Modelling non-demanders in choice experiments

Area of Science:

• Health Economics
• Behavioral Economics
• Public Health

Background:

• Discrete choice experiments (DCEs) are valuable for studying healthcare preferences when real-world data is scarce.
• Hypothetical scenarios in DCEs must accurately reflect real-world decision-making contexts.
• The 'opt-out' option, where individuals decline all available choices, is a common real-world scenario.

Purpose of the Study:

• To explore the methodological challenges of modeling "opt-out" choices within discrete choice experiments.
• To investigate women's preferences for cervical screening services, incorporating the possibility of opting out.
• To enhance the realism and applicability of DCEs in healthcare preference research.

Main Methods:

• Utilized a discrete choice experiment (DCE) design.
• Incorporated an opt-out option to simulate non-demand for services.
• Focused on modeling women's preferences for cervical screening services.

Main Results:

• The study identifies and discusses key issues in statistically modeling DCE data that includes an opt-out choice.
• Findings provide insights into how to better represent non-participation or non-demand in healthcare preference studies.
• The research highlights the importance of accounting for the opt-out option in understanding patient choices.

Conclusions:

• Discrete choice experiments can be adapted to include and model opt-out scenarios, improving their utility in healthcare.
• Accurate modeling of non-demand is crucial for a comprehensive understanding of patient preferences and healthcare service utilization.
• This approach offers a more robust method for eliciting preferences in public health interventions like cervical screening.