[82lsju]

not surprising and I'm sure there will be many more as time goes on

The Chinese city of Wuhan raised its death toll from the novel coronavirus by 50% on Friday, bringing its total to 3,869, amid doubts about the accuracy of China's data on the disease as global cases mount.



The central city where the virus first appeared in humans late last year added another 1,290 fatalities to the 2,579 previously counted as of Thursday, reflecting incorrect reporting, delays and omissions, according to a local government taskforce in charge of controlling the coronavirus.

https://www.nytimes.com/reuters/2020/04/...-toll.html



and likely some uncounted in NYC

But when the final death toll is reported, it may be bigger than current data would suggest. A [/url][color=var(--ds-color-beijing)]recent report from WNYC and [i]The Gothamist[/i], two New York media outlets, notes that the city’s official covid-19 fatality statistics have until recently excluded those who died in their homes, and therefore were never diagnosed with the virus. Exactly how many people are missing from the tally is unclear. One clue comes from the number of people calling 911 about cardiac arrests, a known consequence of covid-19.[/color]



According to data from the city’s fire department, New Yorkers are making four times as many emergency calls about cardiac arrests as they did last year. The most recent figures show that about 200 of these heart-attack victims are being pronounced dead at the scene each day, up from around 30 this time last year (see chart). It is plausible that some of these people were not victims of covid-19.

[url=https://www.economist.com/graphic-detail/2020/04/13/deaths-from-cardiac-arrests-have-surged-in-new-york-city]https://www.economist.com/graphic-detail/2020/04/13/deaths-from-cardiac-arrests-have-surged-in-new-york-city



https://gothamist.com/news/surge-number-...ted-deaths

[82lsju]

not surprising and I'm sure there will be many more as time goes on

The Chinese city of Wuhan raised its death toll from the novel coronavirus by 50% on Friday, bringing its total to 3,869, amid doubts about the accuracy of China's data on the disease as global cases mount.



The central city where the virus first appeared in humans late last year added another 1,290 fatalities to the 2,579 previously counted as of Thursday, reflecting incorrect reporting, delays and omissions, according to a local government taskforce in charge of controlling the coronavirus.

https://www.nytimes.com/reuters/2020/04/...-toll.html



and likely some uncounted in NYC

But when the final death toll is reported, it may be bigger than current data would suggest. A [/url][color=var(--ds-color-beijing)]recent report from WNYC and [i]The Gothamist[/i], two New York media outlets, notes that the city’s official covid-19 fatality statistics have until recently excluded those who died in their homes, and therefore were never diagnosed with the virus. Exactly how many people are missing from the tally is unclear. One clue comes from the number of people calling 911 about cardiac arrests, a known consequence of covid-19.[/color]



According to data from the city’s fire department, New Yorkers are making four times as many emergency calls about cardiac arrests as they did last year. The most recent figures show that about 200 of these heart-attack victims are being pronounced dead at the scene each day, up from around 30 this time last year (see chart). It is plausible that some of these people were not victims of covid-19.

[url=https://www.economist.com/graphic-detail/2020/04/13/deaths-from-cardiac-arrests-have-surged-in-new-york-city]https://www.economist.com/graphic-detail/2020/04/13/deaths-from-cardiac-arrests-have-surged-in-new-york-city



https://gothamist.com/news/surge-number-...ted-deaths

[burger]

These sorts of reports are coming in from all over the world.  I don't have the links, but my rough estimate from looking at a bunch of these was that the final number will end up being 2x the reported numbers. 

Also, whoever made that NYC graph (the economist?) needs to have their license to plot data revoked for having different scales on the X-axes that make it impossible to compare the timing of the peaks.  But, to my eye at least, it looks like the home deaths are still climbing while the official count is decreasing.  That's scary.

[burger]

These sorts of reports are coming in from all over the world.  I don't have the links, but my rough estimate from looking at a bunch of these was that the final number will end up being 2x the reported numbers. 

Also, whoever made that NYC graph (the economist?) needs to have their license to plot data revoked for having different scales on the X-axes that make it impossible to compare the timing of the peaks.  But, to my eye at least, it looks like the home deaths are still climbing while the official count is decreasing.  That's scary.

[2006alum]

Here's a link to where the Economist will keep updating this data as it is made available (paywalled).

