Course Level: Beginner
Format Requirements: This course will involve a combination of didactic lectures, class discussions and examples. It will be useful for those not familiar with simulations and for those who are already experienced in one or two methods, but want to look ‘outside the box’. No previous knowledge is necessary, although pre-meeting readings abstracted from working papers provide useful background and are recommended.
Background: Most courses about modeling in health care focus on a single modeling approach. This course provides an overview of various approaches: 1) Decision trees (DT), 2) Markov Models (MM), 3) Microsimulation 4) Discrete Event Simulation (DES), 5) Agent based modelling (ABM), and 6) System Dynamics (SD). It provides a unique opportunity to gain insight into alternative modeling techniques and discuss model selection with the several experts. This course will also consider the new results and best practice modelling recommendations of the ISPOR-SMDM Joint Modeling Good Research Practice Task Force.
Description and Objectives: • Introduce the underlying concepts and terminology of DT, MM, Microsimulation, DES, ABM,SD
• Describe fields of application
• Discuss advantages and disadvantages of each modeling approach as compared to the other types
• Provide insight into model selection considering recommendations of the ISPOR-SMDM Modeling Task Force.
This course starts with a short introduction to decision-analytic modeling. Alternative modeling approaches will then be introduced in five sections, each followed by an interactive discussion.
Session 1 This session covers DT and MM(cohort simulation), two widely used methods. MMs are based on a set of health states (state-transition models) and have been applied in decision analyses addressing questions about prevention, diagnosis and chronic diseases.
Session 2 The application of microsimulationin decision analysis allows investigators to model individuals and evaluate heterogeneous populations. Approaches range from state-transition models to discrete-event-simulation and equation-based models. This session gives a general introduction based on their applications in the social sciences, health care and politics.
Session 3 DESis a microsimulation method in which entities (e.g., patients) interact and compete for resources (e.g., hospital beds or organ transplants). We will cover the primary components of DES such as entities, attributes, resources, and queues.
Session 4 ABMis a new approach to modeling autonomous, interacting agents. The fundamental feature of an agent is the capability to make independent decisions. ABMs have been used to examine economic issues and questions in the areas of emerging behaviour and epidemiology. We will cover the role of agents as active model components.
Session 5 SD is a powerful modelling method that involves both qualitative and quantitative approaches. It takes a "whole system" view, demonstrating how a small change in one part of a system can have major unanticipated effects elsewhere, an aspect that is particularly suitable for healthcare applications.