Hybrid simulation modelling of emergency departments for resource scheduling
Yinling Liu, Thierry Moyaux, Guillaume Bouleux, Vincent Cheutet
Abstract
This paper addresses the problem of resource scheduling of emergency departments in an uncertain environment. An innovative approach based on Agent-Based Simulation (ABS) and Discrete Event Simulation (DES) is proposed. To do so, this paper first proposes a framework integrating ABS and DES. Subsequently, we provide a detailed description of the DES and ABS parts of the model of the ED. Five combinations of strategies – namely, “FIFO+Random” (i.e. FIFO: the queue of patients operates in FIFO; Random: the patients are randomly assigned to doctors), “FIFO+Centralised”, “Random+Centralised” (“Centralised” implies the assignment of the most appropriate doctor to a patient is only determined by the Platform agent via optimising all the resources), “Random+Random” and “Autonomous” – are implemented to schedule the resources of the ED. Finally, experiments analyse the efficiency of the five combinations of strategies on the duration of patients’ stay. The results show that “FIFO+Centralised” outperforms others regarding the average for the duration of patients’ stay without hospitalisation: 3.75% of the time has been saved; “Random+Centralised” outperforms others regarding the average for the duration of patients’ stay with hospitalisation: 0.57% of the time has been saved.