by
Les May
A FEW days ago I heard an assertion by Tory MP Damian Green
that China needed to be 'brought to account' about its
handling of the Covid19 outbreak.
I would not want to live under the Chinese system of surveillance and make absolutely no apologies for it, but I think it is important to separate our opinion of the political system from how we think and act about the Covid19 pandemic.
If people want to criticise the figures coming out of China they should produce the evidence that they are untrue. There are well understood ways of doing this in a procedure called 'forensic economics'. It has been used to show that stock (share) options which are based upon how the business performs over a period were being backdated so that those receiving the options benefited from the fraud. It has also been used to show that 'spread betting' on baseball games had all the hallmarks of being rigged by bribing players to play less well in some games. It is up to the people who want to say the Chinese figures are phony to do the work needed to demonstrate it.
There
are an abundance of data for the numbers of infections and deaths in
the different provinces of China for such an investigation. Should
we not be at least equally sceptical about the single figures for
testing in the USA which Trump produces at every news
conference?
China has been criticised for not including cases of Covid19 when the person shows no symptoms but shows as infected when tested. The WHO says they should be and China has had to revise its figures to reflect this. But look at the situation in the UK. Using the figures released recently about one in five of the people who are tested are found to be infected with the virus. This is the number which goes into the published figures. But what about the people who show mild symptoms? These people are told to quarantine themselves for 14 days along with anyone else in their family.
China has been criticised for not including cases of Covid19 when the person shows no symptoms but shows as infected when tested. The WHO says they should be and China has had to revise its figures to reflect this. But look at the situation in the UK. Using the figures released recently about one in five of the people who are tested are found to be infected with the virus. This is the number which goes into the published figures. But what about the people who show mild symptoms? These people are told to quarantine themselves for 14 days along with anyone else in their family.
Crucially
no one is officially collecting figures for these people nor are they
being tested. As a result we are underestimating the true number of
infections in the UK and we do not know how much. In
addition it is now known that the first death occurred on 28 February
not 5 March as originally declared. France has also had to make an
amendment to its figures. Admitting such an error does not mean that
from now on we have to doubt everything the UK or French government
says about the scale of the infection. So why single out China?
Yesterday
in what Matt Hancock had to say there was a tiny glimmer of
hope that the fog around the true number of people in the UK who are
or who have been infected with the virus may begin to clear.
Hancock
made a brief mention of an accurate test devised at Porton Down
which detects anti-bodies in a person’s blood if they have been
infected with the virus in the past and the intention to undertake
‘surveillance testing’. This means testing a random
sample of the population, i.e. one that is not biased with respect
sex, age, ethnic origin, location etc. From the results of such
testing it is possible to draw inferences about what proportion of
the population has been infected.
As
it is an estimate based upon a sample of the whole population it will
be subject to a small amount of error, though the larger the sample
the smaller the error. Such estimates are normally published
together with a ‘confidence interval’ or ‘credible
interval’ which can be taken to mean that the true value for
the proportion of the population infected can be expected to lie
within this interval (range).
If
the sample is ‘stratified’ it is possible to make similar
estimates for different age groups, either sex, ethnic origin,
occupation etc.
Ideally
such a test should be administered a month after a person has
recovered from the infection. If anti-bodies are still present it
means that immunity to further infection is long lasting. This is an
important criterion if a population is to develop ‘herd
immunity’ and for the
development of a vaccine.
**********************************
Posts
No comments:
Post a Comment