Excess Mortality Due to COVID
Quinary age & sex standardised monthly all cause and non-COVID mortality for England & Wales 2020 - 2021 (rev 1.2)
Yesterday I revealed the monthly series for non-COVID mortality for England & Wales for the period 1970 – 2021 and compared this with all cause mortality. This morning it made sense to subtract one set of figures from the other and arrive at an estimated of excess mortality due to COVID. In the plot below you’ll see two nicely defined waves, the second being seasonal and the first being distinctly aseasonal. The first looks like an impulse and consequent response, and we may ask how a virus can yield instant results across the width and breadth of two nations like this. The second wave follows what we may expect from something nasty spreading.
The Big Issue
But the big issue here is how we go about defining death due to COVID as opposed to death with a positive test, and I’ve grumbled on about his a fair deal since Mar 2020. Let us start with some definitions that determine what the Office for National Statistics will count.
Definitions
Herewith a screen shot of the definitions tab on the file monthlymortalityanalysisapr2022.xlsx which may be found here.
It is worth recalling that the definition for COVID started out as just ICD10 U07.1 and U07.2 and has since expanded to include U09.9 and U10.9. That is to say, folk who have recovered from COVID but are believed to be suffering from some long-term condition connected to COVID who go on to eventually die are now counted as a COVID death.
Then we’ve got that U10.9 multisystem inflammatory syndrome (a.k.a Kawasaki) that has been around for a while and is thus obviously caused by things other than SARS-COV-2 (and presumably still caused by things other than SARS-COV-2). No matter, we’ll go count those as COVID deaths too, shall we? Needle-less to say any genuine vaccine injuries may well be disguised by these codes.
Whilst we are at it we might as well grumble about U07.2, which also ensures all bases are covered no matter what. But that’s not the end of it -here’s how WHO recommend physicians approach death certification…
"With reference to section 4.2.3 of volume 2 of ICD-10, the purpose of mortality classification (coding) is to produce the most useful cause of death statistics possible. Thus, whether a sequence is listed as ‘rejected’ or ‘accepted’ may reflect interests of importance for public health rather than what is acceptable from a purely medical point of view. Therefore, always apply these instructions, whether they can be considered medically correct or not."
Neither is that the end of it because a fancy bit of software called MUSE 5.8 (Multicausal & Unicausal Selection Engine) decides what eventually gets coded as an underlying cause of death. Until we independently audit this system against the original death certificates we’ll never know whether what a physician decided is what turns up in the statistics. The slippery nature of all this is why eminent bods prefer to analyse all cause mortality in order to ascertain what might have been going on as opposed to what the authorities claim has been going on.
Differences in Counts
We may come to the realisation that we’re swimming in vaguely-fashioned data that we can trust just as much as we can trust Boris Johnson’s hair stylist. This harsh fact of numerical life being accepted you’d think that data authorities would at least get their story neat and straight, but they can’t even do that. What I’ve done is download counts of certified COVID deaths from the UK GOV coronavirus dashboard and compared these with monthly figures declared by the ONS to be due to COVID. Here’s the result.
We observe decent agreement for some months and decently wild disagreement for other months. Quite how these can be so at odds with each other is beyond me. If we calculate the ratio of ONS counts to UK GOV counts we arrive at the following slide, which reveals a strangely cyclical dance. Sometimes I sit back, blow on my hot cuppa and ask if any of this is real.
