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

Estimation of total mortality due to COVID-19

healthdata.org

Post reply on HN
Print view View original

381–387 of 387 posts

Re: Estimation of total mortality due to COVID-19

#381

Post by Blahah »

grumpyautist wrote: post #27
Blahah wrote: post #24

Earlier quoted context omitted.

More generally, there is no universally agreed way to record all the factors that contributed to a death at the time of death or certification. This is why excess deaths is the real metric by which the toll of this pandemic will be measured, at least in countries with reliable records about whether deaths happened at all.

Excess deaths doesn't work either. Gun violence has increased everywhere. So have suicides in my region of the world. I don't know who or why this comment got flagged, it doesn't change reality

Excess deaths in a decent model account for other explanatory factors (war, sociopolitical things, etc.) that are not linked to the factor in question (covid-19 here). I'm not suggesting the crude arithmetical number of excess deaths is the key, but that a nuanced estimate of excess deaths in a model that accounts for other significant factors is a far better measure than anything else we have available.

Re: Estimation of total mortality due to COVID-19

#382

Post by spekcular »

timr wrote: post #380
lazide wrote: post #375

Earlier quoted context omitted.

Not the poster, but happy to answer since I have a decent amount of experience in similar areas. Their point is that while in theory you could blind yourself to the holdout set - that rarely REALLY happens. At least in a truly good way. Almost everyone is going to either peek at the data, massage the data, or try the validation and when it fails tweak the model and try again, or any number of shenanigans that no one…

Yes, exactly. But I'd extend this one part: > Almost everyone is going to either peek at the data, massage the data, or try the validation and when it fails tweak the model and try again, or any number of shenanigans that no one admits happen - but happen all the time. While "peeking" at the data is certainly something that happens all the time, usually what happens for even the most rigorous of model-makers is that…

The point about peeking at the holdout set is fine, I guess. But then the slogan should be "models without a truly blinded holdout set are statistical fiction." Prospectivity doesn't do any work above and beyond blinding there.

I already addressed the point about distribution shift above. You need to assume that the past reflects the future at some point, otherwise you can't justify deploying the model even after the prospective test.

Re: Estimation of total mortality due to COVID-19

#383

Post by grumpyautist »

des1nderlase wrote: post #377
grumpyautist wrote: post #348

Earlier quoted context omitted.

So, the study is funded by groups that admit to existing to sell new vaccines? And you don't see a conflict of interest?

There could be conflict of interest, but someone still needs to show that these numbers are substantially wrong.

No, the burden of proof is on the author. And considering this is a model, not a study, I wouldn't hold my breath on it.

Re: Estimation of total mortality due to COVID-19

#384

Post by knowaveragejoe »

grumpyautist wrote: post #383
des1nderlase wrote: post #377

Earlier quoted context omitted.

There could be conflict of interest, but someone still needs to show that these numbers are substantially wrong.

No, the burden of proof is on the author. And considering this is a model, not a study, I wouldn't hold my breath on it.

The burden of proof rests on your shoulders, in this case.

Re: Estimation of total mortality due to COVID-19

#385

Post by teachingassist »

timr wrote: post #323
spekcular wrote: post #304

Earlier quoted context omitted.

Maybe I'm just ignorant of the relevant statistical methods, but how do you suggest models be validated prospectively ? All the model validation techniques I'm aware of come after the model is fit on the data (or a subset of it), like testing against a holdout set, or k-fold cross-validation, or checking R^2.

> testing against a holdout set Not just a holdout set -- a holdout set of prospective data , where you actually predict the future, blindly, and see how you do. Obviously, you have to wait a little while to gather such data before you make big claims with your model. Doesn't mean you get to skip it. It's not even clear to me that the authors did a cross-validation here, or even bothered to fit the free parameters of…

> a holdout set of prospective data, where you actually predict the future, blindly, and see how you do.

According to your theory, publishing would act as a 'nothing up my sleeve' validation of their model.

Since, you can check this model in the future - you can do it for yourself.

And yet, you seem to be complaining about this publication?

(I still don't agree with this as a principle - since I wouldn't now expect the future to reflect the past, and the authors make no claim to predictive powers when they are identifying that the present is unlike the past)

Re: Estimation of total mortality due to COVID-19

#386

Post by worik »

zhaksbs wrote: post #372
worik wrote: post #339

Earlier quoted context omitted.

Golly. If that is a hill you will not budge from it may be a hill you die on. All the cherry picking data in the world will not help you if you get sick. I am really sorry that some abstract principal becomes a hill to die on.

Ya, i think thats what he meant. Hes willing to do all the reasonable things and all the unreasonable theatrics but he refuses to be at homw-arrest. I cant speak for the OP, but for those of us who have given life and death some thought, lockdowns are worse than death.

I give life and death a lot of thought.

"Lockdowns are a unacceptable imposition on freedom" I can understand. I disagree, but I understand.

"lockdowns are worse than death" strikes me as odd.

Re: Estimation of total mortality due to COVID-19

#387

Post by pbreit »

joshuamorton wrote: post #340
pbreit wrote: post #334

Earlier quoted context omitted.

Hospitals and doctors were basically not handling any non-covid patients.

Alzeimer's currently doesn't have any effective treatments, so I'd repeat the question: how would the fatality rate from Alzheimer's double year over year?

Alzheimers patients are among the worst off in a restricted healthcare scenario.

https://www.wsj.com/articles/coronavirus-pandemic-led-to-sur...

Post reply on HN