Core Guide to the access_lib Framework

access_lib is a Python framework for healthcare accessibility analysis developed based on the Extended E2SFCA (Enhanced Two-Step Floating Catchment Area) model. It supports functions such as demand estimation, transportation network analysis, sensitivity analysis, and scenario simulation, aiming to address the multi-dimensional challenges of quantifying healthcare accessibility and helping decision-makers scientifically evaluate and optimize the allocation of healthcare resources.

Background and Challenges of Quantifying Healthcare Accessibility

Equitable distribution of healthcare resources is a core issue in public health policy, but 'accessibility' involves multiple dimensions such as geographical distance, transportation conditions, population demand, and service capacity, making it difficult to quantify simply. Traditional supply-demand ratio indicators ignore the complexity of spatial interactions and cannot accurately reflect the ease with which residents actually access healthcare services, which led to the development of the access_lib project.

E2SFCA Model and Extended Improvements

Basic Logic of the E2SFCA Model

• Step 1: Centered on healthcare facilities, calculate the supply-demand ratio within the service radius
• Step 2: Centered on residential areas, accumulate the supply-demand ratios within the accessible range to obtain the accessibility index

Extensions in access_lib

• Accessibility calculation for multiple transportation modes (driving, public transit, walking, etc.)
• Dynamic demand estimation (population age structure, disease incidence rate, etc.)
• Service capacity constraints (differences in admission capacity among hospitals of different levels)
• Multi-time slice analysis (weekdays vs. weekends, peak vs. off-peak hours)

Modular Framework Architecture

Data Layer

Supports reading and writing of vector data (Shapefile, GeoJSON, etc.), raster data (GeoTIFF, etc.), network data (OSM data), and attribute data (CSV, Excel, etc.)

Core Modules

• Demand Estimation: Population interpolation, disease incidence mapping, age-standardized weights, etc.
• Transportation Network: Shortest path calculation, multi-mode integration, real-time traffic interface, differentiated travel modeling
• Analysis Engine: Extended E2SFCA algorithm, supporting decay function selection and parallel computing
• Scenario Simulation: Impact of telemedicine, network outage shocks, facility layout optimization (genetic algorithms, etc.)

Application Cases

1. Urban-Rural Healthcare Gap Assessment: Analyze provincial-level urban-rural specialized healthcare gaps to support the construction of county-level medical communities
2. Emergency Network Optimization: Evaluate the coverage efficiency of emergency stations, simulate the location of new stations, and reduce response time by 15%
3. Pandemic Response Planning: Evaluate the accessibility of fever clinics and develop emergency transfer plans

Sensitivity Analysis Functions

• Single-factor analysis: Adjust parameters to observe result changes
• Multi-factor analysis: Monte Carlo simulation to evaluate parameter combination uncertainty
• Scenario comparison: Result differences under different assumptions
• Visualization report: Automatically generate heatmaps, tornado charts, etc.

Helps identify key parameters, enhance the robustness of conclusions, and provide quantitative information on result uncertainty

Limitations and Future Outlook

Current Limitations

• Efficiency of large-scale network analysis needs improvement
• Real-time traffic data interface is not perfect
• 3D spatial accessibility is not supported

Future Directions

• Introduce graph neural networks to improve demand prediction accuracy
• Develop a web visualization interface to lower the entry barrier
• Build an interface for a global open database of healthcare facilities
• Support multi-objective optimization algorithms to handle facility layout problems