Revised arrivals testing figures reveal the staggering cost of every sample sent for sequencing

1 Jul

Although we are not due any more data from the UK arrivals testing programme until next week, I was reading through the small print of the latest weekly release from NHS Test and Trace and discovered a few bomb-shells buried in the small print.

The first thing is that the figures that were already published have been retrospectively revised, explained by the following notes:

"From the publication on 24 June 2021 onward, tests conducted by private providers will also be included in Table 21. Private tests are being introduced in Table 21 ahead of our other MQS tables to ensure alignment with the Joint Biosecurity Centre’s assessments."
"The proportion of tests that are private varies. Since the start of the program there have been roughly even numbers of private tests and tests purchased through NHS Test and Trace, but the proportion of tests that are private has been increasing: in the month of June to date 75% of tests conducted were by private providers."

So it turns out that the government hadn’t been publishing the data that “aligned with the Joint Biosecurity Centres’ assessments”. Up to 75% of it was missing.

Secondly, we now have an additional note to the Table 21 data which seeks to explain why so many of the test results have a country of origin classified as “unknown”:

“Due to an issue with the underlying data, the number of people tested whose recent travel destination is classified as unknown may be counting individuals who did not travel. We have identified the cause of the issue and are looking to resolve as quickly as we can. Tests that are assigned an unknown recent travel destination do not impact the JBC's evaluation of Risk Assessment Status of any individual country”

With the revised data, the number of “unknowns” has gone by a factor of three. If we are to believe the explanation, 25,000 people a week who might not even have travelled are being included in the arrivals testing statistics.

Does it change the story?

Overall, I don’t think the revised data changes that much the story that I told in this blog post. My post was picked up by quite a few of the mainstream media and thank you to those that attributed me as the source. One of the media outlets that didn’t, did come up with quite a nice summary graphic. So to return the favour, they shall remain nameless and I’ll just “take inspiration” from their work..

Here are the figures for the three week period ending June 9, before they were restated:

And this is what they look like after the revisions:

So overall, I don’t think it changes the story that much overall. The number of unknowns looks even more embarrassing. The big change is the volume of tests carried out on amber routes. The testing programme is now shown to have been detecting eight times as many positive cases, but the positivity rate hasn’t changed. 34 cases a day is more than before, but still small, and the testing scheme still hasn’t picked up any variants of concern from amber or green list countries.

Sequencing rates

Something that you can’t see from the previous summary is what has happened to proportion of positive tests that are sequenced. The old stats implied that 38% of positive test results were sequenced. Whilst the number of positive tests has gone up with the revisions, the number of cases sequenced has barely changed. So the proportion of postitive tests sequenced is now only 17%.

*Source: GridPoint analysis of Table 21 data from NHS Test and Trace Statistics (England), published July 1, 2021*

It can also been seen how much the sequencing is biased towards countries already on the red list. 40% of positive tests from red list countries were sequenced, whilst only 2.7% of those from amber list countries were. There were no positive test results from green list countries and therefore nothing to sequence.

The revised figures show that 200,000 people arriving from green and amber list countries over this three-week period had to take and pay for PCR tests after arrival. The main justification given by the government for the need for these tests is to allow them to do genomic sequencing of arrivals into the UK, in order to spot variants of concern. In fact, only 20 of these samples were sent for sequencing and no variants of concern were detected.

The government can’t even be bothered to sequence the tests it is forcing the travelling public to take, at great expense and inconvenience.

What has changed at an individual country level?

With the old statistics, which we now know to have been incomplete, “Unknown” was the biggest country of origin, followed by India and Pakistan. I thought it was odd that red list countries topped the list and we now have quite a different picture. Unknown is still top, but Portugal, which was green at the time, is now shown to be number one in terms of passengers arriving into the UK during the period.

Highest risk countries

Which countries were responsible for the most positive arrivals tests? Before, the top three were India, Pakistan and Bangladesh. That has changed a bit, with Portugal now making into number two spot. The test positivity for Portugal has fallen considerably though. Before the data was revised, the rate was 1.1% and it is now only 0.3%. Most of the missing test results were negative.

Variants

Since the data for the red list countries didn’t change much, there is not much of a change to the final chart that I showed last time. There is one more country which has now been revealed as an origin from which variants of concern were detected. That is Uruguay, which has been on the red list for some time.

In summary

The revised statistics don’t really say anything new about the red list countries.

The increased coverage of the testing data from amber list countries hasn't changed the overall test positivity score, which has remained at 0.4%. What it has done is to bring home how large a volume of PCR tests are being carried out on amber list travellers and how few of them are even being sent for sequencing by the government. And remember that none of them resulted in the detection of any variants of concern.

There were about 380,000 post arrival PCR tests associated with the 200,000 people coming from amber and green list countries in three weeks. In estimating this figure, I’ve allowed for Portugal having been on the green list at the time. At around £30 a test, that adds up to a cost of £11.4m to travellers (Update: it has been pointed out to me that the current rates being charged for PCR tests are at least double this). Only 20 samples were sent for sequencing.

That works out at £570,000 (Update: £1.1m+) in traveller-funded post arrival PCR tests for every sample that the government could be bothered to send to be sequenced.

WTF?

COVID-19UKTestingTraffic Lights

Robert Boyle