CHOOSING THE APPROPRIATE MODELING METHOD- CAUSAL MODELING, HEALTH ECONOMIC MODELING, CONSTRAINED OPTIMIZATION OR SIMULATION?
Author(s)
William H Crown, PhD, OptumLabs, Cambridge, USA; Deborah A Marshall, PhD, University of Calgary, Calgary, Canada; Kalyan S. Pasupathy, PhD, Robert D. and Patricia E. Kern Center for the Science of Health Care Delivery, Mayo Clinic, Rochester, USA; Praveen Thokala, MASc, PhD, The University of Sheffield, Sheffield, UK
Presentation Documents
PURPOSE: There are a variety of methods for approaching health economic evaluation problems. Most ISPOR workshops concentrate on presenting a particular method. This workshop will focus on the higher order issue of choosing the correct method to address a specific evaluation problem.
DESCRIPTION: Researchers tend to have expertise in a specific analytic method, or subset of methods, such as causal inference from health econometrics/epidemiology, health economic modeling, constrained optimization or simulation. These methods are often highly complementary. However, analytic opportunities are sometimes lost because deep methodological domain knowledge keeps researchers locked within their own methodological silos. The goal of this workshop is to consider the characteristics of a range of health economic evaluation problems and then show how these characteristics guide the selection of the best analytic method to be utilized. We will also show that these broader categories of methods are not necessarily mutually exclusive. For example, some constrained optimization problems are best solved by simulation. Similarly, discrete event simulation methods are widely used in health economic modeling. Causal modeling methods are often a precursor to estimate the parameters in health economic models or building the structural equations in a simulation model. The workshop will consist of four brief presentations on model selection criteria (10 minutes each) with respect to causal modeling, health economic modeling, constrained optimization, and simulation. These presentations will also address the selection of specific methods within each broad category based upon the characteristics of the research problem being addressed. Following the presentations, the audience will be presented with several problem scenarios and asked to select the appropriate methods. The distribution of answers will then be used to guide a question and answer session. Presented by the ISPOR Optimization Methods in Healthcare Delivery Emerging Good Practices Task Force
Conference/Value in Health Info
2018-05, ISPOR 2018, Baltimore, MD, USA
Code
W11
Topic
Economic Evaluation, Methodological & Statistical Research