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

expert reaction to latest ONS Infection Survey data

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 bulletin on the ONS Infection Survey is out a day earlier than usual this week because of Christmas, and a few of the usual details have not been provided.  However, there is new data on the percentage of positive swab tests in each region of England that are compatible with infection with the new variant of the virus, that appears to be more infective than the previous versions.  Please note this is referring to the UK new variant, not the South African variant.

“There is some uncertainty about these figures, partly because a full sequencing of the genetic code of a virus specimen is needed to be certain that a sample is really from the new variant.  But a good indication can be found from the results of a standard PCR swab test, as is used in the ONS survey.  This is why the reported percentages are of infections that are ‘new variant compatible’, not ‘new variant’.  The percentage that are ‘new variant compatible’ varies a lot between regions and UK countries.  In London, over 2/3 of the infections found in the survey, 68%, were new variant compatible, In the South East and the East of England, the percentages were almost as high (65%).  The percentages are lower, but still high enough to be concerning, between 38% and 50%, in Northern Ireland, Scotland, and the South West of England.  They are currently quite a lot lower in the regions of the North of England, and in Wales.  ONS warn that there is considerable uncertainty about the estimates in the UK countries other than England, because the number of positive swabs tested there in the survey is lower.

“There is more detail from statistical models showing the trends over time in new variant compatible infections in England and its regions1 and, separately, in the other UK countries2.  These are what is behind the graphs presented at yesterday’s 10 Downing Street briefing (23 December).  We’ll need to keep a close watch on these figures.

“Apart from that, the ONS survey presents data essentially confirming what has already been reported from other sources.  The new bulletin goes up to the week ending 18 December, and the percentage of people in the English community population that would test positive (for any variant of the virus) is estimated at 645,800, almost back to the level in early November.  That’s about 1 in every 85 people.  Because the estimates come from a survey, they are subject to some statistical uncertainty, and ONS give an interval to describe this uncertainty that runs from 1 in 80 to 1 in 90.  Those rates are worryingly high.

“There have already been data from numbers of confirmed cases that show these increases, but it is still important to have the results from the ONS survey.  That’s because the ONS infection survey results are based on swabs taken from a reasonably representative sample of people across the community, who are being tested only for the purposes of tracking infection rates, and not because they have symptoms or work in certain jobs or live in certain places.  The data on new confirmed cases, that are reported every day and appear on the dashboard at, can sometimes be distorted by changes in the numbers and types of people asking to be tested, for whatever reason.  But the ONS survey data confirm that the depressing trends in the confirmed cases data are not misleading.

“Other aspects of the ONS results also broadly confirm what data on confirmed cases are showing.  Infection rates are highest in London, and not much lower than that in the South East and the East of England.  In London, ONS estimate that 1 in 45 people would test positive (with a range from 1 in 45 to 1 in 50 to allow for statistical uncertainty).  That is very worryingly high.  The central estimates for the South East and the East of England are rather lower, but still worrying, at 1 in 75 and 1 in 85 respectively.  Rates in all three of those regions have been increasing rapidly.  The infection rate in the South West has also been increasing fairly rapidly, though it was still much lower than in those south-eastern regions (1 in 190).  In the other English regions, infection rates are between those in the three south-eastern regions and the South West, but have been falling in recent weeks.  I hope that the new virus variant and any extra mixing of people over Christmas does not reverse the positive trends in those parts of the country.

“In terms of age groups in England, rates have been rising in all the school-age age groups, primary as well as secondary, and also (though more slowly) in young adults up to age 34.  In the older age groups, the trends are broadly level.

“The ONS survey shows a trend of rising infections in Wales and in Northern Ireland, with the estimate for Wales for the latest week being particularly high at 1 in 60 of the population.  There is a wider range of statistical uncertainty in Wales, 1 in 45 to 1 in 75, than generally in the English regions, because not so many people are tested in Wales.  Uncertainty levels in Northern Ireland are high too – there the central estimate is 1 in 180 and the credible range goes from 1 in 125 to 1 in 300.  In Scotland, the rate for the most recent week shows a decline on the week before, though there is some statistical uncertainty.  The central figure for the latest week is about 1 in 140, with a range from 1 in 115 to 1 in 180.”





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 the Advisory Committee, but my quote above is in my capacity as a professional statistician.”


in this section

filter RoundUps by year

search by tag