The role of fixed and variable time horizons in the time trade-off method

Aim: The purpose of this study is to assess whether using different time horizons in the time trade-off method yields equivalent outcomes, and if not, whether a non-parametric correction for discounting and loss aversion can solve the disparity in outcomes. Methods: A cross-sectional dataset was obtained through an online questionnaire. Participants underwent self-administered time trade-off utility questions for multiple health states in 10-year, 20-year and SLE time frames. The outcomes were compared using Wilcoxon signed-rank tests, bivariate correlations and multiple linear regressions. Results: We found that time trade-off tasks using 10-year and 20-year time frames yield lower outcomes compared to the SLE time frame. This difference cannot be mitigated by correction for discounting and loss aversion, but the predictive power increases when the correction is applied. Conclusions: The 10-year and 20-year time frames yield valid results with no significant difference between the outcomes. Outcomes obtained using the SLE time frame appear to need a downward correction. We conclude that the use of 10-year and 20-year time frames is adequate for the time trade-off method.