Tuesday, March 31, 2020

Follow the deaths


Yesterday’s numbers (from Worldometers, as of about 7 PM EDT March 30)
Total US confirmed cases: 164,435
Increase in cases since previous day: 21,689 (vs. 18,965 increase yesterday)
Percent increase in cases since yesterday: 15% (vs. 15% yesterday)
Percent increase in cases since 3 days previous: 33% (vs. 37% yesterday)

Total US deaths as of yesterday: 3,175
Increase in deaths since previous day: 686 (vs. 260 yesterday)
Percent increase in deaths since previous day: 28% (vs. 12% yesterday)
Expected* US deaths over course of pandemic:  85,749 (based on 4% case mortality rate)
*This number assumes a) Total cases grow by 33% into the future (= yesterday’s 3-day growth rate in cases); b) New cases drop to zero on April 28, because of the lockdown; and c) case mortality rate = 4%. To consider a 6% mortality rate, multiply each projection by 1.5. For 8%, double it. For comparison, Italy’s case mortality rate is currently 11%.

Projected as of April 7 (7 days from today):
Total expected* cases: 335,830
Total deaths set in stone** over course of pandemic: 201,499
*Expected cases assumes a) Total cases grow by 33% into the future (= yesterday’s 3-day growth rate in cases); b) we wait until April 7 to impose a total lockdown; c) New cases drop to zero on May 5, because of the lockdown
** Deaths set in stone = expected cases X case mortality rate of 4%. To consider a 6% mortality rate, multiply the figure by 1.5. For 8%, double it. For comparison, Italy’s case mortality rate is currently 11%.
Projected*** number of actual deaths on 4/7 alone: 1,700
*** Projected deaths = previous day’s death number, grown by the 3-day percentage growth rate. Note this is calculated completely differently from the deaths set in stone.

Projected as of April 14 (14 days from today):
Total expected* cases: 503,157
Total deaths set in stone** over course of pandemic: 356,852
*Expected cases assumes a) Total cases grow by 33% into the future (= yesterday’s 3-day growth rate in cases); b) we wait until April 14 to impose a total lockdown; c) New cases drop to zero on May 12, because of the lockdown
** Deaths set in stone = expected cases X case mortality rate of 4%. To consider a 6% mortality rate, multiply the figure by 1.5. For 8%, double it. For comparison, Italy’s case mortality rate is currently 11%.
Projected*** number of deaths on 4/14 alone: 15,687 (equal to about five times total deaths on 9/11. And this is the reported deaths in a single day)
*** Projected deaths = previous day’s death number, grown by the 3-day percentage growth rate.

Reported death rate so far in the pandemic in the US:
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%

Date on which 100,000 deaths will be set in stone: April 1
Date on which 500,000 deaths will be set in stone: April 18
Date on which 1 million deaths will be set in stone: April 25
“Set in stone” means that on that date, a total lockdown will be put in place. It will be 100% effective (0 new cases) 28 days later.


I didn’t make any changes in how I’m doing calculations today, but I tried to make the results in the tables above a lot more understandable, and to provide the complete explanation of each number right below it. My goal, of course, isn’t to provide some dry statistics for scholars to review later on, but to try to project what will happen in a way that’s totally transparent. This compares with what you read in the newspapers, which is always what one expert or another says will happen. You don’t know how each expert projected their numbers (unless you’re prepared to read some lengthy paper somewhere, which almost nobody in the general public will do), and you don’t know why one expert’s numbers differ from another’s, or even what precisely the numbers mean.

I’m not developing estimates using anything but simple arithmetic: no modeling of the virus spread, etc. – I can’t do that, and frankly I don’t know how well the epidemiologists can either, since this is such a different situation from before. I’m just trying to explain how I project the numbers and what precisely my terms mean. If you don’t agree with the numbers (and I’m sure many if not most of you don’t), then please let me know your concern, so I can figure out if I’m doing something wrong. Unfortunately, the few people I’ve heard from with concerns have usually just said they’re tired of seeing my gloomy posts – they’re inundated with information and don’t know what to think (which sometimes – but not always - means they’ve made up their minds, so they don’t want to see any more information).

I have developed a standard answer for these people: If you don’t want to read my posts on the pandemic, you have my permission to delete them from your inbox (as if you needed it). If you’re worried you then won’t know about new NERC CIP posts that I put up, you’ll have to check the blog every day to see if one of those has been posted (of course, this applies to Tom Alrich’s Blog, not The Pandemic Blog. I will only post pandemic posts on the latter). But I’m going to write about what I think is important each day. And frankly, I think any human being in the United States should be paying a lot of attention to what’s going on with the pandemic. Even if you’re not infected now, I’m sure it’s already affected your livelihood; those effects alone will continue to increase.

