Some trade-unionists have suggested swab-testing of all workers in each workplace before a return to work.
The Tory government’s focus on the crude total of test numbers as the big thing has boosted this idea.
Full isolation pay for those with symptoms, or identified as contacts of virus-sufferers, and social distancing plus PPE where necessary in the workplace, will help much more. So will regular (instant-result) temperature checks, widely used and effective (so far as we can tell) in South Korea, Taiwan, and Hong Kong.
More testing is surely needed for a “tracking and tracing” policy. When scientists such as Devi Sridhar talk of “mass testing”, what they mean is large test numbers, but targeted on such a policy.
“Mass testing” in the sense of testing the whole population weekly, or whatever, would mean a scale a hundred times bigger than the highest-testing countries such as Israel (or 400 times bigger than S Korea, 1600 times bigger than Taiwan), and would probably not be do-able even with a workers’ government.
One-off pre-return-to-work testing? Employers might be happy to go for it if the government would organise it. But for workers it would be much more a token than a protection.
The tests are much less than 100% reliable. 90% reliability, say, looks good, and is good for some purposes, but has unexpected effects when you’re testing for a small minority trait.
Even if the numbers of those who have been infected at some time between January and May are at the highest end of estimates, most workforces of, say, 100, will have no, or at most one or two, infections at any one point of time.
With those small percentages of “true positives”, “false negatives” and “false positives” become a sizeable problem. The “true positives” testing positive will be swamped by “false positives”, and “true positives” may well test negative.
In Australia, for example, generally reported as having “done well” in the pandemic, if you get a virus test and test negative, you’re regularly instructed to quarantine for 14 days anyway, on the grounds of whatever symptoms motivated you to take the test.
And a one-off test, even if 100% reliable, tells us only who has the infection at the time of the test. Or rather, who had the infection three to five days before that time: the tests detect the virus only from about three to five days into the infection. According to the WHO, the time from infection to symptoms is typically five days. Sometimes as long as 14, but typically five.
Testing all the workers in a workplace and then restarting work a few days later, when the results come back, would pick up very few who wouldn’t have symptoms by the restart.