A Collaborative Blood Distribution System in a Network of Hospitals based on their Normal and Emergency Requests: a Mathematical Model and Solution

Abstract

<strong>Background and Objectives:</strong> A blood distribution network orchestrates distribution of safe blood products to hospitals. Blood shortage and blood wastage are two important factors which may affect efficiency of blood distribution network. Service delivery time is another factor that refers to the time interval between blood request by a hospital and transfusing it to the patient. Collaboration between hospitals can mitigate the three mentioned factors.<br /> <strong>Methods:</strong> Mixed Integer Non-Linear Programming (MINLP) was used to model both current blood distribution network and collaborative blood distribution network mathematically. Two types of collaboration, including preventive and reactive, and two types of blood requests, including normal and emergency, were considered in modeling. The proposed model has been put into effect by considering platelets as blood product and applying real data of three high usage, medium-usage and low-usage hospitals in Tehran, Iran, during a planned horizon of 7, 14 and 30 days of July 2017 using IBM ILOG CPLEX solver.<br /> <strong>Results:</strong> Preventive collaboration led to a decrease in blood wastage at low-usage hospital and a decrease in blood shortage at high-usage one. Reactive collaboration led to a decrease in blood shortage at all types of hospitals. Moreover, service delivery time of emergency requests was decreased using collaboration.<br /> <strong>Conclusion</strong>: Determining the effect of preventive and reactive collaboration on blood shortage, blood wastage and service delivery time regarding both normal and emergency requests is the contribution of this paper. Managers of blood organizations and hospitals can save patients' life and reduce costs of the whole blood distribution network by defining groups of hospitals which collaborate with each other under predefined protocols. The proposed model can be used to predict improvement in blood distribution network by implementing collaboration between hospitals.