2012 HSR&D/QUERI National Conference Abstract
2007 — Agent-Based Simulation in Health Services Research
Rubin MA, VA Salt Lake City Healthcare System;
Agent-based simulation modeling (also called individual-agent modeling) is a relatively new type of computer simulation with unique features compared with other forms of modeling. In an agent-based model, the system being studied is modeled as a collection of autonomous decision-making entities (agents), which exist and interact within an environment based on a set of rules or behaviors. In this sense, agent-based simulation is well-suited for modeling, understanding, and experimenting with complex systems such as healthcare settings. Workshop participants will be introduced to agent-based simulation, with special emphasis on its historical evolution and its unique advantages over other forms of modeling. Attendees will learn about how agent-based simulations are created and how they can be applied to health services research in a variety of ways, such as for cost analysis, what-if experimentation, and even decision support.
This workshop will: 1) Provide an introduction to agent-based simulation modeling, presenting a historical perspective and comparing it with other methods of simulation while summarizing its unique advantages; 2) Present different applications of agent-based simulation to health services research, including its use as a tool for understanding complex systems, guiding policy development, and for novel forms of decision-support; 3) Provide a demonstration of an agent-based model our group has created to simulate and understand hospital transmission of drug-resistant organisms such as Clostridium difficile; and 4) Explore the interest in creating a training workshop series for VA health services researchers and in forming an HSR&D agent-based modeling users group.
All health services researchers interested in learning about the capabilities and applications of agent-based computer simulation, and the potential for using this methodology themselves.
Assumed Audience Familiarity with Topic:
This workshop is designed to be introductory, without special preparation required. Background knowledge of computer modeling or programming is helpful but not required.