A data-driven optimization model to response to COVID-19 pandemic: a case study.

Area of Science:

• Operations Research
• Healthcare Management
• Epidemiology

Background:

• COVID-19 has strained global healthcare systems, leading to hospitalization limitations and increased mortality risks.
• Efficient management of hospital resources, medication distribution, and waste is crucial during pandemics.
• Uncertainty in patient numbers and potential disruptions necessitate robust supply chain network design.

Purpose of the Study:

• To design an optimized supply chain network for hospitalizing COVID-19 patients.
• To efficiently distribute medications and medical supplies while managing hospital waste.
• To address uncertainty in patient demand and facility disruptions through a data-driven approach.

Main Methods:

• A two-phase approach combining Artificial Neural Networks (ANNs) for demand forecasting and K-Means for scenario reduction.
• Development of a multi-objective, multi-period, two-stage stochastic programming model.
• Incorporation of objectives: maximizing allocation-to-demand ratio, minimizing disease spread risk, and minimizing transportation time.

Main Results:

• Identified optimal locations for temporary facilities in high-density areas lacking existing infrastructure.
• Temporary hospitals can meet up to 2.6% of total demand, alleviating pressure on existing facilities.
• The proposed model maintains an ideal allocation-to-demand ratio even during disruptions.

Conclusions:

• The study provides an effective framework for pandemic-related healthcare supply chain management.
• Temporary facilities are vital for maintaining healthcare service levels during demand surges.
• The stochastic programming model successfully mitigates risks associated with demand uncertainty and facility disruptions.