Regulatory authorities in many countries require that societal preferences are used when health (dis)benefits are reported in terms of quality-adjusted life-years (QALYs). In the United Kingdom the NICE reference case, as set out in its published technical guidance, cites EQ-5D as the requisite health-related quality of life (HrQoL) system and Time Trade-Off (TTO) as the preferred method for eliciting societal values. This stipulation is simple to assert, but virtually impossible to justify and/or operationalise in practice. Nevertheless this has been the UK default position for many years and has established a de facto national “norm”. These issues, however, are global in nature and common to economic evaluation of healthcare in all countries. The UK “preference” for TTO is no more than that, for no scientific case has been made for rejecting Standard Gamble (SG), commonly acknowledged to yield systematically different estimates of utility. Both methods cannot be correct - one (at least) must be in error. It is patently absurd to consider them as commensurable equivalents in QALY calculations. In principle, a similar difficulty arises as new value sets are published, as will be the case in respect of the 5-level version of EQ-5D. Cost-utility analysis reported in the literature reveals a 10-fold difference in incremental benefits (change from baseline) when EQ-5D/HUI/SF-6D are used to compute QALYs, sufficient to reverse the location of an ICER with respect to any threshold. The central issue is that of updating the choice of a definitive value set for reference case analysis. This paper argues for a decision-centric approach in which a new metric may only be adopted if its use in measuring incremental effectiveness yields results that are consistent with those based on the existing reference standard. The argument is exemplified through the analysis of EQ-5D in published studies.