known unknowns” webinars, held on 11 February 2021.

These webinars are organised by the British Medical Journal in cooperation with George Davey Smith at Bristol University and the Integrative Epidemiology Unit there. George Davey Smith has also discussed the pandemic a couple of times directly with Solidarity, in July and in November 2020.

The first webinar was a broad survey of “known unknowns”; the second was on Covid and schools; the next two, on 25 February and 11 March, will be about vaccines and “Zero Covid”. Recordings from them are available online here.

Aside from discussing technical data on test accuracy, the 11 February webinar looked at the social context of testing and, as one contributor put it, “testing as done” vs “testing as imagined”.

Research reported in the webinar indicates that in Britain currently 57% of those with Covid symptoms don’t go for tests (often because, for lack of isolation pay, they fear they can’t afford to self-isolate if they test positive). Up to 40% of the contacts of those who test positive are not reached, and a large proportion of contacts who are reached don’t self-isolate. The UK has very high levels of testing, but, because of poor isolation pay and other factors, very poor anti-Covid effects from it.

With current levels of accuracy in rapid-result (lateral-flow) tests, they probably give more “false negatives” (infectious people falsely “reassured”) than newly-discovered infections. The tests seem pretty good now at giving few “false positives”. They can be improved, and from the scientists most sceptical about the current use of tests, a chief call was for improved tests.

The big advantage of rapid-result tests is that, although people who get Covid with no symptoms ever (many children, for example) probably transmit infection much less, those who will eventually get symptoms are most infectious in a span including the days just before they develop symptoms. As much as 50% of transmission may be from people who are not yet symptomatic.

If you get a test result two or three days after developing symptoms then having a test, then probably you’ve already done all the transmitting infection you’ll ever do.

The webinar didn’t discuss particular workplace regimes of rapid-testing all staff, say twice a week, in detail; but the balance of its evidence suggested to me that these are useful and can become more so with improvements in the tests. One estimate given was that good regular rapid-testing prevents one transmission per five quarantine-days, while a local lockdown prevents one per thousand quarantine-days; another, that it can avert 88% of transmissions.

One-off or sporadic mass population-wide rapid-testing, as done for example in Slovakia on 31 October and 1 November 2020, looks much more like ineffectual theatre. That Slovakian mass-testing bout was hailed at the time as a success, but Slovakia’s Covid death rate has since the start of 2021 been mostly higher, often much higher, than neighbouring Czechia’s, Hungary’s, or Poland’s.

Add new comment

This website uses cookies, you can find out more and set your preferences here.
By continuing to use this website, you agree to our Privacy Policy and Terms & Conditions.