Development and validation of a preference weight multiattribute health outcome measure for rheumatoid arthritis

To develop and validate multiattribute measures for patients with rheumatoid arthritis (RA) to report health states and estimate preference weights.

Survey materials were mailed to 748 patients. Factor analysis, an item response theory-based model, and an internal consistency test were used to identify attributes and evaluate items. Two multiattribute preference weight functions (MAPWF) were constructed. Construct validity of the new measures was then tested.

Four hundred eighty-seven patients returned the survey; 24 items on 6 health attributes were selected to form the new outcomes measure. Two MAPWF were derived with preference weights measured with time tradeoff and visual analog scales as dependent variables. All validity test results were statistically significant.

Our results reveal that the new measures are reliable and valid in assessing health states and associated preference weights of patients with RA.