OBJECTIVE. To compare in patients with fibromyalgia (FM)
utilities derived by rating scale and standard gamble methods;
to gain insight into construct validity by relating utility
values to other outcome measures; to assess the sensitivity to
change of utilities.
METHODS. A total of 73 patients with FM
were randomized into one of 3 groups: low impact fitness
training, biofeedback, or controls. At baseline and after 6 mo
the Maastricht Utility Measurement Questionnaire was applied.
By means of both the rating scale and standard gamble method
patients were asked to value their own health status.
Construct validity of patient utility measurements was
evaluated by Spearman correlation and multiple regression of
baseline values with pain, stiffness, patient's global
assessment, Sickness Impact Profile (SIP), modified Health
Assessment Questionnaire and Arthritis Impact Measurement
Scale (AIMS). Sensitivity to change was assessed against
changes in these outcomes.
RESULTS. Rating scale utilities
correlated significantly (p < 0.05) with patient's global
assessment (rs = 0.53), pain (rs = -0.47), SIP (rs = -0.43),
and with 9 of 11 dimensions of the AIMS (rs ranging from 0.23
to 0.62). Standard gamble utilities correlated significantly
with mobility, pain, and arthritis impact of the AIMS scale
(rs from 0.22 to 0.36) and with pain by visual analog scale
(rs = -0.24) and patient's global assessment (rs = 0.32).
Multiple regression analysis showed that patient's global
assessment explained 41% (rating scale) and 10% (standard
gamble) of total variance in baseline utilities. Also, 16% of
the variance in change in rating scale utility values was
explained by changes in patient's global assessment. In
contrast, variance of changes in standard gamble utility
values was not explained significantly by changes in other
CONCLUSION. Rating scale utilities
correlated more strongly with disease outcome measures than
standard gamble utilities. Also, construct validity for the
rating scale was better than for the standard gamble. In FM,
utility measurement is sensitive to the method chosen to
elicit patient priorities.
Bakker C, Rutten M, van Santen-Hoeufft M, Bolwijn P, van Doorslaer
E, Bennett K, van der Linden S