A Comparison of Preference Assessment Instruments Used in a Clinical Trial

Publisher Website
Google Scholar
Abstract
Objectives. To compare preference assessments that were made by using the EuroQol EQ-5D and the Health Utilities Index Mark II. Subjects. 561 patients in a randomized trial of tirilazad mesylate for aneurysmal subarachnoid hemorrhage. Measures . Three preference assessments (a value score for the EuroQol instrument and value and utility scores for the Health Utilities Index) made three months after randomization. The av erages for each of the three scores, stratified by clinical outcomes and attributes of the Health Utilities Index health status classification system, were compared. To eval uate potential sources of difference between the instruments, the authors estimated two alternative Health Utilities Index sconng rules that were based on patient re sponses to the EuroQol instrument. Results . Patients' ratings of their current health made by using the 100-point visual analog scale from the EuroQol instrument were more similar to the utility scores for the Health Utilities Index than they were to the value scores for the Health Utilities Index. The biggest differences between the visual analog scores for the EuroQol instrument and the utility scores for the Health Utilities index were seen at higher levels of functioning. Conclusion. For states representing higher levels of functioning, differences were seen between patients' self-ratings ob tained by using the EuroQol instrument and the patients' utility scores on the Health Utilities Index; for states representing lower levels of functioning, substantial agreement was observed between these two scores. Differences observed at the higher levels of functioning suggest that further research is needed to determine whether the Health Utility Index's assignment of a score of 1.0 to the reference state representing being healthy is appropriate. Key words: preference assessment; health utilities index; EuroQol; clinical trials; subarachnoid hemorrhage. (Med Decis Making 1999; 19:265- 275)