Emergency Medical Services (EMS) delay poses a challenge to public health, especially in ultra-dense cities and aging societies. The potential relationship between the built environment and the ambulance response time (ART) has not been thoroughly investigated at the building-level. This impedes the unraveling of high-risk buildings for EMS delay, and thus the precise optimization of urban EMS systems. Our study develops a quantitative-qualitative approach to construct a building-level ART prediction model and then validated it with ambulance drivers' experiences. A comprehensive theoretical framework is tailored for ART prediction incorporating neglected built environment factors of land use and development intensity. Based on 73,129 ambulance trajectories, a machine learning model is constructed and then employed to predict the ART for each of the 253,475 buildings in central Shanghai under multiple combinations of traffic periods and weather conditions. The results show that the accuracy of high-risk buildings for EMS delay achieves 91%. Moreover, three of the neglected built environment factors, medical POIs density, business POIs density and floor area ratio, rank high in the factor importance. The ambulance drivers’ experiences validate the importance of these built environment factors and emphasize the value of bus priority lanes for ambulance to borrow for swift access. In addition, to illustrate potential applications, we simulated the effects of built environment interventions in case areas predicted to concentrate high-risk buildings. This work provides a new computer-based tool to pinpoint the service blind spots of such urban emergency systems and to develop built environment interventions.
Read full abstract