The Office for National Statistics (ONS) have released the latest data from their COVID-19 Infection Survey.
Prof Kevin McConway, Emeritus Professor of Applied Statistics, The Open University, said:
“The weekly results from the ONS infection survey have come out a day early this week. There’s some good news – ONS report that, in England, the positivity rate is levelling off. Indeed the ONS central estimate of that rate has fallen slightly for the most recent week (ending 21 November) compared to the previous week. Because, like any results based on a survey, there’s a statistical margin of error, we can’t be sure that the actual rate has fallen – but it’s certainly a possibility. There’s evidence that the prevalence of infection is falling in most regions of England, and indeed also in Wales and Northern Ireland. But it seems to be increasing in the East Midlands of England, and in Scotland, and the position in North East England and in Yorkshire & The Humber isn’t so clear. In the different age groups in England, the positivity rate appears to be level or declining in all age groups except the secondary school student age group, where it might well be increasing (though there’s a lot of statistical uncertainty). So mostly good news, but not without some remaining issues – and where there are falls in the rate, they aren’t yet very rapid.
“It might seem disappointing that progress hasn’t been faster, particularly in England where there has been a lockdown since 5 November. But we’ve got to remember that these latest numbers are estimates of the rate of positive tests in the week of 15-21 November, and that they include people who may have been originally infected some time before then. People continue to give positive test results, on average, for at least 10 days after they were first infected, so some of the people who tested positive in the most recent week would have been infected before the English lockdown began. Usually the ONS infection survey bulletin also presents data on incidence of new infections, which it can do because it tests people repeatedly and can therefore distinguish between new infections and ongoing infections in many cases. However, this week the data on incidence of new infections could not be presented because of a technical problem at the large laboratory in Milton Keynes. That’s a pity, because the incidence estimates could have given a clearer indication of the impact of the English lockdown. ONS assure us that this technical problem doesn’t affect their comments on the trends in positivity prevalence, and I trust them on that.
“As always, the ONS infection survey estimates can give information on the course of the pandemic that isn’t subject to some of the biases of data from other sources. The ONS results come from a survey in which a reasonably representative sample of people from the community population of the four UK countries is tested for current SARS-CoV-2 infection. Other sources, such as data from the Test and Trace system and the confirmed cases published on the gov.uk dashboard daily, give numbers that depend to some extent on who is tested, and in turn that depends on whether they have symptoms, which jobs they are in, whether they are in hospital, and to some extent on how keen people are to be tested. Some of those reason will vary from week to week, and that could lead to changes in the numbers that aren’t entirely related to changes in the numbers of infections.
“That said, there is some good news in the data from the Test and Trace system, also published today1. It shows a reduction in the number of people testing positive in England, for the week of 12-18 November, compared to the previous week. That’s not such reliable information on the course of the pandemic as the ONS infection survey results, but good news nevertheless. What’s not so good is that the performance of the Test and Trace system is much the same as in the last week or two, in terms of the numbers of confirmed cases and their contacts who are reached by the system, or in terms of the time it takes to reach people. The performance is still not good enough to make the required impact on the pandemic.”
Prof Sheila Bird, Formerly Programme Leader, MRC Biostatistics Unit, University of Cambridge, said:
“Confidence intervals about ONS Infection Survey’s weekly reported estimates of SARS-CoV-2 incidence for England is, broadly, of increased incidence after the first and second fortnights. We do not yet have convincing evidence that England’s incidence is back down to around 5 per 10,000.
“Re-weighting to each nation’s demography: England’s sample size is now around 25 and 16 times greater than for Wales and Scotland respectively. Consequently, England’s 95% confidence interval would be expected to be about a fifth or a quarter the width of that of Wales or Scotland. When comparing re-weighting percentages testing positive for COVID-19 for England, Wales, Norther Ireland and Scotland, only Wales has significantly lower % than England in the most recent fortnight.
“England’s re-weighting: Comparison can be made only indirectly between “Official reported estimates of the percentage testing positive, England” (see Table 1a) which are derived by use of modelling and the re-weighted estimates for the above non-overlapping 14-day periods. Modelling gave 1.04% [0.98% to 1.10%] for 31 October to 6 November; 1.22% (1.15% to 1.29%) for 8-14 November; and 1.16% [1.10% to 1.23%] for 15-21 November, all three systematically lower than the corresponding re-weighted estimates. Are modelled Official Reported Estimates NOT re-weighted?
“Northern regions of England versus Scotland: See ONS Infection Survey’s Table 1i for corresponding re-weighted information for all English regions, as shown below for the North East and the North West. The population of the North West is about 2.7 times greater than in the North East, hence narrower confidence intervals.
“In the three most recent fortnights, prevalence was higher in the North West than in North East, only marginally so in the last fortnight. And, in each fortnight, prevalence in the North East was markedly higher than in Scotland.
Prof James Naismith FRS FMedSci, Director of the Rosalind Franklin Institute, and University of Oxford, said:
“Todays data release from the ONS reports the community prevalence up to the 21st November.
“The data set presented this week slightly different this week, due to problems in one the labs, so there is no estimate of new infections per day.
“Up to the 14th November ONS estimated 664,700 people (95% credible interval: 628,300 to 701,200) were infected with covid19.
“Up to the 21st November ONS estimated 633,000 people (95% credible interval: 599,200 to 668,200) were infected with covid19.
“The Kings ZOE system estimates around 500, 000 with covid19 and 29,000 infections per day, with evidence for a decline.
“The prevalence is decreasing as social restrictions have their effect.
“There is further good news in that the number of daily new hospital admissions are falling and the number of people in intensive care has levelled off.
“Sadly, as we say yesterday with 696 deaths announced, it takes time for the gains in social restrictions to feed through.
“It is not unalloyed good news, the ONS data suggest that some areas of the country prevalence may have increased. Obviously, the measures were designed to lower the incidence everywhere.
“As we approach Christmas, there is a great deal of uncertainty about what will happen.
“To help guide our thinking, its worth remembering what we have so painfully learned so far.
“In the UK, high infection numbers lead to high death numbers.
“The UK has a finite capacity (beds, nurses, doctors) of intensive care beds, should we exceed the capacity, the deaths will rise dramatically.
“A doubling in the number of cases is much more serious when case numbers are already high. This is the problem with exponential growth.
“Indoor mixing at Christmas will increase the number of cases. The size of this effect, will be determined by the number of cases as we head into Christmas and the extent to which families mix.
“Today is the US holiday of thanksgiving, it has many similarities to a UK Christmas, children travelling ‘home’ and families coming together over dinner.
“The US has gone into this holiday with a high number of infections, I believe it will be valuable to learn from their experience and be willing to adapt our Christmas.
“Although in absolute numbers, the US death toll yesterday of over 2000 was much larger than the UK, per head of population, the UK total was higher.”
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 Trustee of the SMC and a member of the Advisory Committee, but my quote above is in my capacity as a professional statistician.”
None others received.