Letter to the Editor: Translation and Validation of the German New Knee Society Scoring System.

Abstract

To the Editor, I read the study by Kayaalp and colleagues [1] with great interest. While I believe the authors have made a considerable contribution with this work, there are some concerns that I would like to address. The authors analyzed the construct validity of the new Knee Society Score (KSS) using the German WOMAC and the German SF-36. The Licensed User Manual of the Knee Society Knee Scoring System [5] stated that the new KSS was generally consistent with other “knee-specific” scores. In this study, the German WOMAC does not meet all parameters evaluated by the new KSS, which now includes patient expectation and satisfaction parameters. The low correlation values between the patient expectations subdomain and the German WOMAC subdomain support this criticism. Additionally, applying such long questionnaires burdens the patient [3]. Shortening the questionnaires for patients may reduce the burden, but a validity study with a single knee-specific questionnaire may not provide all the information that patients wish to share or that providers need to make good surgical decisions. Another concern is that although the Licensed User Manual of the Knee Society Scoring System [5] states that there is no total score of the survey, the authors calculated the total score and performed a statistical analysis. This situation should be corrected immediately, as this may lead to misleading results and the development of chain errors. The authors performed the study with 100 patients, and they also included 39 patients for test-re-test reliability. That number of patients is low for the validity and reliability of this questionnaire, considering the recommendation that the sample size should be 10 times the number of items [4]. Since only the patient expectations subdomain of the questionnaire is different between preoperative and post-operative versions of the new KSS, performing the analysis of the other subdomains with all patients could make the statistical analysis more valuable like in the Turkish version of the new KSS. [2]. Finally, I would like to mention a minor error. In Table 3, the correlation coefficient (symptoms subdomain of the new KSS and mental health subdomain of the German SF36), which we normally expect to be negative, was positive, whereas the correlation coefficient (between symptoms subdomain of the new KSS and vitality subdomain of German SF-36), which we expected to be positive, was negative. This may adversely affect the validity of the study.