ONS Covid Infection Survey

The Office for National Statistics (ONS) Covid-19 infection survey is based on random testing of a representative sample of the public in England. It is updated once a week. This gives by far the clearest estimate of the overall infection rate within the community ( The data actually use fortnightly results so the weekly figures represents a rolling average of infections). The results for July 8 show that the level of active infections have fallen to roughly 0.04% of the population and seems to be levelling off at that level. About 1 in 4000 people are currently infected with COVID-19 in England. ONS provide an embedded version of their results which I am hoping gets  updated weekly.

The population of England is 56 million people so an infection rate of 0.03% means there are currently ~ 14,000 people with Covid-19  Another way to see this is to look at the risk of anyone in the public catching COVID-19.

This says that on average there is roughly a 1 in 2500 chance of being infected with COVID-19 each  week. An alternative narrative is that with this level of infection it would take 50 years for everyone to catch COVID-19 at least once and  a further 5000 – 10000 years to actually die from it. This assumes there is no herd immunity or that any  vaccine available.

Here are the regional infections for England.

The graphs in this post should update automatically as the ONS updates their survey results each week. If so I will make a new “widget” !

We shall see next week!

This entry was posted in Public Health and tagged . Bookmark the permalink.

3 Responses to ONS Covid Infection Survey

  1. paulski0 says:

    The ONS Survey also suggests 2.8 million people over the age of 16 in England have antibodies. Assuming that’s a proxy for number of infections that would suggest an IFR of about 1.5% (about 1 – 2% range). Presumably not all people infected develop antibodies, though the study discussed here suggests nearly all.

  2. davidcj0053 says:

    Hi Clive,
    kudos to you for your independent analysis of the data and conclusions. But I want to raise a wider point about the raw data. My comments are informed by various scientists including Stanford University professor Dr John Ioannidis please see


    The testing methodology of the ONS Covid Infection Survey is described thus “This is a test for current infection at the time the swab was taken.”
    Therefore it is a single point in time snapshot of a extremely complex database. As and until there are either multiple such point in time snapshots – say weekly – it remains highly problematic how you can make any conclusions as to how many folks have had the disease but no longer have it. You might attempt to model the actual level of population immunity looking backwards using some I% value for immunity after a Covid-19 infection withing 12 months. But for heavens sake where is the spending [very modest required] on randomised populations using actual antibody testing?

    The lack of this evidence about a vital metric after so many months of an obscenely costly and damaging abuse of basic freedoms, seems to this observer to defy logic. So the question one is forced to ask is, why are these antibody tests not being carried out in every country that is a signatory to the WHO and the results published? Think on the ramifications for health and public policy officials if the public knew that the virus was already widespread.

    In such circumstances the only rational policies are those that target the most vulnerable by the Pareto principle and leave the vast majority of healthy and likely unaffected individuals to make their choices in terms of social distancing and other measures like face masks. Such a volteface by PMs and Presidents would see many careers cut short, possibly whole ranks of politicians and officials losing their seats and sinecures. Eventually sanity will return but I fear that a triumvirate of self interest is now locked in a vicious cycle of deception, fear mongering and control perpetrated against the general public.

    best regards
    David C Johnson

    • Andrew Carey says:

      If you search for “Weekly COVID-19 surveillance report published” you will quickly get to a weekly report which includes in the last few pages the results of serology tests on blood donors in England. Next one out today. Alas the proportion of sero-positives is rather low.

Leave a Reply to Andrew Carey Cancel reply