Home Articles Mapping infectious diseases using SARS maps

Mapping infectious diseases using SARS maps

Xun Shi
Xun Shi
Department of Geography
Dartmouth College
Hanover, USA
[email protected]artmouth.edu

When the sizes of the mapped spatial units vary significantly, a choropleth map of absolute numbers can cause seriously biased visual impression. In this paper, the possible misinterpretation caused by the choropleth SARS maps is discussed and a way to map disease risk instead of simple number of cases is proposed

The outbreak of Severe Acute Respiratory Syndrome (SARS) in 2003 provided an opportunity for GIS professionals to demonstrate the effectiveness of modern spatial analytic and mapping techniques in modeling, tracking, and presenting the information about the spread of an acute infectious disease. During the outbreak period, numerous websites were created to offer frequently updated maps showing the dynamics of the disease at region, country, and world levels. Some of these websites even provided functions for interactive mapping and spatial analysis (Boulos, 2004). Thanks to GIS and internet technologies, general public for the first time in the history have so convenient access to the information about how a dangerous infectious disease is spatially close to them and how serious the conditions in other parts of the world.

It is noticeable, however, that most of the SARS maps available on the web are choropleth maps showing the absolute numbers of cases in spatial units. In such a map, each polygon is filled with a uniform color that represents the number of SARS cases occurring within the spatial unit delineated by the polygon. In different maps, these spatial units can be counties, states, provinces, or countries. Examples of this kind of SARS map include World Health Organisation’s (WHO) SARS maps of the whole world (an example can be found at , Nations Online’s SARS map of China (https://www.nationsonline.org /oneworld/china_sars_map.htm), Sina’s interactive SARS maps of China and the world (https://news.sina.com.cn/duihua/sars/default.html), Maptell’s interactive SARS map of the world worldsars.htm), Corda’s SARS map of the world /examples/go/map/sars.cfm), and Mapcruzin’s sequence SARS maps of The United States (https://www.mapcruzin.com/sars-severe-acute-respiratory-syndrome).

Presenting the information about SARS cases in this way has its reasons. First, choropleth maps might be the easiest maps to create with GIS, and they are visually appealing and seemingly effective. Second, they provide simple identification information that general public usually prefer to see. For example, many people may ask the question “Which countries have been affected by SARS?” In a choropleth map, it is easy to identify which countries have been affected and where these countries are. Third, to most map makers data of disease cases are only available at the areal unit level (i.e. only aggregated data are available) and the map makers had to use a single value for the entire polygon.

However, mapping the numbers of SARS cases in this way violates an established rule in cartography. Cartographers consider that choropleth maps are suitable for presenting values that have been normalized by the areas of the spatial units, such as ratio and density, but should be avoided when the values are absolute numbers (Robinson et al. 1995). Especially when the sizes of the mapped spatial units vary significantly, a choropleth map of absolute numbers can cause seriously biased visual impression.