My projected numbers are based on three numbers that I get every day from a site called Worldometers (who gets their numbers from various sources, including the CDC, Johns Hopkins, etc. The numbers are a) total cases in the US, b) total Covid-19 deaths in the US, and c) recoveries in the US. I just project out the first two. However, even though I use the same calculation to project them out (a 3-day growth rate), they have very different significance for what’s most important in the pandemic – the number of deaths (and if this were a pandemic caused by a disease that is unlikely to kill many people, neither I nor most others who write about Covid-19 would be spending the time to do this).

I’m projecting future deaths in two very different ways. The first way I project deaths is to take the total cases and multiply that by an estimated case mortality rate. I project cases by growing the number three days ago by the most recent 3-day growth rate. Of course, that’s accurate when projecting out a day or two, but it loses accuracy rapidly as we go forward. Since there’s no clear trend to the case numbers (recently the rate of increase has been falling, but there’s no assurance that will continue – see below), the best one can do is update the growth rate daily, so it always includes yesterday’s number.

The mortality rate I’m using now is 4%. This compares with 11% for Italy, and 6-8% for France and Spain, as well as 1.4% for South Korea. I discussed the mortality rate in yesterday’s post. I frankly think that 4% might end up being low, once the pandemic is over and all of the counting is done.

However, there is a data point that can provide some evidence for the actual mortality rate. That is the ratio of total deaths to closed cases, which I report above as well. Closed cases equals total deaths so far plus total recoveries. At the end of the pandemic, when all cases are closed one way or the other, this will provide the final mortality rate, and there will be no need to project it anymore. That is, unless you think the US government is fudging one or the other number, or both. China now is being accused of under-reporting both cases and deaths, so I’m not trusting the numbers from them in the future, unless they can clean up the doubts.

So what is the ratio of deaths to closed cases today? It’s 38%. Of course, if this were to end up being the actual mortality rate, it would mean that many tens of millions of Americans would die in the pandemic. Fortunately, this ratio will certainly come down, as more recoveries are listed. But this is why I currently believe the mortality rate will end up being no less than 4%, and very possibly more.

But let’s focus on the case number. How accurate is that? It’s almost certainly way low. The Imperial College of London (whose study was used to justify the administration’s social distancing rules a few weeks ago) estimated the actual cases are between five and ten times that number. Why is this the case? Of course, it’s because of the woefully inadequate level of testing, which continues to be woefully inadequate.

Moreover, testing is likely to continue to be woefully inadequate, since the president said two days ago, on a call with the governors (who were pointing out that they need more test kits) that he hadn’t heard there weren’t enough kits. This is probably a true statement, since it’s likely that the limited set of news sources he listens to isn’t talking about this problem. However, those of us who have to live in the real world hear every day about a) people who clearly are sick, who can’t even be tested, or b) people who finally get tested, but have to wait up to two weeks to get their results, at which point the results are moot – either the person has gotten over the disease or they’ve died. And even though new kits are becoming rapidly available, the number of tests a hospital can do with each kit depends on their having adequate supplies of the cotton swabs required to get the sample, as well as the chemical reagents needed to conduct the tests – and these are in increasingly short supply.

The bottom line is that the number of actual Covid-19 cases in the US is likely to be many times the reported cases for a long time, even if it ever catches up. And since there’s no good way to determine where in the range of 5-10 times the multiplier is now, and more importantly since it’s very likely that the multiplier itself is going up, as the people with unreported cases infect others, I’m just going to stick with projecting out the reported number. And I’ll continue to estimate total deaths over the course of the pandemic based on the reported numbers, as described at the beginning of this post.

But this means that the reported case number is a good representation of one thing only: the availability of tests. If the number of tests available were to double tomorrow, the reported cases would almost certainly double – on top of the normal rate of increase due to an increasing number of total cases. So the projections I’m making are almost certainly going to continue to increase for a long time.  The 3-day growth rate that I’m using to project out has varied between 115% and 33%. 33% is the most recent rate, so I’m using that in my projections, but as I said there’s no trend to that rate – it’s already gone over 100% twice during March.

But you have to remember that the number of cases is more than doubling every three days, even at the 33% rate. This is why the 100 cases reported at the beginning of March had grown to 164,000 yesterday, which in case you have the courage to look at it, means cases have increased 1,639 percent in 30 days. If we project that out over the next month, that means total cases will be 2.7 million at the end of April. And not surprisingly, that’s very close to the 2.8 million I’m projecting, since both numbers are based on the same three-day growth rate. This is why I’m projecting 1.7 million deaths over the whole pandemic, if the current 33% 3-day growth rate continues through April (at which point a total lockdown is put in place). And since the number of actual cases is some multiple of reported cases, if anything both of these April 30 numbers will be too low, not too high.

Of course, all of this comes from projecting out total cases. What about projecting total deaths, which I’m projecting independently? Those are calculated by growing the reported increase in deaths 3 days ago by the most recent 3-day growth rate, which as of today is 87% - and by the way, this rate has been going down since March 17, when it was 117% (where’s the luck of the Irish, anyway? Now that we need it most!).

