select search filters
briefings
roundups & rapid reactions
before the headlines
Fiona fox's blog
The Office for National Statistics (ONS), has published the latest data from their COVID-19 infection survey.
Prof James Naismith, Director of the Rosalind Franklin Institute, and University of Oxford, said:
“Today’s numbers (which report the state of play up to 11 September) tell a mixed story. The good news is that England has if anything shown a slight decline. I would have expected the effect of school to begin to be seen. Around 1 in 80 people are infected with covid19 in England. If the prevalence in England falls again in the next report, it would appear that school re-opening has not lead to a surge. There is no doubt the vaccine campaign has saved lives. The most important thing now it to reach the unvaccinated.
“Worse news is that Scotland has significantly diverged with cases plateauing at much higher level (1 in 45) than other home nations or regions within England. Scotland is touching levels not seen before in previous peaks. Without the vaccine, the health care system in Scotland would have been overwhelmed and deaths reached new horrific highs. At close to twice the prevalence than anywhere else, something appears to be different in Scotland. I would expect cases to fall in Scotland since the peak has been reached, however, as winter arrives we may see a further peak. Certainly Scotland is not well placed for the oncoming winter. With such a high prevalence for so long, the health care system is under strain. I think it is very important to figure out quite quickly why Scotland is in such difficulty. If there have been policy errors, the sooner we learn what they are and avoid them in the future the better.”
Prof Kevin McConway, Emeritus Professor of Applied Statistics, The Open University, said:
“This week’s edition of the regular ONS release on their Covid-19 Infection Survey (CIS) takes the data on prevalence up to the week 5-11 September. Broadly speaking, the overall position across the UK on the estimated percentage of people who would test positive for SARS-CoV-2, the virus that caused Covid-19, is pretty static – if anything, it is overall slightly down on the previous week. But the position does vary between the four UK countries, and also between different regions of England. For England as a whole, ONS say that the trend in the percentage testing positive is “uncertain”, though if anything it may be slightly downward. For the most recent week ONS estimate that 1 person in 80 in the community population would test positive, with a margin of statistical error from 1 in 75 to 1 in 85. The estimate for the previous week was 1 in 70, but the margins of error for the two weeks overlap. For Wales, ONS say that the recent trend is increasing, with an estimated 1 in 60 testing positive in the most recent week. This looks as if the position in the latest week is rather worse than in England’s 1 in 80. But the margin of statistical error is wider in Wales, 1 in 50 to 1 in 80, because the number of people swabbed for the survey there is relatively small, so it’s not definite that the infection rate is worse in Wales. ONS also estimate that the rate of testing positive is trending upwards in Scotland, though slowly, slower than in the previous week, with an estimated 1 in 45 testing positive in the most recent week. The exception to this rather gloomy picture is Northern Ireland, where ONS estimate that the trend is downward, with the infection rate in the latest week being about a fifth lower than the week before. The latest estimate is that 1 in 75 would test positive there, but the margin of error is wide (1 in 55 to 1 in 100) because the sample size is smaller than in the other countries.
“Overall the position is that the estimated rates of testing positive for the most recent week are much the same in three of the UK countries (England, Wales, Northern Ireland), allowing for the margins of error. In Scotland the rate is higher than in the other countries, higher than the previous week, and so again higher than it has been every week since the CIS began there last October. These rates are all still high, though, and, adding up the central estimates across all four countries, 892,000 people would have tested positive for the virus in the week 5-11 September. Vaccination means that these high numbers aren’t feeding into hospitalisations or deaths at the rate that they would have in the past with such high infection rates – but they certainly show that there’s a lot of infection around and that we can’t relax everything.
“This roughly static pattern doesn’t entirely match what’s been seen in the new confirmed cases on the dashboard at coronavirus.data.gov.uk. There, in all four countries, with the possible exception of Northern Ireland, there’s some indication that new cases have been falling slowly since the first few days of this month, on the basis of the numbers classified by when the person was tested rather than when the test was reported as a new case. One reason why the trends in the CIS figures might differ from the trends in the dashboard new confirmed cases is that the dashboard cases depend on what sort of people, and how many of them, are being tested routinely for the virus, and changes in those things could bias the figures in ways that don’t affect the CIS estimates. But another reason is more encouraging for the immediate future. The dashboard figures are for new cases, while the ONS prevalence figures count anyone who would test positive for the virus, whether they have a brand new infection or have already been infected for a few days or even a week or two. If numbers of new infections were increasing, but start to decrease, that decrease would take longer to show up in the figures for the prevalence of all infections, because the prevalence figures include people who have already been infected for a time and were first infected when the rate of new infections was higher. If that’s what is going on, we might expect to see the CIS prevalence estimates start to fall in the next week or two – but of course that isn’t guaranteed.
“ONS do estimate the incidence of new infections from the CIS, though for a set of technical reasons, the estimates are relatively less precise than for the prevalence estimates, and they also take longer to be produced so that the latest ones are for an earlier date than the prevalence estimates. The latest incidence estimates are for the week 22-28 August. For that week, ONS estimate that the rate of new infections continued to increase in Wales and in Scotland, decreased in Northern Ireland, and the trend was uncertain in England. Overall, the central estimate across all four countries for the week 22-28 August is that about 78,000 people were being newly infected every day, or roughly 547,000 in the whole week.
“Back on the prevalence estimates, one reason that the overall trend in England is uncertain is that patterns are different in different regions, and even within most regions the trend is uncertain. ONS consider that the rate of testing positive was rising in the North West in the most recent week (ending 11 September), falling in the West Midlands and the East of England, and uncertain in all the other regions. Estimates regional rates vary between 1 in 60 testing positive in the North East and in Yorkshire and the Humber, to 1 in 120 testing positive in the East of England, though there is a lot of statistical uncertainty about those figures, because the number of people sampled in an individual region is relatively small.
“Trends and estimates are also relatively less certain for individual age groups in England, again because the number of people sampled in a single age group is a lot smaller than for the whole population. ONS estimate, for the most recent week, that the percentage testing positive increased for those of secondary school age (school years 7 to 11) and for the age groups of 50 and over. But rates of testing positive remain highest in young people, with an estimated 1 in 35 testing positive for school years 7 to 11, and 1 in 45 for school year 12 (age 16/17) to age 24 years. For those aged 50-69, the estimate is considerably lower, 1 in 120, and lower still for those aged 70+, at 1 in 150.”
https://www.ons.gov.uk/releases/coronaviruscovid19infectionsurveyuk17september2021
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 its Advisory Committee. I am also a member of the Public Data Advisory Group, which provides expert advice to the Cabinet Office on aspects of public understanding of data during the pandemic. My quote above is in my capacity as an independent professional statistician.”
None others received