HEALTH ECONOMICS Health Econ. 14: 149–159 (2005) Published online 1 April 2004 in Wiley InterScience (www.interscience.wiley.com). DOI:10.1002/hec.889 OUTCOME VALUATION Testing the internal consistency of the lottery equivalents method using health outcomes Adam Oliver* LSE Health and Social Care, London School of Economics and Political Science, UK Summary The standard gamble has a firm basis in the theory of risk and, for this reason, is the preferred method of health state value elicitation for many researchers. However, it is widely recognised that the use of immediate death as the failure outcome in the standard gamble renders the method insufficiently sensitive for the direct valuation of minor and temporary health states. Consequently, the indirect valuation of minor and temporary health states through a process of ‘chaining’ has been recommended. Unfortunately, a number of researchers have observed internal inconsistency in the standard gamble. That is, for health states that are not thought to be subject to the above mentioned insufficient sensitivity, the values elicited from indirect chained questions quite often significantly and systematically exceed those elicited from the direct procedure. A potential explanation for this is the possible influence of loss aversion when people are asked to weigh certainty against risk. The lottery equivalents method, an alternative value elicitation instrument that also has a firm grounding in the theory of risk, modifies the standard gamble approach of certainty versus risk to one of risk versus risk. The absence of certainty offers reason to suspect that the influence of loss aversion will be diminished (or even removed), and that the lottery equivalents method will thus prove to be internally consistent. The study reported in this article tests the internal consistency of the lottery equivalents method. In a manner similar to that observed with the standard gamble, the results indicate that the internal consistency of the lottery equivalents method is also to some extent compromised. Copyright # 2004 John Wiley & Sons, Ltd. Keywords lottery equivalents; standard gamble; internal consistency; loss aversion Introduction Of the methods that are commonly used to elicit health state values (i.e. the rating scale, the time trade-off and the standard gamble), only the standard gamble, in it being implied from the axioms of expected utility theory (EU), has a firm basis in the theory of risk. Since health care decisions very often involve an element of risk, many researchers therefore hold the standard gamble as the gold standard for such elicitation exercises [1,2]. In its most commonly applied format the standard gamble requires the individual to trade off the certainty of remaining in an intermediate health state with a treatment that offers the chance of regaining full health but that also entails a risk of immediate death. When the risky arm of the standard gamble involves the best and worst conceivable health outcomes (which – as above – are usually taken to be full health and immediate death, respectively), the elicited values are some- times called basic reference values [3,4]. [see Spencer A. The implications of linking questions within the SG and TTO methods. Health Econ, forthcoming]. One possible problem with using immediate death in the standard gamble is that many people Copyright # 2004 John Wiley & Sons, Ltd. Received 12 May 2003 Accepted 7 January 2004 *Correspondence to: LSE Health and Social Care, London School of Economics and Political Science, Houghton Street, London WC2A 2AE, UK. E-mail: a.j.oliver@lse.ac.uk