Inconsistency and health state valuations

Area of Science:

• Health Economics
• Psychometrics
• Survey Methodology

Background:

• Comparing health state valuation methods often relies on aggregate scores after excluding inconsistent respondents.
• Assessing overall respondent logical consistency is vital for data validity.
• Factors influencing inconsistency include health state descriptions, questionnaire administration, and respondent characteristics.

Purpose of the Study:

• To analyze inconsistency rates in EuroQol health state valuations using the visual analogue scale (VAS) method.
• To examine the impact of study design variations on respondent inconsistency.
• To identify factors associated with higher or lower inconsistency rates.

Main Methods:

• Analysis of inconsistency rates from two distinct studies eliciting EuroQol health state valuations.
• Utilized the visual analogue scale (VAS) method for health state valuation.
• Compared inconsistency rates across different study designs and questionnaire variants.

Main Results:

• Inconsistency rates were higher in interviewer-administered surveys compared to postal surveys, potentially due to response bias.
• A positive relationship was observed between respondent age and inconsistency rates.
• A negative relationship was found between educational attainment and inconsistency rates.

Conclusions:

• Survey administration method significantly impacts respondent consistency in health state valuation.
• Demographic factors such as age and education play a role in the logical consistency of health state valuations.
• Understanding and mitigating inconsistency are key for reliable health state valuation data.