Sustainable Blood Supply Chain Modeling with Government Intervention: A Three-Level Stackelberg Approach under Two-Sided Supply–Demand Uncertainty and Explicit FEFO Implementation

Abstract

This study aimed to develop and validate an integrated three-level Stackelberg optimization model for a sustainable blood supply chain incorporating government subsidies, simultaneous supply and demand uncertainty, independent age dynamics for red blood cells, platelets, and plasma, explicit First-Expired–First-Out allocation, and transportation-related carbon emissions. A three-level Stackelberg game was formulated in which the government acted as the leader, the Iranian Blood Transfusion Organization and provincial processing centers constituted the first-level followers, and distribution centers represented the second-level followers. Four combined supply–demand scenarios were constructed using five years of historical data. The follower problem was transformed using Karush–Kuhn–Tucker conditions after verifying the linear independence constraint qualification and was subsequently linearized as a mixed-integer linear programming model. Product aging and FEFO allocation were modeled through age-indexed inventory and allocation variables. The model was solved using CPLEX and validated against five months of operational platelet data from the Isfahan blood supply network. Government intervention increased the collection of O-negative blood by 760%, although marginal benefits declined beyond a subsidy budget of 60 units. Implementation of the model reduced immediate-fresh demand shortages from 30% ± 4.2% to 5% ± 1.4%, representing an 83% reduction. The proposed three-product model achieved 14% lower total cost and a 17-percentage-point higher service level than the single-product benchmark. Extending platelet shelf life from three to five days reduced platelet wastage by 61% and generated monthly carbon-emission savings of 12.4 kg. Optimized Big-M values reduced solution time by more than fifteenfold compared with the conventional value of one million. Model predictions showed a mean validation error of 4.1%. The proposed framework provides an effective decision-support mechanism for coordinating government subsidies, blood-product allocation, perishability, uncertainty, service priorities, and environmental objectives while improving cost efficiency, service levels, and operational resilience.