A queuing model for ventilator capacity management during the COVID-19 pandemic

Area of Science:

• Health Systems Research
• Epidemiology
• Operations Research

Background:

• The COVID-19 pandemic placed unprecedented strain on critical care resources, particularly ventilators.
• Effective capacity planning for intensive care units (ICUs) is crucial for managing public health emergencies.
• Understanding the impact of public health interventions on resource utilization is essential for policy decisions.

Purpose of the Study:

• To develop and apply a queuing model for ventilator capacity planning in British Columbia during the first wave of COVID-19.
• To assess the impact of public health measures and social distancing on ventilator access.
• To quantify the relationship between epidemic projections, critical care resources, and patient access.

Main Methods:

• A multi-class Erlang loss model was developed to represent ventilator use by COVID-19 and non-COVID-19 patients.
• Discrete event simulation was used to project ventilator access under various transmission scenarios.
• The model was calibrated and validated using data from the BC Intensive Care Unit Database.
• Numerical approximation methods were compared to develop a hybrid optimization approach for capacity planning.

Main Results:

• Public health measures and social distancing potentially averted up to 50 deaths per day by preventing ventilator capacity from being reached.
• Without these measures, an additional 173 ventilators would have been needed to ensure 95% patient access.
• The model demonstrated that policy interventions significantly influenced critical care resource needs and patient access.

Conclusions:

• A queuing model framework effectively informs ventilator capacity planning during epidemics.
• Public health interventions play a critical role in managing demand for critical care resources.
• The developed model provides a valuable tool for policymakers to estimate resource needs and evaluate intervention impacts.