A serious discussion about re-opening a state's economy (using OH as an example)

Submitted by crg on April 22nd, 2020 at 4:27 PM

Note: This is not meant to be a political (read: partisan) discussion, but a practical discussion about how states are/can/should approach the issues of re-opening.  I believe that the board is (mostly) mature enough to limit the discussions to reasonable facts and opinions and not lapse into a political flame-war... maybe.

 

I just watched today's COVID update press conference from OH governor Mike DeWine (playfully known as "Wine with DeWine Time" around these parts - if only since the man is so bland when talking if nothing else).  The conversation was mostly about his plan to re-open businesses (to some extent... TBD) starting on May 1 and has been a bit... underwhelming.

For context, DeWine (a Republican) has been (possibly surprisingly so) one of the more pro-active governors in the nation about getting ahead of the crisis.  One good example:  on Sunday March 15th he ordered the closure of bars & restaurants statewide (well before most states had taken any significant action at all - I happened to be back in MI visiting family that weekend and they were shocked to see that such a drastic action was taken by anyone, let alone a Republican in OH).  DeWine received a great deal of criticism (in OH and elsewhere) for that action at the time (and other similar actions shortly after), but was vindicated by late March when much of the rest of the country had taken the same steps (he was even called out for praise internationally by the BBC, among other outlets).

 However, his current actions in pushing for a re-opening of non-essential businesses on May 1st seem in stark contrast to his prior string actions to secure public health.  In his press conference today, he (and his Lt. Gov) outlined various economic reasons why the state economy needs to be re-engaged yet failed to address many of the immediate health issues raised in doing so (let alone why the May 1 date was being strictly pushed).  Just a few of the issues that I saw (and some of the reporters called out in the Q & A):

1)  DeWine earlier this week said that OH would adhere to the guidelines issued by the White House, which include 2 weeks of declining daily COVID cases and deaths before a state economy can re-open.  As of yesterday, the 7 day average of both for OH are slightly increasing (https://www.nytimes.com/interactive/2020/us/ohio-coronavirus-cases.html) and thus could not mathematically meet that criteron - and the re-open date is 9 days from now (he said they have "flattened", which is not strictly true, but also his med chief said today that ICU admissions are trending up again).

2)  A question was asked if the supply chain (masks, disinfectant, etc.) would be ready to support businesses and individuals returning to work by May 1 - especially considering that people are having high difficulty in finding masks even now.  He had no answer to this.

3)  Another question was regarding child care and the practicalities of having working parents leave home when child care centers might not be re-opened to sufficient levels (let alone being deemed "safe" health-wise) and with all schools closed until Fall.

4)  Much attention was given to the financial/economic risks of maintaining the shutdown yet absolutely no numbers/models/estimates were provided about the expected increases in cases/deaths that were likely to be incurred with a May 1 re-open (let alone modeling how these increases would change as a function of changing the re-open date).

5)  Coordination/comparison with other states was not really addressed.  He has in the past emphasized coordination with IN and KY (which have lower cases than OH), yet hardly mentions any coordination with MI and PA (which have higher cases) - yet even in today's Q&A a reported remarked that a Jeep plant in Toledo is scheduled to re-open May 4th and a large contingent of its workers commute from Detroit.  DeWine commented that the plant would be monitored by the OH Dept of Health to keep safe operation but made no mention/discussion about the movement of possible cases across state lines (and how to address it).

6)  A day or so ago, DeWine said he had been approached by small businesses that it was greater concern to them about re-opening soon only to be forced to close quickly again upon any cases resurgence - a later re-opening date would be an easier pain for them to accept than a series of starts/stops.  No information/discussion was provided today (or thus far) about how this would be addressed and how a May 1 date is or is not optimal to avoid this.

These are simply the issues my wife & I noticed while watching today (I'm sure there are others), but it is rather concerning that the state officials seem to be pushing blindly (or perhaps half-blindly) toward this re-open date without crafting a more deliberate and coordinated response.  I know other states are facing the same issue (some more agressively, such as Georgia and S.C., and others more cautiously, such as most of the northern eastcoast, WI, IL and CA), but it just seems foolish to have such a patchwork level response to this issue nationally.

