Predictive Modeling of COVID-19 Intensive Care Unit Patient Flows and Nursing Complexity: A Monte Carlo Simulation...

Area of Science:

• Healthcare Management
• Epidemiology
• Computational Biology

Background:

• The COVID-19 pandemic placed unprecedented strain on healthcare systems, particularly intensive care units (ICUs).
• Accurate forecasting of ICU bed demand and nursing complexity is crucial for effective resource allocation during public health crises.

Purpose of the Study:

• To develop and validate a Monte Carlo simulation model for predicting ICU bed requirements and nursing complexity for COVID-19 patients.
• To assess the model's accuracy in forecasting ICU occupancy and its utility for optimizing healthcare resource management.

Main Methods:

• A Monte Carlo simulation model was developed using patient data (age, sex, ICU length of stay, ventilation status) from March 2020 to September 2021.
• Nursing complexity was quantified using a three-level scale.
• The model was run 1000 times per scenario, and its output was compared against observed data and validated using external public health data.

Main Results:

• The simulation model demonstrated a strong fit with actual patient data (R² = 0.998, RMSE = 0.22).
• The model accurately predicted ICU occupancy and allowed for a 7-day forecast.
• Extrinsic validity was confirmed using French Public Health Authority data.

Conclusions:

• Monte Carlo simulation is a valuable tool for forecasting ICU bed demand and nursing complexity during pandemics.
• The developed model can assist healthcare systems in optimizing resource allocation and preparing for patient surges.
• This predictive capability is essential for enhancing pandemic preparedness and response strategies.