I listened to More or Less on BBC Radio 4 today. It was disappointing- quite biased, really. They mentioned that Trump quoted 94% of deaths put down to covid-19 had at least three comorbidities, so only 6% were therefore due to covid. It was a distorted way of looking at it, but there is a valid point here.
A professor from Oxford spoke on the programme to say that the 'comorbidities' were not life-threatening and that the primary cause of death was covid in 92% of the cases. However, this is also a distorted way of looking at things.
These people would not have immediately died from their comorbidities, it is true. But comorbidities are given that very name because they abbreviate lifespan. They are not merely an incidental condition. Loss of life years due covid is less because these comorbidities will take life years away even without covid.
We can say, in these cases, covid is not responsible for a loss of life, but for a loss of life years. This is a very important statement. Covid has not caused death in terms of a complete, otherwise healthy, life.
When we count up the deaths due to covid, we really should be counting the loss of life years. For example, a 90 year old dying is not a loss of as many life years as if a healthy 20 year old were to die.
How do we do this? Look up QALYs - Quality Adjusted Life Years. It may sound a bit clinical, but the death of a 90 year old is sad indeed - but it is not so sad as the death of a 20 year old. Similarly, the death of an obese diabetic with COPD is sad - but it is not so sad as the death of a healthy person of the same age.
All these factors are completely missed by a mere counting of deaths and their (mis)attributions.
Very good points well put Mike. I will investigate the QALYs as you have suggested. This is a subject l believe that throws up an interesting and relevant point. What is the likely number of life years in grand total saved by lockdown, against the the grand total number lost because of the lockdown and its well known consequences e.g. even if 10 80 year olds manage to achieve on average 3 extra years of life due to lockdown a grand total of 30 years of life are saved, although if such individuals already have comorbidities these are unlikely to be 30 quality years. Whereas if only one 30 year old commits suicide due to financial pressures or isolation as a result of lockdown, then a grand total of 40/50 years of mostly good quality years of life are arguably lost. Have the government factored in this into their equation?????
Scott Atlas (MD) of the Hoover Institute talks extensively about the loss of life years caused by lockdown in this illuminating interview.
Another erroneous article on false positives has appeared on the BBC website here
https://www.bbc.co.uk/news/54270373
Again, it is the More or Less programme with Simon Maybin and Josephine Casserly. They have a go at Julia Hartley-Brewer for over-estimating the false positives. But their argument makes no sense.
They argue that the test results should be split up into two communities rather than over the country as a whole. In which case, groups of high prevalence would show fewer false positives and areas of low prevalence would show more false positives. They take PHE community tests with 7% reported positives as an example (which seems high, but never mind). But what goes on in this group does not represent the wider prevalence and the overall false positives. Indeed, in the remainder of the population, false positives will be even higher than if all the tests had been lumped together. I see no reason to separate out groups in this way - unless there is an agenda that I am unaware of.
Here is how I think it should be. I use only ONS data here, the latest being Friday 2nd October. The 7-day average PCR tests (Pillar 1 + Pillar 2) are 228,842. The 7-day average positive 'cases' are 6,273. Cases are only reported for PCR tests. We have a returned prevalence of 2.75%, which will include true and false positives.
On Friday 2nd October, the estimated prevalence from ONS is 0.21% (0.18% to 0.24%). In order to match this, one requires a specificity of 97.42%. (We can take a typical sensitivity of 80%. This is less critical to the analysis).
With these figures, the false positive rate is 93.93%. Using the ONS' maximum estimated true prevalence of 0.24%, we still get 92.99%. So we are talking of only 1 in 15 'cases' being a true positive!
These two reporters argue that a specificity of 99.2% applies. However, if this is used, it results in a true prevalence of 2.45% (vs. cases/tests of 2.75%). This is way above the ONS estimate of 0.21% and does not make any sense. It is more likely that their specificity has been over-estimated. The average specificity from PCR tests is around 97.6%, so the figure of 97.4% that I came up with is nearer the mark.
Therefore, it is the overall prevalence that is important - the total tested population - not some arbitrarily chosen group. And, as demonstrated, this on average is currently producing only 1 true positive among every 15 reported.
I rest my case m'lud.
I see the front page of lockdownsceptics has also exposed the (un)reality check by More or Less
This bogus reporting is paid for by us, the general public, through an enforced licence fee. Yet the BBC seem to be free to publish whatever they wish unchallenged by their paymasters and without being required to publish a correction.
Such articles remain on their website as a source of misinformation. How can we get a retraction?