One of the best threads I've seen on this is from a guy in Italy who's been meticulous about tracking the differences in each Italian town between total deaths this year, deaths formally attributed to COVID-19, and death rates in prior years:
[tweet]https://twitter.com/DellAnnaLuca/status/1250441566172229634?s=20[/tweet]
If you go through his past tweets you'll see lots of stuff like this. I think at least 2x unreported is right, probably closer to 3x in regions with high numbers of old folks in care homes, etc.

[2006alum]

Here's a link to where the Economist will keep updating this data as it is made available (paywalled).

One of the best threads I've seen on this is from a guy in Italy who's been meticulous about tracking the differences in each Italian town between total deaths this year, deaths formally attributed to COVID-19, and death rates in prior years:
[tweet]https://twitter.com/DellAnnaLuca/status/1250441566172229634?s=20[/tweet]
If you go through his past tweets you'll see lots of stuff like this. I think at least 2x unreported is right, probably closer to 3x in regions with high numbers of old folks in care homes, etc.

[BostonCard]

These sorts of reports are coming in from all over the world.  I don't have the links, but my rough estimate from looking at a bunch of these was that the final number will end up being 2x the reported numbers. 

Also, whoever made that NYC graph (the economist?) needs to have their license to plot data revoked for having different scales on the X-axes that make it impossible to compare the timing of the peaks.  But, to my eye at least, it looks like the home deaths are still climbing while the official count is decreasing.  That's scary.

The economist graph is truncated.

BC

[BostonCard]

These sorts of reports are coming in from all over the world.  I don't have the links, but my rough estimate from looking at a bunch of these was that the final number will end up being 2x the reported numbers. 

Also, whoever made that NYC graph (the economist?) needs to have their license to plot data revoked for having different scales on the X-axes that make it impossible to compare the timing of the peaks.  But, to my eye at least, it looks like the home deaths are still climbing while the official count is decreasing.  That's scary.

The economist graph is truncated.

BC

[82lsju]

The economist graph is truncated.

BC

how can you size embedded pictures to not truncate?

[82lsju]

The economist graph is truncated.

BC

how can you size embedded pictures to not truncate?

[BostonCard]

The economist graph is truncated.

BC

how can you size embedded pictures to not truncate?

Click on the insert an image icon at the top of the toolbar, and enter a width less than the column width of the page, which probably is around 1024 pixels; enter a height that is roughly proportional so that the image isn't skewed.  It will enter the code to display a scaled version of the figure.

BC

[BostonCard]

The economist graph is truncated.

BC

how can you size embedded pictures to not truncate?

Click on the insert an image icon at the top of the toolbar, and enter a width less than the column width of the page, which probably is around 1024 pixels; enter a height that is roughly proportional so that the image isn't skewed.  It will enter the code to display a scaled version of the figure.

BC

[burger]

These sorts of reports are coming in from all over the world.  I don't have the links, but my rough estimate from looking at a bunch of these was that the final number will end up being 2x the reported numbers. 

Also, whoever made that NYC graph (the economist?) needs to have their license to plot data revoked for having different scales on the X-axes that make it impossible to compare the timing of the peaks.  But, to my eye at least, it looks like the home deaths are still climbing while the official count is decreasing.  That's scary.

The economist graph is truncated.

BC

It wasn't the truncation that was a problem--it's just a bad graph.  The scale on the x-axis of the right panel (in days per inch) is different from the scale on the other two.  That makes the slopes impossible to compare across plots, and it makes it hard to know where the decline in reported cases fits in relative to the other deaths.

As someone who spends an inordinate (no pun intended) amount of time making graphs, bad graphs are a personal bugbear.

[burger]

These sorts of reports are coming in from all over the world.  I don't have the links, but my rough estimate from looking at a bunch of these was that the final number will end up being 2x the reported numbers. 

Also, whoever made that NYC graph (the economist?) needs to have their license to plot data revoked for having different scales on the X-axes that make it impossible to compare the timing of the peaks.  But, to my eye at least, it looks like the home deaths are still climbing while the official count is decreasing.  That's scary.

The economist graph is truncated.

BC

It wasn't the truncation that was a problem--it's just a bad graph.  The scale on the x-axis of the right panel (in days per inch) is different from the scale on the other two.  That makes the slopes impossible to compare across plots, and it makes it hard to know where the decline in reported cases fits in relative to the other deaths.

As someone who spends an inordinate (no pun intended) amount of time making graphs, bad graphs are a personal bugbear.