Degeling Koen, Karnon Jonathan, van de Ven Michiel, Brennan Alan, Koffijberg Hendrik
Cancer Health Services Research, Centre for Cancer Research and Centre for Health Policy, Faculty of Medicine, Dentistry and Health Sciences, The University of Melbourne, Melbourne, Australia.
GSK, Amersfoort, Netherlands.
Appl Health Econ Health Policy. 2025 Jun 17. doi: 10.1007/s40258-025-00983-8.
Discrete event simulation (DES) provides enhanced flexibility over modelling techniques that have been traditionally used for assessing health-economic outcomes, making it a particularly interesting technique for modelling complex clinical pathways. Discrete event simulation also facilitates consideration of resources and capacity constraints, making it suitable for addressing a wide range of research questions in health care and beyond. However, those unfamiliar with DES often perceive it to be more complex compared to traditional health-economic modelling techniques, such as state-transition modelling. To address this perceived complexity, this tutorial provides a detailed illustration of implementing DES in the open-source R software using the simmer package, through a case study in colon cancer. The tutorial is aimed at those who have a conceptual model that they want to implement as a DES in R, and are looking for practical guidance. It discusses methodological aspects related to DES and individual-level modelling in general that have not been extensively covered in literature, the conceptual model structure and corresponding pseudocode, data analysis, model implementation, and the deterministic and probabilistic analysis of the model. The documented code provides all building blocks required to develop a wide range of DES models in R using the simmer package.
离散事件模拟（DES）相较于传统用于评估健康经济结果的建模技术具有更高的灵活性，这使其成为一种用于对复杂临床路径进行建模的特别有趣的技术。离散事件模拟还便于考虑资源和能力限制，使其适用于解决医疗保健及其他领域的广泛研究问题。然而，那些不熟悉DES的人通常认为它比传统的健康经济建模技术（如状态转换建模）更为复杂。为了解决这种认知上的复杂性，本教程通过一个结肠癌案例研究，详细说明了如何使用simmer包在开源R软件中实现DES。本教程面向那些拥有概念模型并希望在R中作为DES实现，且正在寻求实践指导的人。它讨论了与DES及一般个体水平建模相关的方法学方面，这些方面在文献中尚未得到广泛涵盖，包括概念模型结构和相应的伪代码、数据分析、模型实现以及模型的确定性和概率分析。记录的代码提供了使用simmer包在R中开发各种DES模型所需的所有构建块。
