Exploring accessibility to medical facilities for older adults: Potential vs. Revealed accessibility

Introduction

The issue of population aging is increasingly prominent, presenting unprecedented challenges and pressures on urban medical resources. The equity and convenience of medical services for older adults have emerged as critical concerns for urban planners. While previous research on medical accessibility for older adults has yielded fruitful results, it has lacked a thorough analysis of the quantitative relationship between revealed and potential accessibility across urban spaces, as well as the influencing factors leading to the spatial heterogeneity of these two types of accessibility.

Methods

This study calculates the ratio of revealed to potential accessibility based on multi-source data, thereby identifying the allocation of medical resources in different sub-districts in Beijing. Furthermore, an extreme gradient boosting model is used to investigate the non-linear associations between the potential influencing factors and the ratio of revealed to potential accessibility.

Results

The results revealed a significant spatial disparity between potential accessibility and revealed accessibility and identified the threshold effects of influencing factors contributing to the accessibility ratio of revealed to potential accessibility. The density of bus stops, vehicle ownership, and density of roads are the top three among the influencing factors in terms of their relative importance in influencing the accessibility ratio.

Conclusions

The threshold effect should be considered when implementing planning adjustments, as simply increasing or decreasing the values of specific built environment variables may not result in a corresponding improvement in medical accessibility for older adults. With limited resources, enhancing the planning and construction of transportation infrastructure is more efficient in reducing the discrepancy between revealed accessibility and potential accessibility compared to other methods.