Author(s)
Zhu J1, Liao XZ2, Cao SM3, Hui CX4, Ji Y5, Huang HY1, Cheng X6, Wang L1, Xiao HF2, Liu Q3, Hu XF7, Xie SH5, Guan HJ8, Ren JS1, Li N1, Chen WQ1, Dai M1, Shi JF1
1National Cancer Center/Cancer Hospital, Chinese Academy of Medical Sciences and Peking Union Medical College, Beijing, China, 2Hunan Province Tumor Hospital, Changsha, China, 3Sun Yat-Sen University Cancer Center, Guangzhou, China, 4Inner Mongolia Center for Disease Control and Prevention, Hohhot, China, 5Cancer Hospital, Chinese Academy of Medical Sciences, Shenzhen Center, Shenzhen, China, 6Institute of Medical Information, Chinese Academy of Medical Sciences and Peking Union Medical College, Beijing, China, 7Affiliated Hospital of Inner Mongolia University for the Nationalities, Tongliao, China, 8China Center for Health Economic Research, Peking University, Beijing, China
OBJECTIVES The EQ-5D-3L has been commonly used for years as a generic measurement of health utility. The recently developed Chinese version of EQ-5D-5L was found to be more sensitive in discriminating different health levels in general population in China, when compared with EQ-5D-3L. However, the psychometric properties of the EQ-5D-5L in cancer patients remains unknown in China. This study aimed to compare the performance between EQ-5D-3L and EQ-5D-5L among cancer patients. METHODS A hospital-based cross-sectional survey was conducted in 3 provincesin Sep15, 2016.Patientsof three common cancers (lung, breast and colorectal) were recruited to complete the EQ-5D-3L and EQ-5D-5L questionnaire.Discriminatory power (indicator: Shannon index (H’) and Shannon evenness index (J’)), ceiling effect (the proportion of choosing fully health state) and convergent validity were compared between the two instruments. RESULTS A total of 1 802 cancer patients (lung cancer: 600, breast cancer: 601, colorectal cancer: 601) were included, with a mean age of 55.6 years. The overall utility scores were0.790 (95% CI: 0.778-0.801) of EQ-5D-3L and 0.803 (0.790-0.816) of EQ-5D-5L (P<0.0001).Compared with the EQ-5D-3L (H’=1.10, J’=0.70), an improved discriminatory power was observed in the EQ-5D-5L (H=1.47, J’=0.93), especially onthe dimension of anxiety/depression. The ceiling effect was diminished in the EQ-5D-5L (26.5%) than EQ-5D-3L (34.5%) (P<0.0001), mainly reflected in pain/discomfort and anxiety/depression dimensions. Theconvergent validityof EQ-5D-5L were also observed to be higher than those of EQ-5D-3L. A similar pattern was also observed in detailed cancer-specific analyses, taking breast cancer as an example: the utility wasevaluated as 0.841 (0.825-0.857) by the EQ-5D-5L and 0.821 (0.806-0.837) by the EQ-5D-3L. CONCLUSIONS With greater discriminatory power, convergent validity and lower ceiling, the EQ-5D-5L might be alsoa good choice when EQ-5D-3L is unavailable , but it may result in fewer QALYs gained due to estimated higher utility.
Conference/Value in Health Info
2019-05, ISPOR 2019, New Orleans, LA, USA
Value in Health, Volume 22, Issue S1 (2019 May)
Code
PCN286
Topic
Patient-Centered Research
Topic Subcategory
Health State Utilities, Instrument Development, Validation, & Translation, Patient-reported Outcomes & Quality of Life Outcomes, Stated Preference & Patient Satisfaction
Disease
Oncology