Also, I know that many on this board live/work in OH, may be from OH or have family there, and am curious what their opinions are of DeWine's response and how they would like to see the state government proceed.

Apologies to all if this seems too ranting, but I had been relatively impressed by Ohio's handling of the crisis to date  - yet this seems like a significant risk to take that could cost 1000's of lives unecessarilly.

blue in dc

April 22nd, 2020 at 6:18 PM ^

Was hoping to find something a bit more objective.   https://pdi.voxel51.com/

This was apparently created by a U of M prof and student.   Provides interesting info for other cities, but unfortunately does not go back far enough for Detroit.

I’m pretty sure you can find true anecdotal data that there were people who didn’t follow social distancing requirements in just about every locality that was  impacted.  

LewisBullox

April 22nd, 2020 at 6:17 PM ^

Everyone understands the mechanism is not genetic. Nevertheless, the odds of a black person dying from covid thus far are disproportionately higher than a white person accounting for local demographics.

rposly

April 22nd, 2020 at 5:28 PM ^

Are you not paying attention?  I get your point (if it IS your point) that it's not technically one's race, but rather co-correlated factors.  But that's just semantics.  This thing is very bad for African American communities.  

https://www.propublica.org/article/early-data-shows-african-americans-have-contracted-and-died-of-coronavirus-at-an-alarming-rate

snarling wolverine

April 22nd, 2020 at 5:07 PM ^

When the real mortality rate is exposed to be around .3% 

If that's the "real" rate, how do you explain South Korea and the Diamond Princess?

South Korea, by all accounts, has actually managed to identify nearly all covid cases in their country.  Their daily growth is minuscule and is almost all from foreign visitors.   But from their 8,515 closed cases, 238 have ended in death - 2.8%.  Even if all of their remaining active cases (2179) end in recovery, they will still have a case fatality rate of 2.2%.

For the rate to actually be 0.3%, there would have to be 80,000 asymptomatic South Koreans who have escaped detection and somehow aren't infecting anyone, which would be pretty strange given that this is an extremely contagious virus.

Then there's the Diamond Princess.  Everyone on board that boat was tested.  712 turned up positive.  657 of their cases are closed and they've resulted in 13 deaths - 2.0%, pretty similar to South Korea.  If all remaining active cases end in recovery, it drops to 1.8%.  

Aside from all this, there is the issue that some recovered patients don't actually make a full recovery, they end up with damage to their lungs, heart or other organs.

clown question

April 22nd, 2020 at 6:13 PM ^

These studies have significant issues. If you choose to follow their data your are being ignorant.

Similarly if you think the rate is 3% and assuming Korea has identified all cases and/or ignore cruise ship demographics you are being equally as ignorant.

There is a likely middle ground. It isn't the plague and it isn't a severe flu.

LewisBullox

April 22nd, 2020 at 6:37 PM ^

It could still be flu like, IFR for the recent bad flu year in 17-18 was over 0.1%. But the infection rate should be higher since no vaccine and no prior partial immunity, so the total number of deaths should be higher.

blue in dc

April 22nd, 2020 at 7:15 PM ^

Go look at the data for NYC and figure out how the rate could be as low as 0.1%

at least 10,000 dead (could be 15,000) and people still dying.

population 8.4 million

If 100% of the population has it and we use the lower number we are already at 0.12%

NYC is still getting many positive tests so they can’t be at herd immunity.   Lets say 50% are infected and stick with the low end death estimate of 10,000 and you are now at 0.24%.   Since we know people are still dying, the 10,000 has to be low.   Assume 12,500 and we are up to 0.3%.

 

LewisBullox

April 22nd, 2020 at 7:42 PM ^

First off, the IFR is going to be population specific. One wouldn't expect it to be the same in NYC as Santa Clara. Second if it's 0.3% that's 2x the 2017-2018 flu, so that would be "flu like." Elsewhere I have said it could be much higher, which would not be flu like.

Another concern is that the amount of infected should be higher for corona than flu unless we shelter in place until an effective vaccine comes about.

Special Agent Utah

April 22nd, 2020 at 8:06 PM ^

What part of “The pandemic is still on going” do you not understand?

You’re using today’s numbers as your basis. Do you not comprehend this thing could continue for months or even years before the final death toll is in. 
 