Projecting the 87% rate through April 30 yields 2,183,000 total reported deaths as of that day – this isn’t an estimate of total deaths over the pandemic, as the other number is. So this is in effect hugely bigger than the 1.7 million figure estimated above, since there will be a lot more deaths coming, certainly in the millions.

Yet I’m sure many of you won’t believe the 1.7 million figure (which I also find unbelievable). Great, where did I go wrong? I think many of you will point out that there’s a lot more social distancing going on now, and that has to affect it. It certainly will, but remember that’s been going on for about 3 weeks already, and the case growth rate has already fallen significantly; it almost certainly has somewhat  further to fall. I certainly hope that continues – but then you have the opposite consideration, which is that the reported case rate is a fraction of the actual rate, and as testing ramps up, the number of new reported cases will ramp up as well. Which will win out? I don’t know, of course.

But there is one thing I do know: At the end of the pandemic, the two numbers will end up at the same point. Whether it will be closer to the number estimated based on cases or the number based on reported deaths so far isn’t clear, of course. However, the fact that the actual cases is a multiple of the reported ones only affects the reported case number. The actual deaths depend on the actual cases – people don’t die only if their case has been reported earlier. So at the moment, I have to say the final toll for the pandemic will be well above 1.7 million (perhaps the 2.2 million high estimate of the Imperial College).

Unless, of course, a total lockdown is imposed very soon – certainly before April 30 (of course, my 1.7 million projection for the total pandemic, based on cases, assumes there will be a total lockdown on that day). And it will be, too – although some dead wood will almost certainly need to be cleared out before this can happen. Why am I optimistic about this? Two words: exponential growth. On April 15, there will probably be more than 100,000 reported deaths. If that doesn’t get somebody’s attention, God help us. Probably nothing will.


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














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.















Saturday, March 28, 2020

Death, on a Pale Horse



When the Lamb opened the fourth seal, I heard the voice of the fourth living creature say, "Come and see!" I looked and there before me was a pale horse. Its rider was named Death, and Hades was following close behind him. They were given power over a fourth of the earth to kill by sword, famine, and plague, and by the wild beasts of the earth. (Revelations 6:7-8)


Yesterday’s numbers (as of about 7 PM EDT March 27)
Total US confirmed cases (from Worldometers): 104,256
Increase in cases since previous day: 18,507 (vs. 17,155 increase yesterday)
Percent increase in cases since yesterday: 22% (vs. 25% yesterday)
Percent increase in cases since 7 days previous: 438% (vs. 522% yesterday)

Total US deaths as of yesterday: 1,704
Increase in deaths since previous day: 400
Percent increase in deaths since previous day: 31% (vs. 26% yesterday)
Minimum expected US deaths over course of pandemic: 60,003, based on yesterday’s cases (assumes complete lockdown of US starting today, with new cases dropping to zero 14 days from now)

Projected as of April 3 (7 days from yesterday):
Total expected* cases: 456,641
Total expected* deaths over course of pandemic (based on 4/17 case figure): 262,812
Projected** number of actual deaths on 4/3 alone: 2,150

Projected as of April 10 (14 days from yesterday):
Total expected* cases: 2,000,089
Total expected* deaths over course of pandemic (based on 4/24 case figure): 1,151,115
Projected** number of deaths on 4/10 alone: 11,530
Case mortality rate used to compute total pandemic death estimates: 3%

* Calculated by growing yesterday’s case number by growth rate in cases for last 7 days: 438% (Note: this is the lowest this rate has been since the beginning of March. It’s probably an outlier, but the number I should use (785%, or the average of the 7-day growth rates over the last 7 days) is too horrible to consider at the moment).
** Calculated by growing yesterday’s death number by average of 7-day growth rates in deaths for last 7 days: 536%.

Total Recoveries in US as of yesterday (3/27): 2,525
Total Deaths as of yesterday: 1,704
Deaths as percentage of closed cases (=deaths + recoveries): 40%

I have revamped the numbers I show today (and plan to continue with them, until I revamp them again), since the ones I used before were hard to understand (even for me). There are still two hard numbers driving the others: total confirmed cases (i.e. tested positive) and total deaths. What has changed is the projections I’m making off of those two numbers:

