WW18: THE ROLE OF HEALTH STATE EXPERIENCE IN LONGITUDINAL CONJOINT ANALYSIS OF PATIENT PREFERENCES

Abstract

OBJECTIVES: Estimates developed from preference-based pharmacoeconomic studies may be influenced by a patient's experience with particular health states. This workshop addresses methodologies and practical aspects in designing and analyzing conjoint style pharmacoeconomic studies that control for and evaluate the importance of these factors. PARTICIPANTS WHO WOULD BENEFIT: Researchers who are involved in the design, execution, or commission of conjoint studies of patient preferences. Estimates developed from preference-based pharmacoeconomic studies may be influenced by a patient's experience with particular health states. This workshop addresses methodologies and practical aspects in designing and analyzing conjoint style pharmacoeconomic studies that control for and evaluate the importance of these factors. Pharmacoeconomics research has seen an increasing number of studies that estimate the relative importance patients place on individual pharmaceutical attributes. Attributes may include disease and treatment related health states as well as drug features such as administration frequency and cost. Conjoint analysis is the primary tool employed in these studies. Currently, these studies do not recognize the possible importance of patient experience and current health status on preference estimates. For example, experience with a particular health-state may influence a study subject's willingness-to-pay to avoid that state. Explicitly considering the role of such factors may provide important information such as indications of a study's validity. At worst, ignoring the role of experience in preference formation could bias study results. In this workshop, participants will learn methods for incorporating health-status data into conjoint style stated-preference studies. Participants will consider a variety of potential situations and solutions through an interactive exercise. However, the workshop will focus primarily on employing the longitudinal health-status and preference data that is available when a conjoint survey is administered over the course of a randomized controlled trial (RCT). The presenters will use a recent study to motivate an interactive discussion of methodology, survey design, and analytical techniques in this context.