> The only data that is somewhat reliable in this pandemic is COVID deaths.

I agree,, especially as you included the word “somewhat”. We have pretty good numbers on people who died in hospitals, but people who died at home appear to be seriously under counted.

Additionally, consequential deaths are currently invisible. These are people who died, say, ecause the coronavirus crowded out their access to hospitals. There’s been a drop in ER visits for things like heart conditions. Some of those people would have recovered fine anyway but some may have arterial blockage that will kill them. You don’t Want these numbers in the primary statistics (as they say nothing about control of the pandemic) but you do want these very hard to track numbers to understating future capacity planning, ethical procedures, and long term economic consequences.

You can get part of that with excess death rates. From the weeks overall deaths subtract the average death rate for this week in past years. That should give both people who died at home and consequential deaths.

I'd reckon it's "doable" but not "easy", putting it in the realm of professional epidemiology. People are also dying less in car crashes, pollution-induced asthma, etc. There's quite a bit that may need to be accounted for. I suspect analyses will have decent error bars a year or so from now, but it's unlikely we'll get precise realtime estimates from this methodology. Would love to be proved wrong.

The NYT has been analyzing that.

https://www.nytimes.com/interactive/2020/04/21/world/coronav...

There are a lot of people not going to hospitals either because elective procedures were canceled or they fear getting corona virus. These will contribute to excess deaths but not be caused by coronavirus.

this is not easy, as death rates are going down because people are not out and about.

The number of deaths may be off but that is likely to be dwarfed by the much higher number of total cases that haven't been confirmed. Between this study, the LA study, and the Santa Clara study there could be 10-80x as many cases. That is more significant that the death count being off by a couple percent when it comes to determining the fatality rate.

Most nursing home deaths are not getting tested either.

France started testing systematically in EHPAD (nursing home for retired fully dependent elders). That's why they have the most dramatic fatality rate out of every country now.

Not sure it was a good idea to test and attribute all the deaths to COVID.

It's difficult because you don't want to use test kits on dead people.

Deaths is a lagging indicator though, so even if it's reliable it comes in a bit late to make any policy changes off of it, and nobody likes waiting for 3 weeks to a month after each little change to see how it's reflected in death rates. Maybe it's reliable in the statistical sense, but not reliable in the sense that a layperson or a politician would be able to rely on it to guide decision making.

>> Maybe it's reliable in the statistical sense, but not reliable in the sense that a layperson or a politician would be able to rely on it to guide decision making.

This seems to be what we're relying on in the UK. Hospital admissions for Covid show a good/relatively quick estimate of how measures are working and then a few weeks later a reflection of that trend in deaths confirms it.

Deaths can lie too. What does it take for a death to be labelled as covid-19 death? That the person who died was merely tested positive, or do they do autopsies for all those who tested positive? How about covid-19 deaths that aren't counted as covid-19 deaths, for a reason or another? How about people dying less of regular influenza this year because of covid-19 and lockdowns?

Even looking back at monthly 2020 deaths in comparison to earlier years might not truly reveal the final impact of covid-19.

In the US, if the person tested positive, it is classified as covid19 death.

Also it seems that 98% of covid19 deaths have at least one comorbidity, 75% at least two. So I would expect that in many cases the virus may have just accelerated something that would have happened within a few months anyway. So you kind of need to look at how the full year death rate evolved to have an idea of the net impact of the virus.

That's a misunderstanding of how death works. Everything is a comorbidity on some level.

> just accelerated something that would have happened within a few months anyway

That's absolutely not what this means. Asthma is a comorbidity. Are asthmatics just a few months from death? I have hypertension, am I about to kick the bucket? Sixty percent of the US population (no doubt including many of the posters on this very thread) are overweight, yet somehow we keep coming back.

Almost nothing kills "by itself". Everything adds up, and eventually the body stops working. Other diseases work this way too, yet you don't complain that yearly influenza deaths are inflated. Would you have argued that AIDS deaths were "really" due to the final infection and not HIV?

The study I got these numbers from doesn't seem to classify asthma or obesity as a comorbidity. The largest ones are hypertension, diabetes, ischemic heart disease, atrial fibrilation, active cancer in past 5y, heart failure, COPD, dementia, stroke and chronic liver disease. Not familiar with all items in this list but some seem pretty serious.

So you think that stroke victims and retirees with dementia are just months from the grave? FWIW: I've personally been living with diagnosed atrial fibrilation for almost three decades. Diabetes is, literally, a lifelong illness for its sufferers.

In fact, please cite me any of those factors which will reasonable by expected to kill you in months. Even heart failure and cancer patients tend to live years to decades with treatment.

Of the people who died of this virus, yes, it is likely that a not insignificant percentage were in pretty bad shape. We are not talking about the whole population who have these comorbidities in the country.

This shows the full list of conditions that are high risk: https://www.health.harvard.edu/diseases-and-conditions/if-yo...

A lot of people fall into those categories.

This is tricky to understand, but "all cause excess mortality" is pretty well understood by statisticians. They use it all the time for flu.

Here's one document from the UK that mentions the kind of testing they do to arrive at "deaths to flu" figures. They have international programmes to see what respiratory diseases are circulating, they test people who arrive in hospital for those, they know roughly how many people in the population they expect to die (because RTAs, cancer, heart disease, etc all stay sort of stable from one year to the next). They also do survey work to see if people have flu symptoms but haven't seen a doctor, and they can see how many people attend various healthcare settings for flu.

It's all a lot more complicated than "did someone write flu on a death certificate".

https://assets.publishing.service.gov.uk/government/uploads/...

Many coroners in my state have been quoted as saying they are unable to test the dead, as tests are prioritized for the living. Our state health department says that they do not report suspected COVID-19 deaths to the CDC if the person did not test positive.

I believe, at least in New York, people are being counted as covid-19 deaths if they were in contact with someone who was diagnosed. They can count as a covid-19 death without ever being tested for it.

I would say the number of hospitalizations is reliable as well. Those are the only two numbers I've really been paying attention to, because the number of total cases is really just a factor of testing, and doesn't yet give us an accurate picture of the situation.

I disagree. In order for a hospitalization to count, they have to have a positive test. First, we have to factor in test inaccuracy and then the fact that doctors are still triaging tests even if a patient has symptoms. These numbers can't be considered accurate until we have enough tests to test everyone in the waiting room indiscriminately.

We have enough tests to cover people who are candidates for ventilators or have similarly severe symptoms. That’s still a lagging indicator, but not nearly as delayed as deaths.

The countries that have fought it most effectively do so by contact tracing: whenever someone is found to have it, test everyone they've come into contact with recently.

This works best with really intrusive surveillance of mobile phone locations :(

I disagree that only deaths are reliable, fully tested patient ICU load is both easily measurable, fairly well controlled, and lower latency (roughly 2-4 weeks shorter!). I do agree that testing the dying is useful although it presents its own statistical challenges (dying with after a car accident with COVID).