Calculating Mortality Rates

We have not yet reached 8000 confirmed cases of the Wuhan coronavirus, which means that the current doubling from 4000 cases is taking more than 48 hours to complete. That's up from the 36 hours it took to go from 1000 to 2000 and from 2000 to 4000, but inline with early estimates made when there were less than 1000 reported cases. With 170 deaths, mortality rate remains at about 10%.

However, there are good reasons to believe that the official numbers are incorrect. Not everybody are equally affected by virus infections. Some, like the German case reported a few days ago, will sneeze and cough, run a slight fever, but otherwise be fine. Many of these cases will never be reported. Only severely ill people will show up at a hospital, especially in the midst of an epidemic. People do not go to hospitals without good cause when there's a risk of catching something truly nasty. The official numbers are therefore likely to be based on severe cases only.

Without a measure of how many mild cases there might be, we cannot say anything about the overall mortality rate of the disease. However, we can be pretty sure that the mortality rate among those who get a bad case of this virus is about 10%. This can be calculated by comparing the current number of dead people with the total number of cases that existed or had existed some 2 to 3 days ago. It can also be found from direct measurements of observed groups. Of a group of 40 early cases, 15% died.

From this it is clear that there's no reason for excessive concern. A light cough, some sneezing, and a mild fever is all that most people who catch this will experience, and even if it gets worse, the survival rate is at 90%.

By DoD - http://www.defenselink.mil/pubs/pdfs/App12_Pt12.pdf, Public Domain, Link

Stay Warm and Comfortable

With 6000 cases of the Wuhan coronavirus this morning, we're on track to reach 8000 by tonight. With 130 deaths, the mortality rate seems steady at about 10%. (There were about 1300 cases some 2 to 3 days ago). Numbers from outside China seem to confirm the contagion rate. Thailand has seen a doubling of cases over the last coupe of days, from 7 to 14. However, no-one outside China has died yet, so the mortality rate remains an estimate based on Chinese numbers.

The fact that the decease spreads during the incubation period, which may last up to two weeks, is troublesome. It makes it virtually impossible to know when to hunker down. By the time the decease is visible and evident to all, everybody that will get it has already caught it. Those that want to hunker down must therefore do so immediately, and be prepared to stay isolated for up to three weeks.

The more reasonable strategy is to focus on our immunity system rather than total isolation. Avoid activities and foods that stress our immunity systems. Don't go for walks in the rain and the cold. Don't stay up all night. Don't strain the body. Don't binge on sugars and junk food. Eat fatty, protein rich foods. Eat fruits and greens. Eat well. Sleep well. Stay warm and comfortable.

By Acabashi - Own work, CC BY-SA 4.0, Link

The Wuhan Coronavirus in Germany

The Wuhan coronavirus has reached Germany. A German national, working for a German company, caught the virus from a Chinese woman he was in contact with about a week ago. The woman was in Germany on business. There was a seminar in which the woman instructed an assembly of Germans, including the man who caught the virus.

On her way home to Beijing, the woman fell ill, and a week later, the German national fell ill too. He caught a fever, and was coughing quite badly. However, there was never any great concern. He recovered on his own. But when he learned that the Chinese woman had been diagnosed with the Wuhan coronavirus, he thought it best to check with a doctor to see if he had in fact been suffering from it himself. As it turned out, the tests proved positive.

This story is interesting, because it tells us for sure that the virus is contagious before the infected person is showing any symptoms. It tells us that it spreads through human to human contact, and it confirms the fact that it has the symptoms of a severe flu.

What happens next will be interesting to follow. Will this man's colleagues and family catch the decease as well? And what about the passengers on the plain the woman took back to Beijing? Many of them must have been Europeans. Are we about to see a widespread outbreak of the Wuhan coronavirus in Europe? If nothing much happens over the next couple of days, we're safe. Otherwise, we might see this spread through Europe in much the same way that it has been spreading through China.

By Laurent ERRERA from L'Union, France - Airbus A320-200 Airbus Industries (AIB) "House colors" F-WWBA - MSN 001, CC BY-SA 2.0, Link

Revising the Numbers

We are already at 4000 confirmed cases of the Wuhan coronavirus, and the death toll stands at 100. That's a doubling of cases in 36 hours, with a mortality rate steady at 10%. Again, it should be noted that the mortality rate must be calculated from the total number of cases that existed, or had existed, some 2 to 3 days ago. 100 dead out of a group of about 1000 cases is 10%.

However, the numbers cannot be trusted. There are good reasons to believe that the government is under-reporting the spread and severity of the disease. If they are clever about this, they may currently subtly overstate the doubling rate, so as to get closer to the real figures over time. Instead of an express revision, making them look bad, they choose to publish more aggressive spread rates for a week or two, so that when it's clear to everyone that the disease has reached epidemic levels, the official numbers match what is seen.

Going from 1000 official cases to 2000 cases took 36 hours. Going from 2000 to 4000 took likewise 36 hours. We can therefore expect 8000 cases to be reported late on Wednesday. By then, we'll probably have some data for the spread rate outside China, giving us a hint as to what the real numbers may be.

The rate of contagion will not be the same everywhere. Only densely populated areas will match the high rates seen in China. In places where people are less crammed together, the decease will spread slower.

