McClure, N.

Minimally Important Difference Estimates for the EQ-5D: A Systematic Review
McClure, N.S., Jin, X., Al Sayah, F., Le, C., Baghbaninaghadehi, F., Ohinmaa, A., Johnnson, J.A.

The EQ-5D is a commonly used generic measure of health-related quality of life (HRQL) that has various applications including clinical and cost-effectiveness studies, population health surveys and routine health outcome measurement programs. When interpreting results obtained with the EQ-5D, as with other patient-reported outcome measures, a common approach is to consider a minimally important difference (MID) to help in interpreting the magnitude of difference or changes (improvement or deterioration) in HRQL as measured by this instrument. This assessment of the magnitude at a group level tells us whether or not the observed difference or change in EQ-5D may be considered important to respondents or decision makers. The aims of this systematic review of the literature on MID for the EQ-5D instrument are to: 1. Identify studies that estimated an MID value (or range of values) for any version (3L, 5L and Y) and component of the EQ-5D (dimensions, index and VAS scores) and document the clinical context of the MID. 2. Examine and summarize methods used to estimate MIDs for any version/component of the EQ-5D. 3. Summarize EQ-5D MID estimates and identify methodological strengths and limitations of existing literature, highlighting areas of need for further research.

Multiple journal databases were searched with no date or language restrictions. Primary methodological studies (randomized or observational) that report on the empirical development of an MID estimate for the EQ-5D instrument were identified through first-level title and abstract screening and second-level full-text screening by two teams of reviewers.

In total 2,255 records were identified from database searches and other sources. After first-level screening, 534 full-text articles were assessed for eligibility of which 63 studies met inclusion criteria with an overall kappa between reviewer teams of 0.79. Data extraction is currently in progress.

This study will help support end-users of the EQ-5D.