Reliability and validity of 3 different Chinese versions of the Oxford knee score (OKS)

Abstract

Abstract Background Different Chinese versions of the Oxford Knee Score (OKS) are available for knee arthritis assessment. These include the Malaysian, Hong Kong and Singaporean Chinese versions with slight variations in wordings and use of Cantonese in the Hong Kong Version. This study evaluated the validity and reliability of the different Chinese OKS versions in Mainland China. Methods One hundred ninety four China mainland-based patients participated in the study, each being diagnosed with knee osteoarthritis. The patients were randomly assigned into 3 groups: Group A completed the Malaysian OKS; Group B completed the Singaporean OKS; Group C completed the Hong Kong OKS. Participants also completed the 36-item Short Form Survey (SF 36). The electronic versions of the questionnaires completed by the patients were sent to smart devices via a social media platform. Results Interclass coefficients for test-retest reliability of the groups were 0.917 in group A, 0.921 in group B, 0.824 in group C. Cronbach alpha results for internal consistency of the 3 groups were: 0.912 in group A, 0.896 in group B, 0.846 in group C. Spearson correlation results with individual SF-36 domains were as follows: Group A showed strong negative correlations with bodily pain and physical function domains; group B exhibited moderate negative correlations with the bodily pain and physical function domains; group C revealed strong negative correlations with the bodily pain and physical function domains and weak negative correlations with vitality domain. Conclusions Different Chinese versions of the OKS showed good reliability and convergent validity in mainland samples of patients with knee osteoarthritis, supporting their use in research and other related studies.

Keywords

• Patient reported outcome measures
• Reliability
• Knee osteoarthritis