  1. Going forward, I plan to grow the total cases number by the average of the 7-day growth rates over the past 7 days. Of course, total expected deaths over the pandemic is directly tied to total cases, since expected deaths (as of day X) equal total cases (as of day X) times the case mortality rate (i.e. the number of cases that end in death – this number can only be known after the pandemic is over).
  2. But when I put together the spreadsheet this morning, the projections based on the 7-day average of 7-day growth rate in tests (785%, based on yesterday’s figures) were too horrible to publish here, and I allowed myself to be swayed by the fact that yesterday’s 7-day growth rate in cases was the lowest since the beginning of March (when my data begin). That rate was 438%, while previous 7-day rates were between 522% and 839%. So rather than use the 7-day average of the 7-day growth rates, I used yesterday’s number, on the theory that maybe the limited social distancing that’s been put in place in a number of states was beginning to have an effect. Hopefully, the numbers will bear me out shortly, and this trend will continue.
  3. I’ve been using 3% as the case mortality rate so far, but that may be too low for three reasons. First, the rate worked out to be around 3.9% in China, which is more or less at the end of the pandemic. Second, if you look at the data so far (at the bottom of my numbers above), the mortality rate for closed cases so far in the US is 40%. This is presumably artificially inflated by the fact that closing a case when the person is still alive will normally lag a long time, since no doctor will want to close a case and then have it turn out the patient was still sick. It will come down, but will it come down to 3%? 4%? 10%? Too soon to say. Third and most importantly, the mortality rate is very dependent on the degree of hospital overcrowding – shortage of beds, shortage of ventilators, shortage of masks for the staff members, etc. In the next two weeks, as the full onslaught hits NYC, we’ll see how serious this overcrowding and lack of resources turns out to be.
  4. I was previously publishing a number for estimated deaths across the pandemic, based simply on the day’s case number times the 3% mortality rate. But I realized this morning that this is a cop-out. It assumes that, by some truly miraculous occurrence, there will be no new cases after today. Of course, given the number of people out there who have the virus but haven’t been tested yet (either because they haven’t showed any symptoms, or because they’ve tried to get tested and haven’t been able to), this is completely unrealistic.
  5. My new realistic minimum number assumes that as of the day in question, a total lockdown on the country is imposed – in fact, every member of a family is locked down separately from every other member of that family (as eventually happened in Wuhan, since whole families ended up infected before they did this), and the lockdown works perfectly. Given the 14-day incubation period, this means that by the end of 14 days, the authorities will have identified every case anywhere (because it will have shown symptoms or else it will have gone away). They can then lock down each of these people; they don’t have to trace contacts, since after the person has been isolated for 14 days, they won’t be able to spread the virus because they don’t have it – at least that’s how it works in theory. In other words, everything goes perfectly. This is of course not realistic, but hopefully a close-to-total lockdown would lead to a close-to-total identification of all current cases. We’re currently far away from being able to do either of those, of course.
  6. So I get my new estimate for total cases during the pandemic by finding the total projected cases 14 days later; total deaths during the pandemic are 3% of that case number. This means that, by my previous method, I would have said total expected cases given yesterday’s case number were 3,128. By the new method, total expected deaths as of yesterday are equal to the projected case number on April 10 (2,000,089) times 3%, or 60,003.
  7. For the first time, I’ve started projecting actual daily deaths (i.e. the number you hear on the news – it was 1,704 yesterday), not just deaths over the course of the epidemic. I previously didn’t do this because the only good way to estimate it would be based on medical knowledge of how the virus will progress in each person, etc. But nobody has this knowledge yet anyway, so I decided to grow the daily reported number by – again – the 7-day average of the 7-day growth rates. This yields an estimate for each day going into the future.

Now let’s look at the results, and – even though they are probably all underestimates – they’re truly horrific:

  1. As I already said, just using the actual case figure from yesterday and the projection based on a lowballed growth rate, the minimum number of US deaths over the course of the pandemic are 60,000. And that’s based on the assumption that the entire country was locked down as of yesterday, which obviously didn’t happen. So 60,000 is the bare minimum number of deaths we can expect going forward. This is 2,000 more than the total number of Americans that died of all causes in the Vietnam War. Note that actual deaths reported yesterday were 1,704.
  2. Let’s say we wait until April 3 (next Friday) and do our total lockdown that day. Then, total deaths over the course of the pandemic are 262,812. However, the projected deaths on that day are 2,150.
  3. Now we’ll go to April 10 (two weeks from yesterday), and start the lockdown that day. Then, total pandemic deaths are – are you sitting down? – 1,151,115. Just as horrific, on that day alone, reported deaths will be 11,530.

And believe me, this trend continues. What will stop it? A total lockdown is the only thing that will help. Italy tried a partial lockdown at first (just applying it to the north, where the hotspots were), and found that did very little good. And the reason for that is simple, as I’ve said earlier: Given that growth of the virus is exponential, it will keep growing exponentially outside of the places you’ve locked down.  And with exponential growth, you end up with the same numbers you would have had if you hadn’t locked down at all, just a little later. I read today that Trump wants to lock down NY state, NJ and Connecticut. This isn’t likely to have much ultimate impact, since the other hotspots will continue to grow, and new ones will keep appearing (remember, the virus had plenty of time to spread through the country undetected in February, due to lack of tests. It’s everywhere now). A couple weeks after the total lockdown (which of course isn’t quite total. It never can be), the rate of new infections is falling, but Italy can’t loosen up at all until it’s fallen to zero. If they do, infections will just spring back again – remember, exponential.

