On 29 December 2019, Chinese authorities identified a homogeneous inexplicable pneumonia cases in Wuhan city, Hubei Province, China (Tan et al., 2020). A novel strain of coronavirus (2019-nCoV) subsequently put a patient in quarantine on 7 January 2020 (Tan et al., 2020). Most cases from the initial cluster had epidemiological links with a live animal market, suggesting a possible zoonotic origin (World Health Organization, 2020). Infections in healthcare workers confirm the occurrence of human-to-human transmission, though the extent of this mode of transmission is unclear. On 21 January 2020, the WHO suggested there was possible sustained human-to-human transmission. Additionally, being connected internationally through both direct and indirect flights (Bogoch et al., 2020), Wuhan possibly become the origin of a new global pandemic.
However, current clinical and epidemiological data are insufficient to understand the full extent of the transmission potential of the epidemic. Therefore, it is imperative to tracking those data until professional epidemiologists ascertain the best methodologies to thwart the epidemic from striking in the future. While today sees online dashboards which track real-time data of Coronavirus information from WHO and Johns Hopkins University, I build an online dashboard with some striking differences, as follows.
- Beside the data gathered from WHO, CDC which are similar to the JHU’s dashboard, I also use the data from Tencent which contains one of the most up-to-date public information in China.
- Tracking day-by-day mapping confirmed cases in the world.
- A formal model to predict 5-day-ahead confirmed prevalence.
- Au courant reproduction indicator estimates with distinctive approaches.
- The simplest epidemiological model SIR.
How to use my dashboard
My dashboard is now on the air at Shinyapps.io or can be followed below.
At a glance, the dashboard is divided into five rows.
The first row shows overall statistics of the up-to-date information of COVID-19 where we can see the number of total confirmed cases, total deaths and total recovered cases in the world, as well as the predictive maximum quantity of total deaths as the epidemic reaches its nadir. People can use the information to up to date themselves, whereas some professional practitioners such as financial analysts can exploit it to conduct some stress testing for necessary capitalization purpose.
The second row indicates the world map with some dard-red points which means the sizes of the prevalence across nations. Also if users click Play, they can observe time-varying pattern of those sizes, assisting to visualize the changes by time all over the globe.
The third row demonstrates how our statistics in the first row change over time and the number of infected people which I predict based on my model. Adjacent to that chart is a table showing the details of prevalence cases amount in China provinces which all makes up almost 99 percent of the total. The numbers are also sorted descending order.
Given in the forth row is information on (basic) Reproduction quantity (\(R_0\)), which can be defined as “the expected number of secondary cases produced by a single (typical) infection in a completely susceptible population” (James, 2007). Accordingly, \(R_0\) of 10 means that an infected person will transmit his disease to an average of 10 other people, provided no one has been vaccinated against or already immune to that disease in their community. As can be seen, there are four approaches I use to calculate the indicator; i.e., Exponential Growth, Maximum Likelihood, Sequential Bayesian, Time Dependence, as well the estimate’s 95% confidence intervals. Next to \(R_0\) estimates is time-series of \(R_0\) calculated based on Time Dependence from 20 Jan 2020 till present.
Last but not least, several predictive statistics according to Susceptible-Infected-Removed model have been suggested. The basic idea is quite simple. There are three groups of people:
- those that are healthy but susceptible to the disease (S),
- the infected (I) and
- the people who have recovered (R).
To model the dynamics of the outbreak, we need two important parameters, i.e, the effective contact rate which controls the transition between S and I and the removal rate which controls the transition between I and R. In other words, the effective contact rate means the number of close contacts per day per infected and the removal rate means the fraction of infecteds recovering in a given day.
Please note that there are some assumptions embedded:
- constant population size
- constant rates (e.g., effective contact rate, removal rate)
- no demography
- well-mixed population: any infected individual has a probability of contacting any susceptible individual that is reasonably well approximated by the average.
The clarification document will be attached to the dashboard in the future.
If you’ve read everything up till now, thank you very much! Any comment, question, or general feedback is always. Please feel free to reach out to me through below comment frames.
Bogoch, I. I. et al. (2020) ‘Pneumonia of Unknown Etiology in Wuhan, China: Potential for International Spread Via Commercial Air Travel’, Journal of travel medicine. doi: 10.1093/jtm/taaa008.
Tan, W. et al. (2020) ‘A Novel Coronavirus Genome Identified in a Cluster of Pneumonia Cases — Wuhan, China 2019−2020’, China CDC Weekly. China CDC Weekly, 2(4), pp. 61– 62.
World Health Organization (2020) ‘WHO | Novel Coronavirus – China’. World Health Organization. Available at: http://www.who.int/csr/don/12-january-2020-novel-coronavirus- china/en/ (Accessed: 26 January 2020).