A data-driven approach to determine daily platelet order quantities at hospitals

Area of Science:

• Hematology
• Healthcare Management
• Operations Research

Background:

• Hospitals face challenges in determining daily platelet unit needs due to unpredictable demand and short product shelf-life.
• Effective inventory management is crucial for blood products like platelets to minimize wastage and prevent shortages.

Purpose of the Study:

• To develop and evaluate a linear prediction model for optimizing daily platelet unit orders.
• To assess the model's impact on reducing platelet wastage and shortage rates in a hospital setting.

Main Methods:

• A linear prediction model was developed using historical demand data and hospital information systems.
• The model's performance was evaluated by comparing predicted vs. actual wastage and shortage rates.
• Data from two hospitals in Hamilton, Ontario (2015-2017) were used for model training and performance evaluation.

Main Results:

• The proposed model demonstrated a significant reduction in expiry (wastage) and shortage rates across two hospitals.
• Hospital 1: Expiry rate reduced from 8.67% to 2.54%; Shortage rate from 13.86% to 4.01%.
• Hospital 2: Expiry rate reduced from 2.28% to 0.05%; Shortage rate from 8.48% to 0.44%.

Conclusions:

• Implementing the developed prediction model can substantially decrease platelet wastage and shortage rates.
• This approach offers a data-driven strategy for improving hospital blood inventory management.
• Optimized platelet ordering enhances resource allocation and patient safety.