select search filters
briefings
roundups & rapid reactions
before the headlines
Fiona fox's blog
NHS Test and Trace have released the latest figures for contact tracing in England for the week of 20th to 26th August.
Prof Kevin McConway, Emeritus Professor of Applied Statistics, The Open University, said:
“The first main point in this DHSC bulletin is true, but, I’d argue, gives quite a misleading impression of what the data shows. It says, “6,732 new people tested positive for Coronavirus (COVID-19) in England in between 20 August and 26 August. This is an increase of 6% in positive cases compared to the previous week and the highest weekly number since the end of May. The number of people tested has decreased by 1% in the same time period.”
“Let’s break it down a bit. Yes, the number of positive cases is 6% up on the previous week (13-19 August), but the number the previous week was 6% down on the week before that (6-12 August). These numbers vary from week to week for a whole lot of reasons apart from an actual change in the number of positive cases out there in the country. There will be some random statistical fluctuations. The effort in getting people tested may change, not only in terms of the number of tests carried out but also because the reasons why people are tested will vary too. It’s true that the number of people tested in the more recent week was 1% down on the week before, but the week before’s number was 1% up on the week before that. So really the best summary of the number of people tested over those weeks is that they’ve hardly changed. My interpretation of the data on positive cases is that they have been at roughly the same level since early August, with some fluctuation from week to week which might have nothing to do with real changes in the number of infected people.
“This is all consistent with the results of the Office for National Statistics (ONS) Infection Survey. That involves testing a large sample of people, not because they have symptoms or have had contact with people with symptoms, but because they are a representative sample of people in England. So the Infection Survey results should not be affected by any changes in the availability of tests or in the operation of Test and Trace. For several weeks now ONS have been reporting that numbers of new infections fell until late July, then rose slowly, but that that rise has levelled off. I agree with that assessment (though we’ll know more tomorrow when the next infection survey results come out). Figure 3 in the Test and Trace bulletin, showing positive cases, shows exactly the same pattern – a fall until July, a slow increase after that, but it has now levelled off. That’s what to look at to get a clear impression, not statements comparing just two weeks. (The lowest point of the Test and Trace positive cases is earlier in July than the lowest point from the infection survey, but that’s probably because the number of tests in Test and Trace was still increasing quite rapidly during July, and the more you test, the more cases you find.)
“What about the Test and Trace bulletin saying that the weekly positive cases for 20-26 August is “the highest weekly number since the end of May”? Well, it’s true, but again (I’d say) misleading. The figure from Test and Trace for the very end of May, actually 28 May to 3 June, is 8,842 cases, a lot higher than the latest figure of 6,732. The only reason that DHSC can make the statement that they did make is that the latest week’s figure is a very small amount higher – just 1% (60 cases) higher – than the figure from two weeks before. If the statistical fluctuations had been slightly different, all DHSC could have said is that this week’s figure is the highest since two weeks ago, and what they should be saying is that it’s pretty much the same as two weeks ago. The current weekly level is about what it was in the middle of June. In the middle of June the number of tests carried out under Test and Trace was more than a third lower than it is now, and as I said, the more you test, the more cases you find. If the same number of tests were being done in mid-June as now, there would have very likely been a lot more positive cases under Test and Trace back then.”
Prof Sheila Bird, Formerly Programme Leader, MRC Biostatistics Unit, University of Cambridge, said:
“Despite recommendations in July by the Royal Statistical Society’s COVID-19 Taskforce, see https://rss.org.uk/RSS/media/File-library/Policy/RSS-COVID-19-Task-Force-Statement-on-TTI-final.pdf, Test & Trace has not yet reported how many of the folk it quarantines have actually developed symptoms & tested positive for SARS-CoV-2. We need straightforward answers separately for the two high-risk groups that Test & Trace quarantines:
a) Members of household of symptomatic index cases
b) External close contacts of symptomatic index cases.
“Based on the REACT-1 study, RSS COVID-19 Taskforce (of which I am member) estimated that high risk group A has about 20 times the UK’s community-risk of developing COVID-19 symptoms.
