Original articlesAre Health States “Timeless”? The Case of the Standard Gamble Method
Introduction
The standard gamble method is based directly on the axioms of von Neumann Morgenstern utility theory [1]. A more detailed discussion of gambles and expected utility theory can be found in Hellinger [2]. As currently recommended for use in health care program evaluation, the standard gamble method provides an individual's preference score or “utility weight” for living in a given health state for the rest of the individual's life [3]. Many researchers interpret this value as independent of time. This interpretation is necessary in order to use the values obtained as weights in quality-adjusted life-year (QALY) calculations. These time-independent or “timeless” values are also used in the development of tables in which different health states are organized in declining order, regardless of the time spent in each health state (e.g., [4]).
Pliskin et al. [5] identified three conditions (in addition to the von Neumann Morgenstern axioms) under which QALYs are a von Neumann Morgenstern utility function—constant proportional trade-off, utility independence, and risk neutrality. According to the constant proportional trade-off assumption, the proportion of remaining life that one would trade off for a given quality improvement is independent of the amount of remaining life. This assumption allows us to estimate “QALY-weights” using the time trade-off method and then use them to compute QALYs. Stiggelbout et al. [6] tested the constant proportional trade-off assumption using a time trade-off experiment and found that this assumption was violated. They found that the “utility-weights” for longer time horizons were smaller than those for shorter time horizons.
The utility independence assumption implies that the ratio of the utility of spending a period of time T in two health states does not depend on T. That is where U(Qi,Ti) is the utility of being in health state Qi for a time period Ti. [7]. If the utility independence assumption holds, the response obtained from a standard gamble question will be the same regardless of the time horizon in the question. Similar to the constant proportional trade-off assumption for the time trade-off technique, the utility independence assumption allows us to estimate “QALY-weights” using the standard gamble method and then use them to compute QALYs. We report here the results of an experiment in which we tested the utility-independence assumption using the standard gamble method.
As a part of a larger survey, we administered standard gamble questions to estimate a preference score or utility weight for the severe pain associated with herpes zoster (or shingles). We asked each respondent standard gamble questions using two different time horizons and tested whether the responses to the two questions were different.
Section snippets
Methods and data
To estimate preference scores for severe pain, we administered two standard gamble questions varying in time horizon to respondents between the ages of 65 to 70 years. We focused on this age group because the incidence of herpes zoster increases with age. Before administering the survey, we conducted focus groups with individuals who had experienced shingles to refine our description of severe pain. We then pretested the standard gamble questions and found that individuals did not have
Results
Of the 118 individuals who participated in the survey, 114 (97%) completed the survey and provided valid responses. The interviews were completed over a 1-week period. The average age of the respondents was 68 years. About 12% had prior shingles experience, and more than 80% had an acquaintance who had experienced shingles.
Table 1 presents the results of our experiment. Only 29 (25%) of the 114 respondents provided preference scores that were not dependent on the time horizon specified in the
Discussion
Are the preference scores for the severe pain health state time-independent or “timeless”? Our results suggest that, even though the mean responses to the two standard gamble questions are not significantly different, the preference scores are not independent of time for the majority of respondents. The utility independence assumption concerns utility at the individual level, and not at the population level. Thus, in our study the utility independence assumption is violated by a majority of the
Acknowledgements
Burroughs Wellcome Co. provided funding for this study. The authors thank Suzanne P. Gambin of Research Triangle Institute (RTI) for her survey administration and data management assistance on this study. The authors also thank Josephine A. Mauskopf, formerly of Glaxo Wellcome Inc. and now at RTI, for her useful comments and suggestions throughout the study.
References (16)
- et al.
The ‘utility’ of the time trade-off method in cancer patientsFeasibility and proportional trade-off
J Clin Epidemiol
(1995) Social preferences for health statesAn empirical evaluation of three measurement techniques
SocioEcon Plann Sci
(1976)- et al.
Valuing outcomes in health careA comparison of willingness-to-pay and quality-adjusted life-years
J Clin Epidemiol
(1998) - et al.
Theory of Games and Behavior
(1953) Expected utility theory and risky choices with health outcomes
Med Care
(1989)The standard gamble methodWhat is being measured and how it is interpreted
Health Serv Res
(1994)- et al.
“Utilities and quality-adjusted life-years
Int J Technol Assess Health Care
(1989) - et al.
Utility functions for life-years and health status
Oper Res
(1980)
Cited by (38)
A Head-to-Head Comparison of the Standard Quality-Adjusted Life Year Model With the Annual Profile Model
2021, Value in HealthCitation Excerpt :The alternative anchor state aims to increase the sensitivity of the scale for temporary conditions, but it could introduce bias or error through the rescaling procedure. More importantly, the chained TTO procedure still does not address the dependence between time and utility.8,37 Some methodological issues and limitations need to be addressed.
Utility scores for vesicoureteral reflux and anti-reflux surgery
2015, Journal of Pediatric UrologyCitation Excerpt :As might be expected, we found a substantially lower preference for the postoperative health state, with a utility of 0.71. Short-term health states in children typically have lower utility values compared with chronic health states [11–13]. However, the overall effect of such short-term disutility is low, as scores are averaged out over the entire year; thus, the 6-week postoperative utility score of 0.71 averages out to a net utility score of 0.97 for the full year (assuming perfect health for the remaining 46 weeks).
Estimating utility values for vesicoureteral reflux in the general public using an online tool
2014, Journal of Pediatric UrologyWillingness to pay in dermatology: Assessment of the burden of skin diseases
2012, Journal of Investigative DermatologyCitation Excerpt :Other authors have helped elucidate components of patient preference that may account for weak correlations between WTP and utilities. Bala et al. (1999b) demonstrated the importance of the duration of an illness. The authors compared utilities associated with 10 vs. 20 years of shingles pain and demonstrated that a smaller amount of time with the condition yielded less burden of disease (Bala et al., 1999b).