Thread by @cfishman

US Coronavirus Update for Friday, April 24, 2020
~ A divided day

Fatalities first ~

• Total US covid-19 deaths as of Fri, Apr 24: 51,192

• Increase in last 24 hours: 1,334 (+ 3%)

--> Smallest absolute increase in deaths in one day since Apr 6 (18 days)

2/ Deaths increased 1,334 on Friday — over a previous total of 49,858.

As a percent increase, that's: + 2.68%

In rough terms, that's like having had 50 deaths, and adding just 1.

Friday was the lowest percent increase in deaths, in the US, since the outbreak began.

3/ How is that even possible? In the early days of the outbreak, here's how death totals progressed.

Total US deaths (new deaths in parenthesis):

Mar 1: 1 (1)
Mar 2: 6 (5)
Mar 3: 9 (3)
Mar 4: 11 (2)
Mar 5: 12 (1)
Mar 6: 15 (3)
Mar 7: 19 (4)
Mar 8: 22 (3)

4/ When the number of deaths is that small, every death is a big jump in percentage.

US added 1 death — on a day it had only 12: + 8.3%

US added 3 deaths — on a day it had only 19: + 16%

US deaths haven't increased 16% in 2 weeks, haven't increased 8% in a week.

5/ Recall deaths increased on Thursday (4/23) by 3,092 — a big jump resulting from New York reclassifying deaths & adding several days' worth at once.

Friday's total — 1,334 — was a drop even compared to the running 5-day average, leaving out Thurs spike: 1,916.

6/ Is this good news? Breaking through to a new lower daily death plateau?

One day is not a trend, or even a signifier, as we've learned. It's an alert.

But I looked at death totals for every state, and none of the big states (CO, FL, NJ, NY, TX) had a dramatic fall.

7/ Pennsylvania reported 0 deaths. But it often goes in clusters of days — a low number, then a higher number, as if reporting is bumpy. And PA running < 100 deaths a day.

A glimmer of optimism.

As we have noted repeatedly: A lot of days of 30,000 new cases to work through…

8/ Now, new cases.

• Total confirmed US cases, Fri, April 24: 898,220

• Increase in cases in last 24 hours: 29,977 (+3%)

--> New cases rise compared to Thursday

--> Largest absolute increase in 6 days

--> New cases added are more than 25,000, on 10 of the last 11 days

9/ Deaths fall. Cases rise — and not a little bit.

But with cases, the pattern has been consistent: Consistently inconsistent.

Let's go back to our pattern of arrows, on the next tweet.

10/ Cases up or down, compared to previous day, starting Sat, Apr 11:

4/24 ↑
4/23 ↓
4/22 ↑
4/21 ↓
4/20 ↑
4/19 ↓
4/18 ↑
4/17 ↓
4/16 ↑
4/15 ↑
4/14 ↑
4/13 ↓
4/12 ↑
4/11 ↓

Alternate 8 days in a row. With the exception of 1 day (4/15), 14 in a row

11/ And the new cases aren't alternating up and down, while creeping down slowly.

Trailing 5-day average:

Fri, 4/24: 27,930

Fri, 4/17: 28,511

We may be staying home.
We may be practicing social distancing.
But somewhere, not enough. Because 28,000 people a day are newly sick.

12/ So what's going on?

Some big states, which have lots of cases, have hit a plateau that looks like a miniature of the national one:

CA
MA
NJ
PA
TX

Cases dropped, but aren't dropping anymore.

13/ If your state is on that list, no one should be talking about re-opening. The virus is making its way around as it is, very effectively.

New York has seen new cases rise 3 days in a row. The increase is a couple thousand, day over day. 3 days isn't a trend, but a warning.

14/ Some states with fewer cases are nonetheless also not seeing cases fall:

AZ
CO
IN
MI
MN
MS
NV
NC

These places add 50 to 200 cases a day — small compared to states adding 1,500 cases a day. But that level is persistent.

That virus isn't under control. It's waiting.

15/ Could Arizona or Minnesota become like New York?

No. But the Phoenix metro area has 5 million people. The Twin Cities have 3 million. Those places could have sustained, ugly outbreaks.

16/ And also on the watch-list should be four states with a small number of cases where the new cases are rising steadily, day over day, or on a higher plateau than 2 weeks ago — a month into stay-at-home:

SC
UT
VA
WI

Utah, Virginia & Wisconsin are going the wrong way.

