Background Timely access to emergency patient care is an important quality and efficiency issue. Reduced discharges of inpatients at weekends are a reality to many hospitals and may reduce hospital efficiency and contribute to emergency department (ED) congestion.

Objective To evaluate the daily number of ED beds occupied by inpatients after evenly distributing inpatient discharges over the course of the week using a computer simulation model.

Methods Simulation modelling study from an academic care hospital in Toronto, Canada. Daily historical data from the general internal medicine (GIM) department between 15 January and 15 December for two years, 2005 and 2006, were used for model building and validation, respectively.

Results There was good agreement between model simulations and historical data for both ED and ward censuses and their respective lengths of stay (LOS), with the greatest difference being +7.8% for GIM ward LOS (model: 9.3 days vs historical: 8.7 days). When discharges were smoothed across the 7 days, the number of ED beds occupied by GIM patients decreased by ∼27–57% while ED LOS decreased 7–14 hours. The model also demonstrated that patients occupying hospital beds who no longer require acute care have a considerable impact on ED and ward beds.

Conclusions Smoothing out inpatient discharges over the course of a week had a positive effect on decreasing the number of ED beds occupied by inpatients. Despite the particular challenges associated with weekend discharges, simulation experiments suggest that discharges evenly spread across the week may significantly reduce bed requirements and ED LOS.