On Coronavirus (2019-nCoV)
Photo from Dr Eric Ding's Twitter.
As coronavirus becomes the hot topic among everyone, especially Chinese, this week, I did some data analysis and visualisations related to the virus and on Sars (a similar virus that happened in 2003).
Let's first have a quick look at what people on Twitter talk about China and coronavirus.
I scrapped 400,000 tweets about China and coronavirus and look at the most popular hashtags people use related to these two topics.
Unsurprisingly, the hashtags people used for these two topics are similar. When China is mentioned in a tweet, it is most likely related to coronavirus and Wuhan. The same applies to coronavirus, it is closely linked to China and Wuhan.
Apart from the worried comments and discussions about the coronavirus outbreak, I also heard (and read on the internet) people claiming the impacts of the coronavirus are exaggerated. While way more people died because of, for example, diabetes and heart disease, no one has in actuality give this tremendous amount of attention to Coca-Cola but to Wuhan and coronavirus.
But the issue here is the death rate.
According to the International Diabetes Federation, there are 463 million diabetes patients and 4.2 million deaths in 2019, so the death rate is about 0.9%.
Now, let's look at the death rate of Sars according to the WHO data.
Sars has an average death rate of 16.4%, with Malaysia has the highest of 40% and China has the lowest of 7%. That said, the death rate of Sars is roughly 18 times higher than diabetes. 1 in 5 people who are diagnosed with Sars has a probability to die compared to 1 in 111 people for diabetes. Not to mention the spreading speed of coronavirus, and the existence of super-spreader.
*NB, I removed South Africa from the graph because it has 1 imported case and unfortunately passed away, thus its death rate is 100%.
Now let's look at the statistics of the countries that have the most Sars cases.
Although China's data only available since the end of March, it is the country that has the most Sars cases and deaths. The combination of death numbers in China and Hong Kong constitute more than 80% of the total death count. Although Taiwan and Canada were both known as one of the severely afflicted countries, the infection and death rate were not at as severe as China and Hong Kong.
As we can see at the map below, there are serval imported Sars cases in Europe as well, only France has reported a death case due to Sars. Thus, its death rate is 14% as shown above.
A dynamic graph showing the speed of the growing Sars cases and deaths from 17 March to 11 July 2003.
Doubtless, the growth (both confirmed cases and deaths) in China, Hong Kong and Taiwan are the most obvious.
One of the most striking features I found from the beginning to the end of my project is the cases and death rate in Macao. Although Macao is right next to Hong Kong and extremely near to some severely afflicted countries such as China and Taiwan, it only has 1 Sars cases and 0 death.
Alas, less than 1 month of the 2019-nCoV outbreak, Macao already has 2 cases this time.
Here is a static version of the graph above using the finalised data.
Given the fact that Sars is a disease known as extremely active under cold weather, I find the pattern in Europe worth a discussion. It shows that while some southern (or SW) European countries like Spain, Italy and France have confirmed cases, northern European countries like Finland and Norway are Sars free.
My guess here (with my limited medical knowledge) is since Europe only had a few isolated Sars cases in some countries and had the virus situation in control, its situation was not applicable to the doctrine of coronavirus is more adaptive under cool weather. And most importantly, Europe did not have Sars super-spreader and the cases there were all imported cases, mostly from Asia.
One thing that worried me now is that if coronavirus is adaptive under cold weather, and was also one of the reasons why it stopped spreading before August 2003 in Asia, then 2019-nCoV should be a global concern now. Since, obviously, it is only in January now. And the virus still has a huge possibility to mutate.
Normille (2013).“Understanding the enemy”
And here's a very useful dashboard built by Johns Hopkins Center for Systems Science and Engineering that provides near real-time tracking of the 2019-nCoV.