Network-Based Kidney Allocation Simulation: Evaluating Organ Matching Strategies in Variable Hospital Networks

Abstract

Kidney allocation is a complex and high-stakes process shaped by various factors. Although existing simulations have improved our understanding of transplant policies, they often stem from the perspectives of medical urgency and biological compatibility. Hence, they often fail to account for the geographic and structural variability of hospital networks that heavily influence organ distribution. In this paper, we present a modular network-based kidney allocation simulation designed to function as a standardized benchmark to evaluate organ matching strategies in more realistic hospital networks. Our simulation models hospitals as nodes in a dynamic network that incorporates detailed donor-recipient compatibility criteria and provides an intuitive interface for testing custom matching policies. Then, through two case studies, we demonstrate how varying network structures and matching strategies affect patient outcomes, quantified through transplant success rates and the percentage of positive patient outcomes. Through these case studies, we derive results that closely resemble real-world observations and identify clear biological explanations for the observed trends, aligning with established empirical knowledge. Our results highlight the importance of geography-/network-aware benchmarks and allocation strategies and provide a standardized platform for future research on optimizing kidney distribution in healthcare networks.