Cities around the world differ on countless dimensions. The glow of a sunset. The smell of a sea breeze. The gritty air from a thousand industrial chimneys. The hum of people and animals and machines and elements. As far as public health is concerned, some of these differences affect outcomes for people living in our cities, and where the differences can be changed, there are opportunities for intervention. One very important and tangible feature on which cities differ is their basic footprint: their size, their shape, their overall configuration. Assistant Professor Christopher Morrison, along with international collaborators, published a study in The Lancet Planetary Health to examine how the footprint of cities around the world are related the incidence of injuries due to motor vehicle crashes.
The authors used digital images for the largest 1700 cities from around the world to develop categories based on the layout of roadways, waterways, green space and other such features. They used a machine learning approach in which they trained a computer to recognize 750 images from each city, and then using another 250 images they developed a “confusion matrix” based on instances when the machine mistook the images of one city for another city. From this confusion matrix they identified 9 categories of city footprints, which they gave names such as checkerboard cities, motor cities, and high transit cities. The authors then compared injury incidence per 100,000 population for these cities footprints. They found that high transit cities, meaning cities in which a large proportion of land area was dedicated to rail infrastructure, had around half the crash injury incidence compared to the worst performing categories: cul-de-sac cities, informal cities, and sparse cities. The high transit cities tended to be in western Europe. Cul-de-sac cities were almost all in Indonesia. Informal cities were in India, Africa and the Middle East. Sparse cities predominated in eastern China.
Lead researcher Jason Thompson, PhD, from the University of Melbourne’s Transport, Health and Urban Design Research Hub said the research emphasized the importance of urban design as a big picture determinant of health. City footprints that discourage public transit had greater injury incidence, possibly due to more motor vehicle use. City footprints that discourage safe road use (for example by combining pedestrian and motor vehicle thoroughfares) had similarly high injury incidence, possible due to more chaotic roadways. Preventive intervention is clearly a long term prospect – redesigning cities is not going to happen immediately – but over time, building cities that preference public transit, pedestrian traffic, and safe roadway use will reduce injuries.