So you now know the cost of waiting to impose the total lockdown. If we did it yesterday, there would “only” be 60,000 deaths over the epidemic in the US. If we wait until next Friday, total deaths will be 263,000. The following Friday, we’re at well over a million, plus our daily rate of deaths is now close to 12,000 – which is equal to close to four times total deaths on 9/11. In fact, starting next Friday (April 3), the daily death toll will be above the 9/11 total toll – and that will continue to be true for probably months. Do you remember the trauma of 9/11, and how it reverberated for months and years? How much trauma do you think there will be when there’s a 9/11 every day, then two 9/11’s, then four 9/11’s, then…This is why I said earlier that April will be hell on Earth. And May will be worse. For sure.

I know you’ll find these numbers unbelievable. So do I. But I’ve played with the numbers a lot today (so much for getting any work done on my day job!), and if anything I think they’re too low. But if anyone can point out anything I’m doing wrong – or if you want to discuss any of the assumptions I’ve made – I’d love to do it.

So get ready. For the next 2-3 months at least, Death will be stalking the country. Riding on a pale horse. Reaping a terrible toll. God help us.


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















Friday, March 27, 2020

Relentless


If you’re looking for Tom Alrich’s blog on cybersecurity/NERC CIP, go here.

Yesterday’s numbers (as of about 6 PM EDT March 26)
Total US confirmed cases (from Worldometers): 85,749
Increase in cases since previous day: 17,155 (vs. 13,653 increase yesterday)
Percent increase in cases since yesterday: 25%
Percent increase in cases since 7 days previous: 522%
Minimum expected cases 7 days from yesterday (April 3): 639,146 (up about 128,000 from yesterday)
Minimum expected cases 14 days from yesterday (April 10): 4,763,991 (up about 1,000,000 from yesterday)

Total US deaths as of yesterday: 1,304
Increase in deaths since previous day: 268
Percent increase in deaths since yesterday: 26%
Minimum expected US deaths over course of pandemic: 2,572 (assumes no new cases starting today, and case mortality rate of 3%. Based on yesterday’s figures)
Minimum expected US deaths 7 days from yesterday (April 2): 32,306 (same number as above, but calculated 7 days ahead. Note this means the number of total US pandemic deaths anticipated 7 days from now, using the same assumptions as above. Note it isn’t the number of deaths I expect to see reported 7 days from now – I can’t estimate that through purely mathematical means) This is up about 3,000 since yesterday.
Minimum expected US deaths 14 days from yesterday (April 9): 240,399 (ditto, but 14 days ahead) This is up 98,000 from yesterday.
Just to be clear: The point of both of the above numbers is that, if we wait either 7 or 14 days to totally lock down the US, the cost in additional US deaths over the course of the pandemic will be about 32,306 or 240,399 respectively, vs. if we had acted today.

Total Recoveries in US: 1,868
Total Deaths: 1,304
Deaths as percentage of closed cases (=deaths + recoveries): 41%
Case mortality rate used to compute death estimates: 3%

These numbers pretty much tell the whole story. For those who don’t know, yesterday the US surpassed both Italy and China to become the country with the largest number of coronavirus cases. And if you look at the 25% growth rate in new cases (identical to the last two days), you’ll agree there’s no way any other country could catch up to us, absent a total lockdown in the US by say 5PM today – and even then it will be hard, because the cases will keep growing at their normal rate for two weeks.

Only after two weeks of lockdown will we even have a reasonable idea of total cases, and then we can lock down all of them – of course, that assumes there will be a) enough testing capacity and b) some place to lock these people down individually, since one of the lessons learned from Wuhan (one of many completely ignored by the White House) is that you can’t just rely on locking down families – you need to separate the family members as much as possible. If you don’t, then you’ll very possibly have an entire infected family.

Since the number of cases in the US two weeks from now will be in the millions, we’ll need say five to ten million individual rooms available. I’m sure the Trump administration will figure out how to do that in two weeks, just like they’ve figured out how to deal with the respirator problem: Put on hold a commitment by GM and Ventec (a ventilator company) to build 80,000, because it would cost about $1 billion. While it’s very possible we’ll need 1 million ventilators at the peak of the crisis (which unfortunately may come even before GM/Ventec can produce the 80,000, even if they’d been given the goahead yesterday).

Of course, the administration’s big problem – which was mentioned in comments by somebody at the White House yesterday – is that these ventilators might just be needed for a week. Then the crisis would pass, and there would be all these expensive ventilators sitting in a warehouse. What a tragedy. I agree completely with the WH in this regard: It’s much better to have a million or two more people die because those 80,000 ventilators weren’t available when they needed them (which could easily happen), than it is to have “wasted” $1 billion building up our strategic stockpile for the next time a virus like this hits. You just have to have your priorities in place, and this administration is making theirs abundantly clear.

