What's the real Covid-19 death toll?
March 18, 2020 3:46 PM Subscribe
A death can only be attributed to Covid-19 if a test performed on the patient is positive, right? Are there any charts out there showing deaths due to any cause this year over last year, that might show spikes that could be attributed to untested coronavirus deaths? I'd be particularly interested in data/charts for locations where the disease is farther along than in the US.
Response by poster: I'm really hoping for some international data regarding deaths, not just infections, perhaps from Korea, China, Italy, Iran, France, or anywhere else that is ahead of the US in terms of infection.
posted by QuakerMel at 7:25 PM on March 18, 2020
posted by QuakerMel at 7:25 PM on March 18, 2020
Sorry I can't find a link now, but I remember reading somewhere that the death toll from influenza was way down compared to last year here in Japan, probably because people were taking precautions against coronavirus.
posted by Umami Dearest at 11:21 PM on March 18, 2020
posted by Umami Dearest at 11:21 PM on March 18, 2020
Via Johns Hopkins:
CoVID-19 Global Cases
gives the total number of deaths at ~9,000
posted by From Bklyn at 4:34 AM on March 19, 2020
CoVID-19 Global Cases
gives the total number of deaths at ~9,000
posted by From Bklyn at 4:34 AM on March 19, 2020
I don’t think this data will be available for a long time - right now you could look at all-cause mortality and look for excess deaths compared to other periods but this might include deaths that are not directly from COVID-19, for instance an acute but non-COVID-19 patient who died because there weren’t enough respirators, or a person with a septic infection who was afraid to go to the hospital because of COVID-19. And as From Bklyn points out, some kinds of deaths may be lower. Once things calm down they’ll be able to do tests to see how widespread the disease was in various populations and then we can make some statistical models.
Also it’s my understanding that the genetics very strongly indicate that this virus did not exist in humans before December (Nature Microbiology article about the taxonomy of the virus) so we wouldn’t be looking at the last year, just the last three-four months.
posted by mskyle at 5:14 AM on March 19, 2020 [1 favorite]
Also it’s my understanding that the genetics very strongly indicate that this virus did not exist in humans before December (Nature Microbiology article about the taxonomy of the virus) so we wouldn’t be looking at the last year, just the last three-four months.
posted by mskyle at 5:14 AM on March 19, 2020 [1 favorite]
Worldometers is kept up to date and is quick to load and easy to read. And allows easy filtering and sorting, eg by country. You can dig down on that site and see some results segmented by age and previous conditions etc
But it is a proxy, not an official source of any type.
What you're really asking for more generally doesn't really exist, or wouldn't be useful at any rate. Because "deaths attributable" to COVID-19 is far too hard to pin down at the moment. All we can go off is the numbers coming from the medical community, which are the ones you see in the official death tolls. There is some evidence that Iran has been less than open with its death toll, but I think the numbers for most countries will be pretty good.
If you die from it in a hospital there doesn't seem much room for misdiagnosis. The numbers should be roughly correct with or without testing prior to death. When you get to the state Italy is at (or near) you start to not admit the very high risk groups (eg very elderly) and they may die without their symptoms being observed by a medic. This is a kind of post-disaster triage - if you can't save the person because you know they'll need a respirator and you have non then don't use resources on them, offer the resources to someone you might be able to save. Its a horrible scenario but there are protocols.
Once you reach that level you might start over counting. Every frail person dying is counted. But if the family doctor has done any end of life assistance they'll be able to diagnose, its a distinctive infection in bad cases.
You can't just compare deaths this year to last year because society has completely changed its operating mode. Workplace accidents are stopping as there is no work. Traffic accidents are stopping as there is no traffic. Drunken accidents outside the bars are going to stop. Operations are being cancelled. People arriving at ER with a heart attack might not have the treatment they would have otherwise had. So many variables. The raw comparison will need a lot of unpicking to be useful.
There are going to be major long term attributable changes in mortality too, the numbers will take years to understand. If large percentages of people with chronic health conditions die this year might life expectancy rise for the remaining population? Might the closure of factories help with general health because the pollution stops? There is evidence that past coronavirus related respiratory diseases have left some survivors with long term compromised breathing, or susceptible to other infections. Might we have a co-hort of millions of survivors in their 30s who all develop chest issues in their 50s?
We have no idea.
posted by samworm at 5:48 AM on March 19, 2020 [2 favorites]
But it is a proxy, not an official source of any type.
What you're really asking for more generally doesn't really exist, or wouldn't be useful at any rate. Because "deaths attributable" to COVID-19 is far too hard to pin down at the moment. All we can go off is the numbers coming from the medical community, which are the ones you see in the official death tolls. There is some evidence that Iran has been less than open with its death toll, but I think the numbers for most countries will be pretty good.
If you die from it in a hospital there doesn't seem much room for misdiagnosis. The numbers should be roughly correct with or without testing prior to death. When you get to the state Italy is at (or near) you start to not admit the very high risk groups (eg very elderly) and they may die without their symptoms being observed by a medic. This is a kind of post-disaster triage - if you can't save the person because you know they'll need a respirator and you have non then don't use resources on them, offer the resources to someone you might be able to save. Its a horrible scenario but there are protocols.
Once you reach that level you might start over counting. Every frail person dying is counted. But if the family doctor has done any end of life assistance they'll be able to diagnose, its a distinctive infection in bad cases.
You can't just compare deaths this year to last year because society has completely changed its operating mode. Workplace accidents are stopping as there is no work. Traffic accidents are stopping as there is no traffic. Drunken accidents outside the bars are going to stop. Operations are being cancelled. People arriving at ER with a heart attack might not have the treatment they would have otherwise had. So many variables. The raw comparison will need a lot of unpicking to be useful.
There are going to be major long term attributable changes in mortality too, the numbers will take years to understand. If large percentages of people with chronic health conditions die this year might life expectancy rise for the remaining population? Might the closure of factories help with general health because the pollution stops? There is evidence that past coronavirus related respiratory diseases have left some survivors with long term compromised breathing, or susceptible to other infections. Might we have a co-hort of millions of survivors in their 30s who all develop chest issues in their 50s?
We have no idea.
posted by samworm at 5:48 AM on March 19, 2020 [2 favorites]
« Older Library ebooks for beginners? | Another should I stay or should I go question with... Newer »
This thread is closed to new comments.
posted by RobotVoodooPower at 4:43 PM on March 18, 2020 [1 favorite]