COVID-19: Day 268: Here we go again…

My puzzlement at what is happening pandemically is now over..

  • Things are very obviously getting worse. 
  • But thankfully about five times slower than in the Spring

Let’s look at the data.

Data#1. Prevalence

Since late April the ONS prevalence survey has been randomly testing people each week to look for the virus. They then collate their data into fortnightly periods to increase the sensitivity of their tests. Details of their full results are described methodically in this ‘bulletin‘.

Click for a larger version

The number of people tested and the number of positive tests are given in their table above. They estimate that at the end of the measurement period on 19th September 2020 roughly 2 in 1000 of the UK population were actively infected.

Their data – graphed below – suggest that the prevalence had been below the 1 in 1000 level for several months but that it is now definitely above that level and doubling roughly every 15 days.

Click for a larger version.

The raw count of positive tests was:

  • 163 from 79,901 people tested in the two weeks to 19th September,
  • 56 from 61,023 people tested in the preceding two weeks, and
  • 26 from 46,107 people tested in the two weeks preceding that.

Because the data are changing so rapidly I have fitted an exponential curve to the last three points and this is shown as a black dotted line (– – –) on the graph above. This is not a prediction: it is a guide to say that “if things continue on this trend, this is what will happen.”

The last data point on the graph refers to the fortnight centered around 12th September. In other words, it is almost two weeks out of date. I expect that Chris Witty and Patrick Vallance had early sight of this data when they announced that things were going pear-shaped at the start of the week.

If the trend continues the prevalence will double every 15 days and increase by a factor 10 in roughly 50 days.

Because we expect this rise to be exponential, we can also look at this on a logarithmic plot. This covers a much wider range and we can see that – if the trend continues – the prevalence will exceed 1% of the population sometime in mid-October.

Click for a larger version

Data#2. Tests and Deaths

The graph below shows:

  • the number of deaths per day.
  • the number of people newly admitted to hospital each day
  • the number of positive tests per day on the same logarithmic scale. 

The data were downloaded from the government’s ‘dashboard’ site.

  • The deaths refer to deaths within 28 days of a test.
  • Positive tests refer to Pillar 1 (hospital) and Pillar 2 (community) tests combined.
  • Hospital admissions for the UK nations combined are no longer being reported because the Scottish data are no longer being reported since 17th September. The curve shown is the sum of data for each country excluding Scotland after 17th September. 
  • All curves are 7-day retrospective rolling averages.

Click for larger version.

The graph above shows the data across the period of the pandemic. From May to July, all three quantities have decreased with (very roughly) the same time constant – the same slope on this logarithmic graph.

The graph below shows the data since July.

Click for a larger version.

Looking at this more closely, we that the data tell a clunkingly obvious story:

  • Tests begin to rise, then
  • Admissions begin to rise, then
  • People start to die.

Click for a larger version.

To highlight this I have drawn rough trend-lines against each curve with the same slope. Physically, if phenomena have the same decay rate, it is likely that they are linked, either causally or by an underlying common cause.

We see that the first sign of the viral spread was in early July when tests began to deviate from their previous exponential decline.

About 3 weeks later, Hospital Admissions began to deviate from their previous exponential decline.

The data for deaths are noisy – because the number of people dying was so small, but it looks like there was a 5 week delay before deaths started to rise. This long delay could be plausibly understood if infections only reached vulnerable people indirectly through an intermediate population of non-vulnerable people.


The data are – in my opinion – decisively clear.

The prevalence data are unequivocal: the virus is spreading with a doubling time of 15 days.

  • This is about 5 times slower than the initial rise in March/April.
  • This gives us time to respond but the response needs to be drastic and immediate.
  • In my opinion the current response does not meet this challenge.

The testing, death and hospitalisation data are unequivocal:

  • This story is more complicated than the prevalence data because the way community infections feed through to Hospital Admissions and eventually deaths is complex.
  • There are factors that depend upon community behaviour and age that make predictions difficult.
  • The current doubling time for Hospital Admissions appears to be about 18 days.

Together, the data all point to the fact that the viral prevalence is growing and if unconstrained will lead to more deaths and another inevitable total shutdown.

What Next?

It is important to understand that the current rate of deaths – about 30 per day – arises from Hospital Admissions that occurred a few weeks ago.

Click for a larger version.

The current rate of Daily Hospital Admissions has risen by a factor three since early September. So we are probably already committed to death rates of roughly 90 to 100 people per day in mid-October.

Last week I said:

The key question is:

  • With the viral prevalence we now have, and the mode of conducting our lives that we have now adopted – to what rate of hospitalisation and death have we committed ourselves?

I don’t think anyone knows the answer to that question. But we will all find out fairly soon.

We are finding out now. We have probably already committed ourselves to a period of several hundred COVID-related admissions per day, and around a hundred of COVID-related deaths per day.

Given that around 1700 people die every day in the normal course of events, we may consider this acceptable.

But COVID-related death rates of thousands per day will follow if the Government do not act.

Personally, I would advocate:

  • A series of planned lockdowns from now until the spring of 2021. This will allow workplaces, schools and universities to plan and avoid the current chaotic uncertainty.
  • The required ON/OFF periods could be optimised but I would guess that they would be roughly 3 weeks ON/ 3 Weeks OFF.
  • The intervention should start now.

Come December, we will all desperately need a break!  But without immediate action on this scale, I fear that Christmas will be cancelled.

Leave a Reply

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

You are commenting using your 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

%d bloggers like this: