RELIABILITY AND VALIDITY OF TWO PROXY VERSIONS OF EQ-5D-5L IN JAPAN

Author(s)

Noto S¹, Igarashi A², Shiroiwa T³, Fukuda T³, Ikeda S⁴, Moriwaki K⁵, Saito S⁶, Shimozuma K⁷, Ishida H⁸, Kobayashi M⁹
¹Niigata University of Health and Welfare, Niigata, Japan, ²University of Tokyo, Tokyo, Japan, ³National Institute of Public Health, Wako, Japan, ⁴International University of Health and Welfare, Ohtawara, Japan, ⁵Kobe Pharmaceutical University, Kobe, Japan, ⁶Okayama University, Okayama, Japan, ⁷Ritsumeikan University, Kusatsu, Japan, ⁸Yamaguchi University Graduate School of Medicine, Ube, Japan, ⁹CRECON Medical Assessment Inc., Tokyo, Japan

Presentation Documents

PHP179--strong-u-noto-s-u-sup-1-sup-strong-igarashi-a-sup-2-sup-shiroiwa-t-sup-3-sup-fukuda-t-sup-3-sup-ikeda-s-sup-4-sup-moriwaki-k-sup-5-sup-saito-s-sup-6-sup-shimozuma-k-sup-7-sup-ishida-h-sup-8-sup-kobayashi-m-sup-9-sup-br-sup-1-sup-nii ...

OBJECTIVES: It aims to examine two proxy versions of EQ-5D-5L in Japanese for their reliability and validity METHODS: We compared various patient groups with use of EQ-5D-5L in Japanese version (Self-response version) and two proxy versions. In Proxy 1, a proxy will be asked how he/she would make an evaluation when evaluating a health condition for a subject. In Proxy 2, a proxy will be asked how it should be determined when a subject can communicate his/her own health condition. Proxy version evaluated an occupational therapist in charge. For the statistical processing, reliability and validity were calculated with Cronbach’s α coefficient and Pearson’s correlation coefficient respectively through using STATA14.0.

RESULTS: Total 251 patients became the subjects. The breakdown of disorders was 101 patients for cardiac disorder, 75 patients for respiratory disorder, 41 patients for cerebral tumor, and 34 patients for cervical myelopathy. 159 patients were female (63.3%). The mean scores of EQ-5D-5L were 0.739 (95%CI; 0.718-0.764) for Self-response, 0.735 (95%CI; 0.710-0.759) for Proxy 1, and 0.729 (95%CI; 0.705-0.754) for Proxy 2. α coefficient indicated 0.929 for Self-response/Proxy 1 and 0.956 for Self-response/Proxy 2. When verifying α coefficient with 3 scores by disorder, cardiac disorder (0.951), respiratory disorder (0.985), and cerebral tumor (0.912) indicated higher while cervical myelopathy (0.777) indicated lower. The correlation coefficient was 0.867 for Self-response/Proxy 1 and 0.952 for Self-response/Proxy 2.

CONCLUSIONS: We successfully clarified the reliability and validity for two proxy versions of EQ-5D-5L in Japanese. It was concerned that Proxy 2 may possibly estimate a patient’s QOL to be lower and the reliability would become less depending on disorder. Thus, it is required to give careful attention to those results when using Proxy versions.

Conference/Value in Health Info

2017-11, ISPOR Europe 2017, Glasgow, Scotland

Value in Health, Vol. 20, No. 9 (October 2017)

Code

PHP179

Topic

Patient-Centered Research

Topic Subcategory

Health State Utilities

Disease

Cardiovascular Disorders, Musculoskeletal Disorders, Oncology, Respiratory-Related Disorders

Explore Related HEOR by Topic

Patient-Centered Research

Presentation