Value in Health, the official journal of ISPOR- the professional society for health economics and outcomes research, announced today the publication of research demonstrating that discrete choice experiments (DCEs) are able to predict real-world healthcare choices. The report, "Are Healthcare Choices Predictable? The Impact of Discrete Choice Experiment Designs and Models," was published in the September 2019 issue of Value in Health.
The discrete choice experiment technique, originating from mathematical psychology, is mainstream in marketing, transport, and environmental economics, where it is used to predict individual and collective choices. DCEs have also been introduced in health economics, where they are commonly used for valuing health and nonhealth outcomes, investigating trade-offs between health and nonhealth outcomes, and developing priority setting frameworks. Currently, among other barriers, the lack of evidence about the external validity of DCEs inhibits their greater use in healthcare decision making.
Researchers from The Netherlands and Australia sought to determine whether the number of alternatives in a DCE choice task should reflect the actual decision context, and how complex the choice model needs to be to be able to predict real-world healthcare choices. Six DCEs were used, varying with regard to medical condition (influenza vaccination or colorectal cancer screening) and the number of alternatives per choice task. For each medical condition, 1200 respondents were randomized to one of the DCE formats.
At an aggregate level, the choice mimicking a real-world decision to opt for influenza vaccination and colorectal cancer screening was correctly predicted by a DCE-based model, if scale and preference heterogeneity were taken into account. At an individual level, the use of 3 alternatives per choice task and a heteroskedastic error component model seemed to be most promising, correctly predicting in 93.6% and 97.1% of the cases for vaccination and screening, respectively. Five respondent characteristics consistently explained a part of the observed scale and/or preference heterogeneity: sex, numeracy skill (the ability to understand and work with numbers), decision-making style, general attitude toward the health intervention of interest, and experience with the health intervention of interest.
Our study shows that discrete choice experiments are able to predict choices- mimicking real-world decisions- if at least scale and preference heterogeneity are taken into account. While further research is needed to determine whether this result remains in other contexts, we are pleased that our work has contributed to the evidence regarding the external validity of discrete choice experiments in healthcare decision making."
Esther W. de Bekker-Grob, PhD, Erasmus School of Health Policy & Management, Erasmus University Rotterdam, The Netherlands
Study reveals AI's potential and pitfalls in medical diagnosis