[Paper Review] A Two-Stage Stochastic Programming Model for Blood Supply Chain Management, Considering Facility Disruption and Service Level

Mohammad Arani, Mohsen Momenitabar+2 more|arXiv (Cornell University)|Nov 2, 2021

Blood donation and transfusion practices22 references4 citations

TL;DR

This paper proposes a two-stage stochastic programming model for blood supply chain management that integrates facility disruption risks and service level requirements under supply and demand uncertainties. By optimizing blood flow across donors, collection centers, blood banks, and hospitals, the model enhances network resilience and service reliability, with numerical analysis demonstrating improved performance under stochastic disruptions and demand variability.

ABSTRACT

In this paper, a blood supply chain network, where the occurrence of disruption might interrupt the flow of Red Blood Cells, is dealt with. In principle, the probability of disruption is not the only property confiding the network, but unprecedented fluctuations in supplies and demands also contribute to the network shortages and outdated blood units. Although the consideration of parameter uncertainties is of paramount importance in the real-world circumstances for a decision-maker, she or he would be willing to monitor the network in a properly broader perspective. Therefore, one of the eminent key performance indicators known as service level turned our attention. To tackle uncertainties in the mentioned network - comprising of the four conventional levels containing donors, blood collection facilities, blood banks, and hospitals - we present a two-stage stochastic programming model. Consequently, a toy-example is randomly generated to validate the proposed model. Furthermore, numerical analysis led us to a comprehensive service level analysis. Finally, potential pathways for future research are suggested.

Motivation & Objective

• To address the vulnerability of blood supply chains to facility disruptions and demand/supply fluctuations.
• To incorporate service level requirements as a key performance indicator in blood supply chain decision-making.
• To develop a robust optimization framework that accounts for uncertainties across all network levels: donors, collection facilities, blood banks, and hospitals.
• To validate the model through a synthetic case study and analyze service level trade-offs under various disruption scenarios.
• To provide actionable insights for improving blood supply chain resilience and operational efficiency.

Proposed method

• Formulates a two-stage stochastic programming model to handle uncertainty in blood supply chains.
• Models disruptions at facilities using probabilistic scenarios to reflect real-world unpredictability.
• Incorporates service level constraints to ensure adequate blood availability at hospitals.
• Uses a four-tier network structure: donors, collection centers, blood banks, and hospitals.
• Employs scenario-based optimization to balance cost, reliability, and service level performance.
• Validates the model using a randomly generated toy example and conducts sensitivity analysis on service levels.

Experimental results

Research questions

• RQ1 How can a blood supply chain be optimized under the combined impact of facility disruptions and demand/supply variability?
• RQ2 What is the impact of service level constraints on the overall cost and reliability of blood supply chain operations?
• RQ3 How does the two-stage stochastic programming approach improve decision-making compared to deterministic models?
• RQ4 What are the trade-offs between operational cost and service level under different disruption scenarios?
• RQ5 How can resilience be enhanced in blood supply chains through strategic facility and inventory planning?

Key findings

• The proposed model effectively reduces blood shortages by accounting for facility disruptions and demand fluctuations.
• Service level constraints significantly improve the reliability of blood delivery to hospitals, especially under high-uncertainty conditions.
• The two-stage stochastic approach outperforms deterministic models in maintaining system stability and minimizing waste.
• Numerical analysis reveals that higher service levels require increased safety stock and strategic facility redundancy.
• The model demonstrates robust performance across diverse disruption scenarios, validating its practical applicability.
• The toy example confirms the model’s ability to balance cost, service level, and resilience in blood supply chain networks.