A generic simulation model for planning critical care resource requirements

Intensive care capacity planning based on factual or forecasted mean admission numbers and mean length of stay without taking non-linearity and variability into account is fraught with error. Simulation modelling may allow for a more accurate assessment of capacity needs. We developed a generic intensive care simulation model using data generated from anonymised patient records of all admissions to four different hospital intensive care units. The model was modified and calibrated stepwise to identify important parameters and their values to obtain a match between model predictions and actual data. The most important characteristic of the final model was the dependency of admission rate on actual occupancy. Occupancy, coverage and transfers of the final model were found to be within 2% of the actual data for all four simulated intensive care units. We have shown that this model could provide accurate decision support for planning critical care resource requirements.