Stock management in hospital pharmacy using chance-constrained model predictive control

Area of Science:

• Operations Research
• Health Systems Management
• Control Theory

Background:

• Hospital pharmacy stock management is a critical issue, balancing clinical needs with resource limitations.
• Drug demand is inherently random, complicating inventory control and decision-making.
• Existing methods struggle to efficiently manage complex constraints and optimize resource allocation.

Purpose of the Study:

• To propose a novel chance-constrained model predictive control (CCMPC) for hospital pharmacy stock management.
• To address the dual objectives of satisfying clinical drug needs and minimizing economic resources.
• To provide a robust framework for managing random drug demand under multiple constraints.

Main Methods:

• Development and application of a chance-constrained model predictive control (CCMPC) model.
• Explicitly incorporating multiple objectives and constraints within the control framework.
• Evaluating the CCMPC solution's implementation feasibility in real-world hospital settings.

Main Results:

• The CCMPC approach effectively manages random drug demand and complex constraints.
• The model demonstrates a practical trade-off between inventory conservativeness and operational efficiency.
• Implementation feasibility was assessed in two Spanish hospital pharmacy departments.

Conclusions:

• Chance-constrained model predictive control offers a flexible and efficient solution for hospital pharmacy stock management.
• This method enhances the ability to meet clinical drug requirements while optimizing resource utilization.
• The study validates the CCMPC approach for practical application in healthcare settings.