“The excellent paper from Iceland, published this week in the New England Journal of Medicine (https://www.nejm.org/doi/full/10.1056/NEJMoa2026116), affirms – for Iceland – that 20-fold multiplier because Iceland does what Test & Trace fails to do . . . Iceland monitors how many of those whom it quarantines develop symptoms and test positive for SARS-CoV-2. The paper affirmed: “Household exposure was more likely to lead to infection than other types of exposure, which suggests that people who share a household with an infected person should not have contact during quarantine and that CONTACTS of HOUSEHOLD MEMBERS SHOULD BE QUARANTINED.”
“England does not audit adherence to quarantine for those whom Test & Trace asks to stay at home. Nor do we offer swab-testing on a random-day (or pair of days) during quarantine to assess asymptomatic positivity. If the Icelandic authors mean that the contacts of household members of a symptomatic case should be quarantined even before household member has tested positive, then their advice goes beyond our practice in UK.
“The main point, however, is that Test & Trace should be delivering answers for England. We should not have to rely on Iceland to give us answers that we fail to obtain for ourselves!
“British science can & should do infection control better than Test & Trace is delivering: for example, by paying heed to the recommendations of the Royal Statistical Society’s COVID-19 Taskforce or by explaining why those recommendations cannot be followed – for example, because design-problems mean that data are not held centrally which define the quarantine period and personal identifier for each member of high-risk groups A & B.”
Prof Oliver Johnson, Professor of Information Theory, School of Mathematics, University of Bristol, said:
“The Test and Trace programme is a key part of the Government strategy to allow schools and universities to remain open this autumn. This system doesn’t need to be perfect, and even cutting a significant fraction of onward infections will help in this aim.
“One key test metric is the number of positive tests under Pillar 1, representing some of the most serious illnesses. It is encouraging this has remained constant since last week, at under 90 new such cases per day, which is consistent with current low hospitalization levels. Another significant value reported is the overall proportion of people who test positive each week. While this has risen slightly to
1.5%, it is worth noting this is around a third of the value currently reported in France for instance.”
Prof James Naismith FRS FMedSci, Director of the Rosalind Franklin Institute, and Professor of Structural Biology, University of Oxford, said:
“It is very important that these figures continue to be published. The virus is circulating and we will go into the winter with significant numbers of infections. At the moment, most people with the virus are young. The virus has not changed, it is not suddenly targeting young people. Rather our testing is developed enough to see what we missed at the start of 2020, the virus spreading in younger more social element of society. As a group, they face very little danger from the virus with the majority having no symptoms at all hence without testing we can’t see the virus. However, even perfectly healthy young people have died. The even more serious concern is that unchecked the virus will reach levels where it can once again significantly spread into much more vulnerable populations where we will see illness, hospitalisation and deaths. Whilst I am confident that the rapid scientific and medical advances that have taken place mean we will not see the same levels of death or severe illness as spring, circulating virus in vulnerable populations will bring tragedy. We all need the test, trace and isolate system to work in order to check the virus.
“The report itself has some areas for concern, the number of people with positive tests has increased and at the same time the number of people tested has decreased. The UK still has an excellent ratio of people tested to positive cases, there is no reason to be alarmed at the slight decline but this is a move in the wrong direction. Overall testing capacity has increased which is welcome. The test and trace system overall however appears to be showing some strain, with decreases in positive cases being traced and reduction in the contacts reached. The turn around in satellite centres for test results is deficient. Many scientists cautioned that the challenges that setting up a test and track system from scratch in such a short time were being underestimated by politicians and the media.
“The report also notes that close contacts are predominantly within the home. This might mean that social distancing in the wider world is working as hoped. In her comments around restricting household gatherings, the First Minister of Scotland, referred to data showing mixing of different households in a single house to be a key driver of spread.
“I welcome the moves to support isolation of positive cases, but this has some way to go. Without effective isolation, test and trace, will merely serve to better inform of us of a potential unfolding tragedy. There needs to be some imaginative solutions to help people isolate, especially the young and otherwise healthy and those in multi person households.
“We do not have long before the winter is upon us. We should use every second of that time.”
NHS Test and Trace (England) and coronavirus testing (UK) statistics: 20 August to 26 August 2020
All our previous output on this subject can be seen at this weblink:
www.sciencemediacentre.org/tag/covid-19
Declared interests
Prof Kevin McConway: “I am a member of the SMC Advisory Committee, but my quote above is in my capacity as a professional statistician.”
Prof Sheila Bird: SMB is member of RSS’s COVID-19 Taskforce which issued statement on 23 July on how efficient statistical method can glean intelligence from Test, Trace & Isolate.
None others received