Changes in U.S. Reporting

[CSSEGISandData]

In light of the increasing rate of cases being reported domestically in the U.S., and in order to retain timeliness and accuracy, we have switched from reporting at the county level to state level.

[MarkMMullin]

I've read the FAQs and appreciate the issue - with increasing spread comes a gut busting # of lines to represent detail data within the states. That said, unless JHU is calling that level of data suspect, I hope this level of detail returns. Pure state level modeling is a little grainy.

[reyemtm]

For those wanting county level data, the University of Virginia is posting county level data, it just needs parsed out. http://nssac.bii.virginia.edu/covid-19/dashboard/ I believe the county numbers are just for confirmed cases. The counties are also missing Census codes or lat lngs. For example:

name,Region,Last Update,Confirmed,Deaths,Recovered
Ohio,USA,2020-03-15 03:00:00 * CTY: Cuyahoga 11; Butler 4; Stark 3; Summit 2; Trumbull 2; Belmont 2; Lucas  1; Franklin 1; Lorain 1; Tuscarawas 1,28,0,0

If anyone creates a json parsed version of this data, please post a link or conversion function.

[DavidGeeraerts]

See ticket #558 The data csv files will be posted here: http://blog.lazd.net/coronadatascraper/

[JimBudde]

I haven't read all the comments here but here's my 2c on keeping stats at the state level 1) I completely appreciate the amount of work to capture all this information so adding county level detail, without some level of automation seems daunting. 2) I think it is more important to get as accurate a number as possible in as timely a fashion as possible 3) I believe there are Public Health risks announcing stats at too granular a level at this point in time. We need containment and if people think there are zero or very few cases in their area of interest, it invites complacency 4) I can't imagine State Public Health agencies aren't tracking things at their own county/city level so they can manage resources, etc so adding them here is really for just "us"

[reyemtm]

4 is a dangerous sentiment - public transparency should be the goal here. Also, we have mass complacency, it seems doubtful anything could make it worse. And as far as accuracy, this data is valid in terms of reported, symptomatic cases, but hardly accurate due to extreme lack of testing of asymptomatic carriers. For personal use I may just start tracking my own state from their daily reports and use this data as a sort of reference.

[JimBudde]

Re: #4. I agree with transparency but I have been in the middle of major events and while you want to get as much info out to the public/communications group, your main goal is to return to normal operations as quickly as possible, not keeping a website up to date with data. Hence, groups from around the globe have stepped in.

And as far as accuracy, this data is valid in terms of reported, symptomatic cases, but hardly accurate due to extreme lack of testing of asymptomatic carriers

I don't know what 'accurate' means. There will always be an unknown number of infected people; most of us with regular cold, or even flu just stay home, nurse ourselves back to health and were never tested.

[reyemtm]

An more accurate count of infections will give a higher confidence to the death rate. This can only come about if asymptomatic people or people mild symptoms are tested. That is what I mean by accurate. But in terms of the discussion at hand, if validity and expediency is our main concern, I can see the argument for omitting the county data.

[JHEASTON]

Thank you Thank you Thank you!

[JeremyIglehart]

It's so good to see county level data back today, THANK YOU!!!