From a general, non-expert look at what is being written about testing, or the lack of it, mostly based on US sources, I wonder if there may be a complex of different reasons lying behind the decisions that have been taken, and altered over time, in the UK.
Firstly, I suspect there may be a range of different reasons to undertake ‘tests’ – i.e. testing for different disease management purposes. Is it possible that whilst it may be useful to test in order to identify and track an infected individuals’ contacts and then quarantine when numbers are relatively low, this may not be effective/feasible or the prime purpose of testing at a later stage after the more extensive spread of the pandemic? Intuitively, once support for the ‘herd immunity’ strategy was adopted – for whatever reason – and time was allowed to elapse, the feasibility for effective testing for this former, specific purpose may have been reduced or lost. One question for the epidemiologist/statistician is on the extent that testing and tracking contacts now can make good what might have been achieved by testing then.
I can envisage that testing now and subsequently may be directed more at (a) diagnosis of especially vulnerable people who have fallen ill in order to rule in/rule out the virus; (b) ruling out infection so key workers don’t need to be isolated uneccessarily; and (c) testing in a survey of the wider population as part of a statistical ‘surveillance’ exercise. There may be other purposes that I can’t think of.
The following US source discusses the relative failure to test in the USA and identifies a range of reasons, some highly critical of Federal Government bureaucracy:
The following source by contrast focuses on science: specifically it discusses technical limitations in the testing available in the USA.
“The currently available diagnostic test is a PCR test developed by the CDC (US Centre for Disease Control & Prevention), which looks for RNA from the virus. However, hospitalized patients infected with the new coronavirus can have test results that vary from day to day because the amount of virus produced by the body can change throughout the course of the illness, ….”.
It goes on: ‘Repeat testing may be necessary to determine if a suspected person has been infected or when a patient is no longer infectious. “The take home message is that a test that looks at a single time point is not sufficient to rule out infection”.’
And finally: “Evidence from the case in Washington State also suggests that the severity of the illness does not necessarily correlate with levels of the virus in the body – meaning someone can be very infectious without seeming very sick. “That’s why there’s concern that patients who are minimally symptomatic may be fueling the outbreak simply because they don’t feel sick enough to go to the hospital,…”.
Very far from straightforward issues look to be involved. It would be fascinating to have more insights into the decision making processes within the UK governments as they evolved over time on the nature, purposes and perceived value of testing.