On-line Discrete event Simulation for the management of inpatient beds
11:00 - 11:30, A54 PSC
Effective management of inpatient beds in acute hospitals is one of the most crucial tasks for hospital managers and admissions planners. In hospitals where beds are shared between elective (planned) and emergency (unplanned) patients, a balance must be struck between the number admissions from each patient type. This poses an operational problem which must be managed on a daily basis, given the variation in emergency patient arrivals, for example. The problem is compounded further by the complex dependence structure between inpatient wards during times of high bed occupancy.
Online DES is a relatively new technique predominantly used to inform operational decision making in a manufacturing setting. Short-term estimates of the performance of the real system can be made for each proposed decision, in each planning horizon by periodically updating the simulation to the true system state. In this way, the short-term implications of a particular admissions schedule can be tested using a model which adequately captures the complex dependence structure of inpatient wards.
This talk presents some of the work undertaken so far to produce statistical models of the dependence structure of ten inpatient wards in an Australian General Hospital during times of varying bed demand, with a view to implementing these models in an online simulation for operational decision making.