A genuine question with how Omicron deaths will be measured

[GardenAhead5555]

To start the thread off, I am not a doctor so genuinely asking for opinions and happy to be corrected.

It is my understanding that covid deaths are registered as any deaths within 28 days of a positive covid test result. On main stream platforms such as sky the death figures are announced as deaths within 28 days of a covid positive test coming back measured by the ONS. If the vast majority of the population will contract the Omicron variant this winter, will that mean all deaths within 28 days of a positive test being confirmed will be classified as Omicron of all causes if it is within 28 days of a positive test result? Surely not as that will be millions of people testing positive of Omicron

If around 800 die per day of all causes each day in the winter period, how will they separate these from Omicron deaths?

There are two measures. The one which is reported which you mentioned is those who died within 28 days of a positive test. This is the official one and means accurate daily and weekly trends can be ascertained as it's a simple definition - it also allows comparison with other countries.

The second measure is where covid is thought to have contributed to the death - this is where covid is named on death certificates. Someone admitted with pneumonia who then also gets covid may die of pneumonia, but their doctor may judge that they would have survived had they not contracted covid, so covid is also named in the death certificate.

This second measure is much higher than the first.

The specific variant isnt afaik sequenced when someone dies - it's still covid.

But obviously variants are being tested for now with positive tests (they always were though - that's how new variants are detected).

As all positive cases are likely to be omicron in the next week or two, then yes all deaths within 28 days after that from someone positive would be attributable to this variant. Which it would be.

On your last paragraph I don't follow. If everyone who tests positive has the omicron variant why would they want to separate them from delta or other variants?

Excess deaths is a different measure.

I'm also not sure why you think the vast majority of the population will get the omicron variant - that's very unlikely!

@titchy

There are two measures. The one which is reported which you mentioned is those who died within 28 days of a positive test. This is the official one and means accurate daily and weekly trends can be ascertained as it's a simple definition - it also allows comparison with other countries.

The second measure is where covid is thought to have contributed to the death - this is where covid is named on death certificates. Someone admitted with pneumonia who then also gets covid may die of pneumonia, but their doctor may judge that they would have survived had they not contracted covid, so covid is also named in the death certificate.

This second measure is much higher than the first.

The specific variant isnt afaik sequenced when someone dies - it's still covid.

But obviously variants are being tested for now with positive tests (they always were though - that's how new variants are detected).

Ty for your reply to my question. I still don't understand though if millions of us contract covid, won't the official figures used by the government etc just shows all deaths within 28 days of an Omicron test as "Omicron deaths " which will be a huge percentage of deaths from all causes if millions contract a highly transmissible virus.

What I mean is, in winter thousands of people pass away every year but if the majority of the population have Omicron this winter then will it be classified as a death from Omicron even if it wasn't assuming it was within 28 days of a positive test?

@titchy

I'm also not sure why you think the vast majority of the population will get the omicron variant - that's very unlikely!

yes maybe you are right but they say cases will double up every 2 days so probably a large number

According to most recent ONS stats

The number of deaths registered in the UK in the week ending 3 December 2021 was 13,351, of which 909 involved COVID-19

@AlecTrevelyan006

According to most recent ONS stats

The number of deaths registered in the UK in the week ending 3 December 2021 was 13,351, of which 909 involved COVID-19

Does that mean due to covid or within 28 day of a positive covid test. I sometimes wish the government would be clearer on it as surely if someone goes into hospital for something life threatening ( non covid) but tests positive for covid within 28 days of death it can't be classified as one of those deaths.

It is a good question OP. They are about to start recording reinfections in the daily numbers, maybe they will separate out / redefine some other things at the same time?

It’s a really good question, particularly if omicron — whether because of vaccines/immunity or inherent qualities — turns out to be significantly milder than previous variants. Imagine, for instance, if we deemed everyone who died within 28 days of testing positive for one or another of the common cold viruses to have died from that virus. It would look like the cold had killed quite a few people simply because a lot of people get colds and some of those people (quite coincidentally) die. So whereas the measure of within 28 days of a test has been a blunt but useful tool in the pandemic to date, it may become quite confusing in the case of an extremely prevalent but not especially virulent variant. I’d like to know what statisticians think.

The 28 day measure is a quick and dirty, more up to date - but rough - indication of how many covid deaths are happening.

The count by what a doctor puts on the death certificate is more accurate but is subject to big delays because of autopsies, and reporting lags etc.

Excess deaths are also a useful measure.

www.google.co.uk/amp/s/amp.ft.com/content/05e32f95-0e7e-4d2a-b408-6ec6035dea8e

Does that mean due to covid or within 28 day of a positive covid test. I sometimes wish the government would be clearer on it as surely if someone goes into hospital for something life threatening ( non covid) but tests positive for covid within 28 days of death it can't be classified as one of those deaths.

