Testing Nonmonotonicity in Health Preferences

Area of Science:

• Health Economics
• Decision Science
• Psychometrics

Background:

• Previous research suggests individuals may prefer shorter durations of poor health up to a maximum endurable time (MET).
• Understanding preferences for health duration is crucial for health policy and economic modeling, particularly for quality-adjusted life years (QALYs).

Purpose of the Study:

• To empirically test monotonicity in life duration preferences across various health states.
• To investigate the prevalence of maximum endurable time (MET) preferences and preference reversals.

Main Methods:

• Employed two ordinal tasks: direct choices and rankings of health episodes combining EQ-5D-3L states and durations.
• Recruited 90 participants for experimental sessions to elicit and analyze preference patterns.
• Quantified monotonicity violations and preference reversals between the choice and ranking tasks.

Main Results:

• Significant rates of nonmonotonic preferences were observed, ranging from 49% (choices) to 71% (rankings) within subjects.
• Mean nonmonotonicity rates per health state were 22% (choices) and 34% (rankings).
• Evidence of preference reversals (1.5%-33%) and scarce transitivity violations were found.

Conclusions:

• Nonmonotonic preferences, particularly the MET phenomenon, are prevalent in mid-range health states.
• Observed preference reversals suggest potential limitations in using similar response scales for different tasks.
• Findings challenge the validity of multiplicative QALY models and highlight the need for more nuanced preference elicitation.