LAUSR

Earlier this year (2020), details of a new preference-based, health-related quality of life (HRQoL) instrument—the SF-6D version 2 (SF-6Dv2)—were published in Medical Care [1, 2]. Although the concept of the SF-6Dv2 has been around for a number of years [3, 4], the recent publications provide the first comprehensive reports regarding the development of the classification system [1] and the estimation of a value set (for the UK) [2]. A Simplified Chinese version of the SF-6Dv2 classification system has also been developed [5]. The emergence of any new preference-based instrument raises questions (or should raise questions) about the contribution to the field, psychometric properties, relative performance when compared with other candidate instruments, and the implications for health technology assessment. When the new instrument is closely related to a previous instrument, there are other important considerations, with the most pertinent being whether the new instrument is a replacement for or an alternative to the earlier version. For example, there has been debate as to whether the EQ5D-5L should replace the EQ-5D-3L [6–8] and, to some extent, these decisions are beyond the control of instrument developers. Developers of the SF-6Dv2 (two of whom are co-authors of this paper) have been explicit in stating that the new classification system is an improvement over the original, based on the use of more sophisticated methods for dimension formation and item selection, incorporation of cross-cultural considerations, and simplification of the wording of dimension levels to support valuation [1]. Psychometric comparisons between the SF-6Dv2 and other preference-based HRQoL instruments (including the SF-6D) will inevitably follow. Such assessments are typically dominated by analyses where the index score is the unit of interest, and the UK valuation study for the SF-6Dv2 provides a clear indication that large differences in SF-6Dv2 and SF-6D health state values are to be expected [2]. For example, the difference in the minimum value of the recommended UK value sets for the two instruments is 0.865 (− 0.574, SF-6Dv2; + 0.291, SF-6D), which is greater than the entire scoring range of the UK value set for the SF-6D. However, understanding the differences in the respective classification systems and the consequential implications for how individuals can describe impairment across dimensions are also important considerations in this field of research [9, 10]. In this Commentary, we focus exclusively on the classification systems of the SF-6Dv2 and SF-6D, providing a high-level overview of the differences between the two instruments, and illustrate the implications of these differences for health state descriptions using data collected in a multi-country cross-sectional survey comprising over 8000 participants. Through the provision of insights into how health state descriptions differ between the SF-6Dv2 and SF-6D classification systems, our intention is to offer a useful precursor to further evaluative research regarding the SF-6Dv2. Electronic supplementary material The online version of this article (https ://doi.org/10.1007/s4027 3-020-00957 -9) contains supplementary material, which is available to authorized users.