Should We ‘Drop Dead’ from Health State Valuation?

Article by: Chris Sampson, David Parkin and Nancy Devlin

Resource allocation decisions in health care may involve trade-offs between improving people’s quality of life and improving their longevity. The quality-adjusted life year (QALY) is used to recognise this trade-off. But is it necessary to use ‘dead’ as an anchor in health state valuation? In a new OHE Research Paper, Chris Sampson, David Parkin, and Nancy Devlin argue the case for ‘Dropping Dead’ as the anchor for health state values. And the authors would like your feedback.

By convention, the values developed to accompany generic measures of health-related quality of life, such as the EQ-5D, are anchored on a scale where 1 = full health and 0 = dead. As a result, stated preference methods used to value health states – such as time trade-off and discrete choice experiments – involve consideration of the state ‘dead’. Some health states are judged by respondents to be worse than the ‘dead’ state, meaning that they need to be assigned negative values.

In recent years, numerous researchers have identified challenges arising from anchoring at ‘dead’. Despite considerable efforts devoted to developing new methods, the need to identify negative values continues to cause fundamental problems. For instance, there is no agreement on what the lowest negative value should be.

In our new paper, we challenge the assumption that anchoring health state values at ‘dead = 0’ is a necessary condition for values to be used to estimate QALYs in cost-effectiveness analysis.

We consider five propositions:

  1. anchoring at ‘dead’ is not required by theories of scale measurement and utility;
  2. calculating QALY gains does not require a distinction between states better than and worse than dead;
  3. cost-effectiveness analysis does not require that ‘dead’ has a value relative to health states;
  4. using ‘dead’ as an anchor causes problems that make studies difficult to conduct and their results difficult to interpret; and
  5. there are alternative states to ‘dead’ that exhibit favourable properties for anchoring.

Informed by a narrative review of the literature, we find support for each proposition.

Anchoring health state values at ‘dead’ was an arbitrary choice made early in the development of health state valuation methods. Its use as an anchor has gone unchallenged for a quarter of a century and has become part of the accepted wisdom in health economics. We provide arguments to show that the use of ‘dead’ as an anchor is not only unnecessary but also undesirable because of the methodological and conceptual problems it causes. We conclude that, in valuing health states, researchers should ‘drop dead’.

But what do you think?

To ‘drop dead’ would be a relatively straightforward methodological change, with the ‘dead’ anchor simply being replaced with an alternative state (such as ‘worst health state imaginable’). However, we recognise that such a change raises fundamental questions. Therefore, we would like to hear your views. For example, we would like you to consider:

  • Do you agree that it is time to ‘drop dead’?
  • For any of our five propositions, would you have reached a different conclusion?
  • Could the inclusion of ‘dead’ be important for reasons that we have not considered?
  • Could there be downsides to dropping ‘dead’?
  • What are the key unanswered questions?

Please let us know what you think by visiting ResearchGate and adding a comment to our paper’s webpage.

This project was funded by the EuroQol Research Foundation. The views expressed are those of the authors and do not represent the views of the EuroQol Group.


Sampson, C., Parkin, D. and Devlin, N., 2020. Drop Dead: Is Anchoring at ‘Dead’ a Theoretical Requirement in Health State Valuation? OHE Research Paper, London: Office of Health Economics. Available at:

Related Research

Shah, K.K., Ramos-Goñi, J.M., Kreimeier, S. and Devlin, N.J., 2020. An exploration of methods for obtaining 0 = dead anchors for latent scale EQ-5D-Y values. The European Journal of Health Economics, 21(7), pp.1091–1103. 10/ghmxn9.

Devlin, N., Buckingham, K., Shah, K., Tsuchiya, A., Tilling, C., Wilkinson, G. and van Hout, B., 2013. A comparison of alternative variants of the lead and lag time TTO. Health Economics, 22(5), pp.517–532. 10/gd83zr.

Posted in EQ-5D and PROMs, Health Technology Assessment, Research | Tagged Research Papers