Simulation in healthcare — Part 1

The use of simulation in healthcare is becoming increasingly vast and diverse, to the extent that so many interesting papers were submitted to the healthcare themed issue of JOS that we decided to split the issue into two parts.

Part I of the issue begins with a review of academic literature of the methods and applications of simulation and modelling in healthcare. This paper by Brailsford et al places the use of simulation in healthcare into context alongside other modelling techniques. Based on the papers submitted to this special issue, it appears to be commonplace that more than one technique of analysis is being utilised within healthcare simulation studies. The conceptual modelling themed issue of JOS published in 2007 presented several methods that may be used during the simulation model development process. It is clear that modellers are now not considering simulation as a standalone or isolated technique, but rather that there are a number of adjunctive techniques that can be used to make simulation models more useful. This is apparent from submissions to a Journal of Simulation that do not in themselves describe simulation, but rather tools to improve simulation modelling. This is particularly important in a field such as healthcare, where systems may be so large and subject to such variability that no one person could describe it and where subjectivity, ethical considerations and the concerns of multiple stakeholders all come into play.

A paper by Vasilakis et al presents the use of a Unified Modified Language (UML) that could be used to develop requirements for a simulation model. Within this paper, UML is applied to a case study of older people presenting with fractured neck of femur at an acute care hospital in the UK. In another study, alternative techniques are used alongside simulation. The paper by Gonsalves et al uses optimisation to minimise waiting times according to subjective expectations modelled using fuzzy constraints. Discrete event simulation is used to model the clinic at which waiting times are being minimised. This range of methodologies is used to address the subjective as well as the objective elements in the patients’ evaluation of healthcare services.

In a very different application of simulation, Robinson et al describe a methodology for generating synthetic patient data that has the same statistical properties and interdependencies as an original data set in order to avoid delays within clinical research. Chest pain diagnosis at an Emergency Department is used as an example.

Alongside the development of healthcare simulation modelling and associated techniques, several papers submitted to this issue of JOS also consider issues associated with implementation; one of the key challenges for the use of simulation in healthcare. Bowers et al describe the development of a simulation of an Accident and Emergency department and discuss the successes and failures of the development and use of the model. The paper provides insight into some of the reasons why the results of simulation modelling may not be readily adopted within a healthcare setting. This should be a key consideration for the development of simulation models within healthcare. As Brailsford et al describe within their paper, while there is a wide range of modelling studies being undertaken within healthcare, only a small proportion of those studies that have been reported have been implemented in practice for their stated purpose.

The paper by Eatock et al uses simulation to assess the effectiveness of the use of telemedicine within a generic outpatient clinic, using data generated from clinical trials. The paper outlines the differences between the trial, the model and current reality and the implications of these differences. Interestingly, the paper also highlights the disparity in healthcare practitioners’ views and modellers’ views around the use of modelling to support decision making. It is likely that these differences in views are fundamental to the reasons why the implementation rate of healthcare simulation models is lower than healthcare modellers would generally like. The implementation of simulation modelling in healthcare remains a significant challenge and should be a focus of future research.