17/ Graphs, from the data wizards @washingtonpost — for those three states below.

Check your own trends here:

washingtonpost.com/graphics/2020/…

18/ The national number of new cases isn't dropping because in some important states, cases have plateaued, and in a small number of states, cases are increasing.

Remember: If we stay at home, there shouldn't be an increase in cases over time.

Those states should be asking why.

19/ Now test data.

• New tests performed Fri, April 24, 2020: 223,552

* Third big day of testing in a row *

• Change from previous day: + 32,497

• 5-day trailing average of new tests performed: 203,531

• Total US tests performed as of April 18: 4,883,802

20/ Testing is growing. A real bright spot.

The last 2 days (Thu, Fri) have seen the highest daily totals that didn't happen because of a spike in reporting.

Went through every state's test data each of the last 3 days.

Thu & Fri represent real organic growth in testing.

21/ What does that mean?

If you look back at a state's test data, what do the trends look like?

FL: 20,000 tests a day, 2x their recent average

GA: 6,000 tests a day, 2x recent average

IL: 16,000 tests a day, 2x recent average

NM: 5,000 tests a day, 2x recent average

22/ The daily increment from just those four states comes to 23,000 additional tests a day being done.

If the whole country has only been doing 150,000 a day, adding 23,000 a day — permanently, at a new level — is a good increase in capacity.

23/ Also Friday:

New York state recorded its biggest test day ever — 34,000 tests in 1 day, part of a trend where New York is doing 10,000 more tests a day than its recent average.

Texas recorded its biggest test day ever — 17,000 tests Friday, 8,000 above its recent average.

24/ NY & TX — if trend holds — are adding another 18,000 new tests than typical per day.

Those 6 states are doing 41,000 more tests a day — recently — than they had been.

That's how you get from 150,000 tests a day to 200,000.

Several other states trending up less sharply.

25/ Reminder: 200,000 tests a day is not nearly enough. It's just the next step up from 150,000 tests a day.

The US needs 1 million to 5 million tests a day, depending on how testing is being used.

Testing is just now starting to catch up with the pace at which cases spread.

26/ Finally, let's benchmark Georgia — as Georgia is in the spotlight for starting to open businesses & its economy.

Today, we'll just look at where Georgia stood as of Friday, April 24.

We won't know how the Georgia experiment works for somewhere between 1 week and 4 weeks.

27/ Georgia has a population of 10.6 million people.

• Total covid-19 deaths in Georgia, through Apr 24: 904

• Deaths in the last 7 days: 231 (26% of total)

• Deaths per 100,000 residents: 9

• Rank in death rate of 50 states + DC: 16th

(Deaths trend below.)

28/ One thing to note about deaths in Georgia:

Except for a big spike on April 20, they have plateaued & started to trend down.

Deaths won't be an indicator of the safety of re-opening for many weeks, of course.

Coronavirus has a 7 to 14 day incubation period — to become ill.

29/ Now, Georgia cases:

• Total covid-19 cases in Georgia, through Apr 24: 22,695

• Days in a row of falling cases: 0

• 5-day trailing average, new cases added:

--> Fri, 4/24: 867
--> Fri, 4/17: 976
--> Fri, 4/10: 1,023

30/ Georgia has not started a significant, measurable downward trend in new cases.

On Fri, Apr 24, the day it started to allow people to go back to some workplaces, & to bring in customers, Georgia reported 979 new cases.

31/ New cases Friday at 979 is slightly above the 5-day average for that day — cases rose.

And 979 new cases is above the 5-day average for 1 week ago, and not significantly different than the 5-day average for 2 weeks ago.

Graph of new cases below.

32/ That is not a graph that shows steadily declining cases. And Georgia isn't at a plateau of adding 50 or 100 cases a day statewide.

It's adding 750 to 1,000 on average. As the graph shows, the reporting is a little bumpy.

The epidemiologists said: 14 days of falling cases.

33/ The White House said: 14 days of falling cases.

Is Georgia ready to start having starngers interact?

Is the coronavirus under control — not likely to catch fire again?

Skepticism is warranted.

But that's the experiment.

34/ It would be nice to see the spread of the virus whither as Georgia slowly re-opens to certain kinds of work life and recreation.

That would be confounding in the best way.

We'll watch Georgia. Update in next Saturday's thread.

#