Simulation in healthcare — Part 2

This issue of JOS is the second part of a healthcare themed issue following Volume 3(3). We have actually struggled to include all of the papers we wanted into two parts! In fact, it seems a healthcare JOS could be a journal in itself. So why has there been so much interest? Healthcare policy decisions are usually highly complex, affecting human outcomes on a large scale leading to constraints around real-world experimentation and, of course, as a result of the high demand for healthcare services, there are many queuing systems and bottlenecks which many of us will have experienced. Simulation is therefore well suited for addressing many policy decisions required within healthcare. In addition, healthcare is such a broad ranging subject that there is a wide range of applications for simulation.

Within Part 1 of the healthcare themed issue of JOS, the editorial discussed the challenges associated with implementation of simulation modelling within healthcare. Within Part 2, Naseer et al describe a toolkit to allow the healthcare practitioner to select appropriate methods according to their individual needs. The interactive nature of the approach between the modeller and the healthcare practitioner is a good example of methodology which may facilitate implementation of simulation within healthcare.

The wide variety of healthcare applications for which simulation is being used and the different techniques being employed is apparent from the vast array of papers submitted to the issue. Stevenson et al describe an assessment of the cost-effectiveness of a diagnostic test for detecting thrombophilia. Alongside the case study, the authors provide a review of the use of discrete event simulation (DES) within the Health Technology Assessment (HTA) Programme. While HTA modelling is well established, the application of DES in this field is not. The authors provide some potential explanations for this.

Sobolev et al describe a statistical paper focusing on methodology for calculating the number of simulation runs required and how input factors should vary across the runs. This methodology is applied to a case study of two approaches for booking admission dates for patients undergoing elective surgery. A very different application of DES is described by Gregory et al. A mock circulation loop used to test ventricular assist devices is simulated. This work requires a detailed physical understanding of the human heart, rather than an understanding of a non-biological system such as a hospital.

Gunal et al present a literature review around performance modelling in hospitals, describing which aspects of hospitals have previously been simulated. Within this the authors discuss a long-standing question in healthcare modeling—is it worthwhile to develop a generic model which can be modified for different localities, or should a specific model for each location be developed. The authors also discuss the potential use of other modelling techniques, which may be applied alongside DES to facilitate whole hospital modelling.

Jones et al describe a DES to optimise elective bed availability within a critical care unit to avoid hospital cancellations of elective surgery. The authors test a number of ‘what if’ scenarios. Meng et al describe an application of agent-based simulation to the management of hospital-acquired infection. Interestingly, this paper discusses the shortcomings of using DES (commercial software) for modelling infectious diseases and describes why agent-based modelling is more appropriate for this type of modelling. Finally, a paper by Huang et al describes methodologies for teaching simulation modelling of infectious diseases. The paper illustrates challenges and solutions to building network-based epidemic simulations by considering several case studies including HIV, SARS and influenza—a particularly timely paper given the recent outbreak of Swine Flu.