select search filters
roundups & rapid reactions
before the headlines
Fiona fox's blog

expert reaction to latest data from the ONS infection survey

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:

“This week’s release from the ONS Covid-19 Infection Survey (CIS) takes the data up to the week 6-12 June. The estimated number of people who would test positive for the virus that can cause Covid-19 is up again, compared to the previous week. There’s more evidence that the Delta variant is becoming dominant across much (though not all) of the UK. That all adds to the gloom around the announcement on Monday that step 4 of the roadmap out of lockdown in England was to be postponed. So, is there anything positive to be said about the new results?

“I think there is, indeed quite a lot, though clearly the position could be better. First, though the estimated numbers testing positive increased compared to the week before, the rate of increase was rather less than it has been in previous recent weeks. Overall, the official estimate of the number who would test positive across the UK is around 7% higher than the previous week’s figure. But, before that, the week-on-week increases since mid-May have been in double figures of percentages. For England, ONS certainly consider that the rate of testing positive has continued to increase, and that now about 1 in 520 people would test positive (with a statistical margin of uncertainty from 1 in 440 to 1 in 620). ONS report that the trend in the positivity rate is uncertain in Scotland and in Northern Ireland, where they estimate that, respectively, 1 in 600 and 1 in 610 would test positive in the latest week. Those look like slightly lower rates of infection than in England, but the statistical uncertainty means that actually we can’t be sure that they differ. However, rates in Wales remain lower, with a central estimate of 1 in 1,500. When rates are that low, it’s really difficult to estimate them to a high degree of precision.

“We had another set of estimates of the rate of positivity in England just yesterday, from the REACT-1 survey led from Imperial College. That covers only England. For Round 12 of REACT-1, which ran from 20 May to 7 June, the REACT researchers estimated that about 1 in 670 people would test positive. That rate is rather lower than the latest CIS estimate, but it’s over a period that started earlier, and the average (modelled) estimate from CIS for the same period is very close to the REACT-1 estimate. However, the REACT-1 researchers estimated that the numbers of people who would test positive was increasing at around 6% a day during their Round 12, so that the numbers would double in about 11 days. Currently, over the most recent two weeks, the ONS estimated numbers testing positive in England are increasing much more slowly than the REACT-1 estimate, by considerably under 2% per day. The doubling time, if that rate of increase were to continue, would be somewhere around 7 weeks, rather than 11 days.

“Now, it’s pretty well impossible to estimate rates of increase like this accurately, even from large surveys like REACT-1 and CIS, when the overall rate of infection is low – there’s just too much uncertainty. There could be many reasons why the ONS figures are, now, growing more slowly than the REACT-1 estimate. But one possibility arises because the 11 day doubling time from REACT-1 refers to the whole of their round 12, starting on 20 May. Early in that period, ONS were estimating a considerably faster rate of increase than they are now. The ONS data are in accord with the rate of increase having slowed fairly substantially in the last couple of weeks. I definitely can’t be certain that that is happening, but it’s certainly a possibility, and an encouraging one.

“If it’s true (very big If), then these results can’t tell us why it’s happening. It certainly can’t be because of the announcement on Monday of the postponement of step 3 of the roadmap, because that happened after the end of the most recent week in these data (last Saturday, 12 June). It could have something to do with the measures taken in various parts of the country to try to counter the spread of the Delta variant, and it could be something do to with people being more careful because of the publicity in the last few weeks about increasing cases. Or it could be something else altogether (a statistical fluke, even).

“Also, if infections really are rising more slowly that they were, that isn’t evidence that the decision to delay step 3 was mistaken. Infections are indeed still rising, and if they continue to rise substantially, that could still cause serious problems in a few weeks. One of the reasons for the four-week delay was to give time to gather better data on how the pandemic is progressing, and in particular on the Delta variant. These results are part of that process. Given that the decision to delay has been made, it’s important to consider all the data very carefully over the next couple of weeks.

“In England, the estimated rate of testing positive for the latest week is now at about the level of mid-September last year. (The current rates in Scotland and Northern Ireland can’t be compared in that way, because the CIS hasn’t been running there long enough. In Wales, current rates are considerably lower than they were in mid-September, though still somewhat above the rates in August last year.) Back in September, there was concern about rising numbers of cases, just as there is now. The difference now, compared to then, is that we are many months into a successful vaccination programme, that is continuing apace. The CIS data can tell us only about infection levels and trends, not about trends in actual Covid-19 illness, hospitalisations or deaths – how important the continuing rise in infections is will depend on exactly what the vaccinations have done to break the link between infection and serious illness, and we don’t have complete data on that yet by any means, not for the Delta variant.

“There are a few other positive signs in the CIS results for the regions of England and for different age groups – though ONS rightly warn that these are subject to a lot of statistical uncertainty because the number of people swabbed for the survey in a single region or age group is relatively low, and the overall rate of positive tests remains fairly low. ONS estimate that the infection rate in the North West is the highest of all the English regions, about 1 in 180 of the community population, and they also consider that the rate is increasing there and also in the West Midlands (though the West Midlands rate is considerably lower than in the North West, at about 1 in 600). ONS consider that positivity rates were decreasing in the most recent week in the East Midlands and in the East of England, and there and in the South West positivity rates are now very low (around 1 in 1,000 in the East Midlands and the South West, and about 1 in 2,500 in the East of England, though the margins of error on these low rates are relatively very wide). This, overall, could be something of an indication that the rises from the Delta variant are not spreading everywhere – but the continuing increase in the North West is concerning.

“In separate age groups in England, rates of testing positive remain highest in younger people, particularly those between school year 12 (age 16-17) and age 34, where the rate has increased. ONS consider that there are some signs of a decrease in infections the main secondary school age group (years 7 to 11), but that trends are uncertain in all the other age groups. People in all those younger age groups (below age 34) are much less likely to have been vaccinated, or at least to have been vaccinated long enough before the latest week of data for the vaccine to have had its full effect (and of course those aged under 18 wouldn’t have been vaccinated at all except in a few special cases). For people in age groups over 34, where many more would have been vaccinated, infection rates remain low (1 in 800 for ages 35-49, 1 in 1,050 for ages 50-69, and a really pretty tiny 1 in 1,790 for the 70+ group). All those estimates have wide margins of error because of the low rates and relatively small numbers tested for the CIS, but the fact that they are so low and showing no clear signs of an increase is another indication that vaccination can be effective, even against the Delta variant. And that’s encouraging news.”



All our previous output on this subject can be seen at this weblink:


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 received.

in this section

filter RoundUps by year

search by tag