By Robert Valette - Own work, CC BY-SA 4.0, Link

Narrative, Rhetoric and Action

The latest headline numbers for the Wuhan coronavisus is 2800 confirmed cases, and 80 dead. That's in line with our projections of 4000 confirmed cases by mid day tomorrow. It also confirms our estimated mortality rate of 10%, since the 80 dead is based on the compound number of cases that we had some 2 to 3 days ago. I.e. about 800 cases.

The disease has now spread to places outside of China, which means that the Chinese authorities no longer are in charge of the overall narrative. If the Chinese have misinformed the public, we will know this pretty soon. We will get mortality rates and rates of contagion that either confirm or contradict the Chinese numbers.

What caught my attention a few days back when I started to follow this more closely, was a deviation of narrative from previous scares, such as SARS, Bird flu, Swine flu and Ebola. In those cases, national and international authorities have tended to emphasize the severity of the outbreak while simultaneously claiming that things are under control: That there are vaccines and that everybody will be fine if we just obey whatever government health authorities demand. Importantly, previous cases have had catchy names. In this new case, the decease was simply the coronavirus. But coronavirus is not a name, it's a type. Most colds and flues are caused by coronaviruses.

The current outbreak follows a different narrative. We were repeatedly told that this is nothing to worry about. Everything is fine. The disease is contained. It does not communicate between people. If we just stop eating bat or snake soup, everything will be fine. But the panic laden actions taken by Chinese authorities a few days back was pretty much proof positive that things were far from contained. The rhetoric changed. We were suddenly faced with something severe. Yet there's still no word of any vaccine. There's also a blame game going on. People are pointing fingers. Scapegoats are found, such as bloggers promoting strange and unusual foods. The narrative is rapidly shifting. There's no central control of what people are being told. All of this indicate that the situation is the result of something completely unforeseen. Some accident in a government bio-lab, perhaps.

By Jean-Pierre from Cosne-Cours-sur-Loire (Nièvre, Burgundy region), France - Chemistry laboratory, detail, CC BY-SA 2.0, Link

Dealing With Uncertainties

There are now 2000 confirmed cases of the Wuhan coronavirus, so we are a little ahead of schedule. This means that the outbreak is in its exponential phase. At this rate, we can expect 4000 cases by mid Tuesday.

The mortality rate for this disease is said to be 3%. However, that number is in conflict with our calculations. The death toll is 3% of the current number of infected people. But to say that the mortality rate is 3% based on this is to understate the true number, which should be calculated based on the number of infected people 2 to 3 days back in time. Mortality is after all a trailing number. It takes about 2 to 3 days for a virus like this to go from identification to its peak severity. Mortality rates should therefore be calculated based on the total cases in the past, not the present number. Furthermore, numbers taken from an early sample of 41 patients yielded a mortality rate of 15%, more than our own estimate of 10%.

This illustrates that official figures are rarely accurate as to the true size and severity of an outbreak. However, by applying some simple math, we can still figure out the overall numbers. We may be off by a quite a lot, but the overall calculated trend and severity will be sufficiently accurate to take reasonable precautions. If officials understate the size of the outbreak by half, we're only off by 2 days in our projections. If officials understate the mortality rate, we can calculate our own numbers, or dig down to find more reliable numbers.

We don't have to know exactly how and when things happen, nor do we need to know the exact figures. It's still possible to figure out whether or not an outbreak is contained. We can also figure out roughly how dangerous it is. In this particular case, the numbers indicate that the outbreak is neither contained nor benign.

By I. Columbina, ad vivum delineavit. Paulus Fürst Excud〈i〉t. - 1. Johannes Ebert and others, Europas Sprung in die Neuzeit, Die große Chronik-Weltgeschichte, 10 (Gütersloh: Wissen Media, 2008), p. 197. https://books.google.co.uk/books?id=3DVH8dVGkX0C&pg=PA197 2. Superstock: Dr. Schnabel of Rome, a Plague Doctor in 1656 Paul Fuerst Copper engraving (Stock Photo 1443-1112), Public Domain, Link

Watching out for 2000, 4000, 8000, etc

As of last night, there were about 1000 confirmed cases of the coronavirus. That means that there should soon be 2000 cases. If the infection spreads exponentially, with a doubling every 2 days, we should reach 2000 by Sunday evening. If it spreads faster than this, we'll see 2000 before that. If slower, 2000 will appear later.

The next number to look for after 2000 is 4000. If 4000 is reached quicker than we reached 2000 from 1000, we're faced with an explosive situation. If the time interval is the same, we're still in a regular exponential phase. If slower, we may be looking at an early topping off, in which case there's still hope for containment.

After 4000, we look for 8000. Here again, we'll compare time intervals to judge where in the cycle we are. We can again determine if the disease is progressing explosively, exponentially, or topping off. For every doubling, we look for the next doubling. This way, we stay on top of the situation. We can estimate the rate at which the disease is spreading, and roughly when to hunker down for a day or two.

By Michiel Sweerts - Image: http://collections.lacma.org/sites/default/files/remote_images/piction/ma-31770455-O3.jpg Gallery: http://collections.lacma.org/node/183358 archive copy, Public Domain, Link