Prison Ship

The cruise ship Diamond Princess, quarantined in Yokohama for the next couple of weeks, has about 3,700 people on board. The vessel looks impressive but those confined to inside cabins must feel as if they are not just doing time but have been sentenced to death. Everyone knows how unhealthy cruise ships can be. Forty new cases of novel Coronavirus were declared today, tripling the number of infected.

The World Dream, a cruise ship in Hong Kong with a similar complement of human cargo, has so far had no positive test results. In New Jersey, a small number of passengers on the Anthem of the Seas has been screened and four taken to hospital for further tests.

Whether you consider Diamond Princess a modern prison hulk or a giant petri dish, it is perfect for modelling.

I have logged patient zero’s appearance on 20 January, though he had left the vessel just before that date in apparent good health. The incubation period for this coronavirus is 5 to 14 days, with 6 or 7 days seeming to be commonly accepted now. I have plumped for 7 days. To arrive at a Total Infected figure approximating 61 today, and no fatalities, I have had to set an R0 of 3.9 and a low mortality rate of 1%. In three days from now, this model predicts 75 passengers/crew infected and three deaths.

The good news is – the virus will burn out in just over a month. The bad – 404 people may have died by then.

There haven’t been any fatalities yet from the Princess but if the reality matches this projection it may validate the tool and make my country projections seem more believable. (I really can’t imagine 10 million UK citizens succumbing to the virus before mid-summer.)

A new country has been added to the list on the Johns Hopkins map – called “Others”, it currently comprises just the Diamond Princess. It is the hardest hit “country” of the 28 outside Mainland China.

DiamondPrincess1

Model: Andology Novel Coronavirus Prediction Tool.

2019-nCoV

The novel Coronavirus hasn’t been given a snappy acronym yet, like MERS or SARS. I would favour a name like Mad Scientist Disease or BWF (BioWeapon Flu). Gates of Hell Syndrome has a ring to it. Watching videos of people falling dead in the street in China, or vomiting blood on public transport, is very upsetting. How much worse is it going to get?

I downloaded the coronavirus prediction tool from Andology this morning. It is an Excel spreadsheet in which you can insert country-specific values for the following Assumptions:-

Starting Population

Start Date (Patient Zero)

Initial Infections

Infection Rate (R0)

Incubation Time (Days)

Mortality Rate

Mortality Complicator

Virus Burnout Rate

If you like messing around with spreadsheets you may find this one a joy to use – the pleasure obviously tainted by observing the deaths of millions of people.

I “modelled” China first. There is a suspicion that the Chinese Government is under-reporting infections and deaths but my first set of Assumptions generated much lower figures than those published to date. Adjusting R0 and Mortality Rate brought 450 predicted deaths to 5 February (494 reported), but these derived from only 2744 infections (24,447 reported). I will play around with the assumptions to see if I can bring predictions and reality closer together and thereby get a better idea of the final death rate when the virus burns out.

Each country infected will have a different journey over the spikey bell curve of rapidly increasing illness and death and an equally precipitous fall to zero infection in the population.

I applied the China assumptions to the UK (just two infected people so far) and was surprised that the virus died at about the same time – mid-June this year in China and a couple of weeks later in the UK. In around six months my models predict 392 million Chinese deaths and almost 10 million Brits and Northern Irish. As we wait in the UK for the novel coronavirus to claim its first victim, this outcome seems unlikely.

But the exponential growth of anything bad is the stuff of nightmares.

CHINA_Coronavirus1

With my set of amateurish assumptions, China will lose 27% of its population and the UK 14%. That’s 42% of those infected in China and 16% in the UK. The R0 and initial assumed Mortality Rate are the same for both countries (3.7 and 10%). I’m not clever enough to figure out this difference in the final outcome.

I hope things work out much better than this – but it could possibly be worse. Deaths resulting from the possible collapse of the global economy and social breakdown worldwide have to be factored in somehow.

2020 was going to be an interesting year without this.

Another name for novel coronavirus – WuFlu. You can also keep up with developments at Johns Hopkins. And if you download the prediction tool from Andology you can get up-to-date population figures at Worldmeters.