Meanwhile, the new $2T recovery bill includes $17 billion for Boeing with no US equity position (even though that’s being done with the airlines), even though Boeing’s CEO stated last week that they didn’t need the money. It’s all about priorities…

I added two new numbers. One is total recoveries in the US (which had been published for a few days by Worldometers, then dropped). If you add that to total actual deaths, you get total closed cases so far. Dividing today’s total deaths figure yields deaths as a percentage of closed cases. Currently the number is 41%, but it was even higher when Worldometers first published it.

Compare this with the estimate of case mortality rate that I’m using for all of my deaths estimates, which is 3%. People have been telling me that number is way too high, although there will be no way to calculate it exactly until the pandemic is finished in the US. They often point to Wuhan (where it started, and is now close to finished, although they still haven’t ended the lockdown), where it seems the case mortality rate was 1.4%.

In reply, I say, “That’s a great example, because China beat the coronavirus in Wuhan by completely locking down Hubei province – and people are only now being allowed to leave there. The US needs to do the same thing but for the whole country, since the virus is literally popping up in just about every city now. If we were to do that, maybe we would end up with a 1.4% rate, too. But I certainly can’t assume that now.”

In fact, I’m wondering about the 3 percent number. China as a whole is now pretty stable, with about 82,000 cases and 3,200 deaths, which leads to a 3.9% rate. And they did a great job, after the initial miscues in Wuhan (and I’m not minimizing those, since the right steps at the right time there would have ended the whole epidemic – and even then, they could have grounded external flights so that their citizens didn’t infect the rest of the world. The novel coronavirus would just be something you’d find in some obscure academic journal entry. China has a lot to account for to the rest of the world. At the minimum, they need to explain how they won’t let this happen again). Given the problems we’re likely to have in hospitals across the US (and all ICU beds in NYC will be filled as of today), it might be hard to keep the rate at even 4%.

Finally, take a look at the last three numbers I highlighted in red and the sentence in red at the bottom. Were we to make the decision today to lock down the country (and were it to be perfectly successful, so that in 14 days there would be virtually no new cases), there will be 4,334 total deaths (using the average 7-day growth rate in total cases from the last 7 days. That is going to be my growth yardstick from now on, or at least until I think I've found a better one). If we wait a week to do this, there will be ­32,306 deaths. If we wait two weeks (to April 10), there will be 240,799 deaths. And beyond that, we're into the millions.

A couple people have told me we just need to give Trump more time on this. If I were sure that he’s a) learned his lessons from his complete inaction starting in late January, when many other world leaders did start preparing for the onslaught they knew was coming – while he has golfed every weekend until two weeks ago – and b) brought in someone who both knows exactly what needs to be done and won’t undercut him or her with his own contrary messages and actions (which he’s continued to do through yesterday and probably today, since I don’t do Twitter), I would give him the time. And of course, that person (Deborah Birx?) needs to implement a total lockdown starting tomorrow morning at the latest. If all this happened, I agree that changing horses at this point wouldn’t help.

But let’s say Trump wastes two weeks on his plan to have the whole country up and running by Easter (April 12) – omitting the fine detail that he’d have to force most of the workers to go to work at gunpoint, since they would literally be risking their lives to do so at that point – and on April 10 has a complete Road to Damascus conversion. He immediately implements a total national lockdown with grounding of all nonessential travel (and the lockdown, again, is 100% successful). That two weeks will have cost us 236,465 lives in the long run. I guess that’s not such a big price to pay. After all, Trump needed some time!

As I said, relentless.


11:30AM: You may have noticed that I implicitly wondered whether Deborah Bix might be the person to lead us from Egypt's land (where the plague runs rampant) to the Land of Canaan. However, I dropped that idea when I heard just now that she stated she knew of no shortage of ventilators. Unfortunately, nobody who works for Donald Trump can be trusted at this point. He needs to completely step aside, and someone from outside needs to be given complete control of the response. Preferably today, but tomorrow at the latest. If you don't understand why I said that, you haven't read this post.


Thursday, March 26, 2020

Meanwhile, back at the ranch…



If you’re looking for my pandemic posts, I’ve created a new blog. If you’re looking for my cyber/NERC CIP posts, you’re in the right place.

As I’ve been putting out long posts on the pandemic, some people have wondered what happened to my cyber/NERC CIP posts. I certainly haven’t forgotten about them, but I decided for the moment – and that moment isn’t over at all, would it weren’t the case – that I want to devote more time to my pandemic blog than to this blog. I won’t kid you - that’s not going to change soon, given the way things are rapidly deteriorating on the Covid-19 front.

