Measuring and modelling surgical bed usage

Surgical departments treat two groups of inpatients - the simple and the co mplex consequently a single average fails to describe the use being made of the occupied beds. Using decision support techniques, we show why indicato rs such as the average length, the average occupancy and the average admiss ions mislead. Furthermore, by analysing the fluctuating pattern of weekly a dmissions we show how weekends and the Christmas holiday periods impact on bed usage. Next, we demonstrate that flow process models can be used to des cribe how the in-patient workload concerns two groups of patients. On an av erage day, 71.4% of the beds contained patients who will have an average (e xponential) stay of 4.8 days, and the other beds, 28.6%, contain patients w ho will have an average (exponential) stay of 22.8 days. The article conclu des by demonstrating the short and long-term impact on daily admissions of a 10% change in four different parameters of the model. The data used come from a surgical department in Adelaide, as UK data sets report finished con sultant episodes rather than completed in-patient spells.