That's the example everyone always brings up, and even if it's true, do you think there are a huge portion of people with COVID who are dying in car accidents? That seems very unlikely to make a significant difference in the figures, and for a new virus that is spreading very quickly and isn't yet well-understood, it's probably worth using a definition that errs slightly (but probably not significantly) on the inclusive side.

Car accidents in specific are probably small, but they're representative of similar catastrophic events that will lead to death anyway, that would still be marked as covid.

The most common of which is simply old age. Certainly covid causes death in some people, that's indisputable. But If you catch it in your last 2 weeks while you're on your death bed anyway, it's not really clear that covid even accelerated it.

The death die with covid more often not only because it probably accelerates some old peoples' deaths, but because this very population is undergoing its end-of-life process anyway, and are going to die whether or not they have it.

> But If you catch it in your last 2 weeks while you're on your death bed anyway, it's not really clear that covid even accelerated it.

Okay, but what's unique about COVID where we're supposed to discount that? Surely we care about a murderer who kills an elderly person.

> and are going to die whether or not they have it.

This is true for all people and causes of death.

> Okay, but what's unique about COVID where we're supposed to discount that? Surely we care about a murderer who kills an elderly person.

I'm not sure what you're saying here. If a murder kills someone, the murder was the direct cause of death and we count it as murder. If you die of old age while having covid, it is not necessarily covid that is the killer, so it's not clear that it should necessarily be counted as the killer in all old-age cases.

> and are going to die whether or not they have it

Yes but... you understand I'm talking about the specific case of actually being on your deathbed, right?

> Surely we care about a murderer who kills an elderly person.

That's not what it is though. If you put poison in someone's food, and they die before they eat it, you didn't murder them.

Sure, but if police found out, the investigation and legal process would be roughly the same as if you had successfully murdered them. The legal penalty would likely be less, but both cases are still treated with roughly the same level of seriousness by the legal system.

Car accidents were just the clearest example I could think of.

At https://www.socialstyrelsen.se/statistik-och-data/statistik/... they do calculate the difference:

> Av dessa har 90 procent (5 514 av 6 128) laboratoriebekräftad covid-19 enligt Folkhälsomyndighetens databas över smittade.

90% are due to Covid 19. 10% not.

Edit: I misunderstood, tsimionescu has corrected me, thanks!

That is not in any way what that 90% means.

> "Of these, 90 percent (5,514 out of 6,128) have laboratory-confirmed covid-19 according to the Swedish Public Health Agency's database of infected people."

That is 90% of:

> The statistics show the deceased where the underlying cause of death was covid-19, according to the cause of death certificates received by the National Board of Health and Welfare.

So in 10% of cases, the doctors were sure enough that the cause of death was Covid19 to write a legal document certifying it, even though they didn't order laboratory work to confirm with 100% certainty. This could mean many things, from medical malpractice (doctor lied on a legal document) to simple common sense (patient is husband of person with confirmed case, died of clear Covid19 symptoms).

But there is no way to read that 10% number as meaning what you claimed. In fact, if a person were hit by a car, confirmed Covid19 positive, died of their wounds on their way to hospital, and got a death certificate claiming they died of Covid19, they would be part of 90% number, NOT part of the 10% number.

Is that relevant? How many people happen to have some other acute terminal illness/accident AND Covid at the same time?

Also, do you imagine this is different for Influenza deaths or TB deaths? If anything, Covid-19 deaths are much more accurately counted than deaths from any other major disease, which are often just estimates.

10%, see my other reply.

Which is why I think excess mortality seems like the best high-level number to look at.

That 10% from your other comment has no relationship to Covid19 deaths vs people dying of unrelated causes who happen to be Covid19 positive.