Coronavirus (COVID-19)

Its now 7 days since I started this thread, with the comment that the events unfolding in China were scaring me badly. This remains the case. At the time there was around 2700 confirmed cases in China, plus around 50 across the world. Most provinces other than Hubei had typically 20-60 confirmed cases, each of which I had observed increasing at an 'exponential' rate over the previous days (actually 2nd order polynomial). We're now at over 20,000 confirmed cases in China and almost 200 outside, with the provinces outside Hubei now ranging up to 725 per province. (Data source, BNO News , same link as per my OP which is being updated in realtime )

Provinces vary in size upto 120 million inhabitants, with most a similar size to ourselves / France / Germany / Italy of 40-80m. (Wikipedia) I watched Matt Hancock's statement to parliament this afternoon (which generally came across well) and he made a couple points that I think are worth repeating, firstly that this is "a marathon not a sprint" and will be ongoing for many months, and secondly he has around 50 specialist beds, plus a further 500 that are appropriate (which I guess means isolation beds in other hospitals). So for a population of 65m we have around 550 beds available to treat contagious patients (less those currently occupied by any severe measles cases etc)

As a first order approximation it is probably not unreasonable to assume that the per capita distribution of isolation unit beds won't vary much across the globe. Which means that the confirmed cases in many of China's provinces may be reaching the extent of normal health care provision. Guangdong has one of the largest counts of confirmed cases outside Hubei, and is the province with the border to Hong Kong & Macau, and supplies a lot of their day to day needs, especially food.

The amount of geo-politicking the last 7 days is truly frightening, with more emphasis on implications for trade than health. I fear that is going to bite the world big time over the coming months, fuelled in part by the WHO saying closing borders with China is not necessary. I suspect their view is more nuanced than that, and is really focussed at keeping trade flowing. This appears to have given the green light to countries to formulate their own policy, with HK slow to implement measures and Pakistan restarting scheduled flights to China following the WHO statement. The gulf states also seem reluctant to sever direct flights, and of course the UK is still accepting a few direct flights by Chinese airlines. There is a real danger here of complacency.

The markets are shrugging this off on the whole, yet there are clear downside risks to G7 economies even if G7 countries escape with relatively few mortalities. China is now a major part of global supply chains, and based on current data at risk of a very damaging and long lasting epidemic. The notion that they they can isolate Hubei and deal with a relatively low incidence of cases in the rest of the country has, over the last 7 days probably failed. If normalish daily life were to resume contagion would sky rocket.

Wuhan is a major centre for automotive manufacturing, and off-patent medicines; industrial sectors that have minimal direct relevance to the UK. Most of the Brits out there seemed to be teaching English, and from a quick google the only degree taught in English at Wuhan uni is a medical degree. Whilst Wuhan is not a major business destination for Brits (although I have been once), there are apparently c. 30,000 Brits living in mainland China (plus I should imagine huge numbers more in HK).

I believe the next 7 to 10 days will determine whether the current calmness of the markets is justified, or whether current assumptions on global trade and growth need to be binned.

mrclondon, no coincidence, I don't think his regression analysis stands up to scrutiny. 

I was very sceptical of R²=1.0000 for this kind of application.  It would mean that every data point would fall precisely on that polynomial, which is extremely unlikely for a statistic like a count of viral cases being reported.  So I analysed the data points shown on that graph and I cannot see how they lead to that R² value, or indeed even that polynomial...

I assume 'd' is the number of days counting back from today.  For the function given, namely f(d) = 179.14d² - 3399.86d + 20368.6, we can see immediately that f(0)=20368.6.  That value isn't a data point on the parabola given - and it certainly should be if R² = 1.0000.  On no day were there 20,369 cases reported.  
Similarly f(1) doesn't match any data point, etc, etc.
It's easy to calculate f(10) = 17914 - 33998.6 + 20368.6 = 4,284 .  Yet according to the raw data on the graph, ten days ago there were just 1,287 cases confirmed, and 4,284 was nearer seven days ago.  Something's wrong here.

Now it's been a while since I was fully adept at this stuff, and I'm prepared to accept my analysis may be awry (given advancing years!), either that or I've missed something, but I suspect the finding quoted is somewhat inaccurate.  A parabola (second order polynomial) is a pretty good fit, but certainly not with R²=1.0000.  It's extremely unlikely the number of confirmed cases would ever lead to R²=1.0000 I would suggest.  So much so that if I was told it did, I would hazard a guess that the authorities had cheated by first choosing a parabola, then making up the data to fit it!! 

My own best-fit second order polynomial for the data shown has R²=0.9943.   Yes, it's expanding in a parabolic fashion, but not quite as published on that Tweet. 

 

I think number of isolation unit beds per capita is a really relevant point. I don't have exact data on this but I would assume it is closely correlated with all hospital bed provision which is distributed with very great variance. World Bank database  gives the following per 1000 (all numbers latest data 2012-2013) as follows: China = 4.2; Japan = 13.4; Germany 8.3; France = 6.5; UK = 2.8. 

Simply put, I think the UK is incredibly ill-prepared to deal with anything like this (actually with any non-normal event be that health, economic, whatever). 

Indeed, it would appear that we're at particular risk out of comparable developed world nations.

Source www.kingsfund.org.uk/publications/nhs-hospital-bed-numbers

As we're all acutely aware, there simply is no spare capacity in the NHS.