Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Excess All-Cause Mortality in the US and 18 Comparison Countries

jamanetwork.com

Post reply on HN
Print view View original

41–50 of 95 posts

Re: Excess All-Cause Mortality in the US and 18 Comparison Countries

#41

Post by coldtea »

wombatmobile wrote: post #19
darawk wrote: post #2

They mention it, but a big factor driving this is almost certainly obesity. I'd really like to see this done with proper controls for obesity, age demographics, smoking, etc. I think there'd be interesting stuff to learn from that. It's shocking how difficult it is to do these studies correctly. E.g. https://papers.ssrn.com/sol3/papers.cfm?abstract_id=3674138

Australia has similar obesity rates to USA, and dramatically lower COVID mortality.

Similar obesity rates, similar AS obese people though?

E.g if obesity is > 30 BMI, Australia might have similar percentage to the US, but US might still have much mor people > 35 BMI than Australia which could have mostly 30-35 BMI people.

Re: Excess All-Cause Mortality in the US and 18 Comparison Countries

#42

Post by phreeza »

m0zg wrote: post #36

It's dumb to do this analysis now, when we're in the _middle_ of the epidemic. According to most experts, we're looking at the end of 2021 for the whole thing to play itself out. Take a look at the steep incline the curves have taken on in the countries which previously were "better" than the US and Sweden. In the US deaths are declining (in spite of slight rise in cases), in Sweden hardly anyone is dying at all. Tak…

Treatments are already much better than they were in the early phase of the pandemic, and they are unlikely to get worse over time. Also mortality is likely to rise with hospital load.

So it follows that stretching your infections over a longer time is beneficial.

Re: Excess All-Cause Mortality in the US and 18 Comparison Countries

#43

Post by iso1210 »

m0zg wrote: post #36

It's dumb to do this analysis now, when we're in the _middle_ of the epidemic. According to most experts, we're looking at the end of 2021 for the whole thing to play itself out. Take a look at the steep incline the curves have taken on in the countries which previously were "better" than the US and Sweden. In the US deaths are declining (in spite of slight rise in cases), in Sweden hardly anyone is dying at all. Tak…

Cases started growing in Sweden on Sep15, that's not long enough for them to reach the most vulnerable, to catch it, to go to hospital, and to die

Sweden's GDP in Q2 was down 8%. Germany's was down 10%.

Re: Excess All-Cause Mortality in the US and 18 Comparison Countries

#44

Post by threatripper »

m0zg wrote: post #36

It's dumb to do this analysis now, when we're in the _middle_ of the epidemic. According to most experts, we're looking at the end of 2021 for the whole thing to play itself out. Take a look at the steep incline the curves have taken on in the countries which previously were "better" than the US and Sweden. In the US deaths are declining (in spite of slight rise in cases), in Sweden hardly anyone is dying at all. Tak…

> It's dumb to do this analysis now, when we're in the _middle_ of the epidemic.

Correct. Mortality doesn't say very much now. The true IFR would be a better indicator but since testing availability is very inconsistent it is hard to compare the number of infected across countries or time. The mortality numbers are also inconsistent in many places (which is why we looked at all-cause mortality in the first place).

> It's a virus. It's endemic. _Everyone_ will get it eventually, if not now, then a month or two from now. There's really no way around this simple fact.

This is plainly wrong. Vaccines have decreased the severity of many viral diseases and in some cases even eradicated the virus completely. Even without a vaccine, some countries have driven down the infection numbers almost to zero. Even the countries that could not eradicate it will only see 10-80% of the population infected in the long run.

> Swedish strategy will "win" this one.

For one, it is too early to declare a winner. We have to wait.

The Swedish strategy may work very well in Sweden. But that doesn't mean that it would have worked in countries with a different social structure, population density or climate.

Re: Excess All-Cause Mortality in the US and 18 Comparison Countries

#45

Post by tpoacher »

coldtea wrote: post #39

> While the US had a lower COVID-19 mortality rate than high-mortality countries during the early spring, after May 10, all 6 high-mortality countries had fewer deaths per 100 000 than the US. I'm not sure why the factor this in at all. Why not compare each country to its OWN baseline mortality -- as opposed to confusingly compare between high and low mortality countries?

