SAVING THE SUMMER. SAVING THE ECONOMY. SAVING LIVES

UPDATE 3 JULY ABOUT DATA PUBLISHED 2 JULY

The various data published yesterday support the basic messages in this postL

• England's level of infection is too high to go ahead with easings due tomorrow Saturday 4 July
•  Economy and life in general will eb stronger if we have a short, sharp second lockdown.  Just 3 weeks.
• Supported by better anti-transmission measures and public information about them.  So far the comms have been almost non-existent

See analysis of the latest data here 

UPDATE 1 JULY 2020

PM Johnson talked at PM questions today, 1 July, that daily infections were "below 1000".  This sounds like the total of "pillar 1 and 2" data derived from formal testing that has been routinely shown in the daily press briefings.  This figure is inherently the tip of the iceberg due to:

• Only people who submit to a test.  Not people with symptoms who do not volunteer
• Not those who do not return a  postal test
• 'False negatives' can be 30%-40% due to the difficult of swabbing.  This is especially true of self-swabbing used in postal tests.

The real pillar 1 figure is therefore likely to be at least two or three times the number quoted.

Indeed that is borne out by the studies published last Thursday from the ONS and King's College ZOE that indicate that a more realistic estimate of new infections recently has been 2500-3000 per day for England alone.  The analysis is below.  There have been partial updates to this data which have shown an overall decline nationally, but relatively modest.  Individual regions may still be rising.

Prof Whitty talked back in May (see below) about the true figure needing to be under 1000 before easings can be made.  The true figure is likely to be still in excess of 2000 a day by Saturday 4 July.

We've seen the issue with Leicester, and any other part of the country could be next.  Do we want to live under that cloud all summer?  No.

Infections are simply too high.

Therefore:

1. The plans to make easings on 4 July in England must be deferred
2. The level of infections must be brought down to a much lower level.  Effectively to Alert Level 2, not the current level 3
3. This realistically requires a national lockdown equivalent to Leicester.  Thereby avoid any border issues around Leicester, but more importantly get infections down nationwide.
4. The short sharp lockdown may only need to be 3 weeks, as estimated below.  When this blog was written we could have re-opened on 17 July, in time for state school holidays.  Re-opening more safely and more strongly for businesses and the economy than we would going ahead on Saturday. 
5. Providing the platform to consider re-opening the performing arts, and gain many other benefits as outlined below.

We've now missed that target date.  Every day delayed now is a day later to re-open.

Let's Save the Summer.  Save the Economy.  Save Lives.

It doesn't fit with this government to err on the side of caution,  But we should, given this is a matter of lives and livelihoods.

In any case we must get infection levels right down if we're to re-open strongly in July.


ORIGINAL BLOG

According to the latest ONS data, it looks likely that a second wave of COVID-19 infections has already started in England:

• R looks to be at least 1.0, and probably higher. 
• Daily infections are at least 2,500-3,000 and likely to be rising.  Not around 1,000 as the media typically report. 

This conclusion is supported by data from the King’s College app known as ZOE.

What should we do about this awful situation?


BACKGROUND

On 9 June 2020 I wrote a letter to the Financial Times which they published a couple of days later. It warned that we were on the verge of a second wave, and that we ought to do two things:

1. Change the plan.  Nip the second wave in the bud by having a short, sharp second lockdown.  That would allow the economy and life to re-open more strongly in July, just three weeks after implementation.  When published, potentially by 4 July
2. Develop a stronger list of cheap and simple actions that the public could adopt to keep transmissions to a minimum thereafter, and indeed during lockdown

Nobody took much notice, and we’ve lost two weeks. 

Then on Thursday 25 June the Office for National Statistics (ONS) and King’s College London published the latest results for their respective studies, alongside an update from SAGE on national and regional transmission rates.

The ONS said that the drop in infections had “levelled off” in England, implying the R number had become 1.0, its upper limit before infections rise.  This is confirmed by the King’s College data.  But the ONS data indicated that infections were actually rising, and R was now above 1.0.  A second wave looks like it has started, in several regions of England.

Yet SAGE said that after several weeks that R in the UK, it was still in the range 0.7-0.9.  That prompted a closer look and a thread of tweets.  These have been condensed into this shorter post.


