A New Model for Blood Inventory Management

点击翻译按钮获取中文摘要

Efficient management of blood supplies is of great economic and social importance, as in many countries demand for blood products is growing faster than donations [2]. To date, the vast body of research in this area model the problem as a queuing or inventory system and use classical operations research tools such as stochastic processes, dynamic programming and stochastic programming to tackle the problem. To make the model tractable, most papers simplify the fundamental elements of the blood product supply chain, such as the multi-echelon structure of the system (i.e. inventory holdings at the regional center as well as individual hospitals), blood product substitutability (e.g. receiving O+ instead of A+ blood product type), variable demand volumes by patients, probability of change of patient status over time (i.e. from non-urgent to urgent), multi-level rationing of blood supplies to satisfy future urgent demands, and transportation time between the echelons of the product chain (e.g. from the regional stock to and among individual hospital stocks) and its impact on remaining blood product lifetime.** Recently, Abouee-Mehrizi et al. [1] introduced a new method called "customer composition" and solved several known problems in single and multi-echelon inventory management including multilevel rationing in multi-class M/G/1 make-to-stock systems [1] and prioritization in multi-echelon inventory management with stochastic production and transportation times [2]. They also showed that unlike the dynamic programming, customer composition does not suffer from the curse of dimensionality.** Building on initial results from applying the customer composition method to blood banking, and supplementing the method with the unit-flow approach [6] and fluid formulation [24], the proposed research program aims to analytically solve a multi-echelon multi-class system of blood product supply chain with stochastic transportation times that accounts for inventory rationing, changes in patient status, variable demand volumes, and blood product substitutability.********