Monday, March 30, 2020

Good news! The Apocalypse is postponed by a few days!



 Yesterday, I put up a post on my Pandemic Blog, but forgot to put it up on this site. If you'd like to see it, go here.

Yesterday’s numbers (from Worldometers, as of about 7 PM EDT March 29)
Total US confirmed cases: 142,746
Increase in cases since previous day: 18,965 (vs. 19,525 increase yesterday)
Percent increase in cases since yesterday: 15% (vs. 19% yesterday)
Percent increase in cases since 3 days previous: 37% (vs. 44% yesterday)

Total US deaths as of yesterday: 2,489
Increase in deaths since previous day: 260 (vs. 533 yesterday)
Percent increase in deaths since previous day: 12% (vs. 31% yesterday)
Expected* US deaths over course of pandemic:  97,133 (based on 4% case mortality rate)
Expected* US deaths over course of pandemic:  145,699 (based on 6% case mortality rate)
Expected* US deaths over course of pandemic:  194,266 (based on 8% case mortality rate)

Projected as of April 5 (7 days from yesterday):
These numbers assume we wait until April 5 to impose a total lockdown, and that new cases drop to zero on May 2.
Total cases: 268,079
Total expected* deaths over course of pandemic: 216,452 (324,678 with 6% mortality rate)
Projected** number of actual deaths on 4/5 alone: 2,838 (i.e. almost as high as total deaths on 9/11. Of course, this number keeps climbing)
Case mortality rate used to compute total pandemic death estimates: 4%

Projected as of April 11 (14 days from yesterday):
These numbers assume we wait until April 11 to impose a total lockdown, and that total new cases drop to zero on May 9.
Total expected* cases: 503,157
Total expected* deaths over course of pandemic: 406,258 (609,387 with 6% mortality rate)
Projected** number of deaths on 4/11 alone: 10,463 (equal to about three times total deaths on 9/11. And this is the reported deaths in a single day)

* Based on growing total cases by the 3-day growth rate in cases as of yesterday: 37%. Further assumes that as of the day in question, a total lockdown is imposed and it’s 100% successful, meaning new cases are zero 28 days later.

Total Recoveries in US as of yesterday (3/29): 5,211
Total Deaths as of yesterday: 2,489
Deaths so far as percentage of closed cases (=deaths + recoveries): 38% (actually higher than that, because I forgot to get the Recoveries number this morning, and both deaths are about higher as I write this)

Date on which 100,000 deaths will be set in stone: March 30 (yes, today!)
Date on which 500,000 deaths will be set in stone: April 14
Date on which 1 million deaths will be set in stone: April 20
These are based on the assumption that, on that date, a total lockdown will be put in place. It will be 100% effective (0 new cases) 28 days later.


I decided that I couldn’t live with a very unrealistic assumption I’ve been making for a week or so: That when a decision to totally lock down the country (which is what most of the experts are calling for) is made, it will be in effect just two weeks. At the end of that period, all infected people will have been identified, so new cases will fall to zero. The people identified as infected will be quarantined (alone, not with their families); their contacts will all be traced, and they will be quarantined, too.

I’ll bet you can see lots of holes in this – I always could, too, but I thought this would be a good yardstick. Since the numbers I was reporting were so horrible anyway, I decided it was better to let people take these numbers with a grain of salt and use them as comparisons (as I did, in determining the cost of waiting one day to lock down the economy – which was about 10,000 deaths when I wrote that a couple days ago, but is today at about 17,000 deaths).

I have now extended the lockdown in my assumptions to 28 days (and there are proposals for a two-month total lockdown, which will probably be necessary anyway), keeping the same assumption that it will be 100% effective and new cases will be zero at the end of the 28 days.

This is much more realistic, but there are some big holes in this as well. Probably the biggest is, of course, testing. People are still having a very hard time getting tested, and even then it takes up to two weeks to get the answer (by which time you’ve either fully recovered or you’re probably in the hospital, fighting for your life).