THE TWEETS CONDENSED

The original set of tweets is still available here .  The condensed and edited version is as follows:


On Th 25 June, the government published that the latest R number for the UK remains at 0.7 – 0.9. The limit is 1.0 before infections start to increase exponentially again.

The ONS published their latest “COVID-19 Infection Survey” saying infection has “levelled off” implying an R of 1.0, or likely higher

So what is going on?


Our understanding of COVID-19 infections in the community needs to be the combination of:

• Number of infections, both daily and in total
• R number
• Daily rate of change

Let’s start with the R transmission number where R stands for Reproduction :

• If infections are stable, R is 1.0
• If infections are falling, R is below 1.0
• If infections are rising, R is above 1.0

We don’t want infections to be rising even slightly, as that will kick off exponential growth again

R does not actually stand for ‘rate’, as it has no time aspect.


To analyse the data we need to bear three things in mind:

• Some figures are for UK, some for England.
• We need to carefully distinguish daily new cases of COVID from the total of people infected
• 95% level of confidence ranges are relatively large, but nonetheless allow conclusions

We also need to bear in mind that transmission occurs at least 5 days before symptoms. Hospitalisation and death aren't until a few weeks later.

Ideally we would monitor R daily according to transmission. But for this analysis we are relying on symptoms or tests some days after. 

Any R number that is based on hospitalisations or deaths is at least 2-3 weeks out of date, as SAGE used to acknowledge for the official R.  Now it is not specified.


Our Frame of Mind

When looking at the data, we no doubt wish to be optimistic. But given deaths and other harsh consequences of COVID-19, we need to err on the side of caution.

If R might be above 1 .0, and infections might be rising, we need to regard that as a serious problem.

That sadly is the situation.


Office For National Statistics Data

Let’s now look at the Office of National Statistics (ONS) data from their “COVID-19 Infection Survey” published on Thursday 25 June.  This is just for England.

They say repeatedly that infections have “levelled off”.  Though as you can see here, they are suggesting infections rose by over 82%

The ONS also provide a graph of the total number of infections for England as a whole and the various regions. The graph for England, shown here, and some of the regions all indicate that infections are rising.

Let's look at these totals more closely.

Samples are small, but the ONS has suggested 95% confidence limits are valid.

The bottom of the 95% range is flat, indicating R is 1.0. There's a 2.5% chance infections are falling. But at least 95% chance infections are rising, or to be cautious at least 90%.


So let's pause there.

The ONS has said that infections in England have "levelled off" but the data suggests a 90% chance that infections are rising.

Data from the King's College "COVID Symptoms Study" also indicates R is 1.0 in England, and higher in some regions.

That is not the foundation for easing lockdown any further.

So: 4TH JULY EASINGS MUST BE POSTPONED


But don't despair. There is a solution that would allow the economy and life in general to re-open more safely and more strongly only two weeks later on 17 July if implemented now.

This letter was published in the Financial Times more than 2 weeks ago. At that time 4th July could have been assured.  Any further delay risks missing 17 July, when state schools mostly break up 


ABOUT THE AUTHOR

Before further analysis, a little more about me.

After graduating as an Oxbridge scientist, I qualified as a finance professional, ran the finance department of a fast-growing biotech, and have since specialised in the effective use of IT.

I advise the CEOs and CFOs of FTSE, AIM, multi-nationals and other businesses on management information, financial modelling and business processes.

My career has several examples of being the first to spot issues that need to be solved, like this COVID-19 problem.  By taking a 360 degree view whilst identifying critical detail, I have solved problems that others had not and those problems where senior management thought were impossible to solve.  I don’t believe in can’t.

Even the Irish border issue, though that is an entirely different matter for another time!

Politically I have no association with any registered political party.  Unusually I didn't vote at the last two General Elections.  I simply represent the silent majority, speaking up for our frustration with the government's handling of the COVID-19 crisis, which by now should be far better under control.


BACK TO THE PLOT

More on the ONS Data

ONS households exclude hospitals, care homes and other institutions, so represent the general population.  Test are taken of a random selection of households, whether or not people have symptoms.

Unfortunately swab tests can show a large proportion of 'false negatives', especially for self-swabbing used by ONS.

