Characterizing Physician Referral Networks with Ricci Curvature

Area of Science:

• Network Science
• Healthcare Analytics
• Computational Topology

Background:

• Identifying systemic barriers to quality healthcare access in the US is challenging.
• Key indicators of healthcare efficacy and regional disparities in care delivery require further understanding.

Purpose of the Study:

• To introduce a novel application of network curvature to Physician Referral Networks.
• To identify indicators of healthcare efficacy and regional disparities.
• To present an open-source tool for analyzing healthcare network structures.

Main Methods:

• Applied Forman-Ricci and Ollivier-Ricci curvature measures to Physician Referral Networks.
• Utilized network topology to characterize network structure.
• Integrated network features with local census data and healthcare effectiveness metrics.

Main Results:

• Forman-Ricci and Ollivier-Ricci curvature measures show promise as indicators of healthcare efficacy.
• Network curvature captures significant regional demographic features and variations in care delivery.
• The study identified correlations between network structure, demographic data, and patient outcomes.

Conclusions:

• Network curvature is a valuable tool for analyzing Physician Referral Networks and understanding healthcare disparities.
• The developed open-source tool, apparent, facilitates the examination of complex relationships within healthcare systems.
• Findings contribute to improving healthcare access and quality by identifying regional variations.