Mapping QLQ-C30, HAQ, and MSIS-29 on EQ-5D

Area of Science:

• Health Economics
• Psychometrics
• Clinical Outcomes Research

Background:

• Condition-specific instruments are used to estimate utility values for economic evaluations via mapping onto the EQ-5D.
• The predictive quality of mapping functions varies, and not all condition-specific measures are suitable for this purpose.
• This study evaluated mapping functions for QLQ-C30, HAQ, and MSIS-29 onto the EQ-5D.

Purpose of the Study:

• To compare the predictive quality of mapping functions derived from QLQ-C30, HAQ, and MSIS-29 onto the EQ-5D.
• To assess the statistical and clinical validity of these mapping functions.
• To explore the relationship between instrument correlation and mapping function predictive quality.

Main Methods:

• Utilized four datasets containing both EQ-5D and condition-specific measures.
• Developed ordinary least squares regression equations for mapping QLQ-C30 (multiple myeloma, non-Hodgkin lymphoma), HAQ (early arthritis), and MSIS-29 (multiple sclerosis) onto EQ-5D.
• Assessed predictive quality using root mean square error (RMSE) and mean absolute error (MAE), and evaluated discrimination among clinical subgroups.

Main Results:

• The QLQ-C30 exhibited the highest correlation with EQ-5D items.
• The QLQ-C30 mapping achieved the best average %RMSE (10.9%), followed by HAQ (12.2%) and MSIS-29 (13.6%).
• Mapping functions generally predicted mean EQ-5D utilities accurately and discriminated between clinical groups, with the exception of the HAQ model.

Conclusions:

• Mapping functions appear suitable for estimating EQ-5D utilities in economic evaluations.
• Lower correlations between instruments result in reduced predictive quality of mapping functions.
• Employing additional validation tests beyond statistical error measures enhances the assessment of predictive quality.