UK Covid-19 Testing Figures Untrustworthy

BBC’s More of Less is a real gem, shining a light on numbers bandied about in the news.

Today’s episode again discussed the UK Government’s Covid-19 testing claims.

It demonstrated to me quite clearly why we cannot trust the numbers being presented by the Government (Govt).

I have listened to the extremely informative 5.5 minute segment of More or Less (MoL) covering this topic – from just after 8’30” into the programme to about 14’ in – and taken some notes that I thought were worth sharing:

• On Sunday 10th May Boris Johnson said

“we must have a world beating system for testing potential victims and for tracing their contacts, so that all told, we’re testing literally hundreds of thousands of people every day”

• Matt Hancock has tweeted again on Monday 11th May, claiming “100,490 tests yesterday” (i.e. Sunday)
• MoL have concluded that the Govt did not reach or surpass the 100,000 target.
• Matt Hancock’s figure included 28,000 samples put in the post that day, but yet to be tested.
• Actual tests conducted, that produce results, are “a long way” from reaching its target of 100,000 (let alone “ramping it up”), according to MoL.
• Even by it’s own “somewhat questionable figures”, the Govt has only reached its target twice in May, and MoL don’t think they have ever reached it.
• Govt has not acknowledged that a sample in the post is not equivalent to a test completed.
• Sir David Norgrove, chair of the UK’s Statistics Authority, wrote to the Health Secretary asking that he shows more clearly how targets are being defined, measured and reported, to “support trustworthiness”.  Ouch!
• The Govt won’t publish the actual number of completed tests (positive or negative) from postal samples. Instead they simply add the number of positive tests to the daily number of confirmed cases. Their ‘excuse’ being that they wanted to avoid double counting [my comment: as though this is not possible by other means in this day and age!].
• MoL have no ideas on the actual number of postal tests being carried out, despite repeated attempts to find out.
• It is not just the postal tests that are causing confusion.  
• Since the middle of April, the Govt’s testing data have included tests from other organisations, such as Universities, and their’s are not just swab tests but antibody tests that can show who has had the virus in the past. They are doing this to look at the prevalence of the virus and to answer other research questions, such as how accurate the home testing kits are.
• The Govt say that because the research tests are not for diagnostic purposes they are not included in the daily count of people who are tested. Yet, this week more than 17,500 of these tests are included in the number of tests completed!

“It’s almost as if they don’t care if the number of tests figure is consistent or indeed accurate, as long as it’s big” (Tim Harford, MoL Presenter)

• This leads to another issue. The number of tests carried out is not the same as the number of people tested!
• Now, of course, some individuals may be tested multiple times so we would expect the number of people tested to be lower than the number of tests [my comment: I would say clearly over a period of a week say, for medical staff, but for the general public? surely not].
• But recall Boris Johnson talked of “hundreds of thousands of people every day”.
• On May 10th, almost 70,000 tests were actually carried out (not including the number for those postal tests samples put in the post that day), but the number of people actually tested was 37,000 (as MoL gleaned from Department of Health (DoH) data).
• This is roughly 2 tests per person each day.
• MoL have asked the DoH for an explanation “but they haven’t got back to us” yet.
• MoL are not sure what could be the issue, but wondered if there is an error in the collating and labelling of data. They will keep trying to get an explanation.
• <End of Notes>

After listening to this MoL episode, I was discussing it with my ex-nurse wife who suggested “maybe they are using two swabs”. Duh, of course.

Is this the simple explanation we need? So, I tried to find out what happens at test centres.

The Govt YouTube video on the test centre process (viewed by only about 100,000 people) doesn’t mention 2 swabs; and implies just one.

Then I found Jack Slater’s piece in the METRO (Sunday 15th March 2020)

“One swab will be put in the back of your throat, and another will be placed inside both nostrils.”

So 2 tests are done for each 1 tested person, at least at some test centres; possibly all.

My cynical thought was then: is there some creative ambiguity going on in not distinguishing ‘tests’ from ‘people’. I tend to prefer the cock-up theory of history, but who knows?

Thank you More or Less for again offering a clear interrogation of the Government’s claims on testing.

I would conclude that the Government is in a complete shambles with respect to simply counting the number of actual tests carried out per day, and also, the number of people tested (whose sample or samples have been tested on said day); and clearly distinguishing these numbers.

Unless the Government can demonstrate clarity and accuracy in its presentation of the testing numbers, how can we trust it to implement a coherent strategy to achieve a “world beating system for testing potential victims and for tracing their contacts”?

Or even, how can it execute its basic job of protecting the public?

Currently, one has to conclude that the UK Government’s Covid-19 testing figures are untrustworthy.

(c) Richard W. Erskine