COVID declining across the UK

Latest ONS figures show “R” number falling but still above 1. The King’s College Covid symptom tracker is more optimistic. This app certainly correctly “called” the start of the turnaround at the peak of the first wave before anyone else.

covid.joinzoe.com

According to the ZOE COVID Symptom Study UK Infection Survey figures based on swab tests up to four days ago, the number of daily new COVID-19 cases is now declining across the UK.

Key findings from ZOE COVID Symptom Study UK Infection Survey this week: 

  • There are currently 35,963 daily new symptomatic cases of COVID in the UK on average over the two weeks up to 01 November (excluding care homes) 
  • This compares to 42,049 daily new symptomatic cases a week ago
  • The Midlands is the only region where the numbers are still going up, with the numbers now reaching the same levels are the North East and Yorkshire and overtaking the North West
  • The number of new cases in the worst affected area, the North West, is now at the same levels they were at the beginning of October and have an R value of 0.8
  • The number of daily new cases in the South East and South West have now stopped rising but have yet to start declining. These regions still have significantly less cases than the North West, North East and the Midlands (see full table of regional results and graph below)
  • The UK R value is 0.9 
  • Regional R values are: England, 0.9. Wales, 0.9. Scotland, 0.9.
  • Age: Again this week, infections nationally have stopped increasing in most age groups apart from modest increases in the over 60s group, which needs to be monitored closely (see graph below)

The ZOE COVID Symptom Study UK Infection Survey figures are based on around a million weekly reporters and the proportion of newly symptomatic users who have positive swab tests. The latest survey figures were based on the data from 13,460 recent swab tests done between 25 October to 08 November 2020. 

If you’d like to receive the full daily report for the ZOE COVID Symptom Study app head to: https://covid.joinzoe.com/your-contribution 

Tim Spector, Professor of Genetic Epidemiology at King’s College London, comments: 

“Having peaked at the end of October, cases coming down across most areas of the UK, is good news but the numbers are still high in most areas, and with a prevalence of over 500,000 infected people there is still a lot of virus in the population. Yesterday, we reported that the R value was below 1 across all UK nations for the first time and we need this trend to continue. A caution is that rates are still increasing in the Midlands for reasons that are unclear. We aren’t out of trouble yet, but with numbers falling and the news of a vaccine, it feels like the end is in sight.”

6 thoughts on “COVID declining across the UK

  1. This is perhaps the first case of FAKE NEWS or at least FAKE HEADLINE on EDW!!!

    Whilst the R number is >1 then BY DEFINITION the level of Covid is GROWING not declining.

    Like

    • The headline is taken from the information source – it reflects the difference in opinion amongst epidemiologists about what is actually going on. The problem revolves around lags in sampling. The Covid symptom tracker uses reported symptoms as a proxy measure for infection tracking. As Owl understands it these have been adjusted to account for those with symptoms subsequently testing positive (from first phase of the study). Owl frequently refers to this tracking study for the following reasons. Despite being a proxy measure it is highly correlated to infections, the study, therefore, provides a consistent daily measure based on a much bigger sample than any other sampling/testing being carried out in the UK. Look at the graph at the bottom of the post and observe two things: the tracker study calculates a higher infection rate than that posted by the NHS, but that rate has been falling since the end of October. I.e. this study calculates an “R” value less than one. The one caveat is that the participants, though well distributed across the UK are self-selecting and therefore may not reflect some social populations.

      Like

    • What the original post says is that there is a difference of opinion between the Office for National Statistics and the ZOE COVID Symptom Study UK Infection Survey.

      The ONS statistics are based on objective scientific tests, and appear to take a week to produce. They show that the number of proper Covid tests showing new infections has continued to rise in the week up to 6 November in England, Wales and Scotland and has levelled off in Northern Ireland. It does NOT say anywhere that the number of new infections is starting to decline anywhere, much less the UK.

      According to the ONS web page, the ZOE Covid Symptom Study uses a mixture of symptom reporting by individuals (and so excludes anyone who has Covid but is asymptomatic) and the results of actual Swab tests on those individuals. The ZOE data is slightly more recent (9 November), but is A) subjective, B) is based on app usage and under-represents key demographics that do not use the app (e.g. hospitals, care homes, elderly), C) is not based on the total base of Covid tests (just the subset of tests of people who are using the app) and D) is adjusted i.e. guessed to attempt to correct for B & C. The ONS explicitly states “Unlike the data presented in this bulletin, the COVID Symptom Study is NOT a representative sample of the population.” Nevertheless, assuming that it uses the same methodology week-on-week, a change in ZOE numbers i.e. showing that the rate of new infections has dropped, may well be relevant as an indicator – however because the data is inherently inaccurate, I do not believe that it can be used to calculate an absolute number such as the R rate.

      However, NEITHER of these measurements of new cases are saying ANYTHING about whether Coronavirus infections are increasing or decreasing. They are both measurements of NEW CASES of Covid infection which is a completely different measure to the number of CURRENTLY ACTIVE Covid infections.

      COVID can only really be said to be “declining in the UK” when the number of currently active infections is dropping i.e. more people are being tested as having been cured of Coronavirus than those being newly found to have caught it.

      Comparing the statistics for England in the ONS reports of 13 November (654,000 people infected on 6 November) and 6 November (618,700 people infected on 31 October) it is reasonably clear that Coronavirus was NOT in decline at least until 6 November.

      So I stand by my previous assertion that the headline is FAKE NEWS.

      I have got used over the years to the Owl running a sanity check on stories before simply repeating the headlines. So it is disappointing to see that this one has slipped through the sanity checking.

      Like

    • P.S. The graph included in the above article was NOT from the ZOE web page linked to. Eventually I found it here: https://covid.joinzoe.com/data#levels-over-time

      It does indeed appear to show that the number of active infections is dropping, however again that is not what the graph says it is showing which is “the number of people calculated to have COVID symptoms”. So, A) I am unclear how this is calculated, and B) if we assume a fixed period for CV19 to be cured once symptoms have been detected, I would expect the number of active infections to follow the same graph as new infections, but just later. The WHO says that someone who only has mild symptoms should be free from isolation 14 days after symptoms occur: https://www.who.int/images/default-source/health-topics/coronavirus/isolation-infographic-eng.tmb-479v.jpg. But this does is simply guidance on timescales, not evidence – people who have been tested positive for CV19 should only stop isolating when they have a follow-up test which shows they are now negative.

      The ZOE graph that shows the ESTIMATE of new cases (https://covid-assets.joinzoe.com/latest/incidence%20in%20UK.png) appears to have peaked around 26 October, so it is possible that the total number of active cases has declined, but this graph also shows that after 8 November the number of new cases has risen again.

      So, again, I do not see any real scientific evidence that demonstrates that Covid is declining.

      Like

    • The ZOE Covid symptom study data identified the turning point correctly last time, approx. two weeks before anyone else. So Owl just thinks their results are worth keeping an eye on. The graph showing a downturn in symptom reporting from around the end of October is not fake news, it’s data.

      Like

    • Yes – but the ONS data is significantly BETTER data. And that says Covid is still increasing.

      Like

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.