You often have to have very clear symptoms to even get a test. And even if you have clear symptoms, the hospital might well send you home if they think it’s unlikely you’ll develop serious problems. Even more importantly, many times the hospital won’t even administer a test even if they’re pretty sure you have Covid-19, because they don’t want to waste valuable tests (this has been happening in New York City, ground zero nowadays, although it’s getting competition from New Orleans, Detroit and soon Chicago). This, of course, means these cases won’t even show up in the tested totals count.

So the lack of testing, far from being a problem of the past, is even more acute now, since it’s almost ensuring that the virus will continue to spread far beyond the people infected as of today. And, if not solved by the time a national lockdown is listed, it will mean there are a huge number of holes that will enable the virus to bounce right back (keep in mind that as of March 1, there were about 100 confirmed cases in the US; yesterday there were 143,000 and two weeks from now there will be 689,000. So in a city of a million people that isn’t locked down today and has 100 cases total, it’s very possible that about two thirds of the city will be infected in one month. Do that for five or ten cities, and you begin to see how serious this is.

The other unrealistic assumption I was making was about the case mortality rate – i.e. for every 100 people infected, how many will end up dying? I’ve been using 3%, and some people argued that was too low. But China is at 4%, and their numbers have been steady – no new cases – for the past couple of weeks (there are some serious questions about whether this is in fact true, but since the questions relate to underreporting of both cases and deaths, there’s no way to say now whether this might end up raising or lowering the 4% rate).

For comparison, Italy is now at 11%. Spain is 8.6%. Iran is 6.6%. France is 6.4%. And South Korea (who got hit faster and harder than we did) is at 1.6%. So if anything I’d say 4% is too low for the US, especially since the one thing that drives it higher is an overburdened health care system. Since NYC ran out of intensive care beds on Friday, and there were even last week some heartbreaking stories of care rationing there (the worst I heard is that the rationing decisions aren’t being made at the hospital. They’re being made by the EMTs, who have to take a look at the patient and decide – not being doctors, of course - if it’s even worthwhile taking them to the hospital. The last thing you want to have happen is to bring them in, take a bed that someone with a better chance of survival could take, and have them die within hours), I’d say it’s close to certain the system will be significantly overburdened (and it’s made much worse by the lack of masks for staff members. They’re literally asking whether they should keep working at all, since there’s such a big risk they’ll get sick, and many will die. We can applaud their dedication all we want, but if we can’t get them masks, the applause rings hollow indeed).

For comparison, I’ve put projections with 6% mortality rate in the above, and in one case I did 8%. While I’m probably going to continue with the 4% rate, I might decide it should be higher, especially if the stories of overcrowding keep growing (although I won’t be able to bear reading them at some point).

There’s one piece of good news today, which is that for the eighth day in a row, the total case rate has declined. To capture this a little better, I switched to basing my projections on the 3-day average of the 3-day growth rate, vs. a 7-day average of the 7-day rate as before. What’s causing this decline? Probably the mandatory shutdowns in a number of states, although the national please-limit-your-gatherings-to-ten-or-fewer-people-not-that-we-want-to-cause-anybody-any-bother-or-anything-like-that rule from the Federal government probably helps a little. Since many of the state shutdowns have been in effect for more than a week, it’s likely that a lot of their effect has already shown up in the case numbers. But I’ll continue to monitor this trend.

But speaking of shutdowns, it’s good news that Trump decided yesterday, after what must have been an interesting meeting with Dr. Fauci, to extend his whatever-it-is another month. That will somewhat help, but keep in mind that, with an exponentially growing threat like the coronavirus, any partial mitigation just has the effect of buying a little more time so the country can get better prepared – it doesn’t change the ultimate course of the pandemic. And since in this case I’ve made my other assumptions more realistic, nothing has really changed that much.
For example, if we institute a total lockdown two weeks from yesterday (i.e. April 12), there will be 468,502 deaths, vs. 97,133 deaths if we’d done it yesterday. And these are both with a 4% mortality rate. If you want to see what 8% looks like (which would still be 3% less than Italy’s rate today), double them.

It seems to me that somebody might want to do something about this urgently. But what do I know?


Comments and questions are welcome. You can reach me at tom@tomalrich.com.















No comments:

Post a Comment