Blood platelet inventory management: Incorporating data-driven demand forecasts.

Area of Science:

• Operations Research
• Healthcare Management
• Supply Chain Optimization

Background:

• Platelet products are critical for treating severe diseases like cancer but are expensive with short shelf lives.
• Managing platelet inventory is challenging due to uncertain and highly variable demand.
• Current inventory management strategies struggle to balance minimizing shortages and wastage effectively.

Purpose of the Study:

• To develop and evaluate a data-driven inventory management model for platelet products.
• To incorporate demand forecasts into inventory management to optimize ordering policies.
• To minimize both platelet shortages and wastage within the blood transfusion system.

Main Methods:

• Developed a novel inventory management model utilizing demand forecasts for platelet products.
• Employed forecast-dependent target inventory levels to guide the ordering policy.
• Utilized a large clinical dataset of daily platelet transfusions from 2016-2018 for a centralized blood distribution center serving four hospitals.

Main Results:

• The proposed data-driven policy effectively minimizes both shortages and wastage of platelet products.
• The model demonstrates resilience, tolerating larger forecast errors as system scale increases (e.g., via demand aggregation and inventory pooling).
• Sensitivity analysis confirms the model's robustness and suggests feasibility of less frequent than daily ordering.

Conclusions:

• Incorporating demand forecasts into inventory management significantly improves platelet product supply chain efficiency.
• The developed model offers a practical solution for reducing costs associated with platelet wastage and shortages.
• This approach supports more sustainable and reliable blood transfusion services.