Developing a ‘live’ map of spatial access to health services in Aotearoa New Zealand
Keywords: network analysis, access to services, python, built environment
Abstract. This paper outlines the development of a proof-of-concept real-time model that incorporates current road conditions to estimate spatial access to health services daily, at the address level. National Highway road-closure data was collected from the New Zealand Transport Agency (NZTA) Application Programming Interface and local road closure data was scraped from council websites. An OpenStreetMaps (OSM) road network was modified daily by removing any closed roads. The distance from each address in the Manawa Taki Health Region, through this new road network, to the nearest Hospital was calculated. The program was automated to run each day in January, using current road conditions for that day to estimate hospital accessibility. Daily estimates of hospital accessibility were successfully automated, with variations in spatial accessibility over time noted. However, the importance of data quality for the accuracy of this model is paramount. Reporting structures and formats meant that data obtained from some local councils was found to be imprecise or unreliable. This approach shows potential for quickly estimating access to health services under changing road conditions, such as during and after extreme weather events. NZTA and local councils are encouraged to work together to improve the quality and interoperability of road closure reporting.
