Sunday, February 23, 2020

Covid-19 Maps

One month ago, on 23 January 2020, the John Hopkins' Wuhan Coronavirus (2019-nCoV) Global Cases interactive map was reporting 17 deaths from Covid-19 and 555 total known cases. One month later the John Hopkins map is reporting 2,462 deaths and 78,823 total known cases.

This week there appears to be the first major outbreak of Covid-19 outside of Asia. In the last few days more than 100 cases of Covid-19 have been confirmed in Italy and two people have died from the virus. Last night Italy imposed 'extraordinary measures' to try to halt the spread of Covid-19. Under these new measures the residents of a number of towns in Lombardy and Veneto, in northern Italy, have been asked to stay at home. People are now forbidden to leave or enter this outbreak area without special permission.

As the number of cases of Covid-19 has grown many other institutions across the world have begun mapping incidents of the disease around the globe. Here are some of the other interactive maps which are tracking Covid-19:

Covid 2019 Tracker - this map from the London School of Hygiene & Tropical Medicine shows the number of deaths and total number of cases of Covid-19 in countries around the world. The map also allows you to compare the rise of the Covid-19 outbreak with the 2003 SARS, 2014 Ebola and 2009 Swine Flu epidemics.

COVID-19 Dashboard - a dashboard showing cases of COVID-19 on a 3D globe. The dashboard also includes a graph showing the rise of the outbreak over time, the total number of cases, the number of recoveries and the number of deaths.

Coronavirus Infection Tracker - this map from Japan's Nikkei visualizes the number of Covid-19 cases over time. The map is available in both English and Japanese.

Tracking Coronavirus COVID-19 - an interactive tracking map from the mapping company HERE. It includes both a map and timeline of Covid-19.

COVID-19 Cases and Clusters Outside of Mainland China - this map from the University of Virginia is tracking the spread of the virus outside of China. It includes charts showing a breakdown of Covid-19 cases by age and by recorded symptoms.

How to Read Covid-19 Maps and Charts

All the above maps and charts provide a good general overview of the rise and spread of Covid-19 over time. However you should consider a number of factors when reading the data.

Non-normalized Data

None of the maps show the incident rates for Covid-19 in each country. All the maps linked above show the total numbers of cases and none of them normalize the number of confirmed cases by the country's population. Therefore these maps all show the total number of Covid-19 cases and not the incident rate of the virus in each country.

For example the John Hopkins map is currently showing three cases of Covid-19 in India and two cases in Spain. These numbers are shown on the map using scaled markers of a very similar size. However 2 cases in a population of 49 million constitutes a far higher incident rate of Covid-19 in Spain than the 3 reported cases in a population of 1.28 billion in India. Despite the markers for Spain and Italy being very similar in size the incident rate of Covid-19 (based on reported cases) is far higher in Spain than it is in India.

Data Collection Methods

Maps which show the spread of the virus in different countries are prone to errors from the way that Covid-19 is detected and reported in each country. There have also been changes to the way that the virus is being monitored and recorded since the outbreak was first detected. For example last week China's National Health Commission removed 108 deaths from the overall total, because of a previous double counting of fatalities. At the same time the overall number of deaths and the overall number of cases was also revised upwards as China also changed its methodology for registering Covid-19 cases.

1 comment:

Anonymous said...

If one's concern is to minimize risk by publicizing known outbreak locations to shape decisions about how people move around, the imprecision introduced by sizing the blobs based upon numbers of cases or deaths hinders. A binary decision on confirmed presence of virus across a more granular geography might prove helpful in shaping travel patterns and limiting its spread. For instance, should I make connections in Frankfurt or Dusseldorf?