Healthcare systems face increasing pressure from demand surges caused by pandemics, natural disasters, and seasonal illnesses. When hospitals operate near capacity, patient outcomes often deteriorate and staff burnout increases; however, managing these resources is notoriously difficult due to fluctuating demand, high uncertainty, and the competing goals of operational efficiency versus care availability. This project addresses the critical challenge of hospital surge response by moving beyond reactive, facility-specific measures to a proactive, systems-level approach that leverages the collective capacity of hospital networks.
We have developed a data-driven framework that integrates machine learning for demand forecasting with robust optimization models to guide high-stakes decisions, specifically the allocation of dedicated surge capacity and inter-hospital patient transfers. By accounting for demand uncertainty and real-world operational constraints, our models identify strategies that balance patient loads across facilities, significantly minimizing the need for costly emergency capacity expansion. We have translated this research into practice through the development of interactive decision-support dashboards, which were successfully deployed within the Johns Hopkins Health System during the COVID-19 pandemic to facilitate daily operations and strategic planning.
