Living and (mostly) dying by the numbers

  
Over the past few days, I fell into the same wishful thinking trap that a lot of other people have (including it seems just about everybody who sets foot in the White House these days): taking a temporary dip in an otherwise robust growth rate to mean that the end of that growth is in sight. Yesterday, the 3-day growth rate in the total deaths number, which is what I use to project deaths going forward, jumped from 7% to 9%, after having gotten to 6% two days before that.

However, that 6% number was down from 12% two days before. And that was up from 9% two days before that, which was in turn down from 18% a week before that. You get the idea: The growth rate in total deaths goes up and down.

But is there a longer-term trend? There definitely is. Just looking at the peaks in the 3-day growth rate, they were:

12% on April 30
18% on April 23
47% on April 15
54% on April 8
105% on April 1
159% on March 24
(peaks before that are all lower)

So the peaks in growth in total cases are declining! (and I just noticed that the peaks are all spaced 7 or 8 days apart. I have no idea why this is happening, but it’s very interesting anyway…) Doesn’t that mean total deaths are slowly approaching a zero growth rate, meaning there will be no new deaths? Unfortunately no, and the reason again has to do with the fact that coronavirus cases and deaths are growing exponentially (meaning one period’s value is equal to some number > 1, multiplied by the previous period’s value).

To illustrate what I mean, on March 24, when the 3-day growth rate was 159%, that translated into a total deaths number of 780. However, if yesterday the 3-day growth rate had been 159%, that would have meant there were 177,697 total deaths, vs the actual number of 74,810 (which made the 3-day growth rate 9%). Conversely, if the growth rate on March 24 had been 9% like it was yesterday, total deaths on that day would have been 328, not 780.

So, in order for the death rate in total deaths to get to zero (which is the same thing as saying that new deaths due to Covid-19 will end), it will have to decline at a greater rate all the time, to keep ahead of the fact that the total number of deaths is increasing exponentially (although at a slower rate). I haven’t calculated what the minimum rate of decline would need to be (I could do it if I remembered my calculus well enough – it’s probably a simple problem), but I do know that if the growth rate keeps going up as well as down, we’ll never get to zero. And we definitely won’t get there if the peaks in the growth rate start going up, not down.

Which brings me to reopening the economy. The reason the growth rate in deaths fell so much is that a large portion of the US economy was locked down around 5-6 weeks ago, and this achieved the intended effect of slowing the virus down. Unfortunately, that trend has turned around, and we’ll inevitably see many more deaths as a result of that. I think it’s likely that, while there will continue to be ups and downs in the growth rate, the trend in the peaks of the growth rate will turn around and start moving up again. And given that we’re at a much higher level of total deaths now, that means much higher absolute numbers of people dying.

One new number I added today was deaths per second for each month. George Conway (Kellyanne’s husband), wrote a very good piece in the Washington Post today, where he pointed out that there was a coronavirus death once every 45 seconds in April (his numbers must have been a little different from the ones I used). I’m going to include that in my monthly projections from now on. I find the impact of that much greater than just hearing what the deaths are per day (although it will be a big deal when we go over 3,000 deaths in a single day – that is, more than on 9/11. I’m now saying that will probably start in three weeks).

The numbers
These numbers are updated every day, based on reported US Covid-19 deaths the day before (taken from the Worldometers.info site, where I’ve been getting my numbers all along). No other variables go into these numbers – they are all projections based on yesterday’s 3-day rate of increase in total Covid-19 deaths, which was 9%.

Week ending
Deaths reported during week/month
Avg. deaths per day during week/month
Pct. Change from previous month
March 7
18
3

March 14
38
5

March 21
244
35

March 28
1,928
275

Month of March
4,058
131

April 4
6,225
889

April 11
12,126
1,732

April 18
18,434
2,633

April 25
15,251
2,179

Month of April
59,812
1,994 (= 1 death every 44 seconds)
1,474%
May 2
13,183
1,883

May 9
17,272
2,018

May 16
13,444
2,467

May 23
22,641
3,234

May 30
27,998
4,000

Month of May
88,478
2,854 (= 1 death every 30 seconds)
148%
June 6
31,651
4,522

June 13
41,490
5,927

June 20
51,306
7,329

June 27
58,001
8,286

Month of June
209,582
6,986 (= 1 death every 12 seconds)
237%
Total March - June
361,930


