Discordance in Utility Measurement in Persons with Neurological Conditions: A Comparison of the SF-6D and the HUI3

Abstract

Objectives: To examine the extent of disagreement in estimated utility between the six-dimensional health state short form (SF-6D) and the Health Utilities Index-Mark 3 (HUI3) in Canadians with neurological conditions and how discordance varied by participant and neurological condition attributes.

Methods: The study analyzed cross-sectional survey data from the Living with and Managing the Impact of a Neurological Condition Study. Self-reported data were collected on the burden and impact of neurological conditions on participants' everyday lives. Disagreement was examined by comparing utility distributions, paired t tests of the means, Spearman ρ correlations, intraclass correlations, and Bland-Altman plots. Associations between participant and neurological condition attributes and utility differences were assessed using multiple regression models.

Results: Disagreement between the SF-6D and the HUI3 was substantial, with a mean utility difference of 0.15 (95% confidence interval 0.13-0.17). An intraclass correlation coefficient of 0.41 suggests only marginal agreement. The Bland-Altman plot and regression analysis showed systematic variation in utility difference associated with level of utility. Depending on the level of utility, utility differences between the SF-6D and the HUI3 shift in magnitude and direction. The pattern of disagreement did not vary substantially by participant or neurological condition characteristics.

Conclusions: The SF-6D and the HUI3 provide inconsistent evaluations of utility in persons with neurological conditions. The magnitude and direction of differences in estimated utility are strongly associated with level of utility. Depending on the health status of the sample, the SF-6D and the HUI3 could provide widely contradictory utility estimates. A concern is that utility scores, and hence potential evaluations and health care decisions, may vary simply according to the choice of instrument.

Keywords: Health Utilities Index; Neuro-QoL; SF-6D; discordance; health-related quality of life; neurological conditions; preference-based quality of life; utility measures.