A stochastic programming approach to perform hospital capacity assessments

Area of Science:

• Operations Research
• Healthcare Management
• Stochastic Optimization

Background:

• Accurate hospital capacity assessment is crucial for effective healthcare delivery.
• Existing models often lack the ability to account for stochastic variations in patient treatment and length of stay.
• Strategic planning for healthcare infrastructure requires robust methods to handle uncertainty.

Purpose of the Study:

• To introduce a novel risk-averse stochastic programming approach for strategic hospital capacity assessment (QAHC).
• To incorporate stochastic treatment durations and length of stay into hospital capacity analysis.
• To identify maximum treatable caseloads within a specified timeframe and risk threshold for capacity exceedances.

Main Methods:

• Development of a bespoke risk-averse stochastic programming model.
• Application of sample averaging techniques for probabilistic constraints.
• Implementation of a novel two-stage hierarchical solution combining meta-heuristics and binary search for a complex mixed-integer programming model.

Main Results:

• The proposed approach effectively identifies maximum caseloads under probabilistic constraints.
• The two-stage hierarchical solution is computationally efficient.
• Case study and numerical tests demonstrate the approach's effectiveness in analyzing hospital outputs.

Conclusions:

• The developed stochastic programming approach provides enhanced clarity and insights for hospital capacity planning.
• This method enables better calibration of healthcare infrastructure to future demands, including pandemic scenarios.
• The novel approach offers a significant advancement in strategic healthcare resource management.