AIBU to think the Covid death toll statistics must be a bit skewed?

[AngeloMysterioso]

Been reading the news update about the covid death toll reaching 100k.

What it says on BBC news is “More than 100,000 people have died with Covid-19 in the UK, after 1,631 deaths within 28 days of a positive test were recorded in the daily figures.” (my emphasis).

Does this mean that if someone tests positive for covid but is asymptomatic or recovers quickly, but then gets poisoned or hit by a bus or dies of some other entirely unrelated cause within 28 days, it’s still counted in the covid death stats? Surely that means the statistics are unreliable?

They aren’t inaccurate, there will be a slight degree of error in either direction, but for the most part they are accurate, the excess death data shows us that.

This is the fifteenth thousandth time this has been discussed here

Ah but surely some of those fifteen thousand were actually threads about being hit by a bus and were just marked down as covid threads because someone on page one mentioned a positive test...

But if it’s an over-estimation, or an underestimation, by as much as 10%, then it isn’t a fact, is it?

More people die outside the 28 day window than die within 28 days of testing positive from totally unrelated reasons to testing positive of covid.

It isn't a perfect statistic but overall it is estimated this underplays covid deaths rather than overplays them, particularly considering how many people will have died Feb-April without being tested at all so go unaccounted for.

Where have you been op? Yes, this is an issue which is widely complained about online. Lots of people with relatives with long term health conditions who go into hospital, contract covid and are then pronounced dead "with covid". Their kidney disease, cancer, Stoke etc secondary. Also people in nursing homes, having not even been tested having covid put on the death cert since vague "symptoms matched". Also last spring, Germany was only noting deaths from covid, hence much lower numbers.

Last I heard, couple of weeks ago, fewer than 400 people under 60 had died from covid with no other comorbidities. Average age of death with covid 84 (life expectancy at birth 82). It's a travesty that our children are putting their lives on hold and paying the price to prolong the lives of octogenarians. But yes, let's save the NHS 🙄

@bitheby

It also misses out all the people who are on ventilators for weeks before they die.

According to ONS stats the 28 day figure is likely to be an UNDERestimate.

Thanks for this post.

I was going to ask about that as my understanding is most people aren't admitted until after first week and then take days to a week before ventilation. I thought it could up to 21 days after infection. And news reports seem to indicate most are ventilated for weeks.

No need to be rude... I’m not trying to be adversarial, I’m trying to understand how this is worked out.

My dad 'died of covid' he was actually very unwell with an underlying infection and had fairly advanced dementia, he'd been on end of life pathway twice earlier in the year. He was admitted to hospital (having tested positive a day or two earlier) for observation following an unrelated injury and kept in because his breathing became poor and he needed oxygen and a combination of steroids and antibiotics to combat both the infection and the covid.

He then came off the oxygen and was recovered enough to return to his care home but then deteriorated again a week later (common in covid I'm led to believe) and died shortly afterwards.

His cause of death was Covid with dementia as an underlying factor. Technically the covid was working its magic alongside his other conditions but I agree with the doctor who wrote his death certificate that if it weren't for the covid he would have still died, just not at that point. Therefore I would say it makes sense for it to be his cause of death despite his apparent recovery and return from hospital.

It is as accurate as we have at any given moment and gives a good indication in terms of providing a snapshot. When viewing it you have to keep in mind that a, not all deaths will be recorded yet because some will have been referred on to the coroner which can take months and b, some people do die outside of the 28 day window for the daily reporting from the government but will still eventually be recorded in with the overall covid death figures.

@AngeloMysterioso

But if it’s an over-estimation, or an underestimation, by as much as 10%, then it isn’t a fact, is it?

No one ever said it is an absolute fact, they said those are the numbers of people who died within 28 days of officially contracting covid.

We need some way of assessing the deaths, what other method are you recommending?

A lot of people take longer than 28 days to die, and they are missed off the death statistic, which more than balances out the very small number that die of other causes during that period.

Surely the 28 day time frame needs to be revised then?

There seems to be an awful lot of scope for error considering it’s information that’s being broadcast to the nation on a daily basis

You'd have to be exceptionally unlucky to have Covid and a fatal car accident within 28 days of each other.

