(In Press) On the volatility of blood inventories

Past research underestimates the volatility of blood inventories. Models are calibrated to be correct on average, but average behaviour does not capture the risk of blood inventories to experience outdates or shortages. Volatility in blood inventories does arise from stochastic supply, stochastic demand, perishability and blood issuing policies. However, these are not exhaustive as the logistics or the supply chain can cause dynamic behaviour even without stochastic supply and demand. We use system dynamics to illustrate the response of the blood supply chain to disturbances. This approach captures delays and feedback not previously incorporated into blood supply models. We first show that the structure of the system itself can cause volatility. We then propose a modification to ameliorate volatility. It is hoped that similar changes to real blood inventory systems might improve the robustness of the blood supply chain to shocks, thereby reducing shortages and outdates