FD1 APPLYING CONJOINT ANALYSIS AND DISCRETE CHOICE EXPERIMENTS IN HEALTH

Sunday, October 23, 2011: 9:00 AM
Columbus Hall IJ (Hyatt Regency Chicago)
Course Type: Full Day
Course Level: Intermediate

Format Requirements: This hands on, full day workshop will inform participants of the fundamentals of conjoint analysis methods, and will include an overview of recent advances in experimental design and statistical analysis. The day will start with an overview of the theory underpinning conjoint analysis; this will be followed by a review of the methods necessary to successfully execute a study. Particular attention will be given to basic aspects of experimental design and statistical analysis. The workshop will conclude with a practical application of estimating willingness to pay and its use in technology assessment. Participants are not required to bring personal computers. No particular software will be endorsed, although applications will be examined using excel, Stata and SAS. This workshop is intended for researchers, practitioners and policy makers who have a working knowledge of statistics and health outcomes research, and who are interested in developing patient-focused preference-based values of health outcomes.

Background: The application of stated-preference conjoint analysis and discrete choice experiments in health has rapidly increased over the past decade. Furthermore, health care regulators are increasingly requiring scientific methods to quantify the values patients and stakeholders attach to health care goods. Conjoint analysis and discrete choice experiments elicit preference data from scientifically designed choice tasks that describe competing scenarios, where each scenario is defined on a number of attributes and levels. Statistical methods are then used to decompose respondents’ preferences into marginal valuations for the attribute levels. From an applied perspective, these methods have been used to estimate preference weights for patient reported outcomes, and to explore the tradeoffs between attributes, including willingness to pay and equivalences between benefits and risk. Conjoint analysis has also been used to study medical decision-making by examining the factors that determine clinicians’ recommendations and the concordance between the values of patients and clinicians.

Description and Objectives: This workshop will provide participants with the basic theoretical and practical background necessary to understand, evaluate, and interpret a stated preference choice experiment. The practical, applied aspect of the course will use recent examples of conjoint analysis and discrete choice experiments to trace the key phases that are essential to conducting a successful conjoint analysis study. Specifically, the course will take participants through the elements of i) defining the research question; ii) specifying the attributes and levels; iii) constructing the choice tasks; iv) experimental design to construct hypothetical stimuli; v) preference elicitation format; vi) instrument design; vii) data collection plan; viii) statistical analyses; ix) results and conclusions; and x) study presentation. By the end of this course participants will:

  • Become familiar with the foundations of conjoint analysis and gain an understanding of the theory underpinning choice based methods; 
  • Gain a practical grounding in the necessary steps for conducting a conjoint analysis or discrete choice experiment, including experimental design and statistical analysis; and
  • Gain practical experience in calculating welfare estimates that can be used in formal technology assessment, including the calculation of willingness to pay measures used in cost benefit analysis.

 

Course Director:
Dean A. Regier, PhD
Course Faculty:
Deborah Marshall, PhD and John F.P. Bridges, PhD