Quote:
Originally Posted by nffc
Chris, let us not also not forget that the official metric is still "death within 28 days of a positive test".
So anyone who dies, but has tested positive for covid in the preceding 28 days, counts as this. Even if their covid illness was mild or asymptomatic, and irrespective of how they have actually died.
It isn't surprising that with 100k+ positive tests a day for the last 2-3 weeks we're now seeing an increase in "deaths" but how many of those deaths were caused by covid?
|
As much as everyone loves a good OF Covid or WITH Covid debate the number of “incidental” deaths can reasonably be calculated.
https://www.ons.gov.uk/peoplepopulat...31december2021
Quote:
|
In the week ending 31 December 2021 (Week 52), 8,477 deaths were registered in England and Wales; this was 4,536 fewer deaths than the previous week (Week 51) and 6.6% above the five-year average (523 more deaths).
|
In the UK between 1st and 28th December there were 2.4m positive cases on the UK dashboard. If you take off 15% for Scotland and NI (actual figures are available) the chances the proportion of the population of England and Wales testing positive is 2m(positive)/58m population give or take. Statistically the chances of incidentally dying within 28 days of a positive test for those dates is about 1 in 30. Of those 8477 deaths it’d be expected that about 1 in 30 would die incidentally following a Covid test - or 40 a day (280 a week E&W). In the same period 767 (UK) were recorded on the dashboard subject to the arbitrary 28 day cutoff.
It remains the case that the vast majority of people dying within 28 days of a Covid test appear - statistically at least - to have that as some factor within their death.
Greater accuracy could be achieved - for anyone with the time or inclination - by comparing age/sex stratified infection rates with death rates. E.g. if Covid infection rates were significantly higher in younger age groups (which I believe was the case) but people were mainly dying in older age groups (reasonable assumption) this would record more incidental Covid deaths in my model.