Walking is a popular activity for both recreation and transport [1], and its health benefits are well known. People who engaged in walking and cycling for travel purpose were less likely to be overweight or obese than those who travelled by private motor vehicle [2]. Other benefits of walking include improved social cohesiveness [3], traffic decongestion [4], and reduced vehicle operating costs [4], infrastructure expenditure [5] and environmental impacts [4]. The World Health Organization's Physical Activity Guidelines present walking as an example of recommended leisure time and transportation physical activities. Researchers, policy makers and public health practitioners have explicitly advocated more walking and cycling for daily travel as the most affordable, feasible, and dependable way for people to get the additional exercise they need [6]. For example, in Australia, Walk to Work Day and National Walk Safely to School Day are annual events to encourage active transport.

Scope for increasing participation in walking for transport is large. A study based on the Sydney Greater Metropolitan Area Household Travel Survey reported that the majority of the car trips were short (25% were 5 minutes or less) such that some of these trips may be potentially replaceable by walking or cycling to achieve health benefits [7].

Key to encouraging walking is the design of the built environment to facilitate walking. Walkability scores have been developed to measure how well the characteristics of the physical environment support walking by providing for pedestrian comfort [8]. Many of the available walkability scores comprise of three or four components related to walkability, i.e. proximity of destinations, population density, street connectivity and land use mix [9, 10]. High population and road density, mixed land use [11], and more connected streets [12] are related with more walking.

However, some of these characteristics or characteristics associated with them, may also be associated with increase in pedestrian-related injury. Kim et al found that areas with high levels of commercial activity were positively and significantly associated with pedestrian-related accidents [13]. A systematic review of child pedestrian injury reported that high road density, population density, presence of schools, land use mix and close proximity of services were associated with increased walking but more child pedestrian injury [14]. Another study found that proportion of area zoned for commercial use was positively associated with vehicle-pedestrian injury collisions [15]. Dai suggested that areas surrounded by business centers, entertainment hubs and residences may become ‘black spots’ for pedestrian crashes due to these areas being highly congested and heavily trafficked [16]. Some environmental features, for example, street parking may encourage more walking but are also associated with high exposure to traffic [14] and thus increase the risk of pedestrian injury. On the other hand, street connectivity such as intersection density, one of the walkability index component, may promote walking and is associated with low risk of pedestrian collisions [17].

For areas to be considered “walkable” they should also provide for pedestrian safety, and this should be reflected in measures of walkability. Improving pedestrian safety is also key to encouraging walking because perceived traffic safety is a key determinant of walking for both transport and leisure [18]. Unfortunately, pedestrians are among the most vulnerable road user groups [19]. In the event of a crash, they have little to no protection compared to other road users [19]. The risk of injury per km travelled by walking and cycling is 5-10 times higher than driving a car [20].

However, current indices that quantify the extent to which environments support walking (and therefore guide planning) incorporate pedestrian safety to only a limited extent. The few walkability scores that have taken pedestrian and traffic safety into account are applicable only at street-level. For example, the UK-based ‘Rate My Street’ online application provides a walkability score based on users’ ratings of their local streets’ walking environment (e.g. ease of crossing the street, pavement width; trip hazards) including safety from crime and traffic [21]. The walkability score developed for the ‘Safe routes to school’ study in the USA includes perceived safety from crime and traffic (as well as accessibility to school) [22].

Walk Score, a publicly available website, has been widely used and validated for estimating walkability of a neighbourhood [23]. It incorporates population density, street connectivity, block lengths and proximity to amenities. Walk Score has shown positive associations with walkable amenities or destinations [24], residential density [25], street connectivity [25, 26], access to public transit [24, 26], and intersection density [26]. Walk Score has also shown positive associations with health indicators [27] and physical activity [28]. Information on the risk of pedestrian injury is not incorporated into Walk Score. As far as we know, there is no study exploring the relationship between Walk Score and pedestrian injury.

The aim of this study is to explore the association between the widely-used neighbourhood-level “Walk Score” and hospitalisations for pedestrian-related injury, and the severity of pedestrian-related injury in those who are hospitalised