BTW, even though I thought NERC would push back the CIP-013 compliance date, they haven’t done it yet (note on 3/27: Kevin Perry pointed out to me that I should mention that NERC can't unilaterally push the CIP 13 date back; FERC officially needs to do that. I had discussed that when I originally called for the pushback, but I'd left it out here. FERC only has three Commissioners now, but that's still a quorum. Presumably, they have a way they can do the required meeting remotely). I still think they will, but it wouldn’t be a good idea not to keep pursuing that – the main thing that has to be done is develop your supply chain cyber security risk management plan, required by R1.1. Of course, if you’re needed to help your employer keep the lights on during the crisis (and can you imagine what a mess it would be if there were a long outage when so many people are working at home? And if hospitals were blacked out – well, I’d rather not think about that. They’re going to have a lot of problems as it is, starting this weekend in New York City), then by all means continue to do that. It would be a good idea to declare CIP Exceptional Circumstances, even though CIP-013 isn’t directly covered by CEC.

My offer at the end of this newsletter of a free webinar for your organization, describing my methodology for CIP-013 compliance, is still valid! But don’t ask me to come onsite to do it. I’m holed up in my apartment in Evanston, Illinois, and I’m not planning any trips outside for a while. I’m part of what’s known as a high-risk demographic, although I’d probably do this even if I weren’t. This is nothing to dismiss – some people are doing that, and a good percentage of them will be very unhappy they took that course.

But I do want to bring three things to your attention:

  1. I’ll be delivering my first of two webinars for the NERC Supply Chain Working Group next Monday at 1 PM Eastern Time. The topic will be Supply Chain Cyber Security Risk Management Lifecycle. You don’t have to sign up for the webinar (and they’re being given each week until early May – same time each week), but the instructions are here. I hope you’ll attend. The webinars are all being pre-recorded, but the Q&A will be live. Both the webinar and Q&A recordings will be posted fairly soon.
  2. There’s a video posted of all the presentations at the Protect our Power conference in January (which seems like 2015, the world has changed so much since then), including mine (on CIP-013 and supply chain security, natch). You can find them all here. Mine is decent, but I highly recommend Monta Elkins’ (which BTW is on supply chain security as well, although focusing on the hardware side).
  3. Last but not least, I want to call your attention to this E&E News article on supply chain attack on a supplier to Ameren, which resulted in data on two plants being breached; however, it doesn’t appear any critical data were accessed. But it just goes to show…CIP-013 is there for a reason! BTW, this article is by Christian Vasquez, who has replaced Blake Sobczak as the main electric power cyber writer (although Pete Behr is still pitching in). Blake has now been elevated to a more general editing role, although Blake’s still publishing his weekly newsletter on energy cyber (which is where I saw the link to this article). Christian is a good writer and quite thorough – he clearly wants to get to the bottom of the story. It’s nice to have him on the beat!


Any opinions expressed in this blog post are strictly mine and are not necessarily shared by any of the clients of Tom Alrich LLC.

If you would like to comment on what you have read here, I would love to hear from you. Please email me at tom@tomalrich.com. Please keep in mind that if you’re a NERC entity, Tom Alrich LLC can help you with NERC CIP issues or challenges like what is discussed in this post – especially on compliance with CIP-013. My offer of a free webinar on CIP-013, specifically for your organization, remains open to NERC entities and vendors of hardware or software components for BES Cyber Systems. To discuss this, you can email me at the same address.


We need to act!

Comment from Tom at 5:20 PM. The US now has 82,547 Covid-19 cases. This surpasses both Italy and China, and makes us number one in the world. And I don't see anyone coming close to us in the future, since this will continue to grow exponentially.

Yesterday’s numbers (as of about 6 PM EDT)
Total US confirmed cases (from Worldometer): 68,594
Increase in cases since previous day: 13,653 (vs. 11,160 increase yesterday)
Percent increase in cases since yesterday: 25%
Percent increase in cases since 7 days previous: 641%
Minimum expected cases 7 days from yesterday (March 29): 460,944 (up about 44,000 from yesterday)
Minimum expected cases 14 days from yesterday (April 5): 3,097,491 (down about 140,000 from yesterday)

Total US deaths as of yesterday: 1,036
Increase in deaths since previous day: 252
Percent increase in deaths since yesterday: 32%
Minimum expected US deaths over course of pandemic: 2,058 (assumes no new cases starting today, and case mortality rate of 3%. Based on yesterday’s figures)
Minimum expected US deaths 7 days from yesterday (April 2): 13,828 (same number as above, but calculated 7 days ahead. Note this means the number of total US pandemic deaths anticipated 7 days from now, using the same assumptions as above. Note it isn’t the number of deaths I expect to see reported 7 days from now – I can’t estimate that through purely mathematical means)
Minimum expected US deaths 14 days from yesterday (April 9): 92,925 (ditto, but 14 days ahead. This is up 25,000 from yesterday)


Note I’ve inserted the word “minimum” in front of “expected deaths” wherever it appears above. This has always been true, but I’m trying to emphasize that this number is based on a wildly optimistic assumption: that on the day in question (either the day before the post was written, or 7 or 14 days after that), a minimum 14-day total lockdown is imposed on the country and it’s 100% successful – meaning the virus is stopped cold in its tracks.

