21 September 2020


Scepticism.  It is absolutely right to question what we are being told.  Professionally I have to be sceptical all the time, whether that is evaluating business software or, back in the day, auditing stock takes.

Scientifically I will always remember a school textbook saying that in the early days of microscopes, 'scientists' believed they could see a tiny man in sperm, just as Aristotle had suggested.  That became the wisdom of the day, but we now know that is incorrect. 

So it is no surprise that people question what they are being told, and have indeed been challenging me and what I have been saying.  So let's get to some of the facts.

Death is one of 6 distinct risks of COVID-19, so let's look at that now.  Did the lockdown actually increase deaths?   What age groups were affected?

We looked at deaths up to week 36, 4 September, recently.  But let's expand on that analysis, using the same weekly data set from the Office for National Statistics.


"Excess deaths" is the number of deaths over (or indeed often under) the average for the last 5 years.  In 2020 this is:
  • Deaths due to COVID-19
  • Plus deaths indirectly caused by COVID-19, such as due to delayed treatment of other medical issues such as strokes
  • Minus deaths saved by COVID-19, such as lower road traffic accidents as might be expected due to lower traffic during the lockdown
  • Plus or minus usual annual variations

For 2020, the ONS has also disclosed the number of deaths where COVID-19 has been mentioned on the death certificate:
  • Primary cause
  • Secondary cause, once doctors became clear on what was expected
Comparing the two on a cumulative basis:

 What this graph shows is:

  • Deaths in January and February were below normal levels until COVID-19 came along
  • The 23 March lockdown was just after the first deaths were recorded, but these are some weeks after original infection, probably in February.  It was clear in mid-March that infections were rising rapidly.
  • The lockdown starting on 23 March would not have had an impact on deaths for some weeks:
    • On a rising trend in new infections, confirmed cases didn't return to the 23 March level until more than a week  
    • Infections in late March wouldn't have resulted in a significant number of deaths until late April at least.  
    • Lockdown had reduced excess deaths to be near zero by around week 20, mid-May
  • Excess deaths initially ran ahead of official COVID-19 deaths.  But this was likely due to confusion on how to complete death certificates in co-morbidity situations, where there was also some other cause of death
  • Both excess deaths and COVID-19 deaths had levelled off by early September.
  • The excess deaths total some 53,000 so far, almost exactly matching the official number of COVID-19 deaths
  • If the shortfall in official COVID-19 deaths in April and May compared to excess deaths  represents an 'error' in co-morbidity situations, then:
    • COVID-19-related deaths actually exceed the excess deaths
    • Causes that reduced deaths exceed the causes increasing them.  So for example, any reduction in situations such as road traffic accidents more than compensated for any increase through health treatment not being available
  • If there hadn't been a lockdown:
    • Deaths would have continued to rise to far higher levels 
    • The impact on the health service to provide non-COVID diagnosis and treatments would have been at least as severe as lockdown


Professor Whitty disclosed in the press briefing this morning, 21 September, that influenza kills around 7,000 a year, with some 20,000 in a bad year such as three years ago.

Deaths from COVID-19 of 53,000 in only some of the year represents a far bigger problem.  That is due, at least in part, to the lack of a vaccine as is used for influenza.


The comparative data for 2015-2019 is only available in a small number of age ranges.  Calculating the excess deaths for 2020 on that basis gives this graph:

We hear that the death problem from COVID-19 is mainly for the over 50s.  But this shows that the problem with excess deaths is for the over 44s, with lower deaths than expected for the under 15s.

That leaves the rather wide 30-year age range 15-44 having excess deaths peaking at 5% of normal.  But what if we were to look at smaller age ranges?  The relevant data is not available publicly, but we can look from a different angle with the data that is available. 


Whilst we can't look at excess deaths, we can look at the absolute number.  If we assume the number of people in each age band is broadly similar, then the different levels indicate how much COVID-19 affects each age group:

In broad terms this shows that you are more likely to die as you get older.  But is there any evidence for increased deaths from COVID-19?

Let's look more closely at the middle section between weeks 10 and 20.  Week 20 was the peak of excess deaths for older year groups in the graph above.  An increase in the slope would indicate the death rate had increased.  We can only see a very slight increase in the slope for the 35-39 and 40-44 age groups:


There's been around 53,000 deaths from COVID-19 in the first 8 months of the year, whether it's based on excess deaths or death certificates.  That is so much worse than 'flu, which is at worst 20,000 in a bad year.

This is mainly because of the lack of a vaccine against the virus behind the COVID-19 disease.  There won't be a vaccine proven safe and effective, and available in significant quantities, until 2021.  In the meantime there have been some improvements in treatments, but a resurgence of COVID-19 infections in 2020 will result in many more deaths.

Looking at age ranges:
  • Death from COVID-19 is a clear risk for those aged 45+, not just those over 50
  • There is some risk of death for those 35-44
  • There is little risk of death for the under 35s.
That doesn't let the under 35s off the hook though:
  • There are five other significant risks of COVID-19 that affect all ages, which we'll look at in more detail in a separate posting
  • The under 35s can be carriers, infecting other younger adults who in turn can infect older and more vulnerable people

No comments:

Post a comment

Popular Posts