Their data should therefore be considered to be the minimum levels of infection.

When PM Johnson announced 4 July easings last Tuesday, he said "...in the first half of June [the average number of people in the community in England with COVID-19] was 1 in 1,700" Consistent with the ONS graph showing infections down to 0.06% in total. Since risen to 0.1%


The ONS now says "During the 14-day period from 8 June to 21 June, there were an estimated 4 new COVID-19 infections for every 10,000 individuals per week in the community population in England, equating to an estimated 22,000 new cases per week...." Over 3,000 per day.


The ONS goes on to say there is "95% confidence interval: 10,000 to 49,000". That's 1,400 to 7,000 new infections per day. This has been going on for at least a fortnight. But what does the King's College "COVID Symptoms Study" powered by ZOE say?

The King’s College London “COVID Symptoms Study”

The King’s COVID Symptoms Study involves almost 4 million participants providing symptoms and other information on a daily basis using the ZOE smartphone app.

Whilst the sample is much larger than ONS, it is restricted to smartphone users, so actual figures may be higher.


Here is the graph published by King’s College for the period to 23 June. This shows new cases of symptoms across the whole UK fluctuating around 2,500 from around 11 June. That's new infections in reality since around 4 June, some 3 weeks ago. R around 1.0.

 But what does this mean for England? On the news we now hear that Scotland has had no deaths from COVID-19 for two days, but England's has gone up. This implies most of the 2,500 new daily cases are outside Scotland, and indeed mostly England.


Given that the King's figures are likely understated, it's looking like there have been at least 2,500 new cases of COVID-19 in England for the last 3 weeks. That is consistent with at least 3,000 being suggested by ONS.

So at least 2,500-3,000 new cases of COVID-19 a day in England.


Why ONS and King’s R figures are higher than official R for UK

The R in England of at least 1.0, probably higher.

King's College R figures for the English regions are in the range 0.8-1.2, with Scotland at 0.9 and Wales 1.0

So England is higher than Scotland. Explaining why the "latest R" of 0.7-0.9 for UK is lower than for England.  This SAGE figure also looks optimistic, just below the limit.


Why ONS and King’s R figures are higher than those from official testing for UK

But why are daily infections something like 2,500-3,000 in England alone, whereas the testing stats in the Downing Street press briefings has suggested 1,000-1,200?  Less than 1000 in recent days.

The 7-day average on Tuesday 23 June was quoted for the whole UK as 1,147.

There are 3 principal reasons:

1. Not all people with COVID-19 request a test
2. Not all tests sent out get returned 
3. Swab-testing, especially self-swabbing, is notoriously unreliable, resulting in up to 40% false negatives.

ONS and King's estimates are understandably 2-3 times higher, and sound more realistic.


FINDINGS IN SUMMARY

So what have we found?

1. R for England has been around 1.0, more likely higher
2. Above the 0.7-0.9 UK because Scotland has been lower
3. New infections in England have been running at around 2,500-3,000 per day for the last 3 weeks
4. Regional Rs in England are up to 1.2, indicating a second wave in some regions

The R of 0.7-0.9 for UK from the government was accompanied by regional Rs. They are all more optimistic than those suggested by the King's study.

For regions where rates are up to 0% per day, R should be up to 1.0 as suggested for NW. That is also London, Midlands and SW

WHAT CAN WE CONCLUDE?

(1)    In England the R number (at least 1.0) and daily infection rates (at least 2,500-3,000 per day) are higher than typically reported by the media, and likely rising.

(2)    Official figures from SAGE appear overly optimistic.

(3)    We should err on the side of caution.

(4)    Just like Apollo 13, "London we have a problem".


RECOMMENDATIONS

R in England looks at least 1, indicating the start of a second wave in several regions.  With infection rates already several times higher than on the continent.  So what should we do about it?

Firstly this situation is not the foundation for further easings. Those planned for 4 July must be postponed.

Indeed we must nip this second wave in the bud.

The solution was published in the Financial Times over two weeks ago.

The only reliable way is a short, sharp, second lockdown. In three weeks infections could be reduced to more like 400-500, comparable with Spain, France and Germany (ignoring local flare-ups).


