Estimating preference-based single index measures for dementia using DEMQOL and DEMQOL-Proxy.

Area of Science:

• Health Economics
• Psychometrics
• Gerontology

Background:

• Condition-specific measures in dementia care lack preference-based utility values needed for cost-per-quality-adjusted life-year (QALY) analyses.
• A gap exists in economic evaluation methods for dementia, hindering comprehensive cost-effectiveness assessments.

Purpose of the Study:

• To estimate two preference-based single index measures: DEMQOL-U (for mild-to-moderate dementia) and DEMQOL-Proxy-U (for all dementia severities).
• To address the limitation of condition-specific measures in dementia by incorporating societal preferences for economic evaluations.

Main Methods:

• Valuation studies involving 593 general population members using the time trade-off technique.
• Multivariate regression models fitted to valuation data to derive preference weights for DEMQOL-U and DEMQOL-Proxy-U.
• Application of estimated weights to a large, clinically representative sample.

Main Results:

• Observed time trade-off values ranged from 0.18 to 0.95 for DEMQOL-U and 0.33 to 0.96 for DEMQOL-Proxy-U.
• Main effects models using individual-level data demonstrated superior performance.
• Inconsistencies in coefficient estimates for DEMQOL-Proxy-U were identified and addressed through model refinement.

Conclusions:

• Developed preference-based single index measures (DEMQOL-U, DEMQOL-Proxy-U) facilitate economic evaluations using QALYs for the full spectrum of dementia severity.
• These measures enhance the utility of cost-effectiveness analyses in dementia care.
• Future research will focus on integrating these utilities into cost-effectiveness models.