A Practical Method for Estimating Generalized Risk-Adjusted Cost-Effectiveness Utilities and Willingness-to-Pay...

Area of Science:

• Health Economics
• Decision Analysis
• Biostatistics

Background:

• Standard health utility measures often neglect risk preferences.
• Generalized Risk-Adjusted Cost-Effectiveness (GRACE) incorporates risk preferences but typically requires Visual Analogue Scale (VAS) health state utilities.
• VAS measures may not always be available to analysts.

Purpose of the Study:

• To develop an empirical method for deriving GRACE utilities from readily available Time Tradeoff (TTO) utilities.
• To enable the use of GRACE in cost-effectiveness analyses when only TTO data exists.
• To provide a practical solution for adjusting willingness-to-pay thresholds based on TTO-derived GRACE values.

Main Methods:

• Utilized nationally representative patient-level data containing both VAS and TTO health state measures.
• Estimated a linear regression model to map TTO utilities to VAS utilities.
• Applied published GRACE utility estimates and the derived TTO-to-VAS mapping to create TTO-to-GRACE utility conversions.

Main Results:

• A linear model demonstrated the most effective mapping between TTO and VAS health state utilities.
• TTO utilities closely approximate VAS utilities for moderate health states but show greater error for less moderate states.
• TTO utilities systematically underestimate VAS health utilities for severe health states (TTO < 0.4).

Conclusions:

• A simple linear mapping allows estimation of GRACE utilities using conventional TTO measures.
• Existing studies relying solely on TTO utilities may underestimate the GRACE value of health improvements in severely ill populations.
• This approach provides a method to calculate GRACE and adjust willingness-to-pay thresholds without generating new patient data.