Eliciting benefit-risk preferences and probability-weighted utility using choice-format conjoint analysis.

Area of Science:

• Health Economics
• Decision Analysis
• Patient Preferences

Background:

• Benefit-risk analysis is vital for medical interventions.
• Non-expected utility frameworks are needed to capture complex patient preferences.
• Understanding patient risk tolerance is key to treatment decisions.

Purpose of the Study:

• To apply conjoint analysis to estimate health-related benefit-risk tradeoffs in a non-expected-utility framework.
• To test and estimate nonlinear weighting of adverse-event probabilities.
• To explore implications of nonlinear weighting on maximum acceptable risk (MAR) measures.

Main Methods:

• Conjoint analysis with preference data from 570 Crohn's disease patients.
• Web-enabled conjoint survey with choice tasks involving efficacy benefits and mortality risks.
• Conditional logit maximum likelihood estimation using categorical, simple-weighting, and rank-dependent utility (RDU) models.

Main Results:

• One-parameter probability weighting functions generally performed better than two-parameter functions.
• The simple-weighting model allowed estimation of risk-type-specific weighting parameters.
• The RDU model with a single-parameter weighting function provided robust MAR estimates, ranging from 2.6% to 7.1% for moderate to mild symptom improvement.

Conclusions:

• Quantitative benefit-risk analysis for medical interventions must explicitly account for nonlinear probability weighting.
• Patient preferences exhibit nonlinear weighting of adverse event probabilities.
• Nonlinear weighting significantly impacts measures of risk tolerance like MAR.