Time-varying spatial accessibility of primary healthcare services based on temporal variations in demand, supply, and traffic conditions: A case study of Seoul, South Korea

Abstract Background: Although demand, supply, and traffic conditions are interdependent, scholars have considered them separately when measuring the accessibility of healthcare services. Additionally, the temporal variations in these three factors is given relatively less attention in healthcare accessibility research, which mislead the wise policy decisions of healthcare planners. This study aims to examine the spatiotemporal variability of accessibility to primary healthcare services and identify areas with spatial discrepancy between time-varying and static accessibility models in Seoul, South Korea. Methods: Using the generalized two-step floating catchment area method, we measure time-varying spatial accessibility with de facto population from hourly-collected mobile-based data, available healthcare facilities, and actual traffic conditions in Seoul, South Korea. Additionally, the bivariate Local Moran’s I allowed for the identification of the areas with discrepancy between the two accessibility models.Results: The results show that a geographical pattern of time-varying accessibility significantly differs with that of static accessibility. Time-varying accessibility is lower in commercial areas and higher in residential areas, compared with static accessibility, whose accessibility is higher in central Seoul. The result of the bivariate Local Moran’s I analysis highlights that ignoring time variations of the three factors will result in overestimation in commercial areas and underestimation in residential areas. Based on the discrepancy between the two models, we typified overestimated areas and underestimated areas as attention needed and false attention areas, respectively.Conclusion: We suggest areas in which the healthcare policy needs to pay more attention by identifying areas with spatial discrepancies between time-varying accessibility and static accessibility. Considering the time variations will provide a more realistic impression of spatial accessibility to primary healthcare.