Eliciting Health State Utilities Using Paired-Gamble Methods: The Role of the Starting Point.

Area of Science:

• Health economics
• Decision analysis
• Psychophysics

Background:

• Standard gamble methods can be affected by the certainty effect.
• Paired-gamble methods were developed to mitigate the certainty effect.
• Starting-point effects are a potential bias in decision-making research.

Purpose of the Study:

• To investigate the influence of starting-point effects on paired-gamble methods.
• To determine how derived utilities vary with the probabilities of stimulus lotteries.
• To assess the robustness of paired-gamble methods against probability variations.

Main Methods:

• A sample of 455 individuals from the Spanish general population participated.
• Health states were valued using face-to-face interviews and paired-gamble techniques.
• Subjects were randomly assigned to subgroups with varying stimulus gamble probabilities.
• Nonparametric tests and interval regression models were employed for analysis.

Main Results:

• No significant differences were found in the probability of a health state being considered worse than death across subgroups.
• Variations in the stimulus gamble significantly altered utility distributions; higher full health probability led to higher elicited utility.
• Regression analysis indicated starting-point effects under expected utility.
• The impact of starting-point effects varied depending on the reference point used (death, evaluated health state, or full health).

Conclusions:

• Paired-gamble methods may be susceptible to starting-point effects.
• These effects are generally small and can disappear when prospect theory is applied.
• Using full health as the reference point in prospect theory analysis mitigates observed starting-point effects in paired-gamble methods.