A Real World Comparison of Utility Values Derived from a Discrete Choice Experiment Versus Patient Reported Outcomes in Clinical Trials
Author(s)
Patten N1, Brydon S2, Mulhern B3, Peacock A4, White B5, von Butler L5, Taylor C6
1Bristol-Myers Squibb Australia, Mulgrave, Australia, 2Bristol-Myers Squibb, oakleigh, VIC, Australia, 3University of Technology Sydney, Sydney, Australia, 4Health Technology Analysts, Sydney, NSW, Australia, 5SurveyEngine, Berlin, Germany, 6The George Institute for Global Health, Australia, Sydney, Australia
OBJECTIVES Health state utility (HSU) inputs are key parameters in generating a cost per quality-adjusted life-year (QALY), which has a strong influence on reimbursement decisions in Australia. Discrete choice experiments (DCEs) provide more country-specific detail on preferences compared to trial-based utilities and are increasingly used to estimate HSUs. The purpose of this analysis is to compare Australian general-population HSU values with treatment specific health states from clinical trials. The intent of this research is for consideration in future Australian reimbursement applications in renal cell carcinoma (RCC). METHODS A DCE study was designed to elicit HSU estimates for treatment profiles in the adjuvant setting (nivolumab+ipilimumab [nivo+ipi]) versus watch and wait [WW]; and nivo+ipi versus standard of care [SOC] of sunitinib across progression-free (PF) and progressed disease (PD) states in the first-line treatment setting. Healthy respondents, representative of the Australian general population across age and gender, were recruited from an online panel. Results were compared to HSU estimates from the phase 3 trial CheckMate 214, which included EQ-5D responses converted to index-based scores using an Australian algorithm. RESULTS A total of 2002 participants completed the DCE survey, with a response rate of 49%. Demographics were broadly aligned with the Australian population. In the adjuvant setting, there was little difference between HSU values for active treatment (nivo+ipi) versus WW. Estimated progression-free and post-progression utility estimates were lower than those reported in the 1L CheckMate 214 trial. CONCLUSIONS Compared to PRO, a DCE provided lower HSU value estimates and a larger difference between immunotherapy and SOC. Differences may be due either to the use of the general population volunteers in the DCE, or due to the specific descriptions provided in the DCE, allowing participants to make more informed trading decisions. Future research could include conducting a similar DCE experiment in an RCC population, rather than healthy volunteers.
Code
PCN42