Here's the 909 covid deaths in question:

I think the keywords there is " a cause of death " rather than the cause of death. I believe death certificates have often several conditions which may contribute to death, covid is listed in those cases you mention but it may / may not be the reason. Having just looked up ONS data, it appears if a patient tested positive for covid within 28 days of death that the dr is duty bound to include it as one of the causes.

Taking 500 deaths a day as an average in the UK which is standard pre covid. If over a 28 day period 14,000 people pass away of all ages and of that group of 14,000 anyone that has had a positive test for covid with the UK one of the most tested nations on earth dying of any cause being labelled a " covid death", I think it is reasonable to question the stats read out at some of the press conferences with Boris where figures are trotted out given how they are obtained.

If a doctor is writing a death certificate and told to include covid 19 as one of the causes just based on the above, the same question could be asked in order to ascertain who has died due to covid rather than merely testing positive for it within a certain time frame before death. I'm not really sure what method of testing would be more accurate to get accurate figures, but this is clearly a very crude way to determine death figures.

Even the excess deaths is open to debate as for whatever reason there were over 70 thousand excess deaths at UK homes since the start of the pandemic according to the ONS and 88% were not covid related. I'd imagine due to healthcare being inaccessible etc for many

news.sky.com/story/england-and-wales-records-more-than-70-000-excess-deaths-in-private-homes-since-start-of-covid-19-pandemic-12420482

I get what you're saying OP.

I think firstly we need to stop counting them as a "running total" and reset the clock on the 1st of January. That will be the 3rd year, it's ridiculous that we're still counting them all together and we're going to lose all perspective.

As for what counts... there's surely a better way? I know only one person who "died of covid" and they're of an age where there are very few deaths in that category. However the person was in such poor health prior (had been given weeks to live way before) that anything from a virus to a strong gust of wind would have sadly done it. The cause of death was put down as covid. It's totally wrong imo, and I've heard many others tell similar stories.

I think it should be counted as a 'covid death' if covid was the thing that actually did it and they wouldn't have died otherwise.

I also think this needs to be sorted out. With a variant that appears to be significantly more infectious but less lethal, there is a real risk of getting into a situation where a high % of people dying of something else will just happen to test positive for omicron too, and this will create distorted ideas of how many deaths omicron is actually causing.

In the meantime, excess deaths/excess hospitalizations for any reasons, are probably the metrics we need to be keeping an eye on.

@rarityroast

I get what you're saying OP.

I think firstly we need to stop counting them as a "running total" and reset the clock on the 1st of January. That will be the 3rd year, it's ridiculous that we're still counting them all together and we're going to lose all perspective.

As for what counts... there's surely a better way? I know only one person who "died of covid" and they're of an age where there are very few deaths in that category. However the person was in such poor health prior (had been given weeks to live way before) that anything from a virus to a strong gust of wind would have sadly done it. The cause of death was put down as covid. It's totally wrong imo, and I've heard many others tell similar stories.

I think it should be counted as a 'covid death' if covid was the thing that actually did it and they wouldn't have died otherwise.

Agreed. Almost two years in, the excess deaths statistically analysis is much more useful. And more startlingly horrific if you look at last year's. Yet the downplayers still insist that the "of/with" chestnut means the statistics are being manipulated to catastrophize Covid. Does beg the question on how those dead people were manipulated into dying to help the stats out if that's the case.

Interesting 'of' vs 'with' question. Nice to see sensible discussion about this too.

I agree the excess deaths is probably the most useful figure.

Yes the measure used in the daily dashboards gives the lowest number of deaths.

I disagree we should only count where covid is the main cause (although of course that data is collected). If someone dies and covid is likely to have contributed, even though not the primary cause, it should be included. If they hadn't had covid they would have survived.

I'm not aware of Doctors having to add covid to death certificates where it was not a contributory factor - if someone dies as the result of a car crash, then the fact that they tested positive a week ago shouldn't mean covid is named as a factor.

0The second measure is where covid is thought to have contributed to the death - this is where covid is named on death certificates. Someone admitted with pneumonia who then also gets covid may die of pneumonia, but their doctor may judge that they would have survived had they not contracted covid, so covid is also named in the death certificate.

The above is slightly wrong. Death certificates have two parts, part 1 the underlying cause of death, part 2 any conditions that may have contributed to the death. The conditions mentioned in part two must be known or suspected to have contributed to the death, not merely be other conditions which were present at the time. So CoVid on a death certificate can be the sole cause of death or suspected to have contributed to the death,

I haven’t looked recently but last year ONS reported approx 87% have CoVid recorded as the underlying cause of death, remaining 13% contributory.

Sorry I worded above wrong, as said certificates have two parts, part 1 the underlying cause of death, part 2 any conditions that may have contributed to the death.

CoVid on a death certificate can be the sole cause of death or suspected to have contributed to the death if other conditions were present, ie any other conditions would not have caused death alone but that condition combined with CoVid caused death. Again as said majority though have CoVid as the sole cause.

Colin Angus who tweets as @VictimOfMaths has some really interesting threads and tweets that cover this on Twitter if you want to check out some cool graphs and nifty analysis.

With a ruler