Honestly, you people. Do you walk out of a UM football game after 1 quarter and say “Well, we’re up 7-0, so it’s clear this game is in the bag”?

MichiganStan

April 22nd, 2020 at 9:35 PM ^

You guys had no issue at all with posting the 3-5% mortality rate studies near the beginning of the outbreak when hardly anybody was being tested except those with serious symptoms but now youre hyper critical of new studies after more testing has been done 

-_-

 

xtramelanin

April 22nd, 2020 at 5:23 PM ^

aren't there some recent studies out, for instance one from stanford, that indicates that the rate of infection may be as high as 80x more than originally thought?  caveats apply of course, but even if that number is 20x or 40x more, that still is a sign that the mortality rate is much lower than feared.

speakeasy

April 22nd, 2020 at 11:01 PM ^

Agree that's not random recruitment in any generally accepted way. That said, given what they were testing for, which almost by definition the individual could not know and therefore bias their response, is the lack of true random sampling a meaningful design flaw?

 

I am not a doctor, statistician, or anything else relevant to answering the question, I just think it's an interesting question.

clown question

April 22nd, 2020 at 6:22 PM ^

In addition to recruiting via facebook, it also:

- relied on a not-so-great antibody test that hasn't been evaluated externally

- assumes antibody = past infection

- weighted different demographics in a way that brought it up a lot.

 

The LA study is better as they used a third party for recruitment but even that isn't as perfect as they make it out to be.

So yes these studies were good news and suggest the mortality rate may be lower. However the numbers they suggest don't fit with other studies or observed deaths in hot spots. It is likely somewhere in between.

 

 

ndscott50

April 22nd, 2020 at 5:55 PM ^

There is no way in hell that South Korea has identify nearly all covid cases in their country. They have tested 10,862 people for every million or 1.08% of the population. We have tested a higher percentage of our population than the South Koreans.  With what we know about the number of asymptomatic cases as well as the larger number of cases with mild symptoms a large number has likely gone unnoticed.  There are many studies that show undetected cases are many multiple times higher than what has been discovered.

blue in dc

April 22nd, 2020 at 6:02 PM ^

Is the Diamond Princess really a representative sample of the population?   If the population of a cruise is skewed more towards the elderly that would not be a representative sample.  

As for South Korea, those asymptomatic people do not spread the virus forever.   They could have been infected and asymptomatic 6 weeks ago.    

slaunius

April 22nd, 2020 at 8:48 PM ^

I looked at this a few days ago: the Diamond Princess population definitely skewed old, but not absurdly so. Extrapolating from This CDC document, it looks like the median age of passengers + crew was around 60. That's definitely gonna skew the numbers upwards, but probably not 20x (assuming all active cases recover), and they also had the benefit of being closely monitored, so they likely received prompter treatment than the general population.

LewisBullox

April 22nd, 2020 at 6:10 PM ^

Regarding the Diamond Princess, you failed to mention the demographic bias of those who go on cruise boats or dosage impacts of people exposed to a virus on a cruise ship vs. typical every day life. The Diamond Princess is interesting though since it's closed set of people.

You say SK has managed to identify all covid cases, "by all accounts," and I'd like to see your sources since it appears they have tested 1% of the population.

If you want to mention these cases, you should mention some of the recent serosurveys with lower estimates.

From what I have gathered, no one knows the IFR better than with about an order of magnitude uncertainty. Roughly between 0.15% to 1.5%.

Swazi

April 22nd, 2020 at 6:40 PM ^

I'm curious as when people that are pro-open up the country keep citing that Stanford study, and promptly raise the number of cases by that number, but oddly enough leave the death numbers stagnant.  

 

Like magically that extra 30-50% of cases all recover.  If they never got tested and die of it, that death doesn't get factored into the total either, right?

ScooterTooter

April 22nd, 2020 at 8:32 PM ^

True. 

I actually think South Korea is a good case study in how we approach this going forward. 

This data is a little older, but as of 4/10:

70+ cases: 1168

0-49 cases: 6042

70+ deaths: 163

0-49 deaths: 4

CFR for 70+ as of 4/10: 13.9%

CFR for 0-49: 0.06

So for those under 50, this is...just the flu (perhaps with more cases of serious after effects)?

But for those over 70, this is an extremely serious disease.