However many people will die from a system weakened by Covid, even if the direct cause was for example heart failure and they had mainly recovered from Covid. Would they have died of heart failure this year if they hadn't had Covid? Were they left so weakened by Covid, that some opportunistic infection carried them off?

@AngeloMysterioso

Been reading the news update about the covid death toll reaching 100k.

What it says on BBC news is “More than 100,000 people have died with Covid-19 in the UK, after 1,631 deaths within 28 days of a positive test were recorded in the daily figures.” (my emphasis).

Does this mean that if someone tests positive for covid but is asymptomatic or recovers quickly, but then gets poisoned or hit by a bus or dies of some other entirely unrelated cause within 28 days, it’s still counted in the covid death stats? Surely that means the statistics are unreliable?

There are also people who last longer than 28 days with the illness before they die. They are excluded, and tbh that is much more likely to happen than being hit by a bus.

The statistics that I place most trust in is the number of death certificates giving COVID as the cause. Doctor's judgement is better than an arbitrary cut off IMO.

The excess deaths are greater than the deaths within 28 days, so it is likely an underestimate of covid deaths.

I'm sure I am bias as a clinician, but I wish they'd just use clinician's judgement and use what's on the bloody death certificate rather than invent an arbitrary number of days from a test. Give us some credit.

The statistics that I place most trust in is the number of death certificates giving COVID as the cause. Doctor's judgement is better than an arbitrary cut off IMO.

Exactly. Although working that data into a statistic would obviously take longer, but in all honesty I don’t see the need for the Daily Death Toll anyway. It’s bad. We get it.

The question I'm asking is not whether ours are skewed, but why other countries with quite large populations have such a low death rate. I know we have an ageing population and are in quite poor health, but it still does quite add up. There is a lot of number fudgery going on in the world or they just don't have the means to count the deaths/gather the data.

@Taciturn

Where have you been op? Yes, this is an issue which is widely complained about online. Lots of people with relatives with long term health conditions who go into hospital, contract covid and are then pronounced dead "with covid". Their kidney disease, cancer, Stoke etc secondary. Also people in nursing homes, having not even been tested having covid put on the death cert since vague "symptoms matched". Also last spring, Germany was only noting deaths from covid, hence much lower numbers.

Last I heard, couple of weeks ago, fewer than 400 people under 60 had died from covid with no other comorbidities. Average age of death with covid 84 (life expectancy at birth 82). It's a travesty that our children are putting their lives on hold and paying the price to prolong the lives of octogenarians. But yes, let's save the NHS 🙄

You know that "having comorbidities" can mean as little as that the person was overweight? A large proportion of the population has underlying conditions of some kind.

Just because someone under 60 "had comorbidities" doesn't meant they were at death's door or even ill.

And plenty of children and teenagers "have comorbidities" too.

The data from death certs is available on the ONS website.

I love how those with 'co morbidities' aren't important. So we shouldn't worry if someone died of covid because they also had psoriasis?? (Yes, that is on the list of co morbidities). Someone with cancer who may live another five years, but gets covid and dies shouldn't be counted because they had cancer anyway. 🙄🙄🙄

@NoOpinionNoProblem

The question I'm asking is not whether ours are skewed, but why other countries with quite large populations have such a low death rate. I know we have an ageing population and are in quite poor health, but it still does quite add up. There is a lot of number fudgery going on in the world or they just don't have the means to count the deaths/gather the data.

I agree. Different countries are counting differently and some are also fudging I suspect. The only way we will know is with the benefit of hindsight when we can look at excess deaths

hit by a bus

That old chestnut. It's the epidemic of bus crashes I'd be worrying about the most

Sure, so it's okay to lockdown the whole country, wreck the economy and leave millions destitute because some people are overweight???

The link is here if anyone is interested: www.ons.gov.uk/aboutus/transparencyandgovernance/freedomofinformationfoi/covid19deathswithnounderlyinghealthconditionsbrokendownbyage?:uri=aboutus/transparencyandgovernance/freedomofinformationfoi/covid19deathswithnounderlyinghealthconditionsbrokendownbyage

I don't know. I don't think anyone knows the true numbers, quite impossible to tell really.

What I do know is that the BBC only headline it if it goes up. It's down the bottom of the page or as an addendum to another article if it's gone down.