Nowadays, many of the healthcare systems are large, complex environments and quite dynamic, specifically Emergency Departments, EDs. They are opened and working 24 hours per day throughout the year with limited resources. EDs are usually the main entrance to the hospital, and a key component of the whole healthcare system. The original mission of EDs is to primarily handle only emergency situations. However, ED visits include a wide range of illnesses and injuries, from truly emergencies to non-urgent cases. As a consequence, EDs are overcrowded. Thus, is mandatory to use extensively computer simulations of EDs to evaluate output responses. The choice of optimal simulation parameters can lead to improved functioning, but choosing a good configuration remains a challenging problem. This improvement can be achieved by modelling and simulating EDs using Agent-Based Modelling and simulation. Optimisation via simulation is an emerging field which integrates optimisation techniques into simulation analysis. _x000D_ In this research a two-phase optimisation methodology for optimisation via simulation for healthcare Emergency Departments is proposed. The first phase is a coarse grained approach consisted in a global exploration step over the entire search space. This phase identifies promising regions for optimisation based on a neighbourhood structure of the problem, using either a pipeline scheme approach of an Emergency Department or the Monte Carlo heuristic plus the K-means method, or both. This first phase returns a collection of promising regions. The second phase is a fine grained approach that consists in seeking the best solution, either the optimum or a sub-optimum by performing a “reduced exhaustive search” in such promising regions. _x000D_ This work optimises the sanitary staff configuration of an actual ED. The sanitary staff configuration comprises: doctors, triage and emergency nurses, admission personnel, and x-ray technicians, the amount, and sort of them. Staff configuration is a combinatorial and multidimensional problem, that can take a lot of time to be solved. In order to do optimisation, objective functions to minimise or maximise have to be set. Three different indexes were set: minimise patient length of stay (LoS); maximise number of attended patients per day (Throughput); and minimise a compound index, the product of the cost of a given sanitary staff configuration times patient length of stay (CLoS). HPC is used to run the experiments, and encouraging results were obtained. However, even with the simplified ED used in this work the search space is very large, thus, when the problem size increases, it is going to need more resources of processing in order to obtain results in a reasonable time.
Optimisation via simulation for healthcare emergency departments.
Cabrera Flores, E. C. (Author). 19 Nov 2013
Student thesis: Doctoral thesis
Student thesis: Doctoral thesis