Inventory Sharing for Perishable Products: Application to Platelet Inventory Management in Hospital Blood Banks.

This article focuses on optimizing inventory sharing and transshipment policies to reduce wastage of perishable products, specifically platelets, in a two-location hospital system. Motivated by the Emory University Hospital System’s platelet management challenge, the authors develop a Markov decision process model and prove that the direction of optimal transshipment depends solely on comparing the age of the oldest inventory at each location after demand fulfillment. They propose a simple, myopic transshipment policy that is optimal for special cases—including a two-period lifetime and an asymmetric demand scenario—and serves as a lower bound in general settings. Empirical results using both synthetic and real data demonstrate that implementing this policy at Emory led to approximately a 20% reduction in platelet outdates without increasing shortages, and that inventory sharing can increase optimal inventory levels for perishable products, contrasting with established findings for nonperishable goods. The study highlights that conventional insights from nonperishable inventory management do not fully apply to perishable products, underscoring the need for tailored policies in healthcare supply chains.