ShortageSim: Simulating Drug Shortages under Information Asymmetry

2025-09-01

AI Summary

Global pharmaceutical shortages are exacerbated by supply chain information asymmetry, while existing regulatory interventions lack systematic evaluation. This paper introduces the first large language model (LLM)-based multi-agent simulation framework for modeling dynamic, boundedly rational strategic interactions among manufacturers, procurement agencies, and regulators. The framework integrates a sequential production game mechanism with 2,925 historical FDA drug shortage event records to enable multi-period dynamic simulation. Its key contributions are: (1) pioneering the application of LLMs to pharmaceutical shortage modeling, thereby relaxing the assumption of perfect rationality and capturing realistic, adaptive decision-making; and (2) achieving an 83% reduction in simulated disclosure delay for stockouts—substantially improving alignment between simulation outputs and empirical observations in historical case validation. The framework implementation and curated dataset are publicly released.

Technology Category

• Search and Optimization: Sampling/Simulation-based Search
• Game Theory and Economic Paradigms: Imperfect Information
• Planning, Routing, and Scheduling: Planning with Language Models

Application Category

• Semantics and Knowledge: Data modeling to support human-machine intelligence, including LLMs agents, intelligent system behavior, explanations, and user-friendly interactions
• Economics, Online Markets and Human Computation: Uses of LLMs and GenAI for marketplace design, bidding, and strategic interactions
• User Modeling, Personalization and Recommendation: Large Language Models (LLM) for user modeling and recommendation

Abstract

Drug shortages pose critical risks to patient care and healthcare systems worldwide, yet the effectiveness of regulatory interventions remains poorly understood due to fundamental information asymmetries in pharmaceutical supply chains. We present extbf{ShortageSim}, the first Large Language Model (LLM)-based multi-agent simulation framework that captures the complex, strategic interactions between drug manufacturers, institutional buyers, and regulatory agencies in response to shortage alerts. Unlike traditional game-theoretic models that assume perfect rationality and complete information, extbf{ShortageSim} leverages LLMs to simulate bounded-rational decision-making under uncertainty. Through a sequential production game spanning multiple quarters, we model how FDA announcements, both reactive alerts about existing shortages and proactive warnings about potential disruptions, propagate through the supply chain and influence capacity investment and procurement decisions. Our experiments on historical shortage events reveal that extbf{ShortageSim} reduces the resolution-lag percentage for discontinued-disclosed cases by 83%, bringing simulated durations more aligned to ground truth than the zero-shot baseline. We open-source extbf{ShortageSim} and a dataset of 2,925 FDA shortage events at https://github.com/Lemutisme/Sortage_Management, providing a novel computational framework for designing and testing interventions in complex, information-scarce supply chains.

Problem

• Simulating drug shortages under information asymmetry in pharmaceutical supply chains
• Modeling strategic interactions between manufacturers, buyers, and regulatory agencies
• Testing effectiveness of FDA shortage alerts and warnings on supply chain decisions

Innovation

• LLM-based multi-agent simulation framework
• Simulates bounded-rational decision-making under uncertainty
• Models FDA announcements' impact on supply chain