A review of the psychometric properties of generic utility measures in multiple sclerosis.

OBJECTIVE: The reliability and validity of generic utility measures have not yet been summarized in people with multiple sclerosis (MS). It is important to assess the psychometric properties of these measures, to ensure that the values obtained by the scoring system are valid for interpretation and utilization by clinicians, researchers and policy makers. Therefore, the objective of this review was to summarize the evidence from published literature on the psychometric properties of generic utility measures in MS.

METHODS: A structured literature search was conducted by using multiple electronic databases. All potentially relevant abstracts and full-text articles were read to identify publications that may be eligible for inclusion in the review. A meta-analysis was conducted to combine correlation coefficient values for convergent validity. The Schmidt-Hunter method, a weighted mean of the correlation coefficient values, was used. Heterogeneity, the percentage of total variation across studies that is due to between-study differences rather than chance, was assessed using the I (2) statistic.

RESULTS: The following generic utility measures were identified: the EQ-5D (n = 9)/EQ-5D-5 Level (EQ-5D-5L) (n = 1), followed by the Health Utilities Index Mark 3/2 (HUI2/HUI3) (n = 3), the SF-6D (n = 2), the Assessment of Quality of Life (AQOL) (n = 2), and the Quality of Well-Being (QWB) scale (n = 1). Ceiling and floor effects were present for the EQ-5D and the SF-6D, but not for the HUI3. The EQ-5D, the SF-6D and the HUI3 demonstrated excellent reliability. In terms of discriminative ability, the SF-6D and the QWB scale were not able to differentiate between moderately and severely disabled MS patients, and the EQ-5D was not able to differentiate between those who were mildly and moderately disabled. The AQOL and the HUI3, on the other hand, demonstrated good discriminative ability, as both measures were able to differentiate between all levels of disability. As for convergent validity, the HUI2/HUI3 were highly correlated (r = 0.7) against measurement instruments that evaluated impairments such as disease severity, ambulation and manual dexterity. The EQ-5D, SF-6D and the QWB scale demonstrated small to moderate correlations (r = 0.4) against instruments evaluating impairments, and slightly stronger correlations against measures of activity limitations/participation restrictions and health-related quality of life (HRQL) (r = 0.6).

CONCLUSION: To our knowledge this is the first study to review the validity and reliability of generic utility measures in MS. The HUI3 demonstrated the strongest psychometric properties when compared with other utility measures. However, the HUI3 only measures impairment and excludes important components of HRQL such as participation restrictions. The EQ-5D, the SF-6D and the QWB scale, on the other hand, do include items on participation. However, these measures demonstrated a lack of content validity in MS by missing certain domains that were important to the disease, as well as difficulty in differentiating between different levels of disability. The addition of MS-specific 'bolt-ons' to generic utility measures and the development of an MS specific utility measure are possible areas of exploration for future research.