Supported by developing a stronger set of cheap anti-transmission techniques that the public can use to keep infections down pn an ongoing basis.

That will require a far stronger public information campaign. "Common sense" is not enough for this new type of infection.


So why might a short, sharp, second lockdown work?

In the graph of various R rates above, an R of 0.5 allows infections to stop growing in total in about 21 days. That means daily infections down to near zero.

In the real world of the first lockdown, rates fell to around 16% in 3 weeks.

So set an R around 0.4 - 0.5 now and we can get infections down much lower than they are currently by 17 July, comparable to the continent

An important date because that's when state schools typically break up, only a week after many private schools.



A summer lockdown can be lighter than the first, as windows are open to provide ventilation and people can be out of the house more.

• Perhaps retaining outdoor groups of 6, and the new dual-household bubbles.
• Even allow pub gardens and outdoor cafes to open from 4 July as planned.  
• Possibly overnight accommodation from 10 July where pre-booked by private school families. 
• Though non-essential retail likely to need to close briefly, before re-opening more strongly just three weeks later

Anti-transmission Techniques

It is important that a stronger set of cheap anti-transmission techniques are developed for the public to follow. 

No reliance on “common sense” as this is a new type of infection outside people’s experience.  A fully-fledged public information  campaign is needed.  Why haven’t we had videos like those of the 1950s, when we have such better technology?


Transmission is primarily by two methods:

• Person to person via the air
• Person to person via surfaces

Techniques would include at least:

• Better hygiene.  We’re used to anti-bacterial, but anti-virus needs to be more stringent, including
• Washing hands and surfaces with soap and water to destroy the fatty coating.  Anti-bacterial are useless unless contain anti-viral ingredients
• UV light to treat surfaces when safe to do so
• Still maintaining social distance.  No hugging or kissing outside family or household
• More extensive use of masks to protect other people should you be infectious.  It is a civic duty.  I wear mine every time I leave the house.  
• Can the government give out masks free of charge, like is done in some places abroad?
• Whatever other techniques can be added

Ideally people should be educated and then left to follow guidance of their own free will, as precautions are required for the foreseeable future.  That i  at least until vaccines and medicines have been developed and made widely available. The assumption has to be an 'if' not 'when', as nothing is guaranteed. 

Test/Trace/Isolate (TTI) can play its part, but is not a magic bullet.  Great in theory.  But given its many practical, technical and cost problems, the set of anti-transmission techniques would need to be developed with little reliance on TTI at the national level.
 
Though TTI is really useful to tackle local outbreaks if rapid testing facilities can be wheeled in, and practical/financial support given to anyone asked to self-isolate.



WHAT ARE THE BENEFITS OF A SHORT, SHARP SECOND LOCKDOWN

There are a multitude of benefits:

(1) We need to make sure any second wave is nipped in the bud, which would require a second lockdown anyway if infections take off.

(2) Get the lockdown over and done with before schools break up and people have booked to go on holiday.  Take control of timing.

(3) With infections lower in any case, people will be more confident to go out when lockdown ends and can be substantially eased on 17 July (or 3 weeks after a later implementation).

That is better for the economy, better for jobs, better for the self-employed, and thus better for the Treasury too.

For the Treasury:

• Lower COVID-focussed support payments
• Lower social security payments from fewer job losses
• The restart of VAT and other tax receipts

(4) Schools can safely re-open in September. Even summer schools

(5) Those people who have been shielding will be able to leave their homes at far lower risk

(6) Easier for "air bridges" and "travel corridors". The likes of Spain and France will see the UK will have infection rates comparable to their own. Other destinations too.  Even Scotland, which is now threatening to close their border with England!

(7) Provide a stronger platform for allowing live music, theatres and other performance art to re-start.

(8) And of curse any lower level of infection will save lives.  In this case thousands saved from death or serious medical complications.


This proposal for a short, sharp second lockdown is therefore better for everyone.  Every person and even the Treasury.

Even any businesses that have to temporarily close, as they can open more strongly just three weeks later.

That is if the government is prepared to implement now. It looks like the data gives them no choice!


Saving thousands of lives, and saving thousands from ongoing medical complications..

Saving the summer. Saving the economy. Saving lives.