I’ve also changed the way I project total pandemic cases (and thus deaths) 14 days out. I have been computing the growth rate over the past 7 days, and assuming that will go forward for at least two weeks. This might make sense if there were some sort of trend in those growth rates, but I’ve found there really isn’t. Since I started measuring them on March 9, they’ve moved up and down, between a low of 545 and a high of 839. I decided to take an average of all the 7-day rates since March 9, and project that forward. The numbers above reflect this.

To be honest, even if I used a somewhat lower rate to project growth forward, it wouldn’t make much difference, because when you’re talking about exponential growth, any decline in the rate of increase just pushes back the date at which a certain level will be reached – it doesn’t mean that level won’t be reached.

Let’s say the 839% 7-day growth rate (the highest of the observed rates) continues through the end of April. This leads to our having more total cases in the US than population (obviously an absurd result) on April 23. But if we set the rate at 545% (the lowest of the observed rates), that date advances to April 28. So we’ve given ourselves five days before we reach Armageddon. Big deal, huh?

But let’s say the rate drops to 300% and stays that way. Then the cases surpass the population on May 18. Of course, total cases doesn’t change, since the exponential growth in cases continues. And total expected deaths over the course of the pandemic (in the US) also doesn’t change. But given that we now have a few additional weeks to play with (and more if we could get the rate of increase below 300% a week), we can get the hospitals more prepared for the surge in cases than they are now. We’ll also “flatten the curve”, so that hospitals are less likely to get overwhelmed.

And speaking of overwhelming, New York City’s 1800 Intensive Care Units are going to be full tomorrow, and the real onslaught of cases will start next week. So it looks inevitable that there will be a lot of deaths due to this situation (and of course, the fact that they need 30,000 more ventilators than they have now). In fact, I’m sure there are unnecessary deaths being caused now, to judge from stories of hospitals in the Bronx and Brooklyn in the NY Times today.

How do we flatten the curve? By massive social distancing, much more than we have now. We need a nationwide lockdown, as well as a steep curtailment of transportation – basically limited to vital shipments or trips only. We would have had other options (isolating all cases and close contacts of cases) if we’d had widespread testing available in early February, but we didn’t. This is the only possible course now.

Italy has given the world a good example of the cost of not doing this right away. They initially tried some halfway measures, but the virus just blew them away – again, exponential growth. So they moved to a total lockdown of the whole country a couple weeks ago (even though the outbreak was mainly in the north – they wanted to keep it that way), and now the growth rate of new cases is declining – although it’s certainly not zero yet.

That’s why we need an immediate change of direction. If we put in place a total lockdown today and it’s 100% successful, we’ll have 2,058 total deaths. If we wait a week to do that, we’ll have 13,828. If we wait another week, we’ll have 92,925. If we wait another week (to April 15), we’ll have 624,444 expected deaths over the pandemic. If we wait until April 22, the number of deaths is over 4 million. And, of course, it goes up from there.

And my deaths figure is based on a 3 percent case mortality rate, which assumes some amount of hospital overcrowding. Maybe it can be somewhat less than that – it ended up at 1.4 percent in Wuhan, so if you want to use that number, you can; then just divide the above death figures in two. But remember, Wuhan put in place a total lockdown – much more total than we’ll ever get in the US; it’s just being lifted (partially) now. Bottom line, if we do a pretty good lockdown in the next week – but not perfect, which is very likely – we won’t reduce the growth in total cases to close to zero, but maybe we’ll keep it under 100,000. That’s if we move in not much more than a week.

Now to Mr. Trump, since I called for him and Mike Pence to be pushed aside from the virus response effort yesterday. There are two data points this morning that I find very relevant:

First, the Administration announced that they’re working on a plan to loosen up the already very inadequate social distancing in place now by Easter (April 12). If you think that’s good news, you haven’t read the above.

Second, Fed Chairman Powell said that getting the economy started again requires defeating the virus first. And it just makes sense: nobody is going to want to come back to an office or factory if they think there’s a decent chance they’ll catch a disease that will kill them; and no corporate counsel will let an office reopen unless there’s no real chance that multiple people will develop Covid-19 and then sue the company for opening too quickly.

People worry about killing the economy, but the economy has mostly been killed already. We need to keep it on ice (a medically-induced coma, as someone called it) so we can identify everyone who is infected, and then quarantine them and all of their contacts. Remember, we wouldn’t have to do this if we’d had adequate tests available early enough. And we won’t be able to reopen the economy until there are adequate tests available. Trump should be focusing all his efforts on getting tests made and finding more ventilators, not on planning for an Easter reopening which is never going to happen.

But as I said yesterday, I don’t see him doing that until he’s forced to. So the death total (for the whole pandemic, not just deaths so far) is going to keep climbing exponentially, until someone with sense forces him to completely step aside from the virus response and bring in someone who can at least change course from the disastrous path we’re on now. I heard the suggestion of Bill Gates for that position, and I think that would be a good choice, but there are others as well.