Platelets inventory management: A rolling horizon Sim–Opt approach for an age-differentiated demand

ABSTRACT

Platelets are highly perishable components of blood. The shelf life of platelets is very short during which the platelets should be transfused, otherwise spoiled. It is evident that some medical conditions require platelets of specific ages. In this study, a stocking model of platelets is considered under age-differentiated and stochastic demand. The model aims at finding the optimal quantities by minimising the overall inventory costs. A hybrid of simulation and optimisation is established to optimise the decision variables. This work contributes to the existing literature by discriminating platelets by five ages, employing a new inventory policy of a rolling demand, and using simulation-optimisation framework. Shelf lives of 3 and 4 days were considered for the comparison with literature. The proposed model could outperform the selected policies by demonstrating lower costs. Moreover, the model resulted in lower cost per platelets’ unit compared to the actual costs of some selected hospitals.

KEYWORDS:

• Blood platelets
• perishable inventory
• simulation
• optimisation
• simulated annealing

Acknowledgements

The authors would like to thank the Kuwait Central Blood Bank for the valuable information provided to complete this study. Deep gratitude also goes to the aforementioned hospitals whose collaboration made this study possible.

Disclosure statement

No potential conflict of interest was reported by the authors.

Statement of Contribution

It is evident nowadays that different medical conditions may require blood platelets of specific ages to avoid further risks and complications. For instance, oncology and transplants patients usually request the youngest platelets, while traumatology patients have no preference. Platelets are highly perishable components of blood which are voluntarily supplied by donors. The shelf life of platelets is very short (around 5 days) during which platelets should be transfused, otherwise, spoiled. Shortage and Expiration costs are considerably high in such an inventory of live cells. Due to platelets short shelf life and the stochastic age-differentiated demand, no elegant analytical models exist for solving the optimal production quantities. Consequently, the main contribution in this study is to find the daily optimal order quantities that will minimise the shortage, expiration, holding, and age mismatch costs under stochastic and age-differentiated demand via simulation–optimisation hybrid frameworks. Age-differentiated demand is relatively new topic in perishable inventory.

In contrast to existing literature, a rolling horizon of weekly order quantities is employed to replenish the inventory in blood banks. Different settings are used such as compatibility between different ages, replenishment of only fresh units, older-and-closer-in-age policy, and preference in age mismatch which make the problem more realistic, yet more challenging in terms of operations research modelling.

A hybrid of simulation and optimisation (Sim–Opt) frameworks are established to optimise the inventory decision variables. The goal is to minimise the costs by finding the optimal order quantities subject to standard transfusion guidelines and constraints. The presented model is benchmarked with the most recent literature studies. For this set-up, four literature policies are considered, namely, NIS, (S, c), 1D, and 2D. The proposed model outperforms the four policies by presenting less cost. It also demonstrates less cost as compared to those incurred in the selected hospitals.