The Singaporean English and Chinese versions of the EQ-5D achieved measurement equivalence in cancer patients

• Published in: Journal of clinical epidemiology 2009-02, Vol.62 (2), p.206-213
• Main Authors: Gao, Fei, Ng, Gim-Yew, Cheung, Yin-Bun, Thumboo, Julian, Pang, Grace, Koo, Wen-Hsin, Sethi, Vijay-Kumar, Wee, Joseph, Goh, Cynthia
• Format: Article
• Language: English
• Subjects: Adult; Algorithms; Asian Continental Ancestry Group - ethnology; Bilingualism; Biological and medical sciences; Cancer; Cancer therapies; Chinese; Communication Barriers; Confidence intervals; Epidemiology; EQ-5D; Equivalence; Female; Health Status Indicators; Humans; Internal Medicine; Language; Male; Medical research; Medical sciences; Middle Aged; Neoplasms - psychology; Oncology; Psychometrics - methods; Quality of Life - psychology; Questionnaires; Regression analysis; Reproducibility of Results; Rheumatic diseases; Sensitivity analysis; Singapore; Studies; Surveys and Questionnaires - standards; Tumors; Utility; Validation; Validity
• ISSN: 0895-4356; 1878-5921
• DOI: 10.1016/j.jclinepi.2008.03.007

Abstract Objective To assess measurement equivalence of the Singaporean English and Chinese versions of the EuroQol Group's 5-domain questionnaire (EQ-5D) in cancer patients. Study Design and Setting Seven hundred and seventy-one ethnic Chinese patients in Singapore were recruited, and they answered either an English or a Chinese version of the EQ-5D. Seven days later, a similar questionnaire in the same language was mailed to the patients. Regression analysis was used to assess equivalence of the mean values obtained by using the two language versions. The validity, responsiveness to change, and reliability of the two versions of the EQ-5D were assessed and compared. Results Based on the prespecified equivalence margin of ±10% for binary outcome, ±0.05 for utility index, and ±5 points in the visual analog scale, the two language versions of the EQ-5D gave equivalent mean values at item and scale levels. They also showed similar characteristics in validity, responsiveness, and reliability. For example, the test–retest reliability values for the EQ-5D utility index in the two language samples were identical in two significant digits: 0.79. Conclusion The Singaporean English and Chinese versions of the EQ-5D were validated in cancer patients and were shown to achieve measurement equivalence.