These targets are arbitrary but of the Government’s making.
A correspondent has taken great trouble in providing Owl with the following article on how the figures on Coronavirus test are being “massaged” to fit a political narrative. The number of tests should be a straightforward matter of fact. Unfortunately they are not.
This is particularly relevant tonight, as Boris Johnson “surprises” the nation with an earlier than expected announcement of the start of a “test, track, trace and isolate” system. Owl believes this is to distract us from Dominic Cummings – “nothing to see here, time to move on”.
The launch looks to be premature: the NHSX phone app is still under development; testing capacity cannot possibly meet Boris’ new target of 200,000 by June 1, and the results are taking far too long to process. And there are accounts of those recruited to track and trace having little or no training.
Regrettably, It is yet another example of the Government launching new initiatives by megaphone then failing to deliver, when what we need is a Government we can believe in and trust.
People might react to messaging and spin but a virus is completely unaffected.
The correspondent writes:
I regard More or Less as one of the best programmes on Radio 4 because it’s devoted to an accurate analysis of various statistical claims. However, there is one that is particularly interesting at the moment because it takes apart the government’s claims to have reached the declared target of 100,000 Coronavirus tests per day. (All the material included below was taken from More or Less BBC Radio 4 20 May 2020.)
On 15 May the government claimed that 136,486 Coronavirus tests were completed. I think most of us would read that as meaning that 136,486 people were tested. Unfortunately that is not the case: only 43,298 people were tested for diagnostic reasons. The other 93,188 claimed tests included home testing kits sent out, whether or not they were completed; tests done for research purposes; and multiple tests done on the same patients – multiple tests done for very good reasons but nevertheless carried out on one patient. In fact, each person tested at a drive-in centre was counted twice in the government’s daily figures because each of those tests involved spitting into a container (one test) and taking a swab from inside the same person’s nose (one test). So these represent two tests within the government’s figures.
The programme shows that the real figure of tests done is less than one third of what the government was suggesting with its figure.
It’s a 30 minute programme which deals with other statistical claims but if you’re only interested in the conclusions relating to the number of tests claimed to have been done each day, they are around the 16 minute mark. The reporter explains some of the key points.
Why are non-diagnostic tests a source of distortion? “While these tests are potentially giving vital information about the prevalence of the disease, the Department of Health and Social care say they aren’t being used to diagnose cases. The government itself doesn’t count people who receive these tests as people who have been tested. Nevertheless, these non-diagnostic tests are being added to the figures and they’re what’s making it possible for the government to keep claiming that the target has been hit. […] If you remove the posted kits because we don’t know how many were actually carried out and the research kits because the government itself doesn’t count those as people tested, then on the fifteenth of May we go from the government’s figure of 136,486 tests to only 69,900 which we know were actually carried out for diagnostic purposes.” What about the number of people tested? “Of the 69,900 tests we know were done on the fifteenth of May, they were done on 43,298 people.”
So the government’s figures are a combination of sleight of hand (including testing kits sent out whether or not they were used and multiple tests on individuals) combined with straightforward misdirection (including in the reported figure tests which the government itself does not regard as people tested).
The full programme is available here: https://www.bbc.co.uk/programmes/p08dnd81