select search filters
briefings
roundups & rapid reactions
before the headlines
Fiona fox's blog
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 ONS Coronavirus Infection Survey (CIS) is really the only reliable source we have, these days, for trends and levels of infection in the UK with the virus that can cause Covid-19, regardless of whether people have symptoms. Its results are based on PCR tests on a representative sample of the UK population (aged 2 or over, living in private households rather than institutions), who are tested only for the purpose of tracking patterns of infection.
“The numbers of cases on the Government dashboard depend very much on who gets tested (and reports the test results to the health services), and testing is much less common that in used to be. The other representative population survey that used to take place in England, REACT-1, stopped after March this year. Estimates based on data from volunteers who report to the ZOE app are still published regularly, but cover only infections that cause symptoms, and many infections don’t cause any symptoms but can still sometimes be passed on to others.
“This week’s data release from the ONS CIS takes the picture forward to the week ending 14 September for England and Wales, 13 September for Northern Ireland and Scotland. The picture is mixed, but not entirely encouraging, though I don’t yet find it particularly concerning. The official estimates for the four UK countries for the most recent week add up to about 928,000 across the whole of the UK, which is about 47,000 more than the previous CIS publication for the week ending 5 September. That’s an increase of about 5% in just over a week – not a huge increase, and we’ve got to bear in mind that these estimates come from a survey and are subject to a certain level of statistical uncertainty.
“The pattern, however, does differ from one part of the UK to another, and between age groups. ONS estimate that infections increased over the most recent week in England and in Wales, and decreased in Scotland and in Northern Ireland. It’s harder to be sure of the trends in the three devolved countries than it is in England, because there’s more statistical uncertainty about the infection levels there. That’s simply because the numbers of people swabbed for the survey is considerably smaller than in England. The estimated increase in Wales over the most recent week is fairly substantial, about 20% between September 7 and September 14, and the estimated decrease in Northern Ireland is also large, about a quarter fewer infections on September 13 than on September 6, a week earlier – but both these figures are subject to rather wide margins of uncertainty. In England the rise was, proportionally, rather smaller than in Wales, with the estimated numbers infected on 14 September about 10% higher than on 7 September. Because many more people are swabbed for the survey in England than in Wales, the trend is much more precisely estimated in England. In Scotland the decrease between 6 and 13 September was estimated as about a fifth (about 20%).
“It’s important to note that these changes for England, Scotland and Wales basically put the estimated level of infection back to where it was two or three weeks earlier, in mid to late August. (In Northern Ireland, estimates have varied around a relatively low level for some weeks longer.) There is certainly neither a really clear wave of increase yet in England and Wales, nor a clear declining trend in Scotland. These trends may continue for more than a week or two, or they may not. Also, this survey can’t directly tell us the reasons for any trends, though some of the detail can give grounds for some speculation on why they might occur.
“The CIS report gives estimates for specific age groups in England (where the number of people swabbed is large enough so that these estimates are not too imprecise). There is pretty clear evidence of an increasing trend over the most recent week in two age groups – children of secondary school age (school years 7 to 11) and people aged 25-34. ONS regard the trends in all the other age groups as uncertain. It wouldn’t be surprising if there were some increase in infection as people come back from summer holidays and as the schools reopen, and maybe the increase in the secondary school age group in England is part of this – though I must emphasise that I can’t tell from these results whether that’s definitely the cause. But, even if it is, there’s certainly no clear indication yet that it will continue. In Scotland, where schools go back about a month earlier than in England and Wales, there was a slight upward bump in the estimated infection level a couple of weeks after schools restarted, but it was very small and the estimated infection rate has now fallen back to below its level just before schools went back. Maybe the same sort of thing will happen in England, but there’s simply no data yet to tell.
“It’s also important to point out that, in England, the estimated infection rates for all three age groups that contain school age children (age 2 to school year 6, school years 7 to 11, and school year 12 to age 24) are rather lower than for all the age groups aged 25 and older. Currently the highest estimated rates are for those aged 50-69 and those aged 70 and over.
“The trends in infection rates differ between English regions, with ONS estimating increases over the most recent week (to 14 September) in the North East, Yorkshire and The Humber, West Midlands, and London, a decrease in the South East, and uncertain trends in the other four regions.
“Of course I’d be happier if infections were decreasing across the whole country, and the fact that hospital admissions connected with Covid-19 have been increasing in England also does not make the happiest reading. But I’m not hugely concerned, not yet anyway, because these very recent trends have certainly not yet taken infection levels very high in England or in Wales, and we can’t yet be sure how long they might continue. Also, despite the statistical uncertainty, the falling trends in Scotland and in Northern Ireland are encouraging. I’m keeping a closer eye on developments in infection rates, but I’m certainly not going to lose any sleep over them yet.”
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. My quote above is in my capacity as an independent professional statistician.”