Saturday, April 4, 2020

Surviving the Coronavirus: Part 9

Good News! Bad News! Good News! Edition

The good news is that I was wrong, oh so glad to be wrong, about when we would reach a quarter million infections in the good ol' USA.

In Part 6 of this august series, the Don't Die! Edition, I reported that the number of infections in the US was 53,478 as of Tuesday, 24 March 2020. As a pop quiz in exponential growth, I challenged readers to guess how many cases there would be in a week, as of Tuesday, 1 April 2020. I predicted somewhere between a quarter million and a half million.

The good news is that I was wrong, oh so glad to be wrong. According to the wonderfully named worldOmeter, the number of infections as of Tuesday, 1 April 2020 was "only" 215,003.

The bad news is that I was wrong by only two days. According to the now world-O-famous worldOmeter, as of Thursday, 3 April, there were 277,161 cases in the good ol' USA. As of today, this moment, as I write type this, at 1612 GMT, at 9:12 AM Pacific Daylight Savings Time, there are 291,545 cases.

Would anyone like to guess how long before we get to a million US infections? I predict we'll exceed that dramatic but no-more-meaningful-than-any-other number by the end of this month, with a few days to spare. I sincerely hope that I will once again be proved wrong, and this time by a wide margin.

The good news is that the doubling time is increasing. The doubling time is the number of days required for the number of cases to double. If there are 10 cases on day 3 and 20 cases on day 5, then the doubling time is 2 days. Similarly, if there are 100 cases on day 30 and 200 cases on day 32, then the doubling time is 2 days. Small doubling times are bad (for pandemics, not stock portfolios). Very small doubling times are very bad.

What we want to see are very large doubling times. When the doubling time reaches infinity, the number of infections has reached its peak, and the numbers will begin to drop.

We can, and I will, extract approximate doubling times from the daily infection numbers. Before I do that myself, I will present a quick mathematically based (don't tell anyone) summary on how you might do it yourself.

TRIGGER WARNING! Look away for a few lines if you are offended by math.

We'll call the doubling time Td.

We'll call the number of cases on a given day Nt (for Number today) and the number of cases on the previous day as Ny (for Number yesterday).

We'll call the rate at which infections are increasing Ri. We can approximate Ri by taking the average of two infection levels (one day after the other) then using that average as the divisor for the difference in infections between the two days. In other words:

     Ri =  (Nt - Ny) / [(Nt + Ny) / 2]

Now the easy part. For a good approximation of the doubling time, apply the incredibly simple equation below:

     Td = 0.7 / Ri

That's it. That's how I will use the worldOmeter data to calculate doubling times below.

If, in the extremely unlikely chance that you would like a more thorough discussion, see the Wikipedia article on doubling times.

TRIGGER WARNING CANCELLED!  You can now look back.

Below, I present the trend in doubling times for USA coronavirus infections in the good ol' USA on a daily basis, based on the good ol' worldOmeter data, in a handy dandy Excel chart.


On 13 March, the Feds declared the pandemic to be a national emergency. Governors began mandating social distancing practices, slowly at first, then more aggressively. Ten days later (or so) the doubling time for coronavirus infections began increasing from around 2.5 days up to nearly 6 days as I write type this sentence. The doubling time will continue to increase, but we have a long, long way to go.

There is no way that the coronavirus will not extract a heavy toll on our country, and on other countries. We need to be aggressive about social distancing, particularly the geezer community among us. We need to get masks and gloves and other protective gear to those who are most exposed: the medical personnel and other first responders, the grocers, those who work in Amazon warehouses, those who deliver our goods. We need to develop and distribute an effective treatment quickly, regs be damned, at least reasonably damned. An effective treatment will allow us to relax, but not remove, the social distancing measures until our society reaches herd immunity with a combination of infection survivors and vaccinated citizenry. That herd immunity, approximately 60% of the population, might be more than a year away.

This is not the time for us to squabble, to deny reality, or to blame others for our own national circumstance. There will be plenty of time to return to those well-worn comfy shoes after we have won this war.

This, instead, is the time for each of us to point at ourself and ask "How can I help rather than hinder?"

No comments:

Post a Comment