OK, three short topics today.

First, an update on the ‘delay’ topic from Update 4 in which I offset the UK, US, and German curves in relation to the Italian timeline. I have continued the plot and it now looks like Germany and the US are accelerating faster than Italy did. The UK remains aligned with Italy’s trajectory.

A steeper curve means faster transmission, or more realistic numbers due to better testing. Given the fact that the US is still testing per capita at only 2.5% the rate of Italy, I would guess that it means the disease is spreading more quickly stateside.

And that brings me on to the second topic. Given that testing in most countries is only being provided to those with symptoms, what can we say about the the true number of of cases?

Well, as explained in this excellent video, we can use just a few data points and some logic to form a relatively accurate picture of the true situation within any geography.

Let’s start with Italy :

• The first death was reported on the 21st of February
• At that moment the official number of cases was 21
• Evidence from many validated sources show that symptoms require about 5 days to appear, and death (if it occurs) follows about 2 weeks later. So the time from infection to death is roughly 20 days
• We have seen average mortality rates as low as 0.6% in Singapore and as high as 5% in Iran, but Italy has world class hospitals, so let’s say it’s likely around 1%
• This means that the first victim in Italy was infected around the 1st of February along with 99 others, of whom only 21 were identified.

• Excel’s trendline feature gives us the information we need to calculate the final piece of the puzzle. Ln(2)/0.2343 = 2.9 days for cases to double in Italy.
• So, 3 days after the 1st, there were 200 cases. 400 by the 7th, 800 by the 10th, and today the total number is likely closer to 4 million (against the 35,713 who have tested positive).

The calculated cases in orange are measured on the left hand axis. The reported cases on the right hand scale. If we use the same scale then you get a better idea at the size of the discrepancy :

It’s for this reason that the number of daily deaths in Italy haven’t slowed – despite 60 million people being quarantined for more than a week :

Now, if we make the same calculations for our favorite countries :

KPIs	Italy	UK	USA	Germany
First Death	21/02/2020	05/03/2020	29/02/2020	09/03/2020
Estimated Real hundredth Case	01/02/2020	14/02/2020	09/02/2020	15/02/2020
Estimated Mortality	1%	1%	1%	1%
Estimated Days from Infection to Death	20	20	20	20
Days to Double Cases	2.9	3.0	2.5	2.7
Current Reported Cases	35713	2626	9259	12327
Estimated Real Cases	3932160	204800	3801088	696320

Perhaps the most interesting thing is that the US may already have as many cases as Italy :

…so to my friends in the US, every 2.5 days that you delay drastic isolation measures, you exacerbate the problem exponentially. Take heart though; there are now 39 countries with more than 100 cases, but those that took fast action are starting to see their trajectories flatten out :

OK, the final thought I have today comes from Imperial College London regarding a possible ongoing social distancing model that we should consider adopting until a vaccine is developed. The proposal foresees regular quarantines lasting 2 months interspersed by active periods of 30 days to allow new infections to work their way through the hospital system.

With the relative certainty of a second wave of infections coming once lockdowns are relaxed, this seems an intuitively sensible approach – even if the idea of 18 months of rolling quarantines is hard to bear.