[Spatial accessibility and resource allocation of prehospital emergency medical services in Guangzhou]

Objective: To evaluate the spatial accessibility and resource allocation of prehospital emergency medical services (PEMS) in Guangzhou and provide evidence for network optimization and resource allocation. Methods: The data from 2 842 communities/village committees and 150 "120" network hospitals, including emergency medical service (EMS) stations, were used. Spatial accessibility was evaluated by using the nearest-neighbor (minimum travel time/distance) approach and the enhanced two-step floating catchment area (E2SFCA) method; spatial clustering was tested with global and local Moran's I, and emergency resource allocation was evaluated by using Lorenz curve and Gini coefficient. Results: By the nearest-neighbor analysis, the shortest travel time from communities in the central urban area to the nearest EMS station were 7.93-9.90 minutes, whereas it exceeded 12.00 minutes in many peripheral districts (e.g., Nansha District). E2SFCA indicated higher accessibility in parts of Conghua and Zengcheng Districts, but lower accessibility in some communities/village in Huangpu, Tianhe and parts of Panyu Districts. The global Moran's I was 0.57 (P<0.001), demonstrating significant positive spatial autocorrelation; high-high clusters were observed in Conghua, Zengcheng and Yuexiu Districts, whereas low-low clusters were mainly in the central urban area and western Panyu District except Yuexiu District. Population-based Gini coefficient for ambulance and EMS stations were <0.30, in contrast, the area-based Gini for ambulance was 0.44. Geographically weighted regression results showed that the associations between accessibility and both the number and proportion of residents aged ≥60 years were usually positive, whereas the directions and intensity of the effects of public/medical facility supply variables exhibited spatial heterogeneity. Conclusions: The PEMS generally met response requirements in Guangzhou, however, spatial accessibility showed significant district specific difference and clustering pattern. Travel time was shorter in the central urban district but low-low clustering persist, whereas peripheral district tend to have longer travel times, while parts of Conghua and Zengcheng Districts exhibited relatively higher accessibility. Resource of PEMS allocation was relatively equitable based on population size but imbalanced based on district area. The coverage of PEMS in peripheral districts and the districts with low population density was still limited.