Simulation of Healthcare Processes: Challenges, Solutions, and Benefits

The emergency department of a hospital plays a key role in the incoming patient management. Therefore, it is crucial to ensure an adequate level of organization and efficiency. In this project, we aim to analyze and improve the processes at emergency departments using business process simulation. Through simulation experiments, various ‘what-if’ scenarios can be tested, and redesigning alternatives can be compared with respect to some key performance indicators. The input for business process simulation is a process model extended with additional information for a probabilistic characterization of the different run-time aspects (case arrival rate, task durations, routing probabilities, roles, etc.). It is thus critical that the business simulation model is accurate so as to ensure that the simulations and the various ‘what if’ scenarios reflect credible alternatives with realistic outcomes. This project aims to create emergency-department simulation models on the basis of the actual executions that are recorded in so-called event logs, which are typically extracted from the information systems that support the execution of processes at the hospital. The process model and the simulation parameters are extracted using different techniques from the field of Process Mining, which builds on an analysis of the event logs and aims to gain insights into how processes are actually carried out. In particular, this project has focused on the analysis of the emergency department at a hospital in Tuscany. The quality of the extracted logs greatly influences the analysis that can be carried out. For example, to tackle the reduction of waiting times for patients and to optimize the resources, two of the most critical emergency-department challenges, it is necessary that the data report when each activity has started and completed, and how hospital staff participated to the execution of each activity. Unfortunately, the event logs extracted from the information systems of the Tuscany’s hospital missed relevant information, including the timestamps when activity started, thus diminishing the realism of the business simulation model. To overtake this issue, the project extended a previous technique to estimate the missing timestamps of when the activities started, and overtook some of its limitations. The project assessed the extended technique on the emergency department of the Tuscany’s hospital, and has shown on this case study how healthcare can leverage on process mining to simulate different ‘what-if’ scenarios, on the basis of which decisions can be made on how to improve real processes.