Testing the feasibility of eliciting preferences for health states from adolescents using direct methods

Abstract

BackgroundMeasuring adolescents’ preferences for health states can play an important role in evaluating the delivery of pediatric healthcare. However, formal evaluation of the common direct preference elicitation methods for health states has not been done with adolescents. Therefore, the purpose of this study is to test how these methods perform in terms of their feasibility, reliability, and validity for measuring health state preferences in adolescents.MethodsThis study used a web-based survey of adolescents, 18 years of age or younger, living in the United States. The survey included four health states, each comprised of six attributes. Preferences for these health states were elicited using the visual analogue scale, time trade-off, and standard gamble. The feasibility, test-retest reliability, and construct validity of each of these preference elicitation methods were tested and compared.ResultsA total of 144 participants were included in this study. Using a web-based survey format to elicit preferences for health states from adolescents was feasible. A majority of participants completed all three elicitation methods, ranked those methods as being easy, with very few requiring assistance from someone else. However, all three elicitation methods demonstrated weak test-retest reliability, with Kendall’s tau-a values ranging from 0.204 to 0.402. Similarly, all three methods demonstrated poor construct validity, with 9–50% of all rankings aligning with our expectations. There were no significant differences across age groups.ConclusionsUsing a web-based survey format to elicit preferences for health states from adolescents is feasible. However, the reliability and construct validity of the methods used to elicit these preferences when using this survey format are poor. Further research into the effects of a web-based survey approach to eliciting preferences for health states from adolescents is needed before health services researchers or pediatric clinicians widely employ these methods.Electronic supplementary materialThe online version of this article (10.1186/s12887-018-1179-7) contains supplementary material, which is available to authorized users.

A Head-On Ordinal Comparison of the Composite Time Trade-Off and the Better-Than-Dead Method

“…Other studies show varying amounts of test-retest reliability in (c)TTO studies. 1,19,20 Our study showed a test-retest reliability in cTTO data comparable with that of Purba et al 20 (95%), although different definitions of test-retest reliability were used.…”

Exploring the Issues of Valuing Child and Adolescent Health States Using a Mixed Sample of Adolescents and Adults

“…There are concerns around whether it is acceptable and appropriate to administer preference elicitation tasks due to the possibility that consideration of dead or trading between quality of life and quantity of life may cause upset or distress for adolescents [41]. The existence of TTO studies that have been undertaken with adolescents using preference elicitation tasks for both their own health and/or for hypothetical health states that involve consideration of being dead (potentially framed as trading of years left to live) [42][43][44][45][46] suggests that these tasks may be acceptable and appropriate. However, the authors are not aware of any studies designed to explicitly test this.…”

“…The plausibility of a life expectancy of 10 years from today for adolescents and younger adults is questionable (and may be implausible for many adults), and potentially this could impact on preferences (though the authors are not aware of any studies examining this). For this reason some studies have used different time frames, for example 60 years [43]. However, increasing the time frame beyond 10 years would be implausible for the elderly participants in the sample.…”

Meeting the Challenges of Preference-Weighted Health-Related Quality-of-Life Measurement in Children

“…Several groups are deriving utility weights for health states described by the generic non-preference-weighted PedsQL via mapping to the CHU-9D and EQ-5D [ 43 – 48 ]. Another promising area of research is using direct elicitation methods including discrete choice experiments to value improvements in children’s health [ 32 , 49 – 51 ].…”