The data - an observation

[matt819]

I generally check on the Johns Hopkins site for updated info on COVID-19: https://gisanddata.maps.arcgis.com/apps/opsdashboard/index.html#/bda7594740fd40299423467b48e9ecf6

We are all well aware of the shortcomings of sites like this and the analysis of data from them. And those shortcomings are driven by the data being provided or scraped. Do we believe that the virus is under control in China? Does anyone believe the data coming from Russia?

And then there's the US.

On March 29, to use one example (because it's a screenshot I have stored), there were 142,106 cases and 2,479 deaths, a case fatality rate of 1.7%, which was lower than the worldwide figure of 4.7% the same day.

Re the US figure, I think many assumed that the fatality rate was probably even lower because in the absence of widespread (or at least wider-spread) testing we believed the number of infected was much higher. It may have even been reasonable to assume that. (Though even a moderately intelligent 12-year-old knows to prepare for the worst and hope for the best.)

My point. Take a look a the numbers today: US infected 399,979 and US deaths 12,912. Just using these numbers, that fatality rate is nearly double that of 2 weeks ago, 3.2% (with the worldwide rate at 5.7%). We've all read the reports that the infection numbers are probably higher than being reported and, this is important as well, the the deaths are probably higher as well because deaths due to COVID-19 are not being attributed to this virus. Of course, we're back into assumption mode. How do we deal with the belief that the infection and death numbers are likely higher than reported. So let's take those out of the picture for now and use the numbers that are being publicly disseminated.

So, my observation is that I find it interesting that the US fatality rate is moving toward the worldwide figure. I find that terrifying, especially with the government we have and its abject, criminal failure to respond properly.

BTW, WHO, as of 7 April, shows 333,811 US cases and 9,559 deaths, a rate of 2.8%. https://www.who.int/docs/default-source/coronaviruse/situation-reports/20200407-sitrep-78-covid-19.pdf?sfvrsn=bc43e1b_2

And worldometers shows 404,156 cases today, and 12,988 deaths, a rate of 3.2%: https://www.worldometers.info/coronavirus/

[genxlib]

1. You have to factor in time

It takes some time (say 10 days) for a case to get to resolution.

So today's death totals are reflective of the new cases that showed up at the end of march.

Current case totals won't really have a resolution rate until they have been in the system awhile.

[matt819]

2. Right

Good point. I usually refer to those but forgot to this time.

You're right about the lag time, and I also wonder about accuracy in reporting recovered numbers.

If you look at China, the recovered and deaths almost add up to the total infected. Which means some are at what one would hope is the tail end of the infection.

In the US, there are still 370,000 ill, neither dead nor recovered. I guess it would be instructive to look at the numbers from China at the same point in the spread of the virus that the US is at now.

[Mike 03]

3. In Italy one way they attempted to reconcile discrepancies was noting an increase

in death notices (maybe newspapers or some more official data). An unusually large number of people were dying beyond what was accounted for by the case fatality figures provided by hospitals. It wasn't intentional, it was just that too many people were dying so quickly.

One thing I'm really worried about is people dying alone in their homes, trying to "tough it out" or otherwise afraid to go to the hospital because they aren't insured, don't have a doctor, or some other hindrance. They'll only be discovered gradually as time goes on or people notice they aren't getting their mail, haven't been seen in some time, etc... Will they be classified as official coronavirus deaths? Will there even be autopsies?

Some hospitals are "assuming" patients have COVID and not testing, and I don't know if they become official statistics since they haven't been tested. When they die of pneumonia, are they officially categorized as virus fatalities or "unknown, pneumonia"?

You raise crucial questions.