It makes sense to do so politically. :p

Re: Excess All-Cause Mortality in the US and 18 Comparison Countries

#46

Post by mlang23 »

coldtea wrote: post #39

> While the US had a lower COVID-19 mortality rate than high-mortality countries during the early spring, after May 10, all 6 high-mortality countries had fewer deaths per 100 000 than the US. I'm not sure why the factor this in at all. Why not compare each country to its OWN baseline mortality -- as opposed to confusingly compare between high and low mortality countries?

Because it is obviously a competition between countries. "We managed to get through with less casualties then country X, so we are superior."

Re: Excess All-Cause Mortality in the US and 18 Comparison Countries

#47

Post by nickff »

evgen wrote: post #30
pattusk wrote: post #21

Earlier quoted context omitted.

I'll probably get downvoted for this, but anyways: Maybe I should have added reliable to data. I'm highly skeptical of the ~4000 deaths figure reported as the death count for China and I have the sense that the authors of the study were too or they would have included it in their study. As for India, when you factor in rural areas, I don't think the reporting and data would be as reliable as for the US. [1] To go bac…

While I share the skepticism in the numbers from China and India, I believe that if they were anywhere near a similar per capita rate as the US there would be visible evidence. China attacked the problem hard and by all accounts seems to have succeeded. For the most part India seems to have just gotten lucky; maybe India's numbers will be adjusted over time to reveal some rural devastation, but I will want to wait fo…

China was trucking in loads of urns every day that were more than ten times larger then their reported death rates; doesn't that count as visible evidence?

Re: Excess All-Cause Mortality in the US and 18 Comparison Countries

#48

Post by threatripper »

phreeza wrote: post #40
threatripper wrote: post #27

Earlier quoted context omitted.

Australia counts 0.1% of the population as infected while the US counts 2.4% of the population as infected. Where in the study did they control for that when looking at the mortality (which is counted per population and not per infected)? EDIT: My point is that the epidemic is not over yet and countries with a low immunity could be hit harder later on. It's just too early to compare mortality numbers.

Why would you want to do that? At the end of the day, successful pandemic response is a question of how well you prevent deaths overall, not how likely infected people are to die.

> At the end of the day

Yes. Once the epidemic is over. It's not over yet.

Having low numbers now could mean that the following outbreaks are more severe. Having lockdowns now could mean that there is no more money for a second, third or fourth lockdown to keep the following outbreaks under control. It could be but it does not have to be. We have to wait until the pandemic is finally under control to really count numbers.

Re: Excess All-Cause Mortality in the US and 18 Comparison Countries

#49

Post by pjc50 »

darawk wrote: post #2

They mention it, but a big factor driving this is almost certainly obesity. I'd really like to see this done with proper controls for obesity, age demographics, smoking, etc. I think there'd be interesting stuff to learn from that. It's shocking how difficult it is to do these studies correctly. E.g. https://papers.ssrn.com/sol3/papers.cfm?abstract_id=3674138

There was a paper (which I of course can't find now) that claimed that you also needed to control for the extent to which obese people were given worse treatment or de-prioritised by medical professionals. It's not a blindable variable.

Re: Excess All-Cause Mortality in the US and 18 Comparison Countries

#50

Post by phreeza »

threatripper wrote: post #48
phreeza wrote: post #40

Earlier quoted context omitted.

Why would you want to do that? At the end of the day, successful pandemic response is a question of how well you prevent deaths overall, not how likely infected people are to die.

> At the end of the day Yes. Once the epidemic is over. It's not over yet. Having low numbers now could mean that the following outbreaks are more severe. Having lockdowns now could mean that there is no more money for a second, third or fourth lockdown to keep the following outbreaks under control. It could be but it does not have to be. We have to wait until the pandemic is finally under control to really count num…

I agree, but controlling for reported incidence as suggested by OP does not help with this.
Post reply on HN