Red = projected numbers

I. Total deaths
Total US deaths as of yesterday: 74,810
Increase in deaths since previous day: 2,525 (vs. 2,360 yesterday)
Percent increase in deaths since previous day: 3% (vs. 3% yesterday)
Yesterday’s 3-day rate of increase in total deaths: 9% (This number is used to project deaths in the table above. It was 7% yesterday)

II. Total reported cases
I no longer pay any attention to the reported case number. It is a huge underestimate of actual cases, which is probably 25-50 times reported. This is because of the huge shortage of testing capacity. If reported cases were to be anywhere near actual cases, we would need to be doing 20-30 million tests a week. I believe the US has done about 6 million tests since the start of the pandemic.
Total US reported cases: 1,263,243
Increase in reported cases since previous day: 25,160
Percent increase in reported cases since yesterday: 2%
Percent increase in reported cases since 3 days previous: 6%

III. Reported case mortality rate so far in the pandemic in the US:
Total Recoveries in US as of yesterday: 213,109
Total Deaths as of yesterday: 74,810
Deaths so far as percentage of closed cases (=deaths + recoveries): 26% (vs. 26% yesterday) Let’s be clear. This means that, of all the coronavirus cases that have been closed so far in the US, 26% of them have resulted in death. Compare this with the comparable number from South Korea, which is 3%. The main reason this number is so high is that total recoveries are so low. I’ve been assuming since March 26, when the recoveries number was first published, that it would rise, so that the estimated case mortality rate (which was 41% on March 26), would be far lower than it is now. If this number really stays at 27%, this means the actual case mortality rate in the US is very high, comparable with say Italy.


I would love to hear any comments or questions you have on this post. Drop me an email at tom@tomalrich.com

Comments

Post a Comment

Popular posts from this blog

The tragedy in India

For a great account of what’s going on in India, read this WaPo article. And read this Atlantic article for a good commentary on how India’s wealthier class ignored the lessons of the 1984 Bhopal disaster and let the health care system stagnate, leaving the country unable to handle this sudden surge in the virus. Although it's wrong to say the system stagnated. For the wealthy, it became (and remains) world-class. Does this remind you of another health care system that, at least in the early days of the pandemic, was overwhelmed?  Both articles call attention to the Modi government’s failures, but they both try to spread the blame around. Yes, others are responsible, but the fact that – after a terrible response to the first wave of the virus – the government got lulled into thinking they’d beaten it, so they could open up completely (including holding big political rallies and allowing the annual Kumbh Mela to take place – this year in Haridwar, which I’ve visited and can ...
Read more

The Indian variant

The numbers keep improving every week. Both new cases and deaths grew at the same 0.7% rate in the past week, vs. 0.9% for both cases and deaths the previous week. And average daily deaths in the past week were 573. For perspective, the last time deaths were this low was March 31, 2020. But given that daily deaths hadn’t gone above 10 until March 15, the fact that they’d accelerated to 468 on March 31 was a bad sign (at the time, my big concern was that total deaths would end up in the low tens of thousands. In fact, they passed 600,000 last week). This morning, WaPo ran a story on the new Indian variant’s arrival in Britain, where it passed 2,000 cases recently. It seems to be at least 50% more transmissible than what we used to call the UK variant, but which is now the dominant variant in both the US and Britain. And that variant in turn easily displaced what had been the previous dominant variant in both countries, since the beginning of the pandemic (in fact, I remember a lot o...
Read more

Deaths and Recoveries

I’m not bothering with my usual Numbers section today, since the 1-day and 7-day rates of increase in deaths and cases haven’t changed at all, and the ratio of deaths to closed cases remains at 17%. The projected deaths numbers are close to what they were yesterday. I haven’t discussed the ratio of deaths to closed cases for a while, although I’ve been reporting it. To refresh your memory, this is the ratio of total Covid-19 deaths so far to “closed cases”; closed cases is the sum of deaths so far and “recoveries” so far. The fact that the US rate is 17% means that, of all closed cases so far in the pandemic, 17% of them have resulted in death. This is down substantially from the initial value of 41%, in late March, when Worldometers started reporting recoveries. But it is still quite high – it seems to mean that, if you’re determined to have a “case” of Covid-19, the odds are close to 1 in 5 that you will die. I’ve been periodically comparing the US performance in this ar...
Read more