OBJECTIVES : EORTC QLU-C10D is a preference-based measure derived from the widely used profile-based measure, EORTC QLQ-C30. This study aimed to compare the responsiveness of EQ-5D-5L with QLU-C10D in cancer patients.

METHODS : We recruited patients with cancer being treated or followed at National Cancer Centre Singapore. Each patient was interviewed twice face-to-face during two different outpatient visits using EQ-5D-5L and EORTC QLQ-C30 questionnaires. Overall health and QOL at baseline and follow-up surveys were assessed using a single-item scale. Multiple country-specific value sets were used to calculate the EQ-5D-5L index. QLU-C10D index was derived using its Australian value set. Responsiveness of the indices was assessed using the standardized effect size (SES), standardized response mean (SRM) and F-statistic.

RESULTS : The analysis included 158 cancer patients (mean age: 58 years; male: 52%) who completed both baseline and follow-up visits (median 62 days apart). At the follow-up visit, 39% and 23% reported improvement and deterioration in overall health, respectively. Based on data from deteriorated patients, the SES, SRM and the F-statistic values for QLU-C10D were -0.69, -0.48, and 7.4, respectively; those values for EQ-5D-5L indices were much lower (SES range: -0.02 to -0.05; SRM range: -0.02 to -0.06; F-statistic range: 0.02 to 0.11). Based on data from improved patients,, the SES, SRM and the F-statistic values for QLU-C10D were 0.28, 0.35, and 7.2, respectively, which were either similar or slightly larger than those for EQ-5D-5L indices (SES range: 0.21 to 0.30; SRM range: 0.21 to 0.31; F-statistic range: 2.6 to 6.0). Similar results were observed when improvement and deterioration were defined by change in overall QOL at the follow-up visit.

CONCLUSIONS : It appears that EQ-5D-5L is non-inferior in terms of responsiveness to improvement in health and QOL compared to QLU-C10D. However, QLU-C10D may be more responsive than EQ-5D-5L for assessing deterioration in health and QOL.