Estimation of Quality-adjusted Life Years alongside clinical trials: the impact of ‘time-effects’ on trial results

Abstract

Objectives: The objectives were to investigate the impact of ‘time-effect’ on the estimation of quality-adjusted life years (QALYs) along prospective clinical trials’ outcomes using an assumed fixed time duration versus the actual time durations for each case. The ‘time’ duration is the length of time in a health state. Methods: Two methods were used in the estimation of QALYs based using EQ-5D 3L scores collected at specific time-point intervals. One method used the actual time durations for each case based on CRF records, and the other used an assumed time duration and globally applied it to all the cases. Using SPSS® software program, we used paired-sample t-tests to assess whether the ‘time-effect’ can potentially affect trial results using CONSTRUCT trial data as reported in the trial results publications. The trial compared use of Infliximab with Cyclosporine for patients with Ulcerative Colitis and it involved some 270 participants. Key findings: The results largely indicate statistically significant differences between the two methods of QALY estimations. QALYs at the respective time-points indicate no statistical difference between the two approaches. However, the difference in terms of total QALYs between the two QALY estimation approaches is statistically significant with considerable impact on costs/QALY. Conclusions: Considering the possible impact of the time-effect on QALY estimations, the result implies that it can have significant implications for resources allocations decisions. In this respect, researchers have to pay due considerations to the approach they use and where possible, actual time durations must be used in QALY estimations along prospective clinical trials.