Limitations of the Knee Society Score in evaluating outcomes following revision total knee arthroplasty.

BACKGROUND Traditionally, the results of revision total knee arthroplasty have been determined with use of surgeon-based measures such as the Knee Society rating system. Recently, outcome and quality-of-life measures have shifted toward a greater emphasis on patient-based evaluation. The aim of our study was to determine the validity and responsiveness of the Knee Society rating system compared with the Short Form-36 health survey (SF-36), Western Ontario and McMaster Universities Osteoarthritis Index (WOMAC), and a four-question 4-point Likert scale satisfaction questionnaire following revision total knee arthroplasty. METHODS A total of 152 patients underwent revision total knee arthroplasty at our institution, between August 2003 and January 2007, and had a two-year follow-up evaluation after revision surgery. The SF-36, WOMAC, Knee Society rating system, and satisfaction scores were completed preoperatively and postoperatively. Spearman correlation coefficients were calculated to determine the degree of correlation for each outcome scale. The SF-36, WOMAC, and patient satisfaction were correlated with the Knee Society rating system. RESULTS Both before and after surgery, the correlation among items of the Knee Society rating system displayed low to negligible levels of association. The Knee Society rating system pain score showed modest levels of convergent construct validity with the WOMAC and SF-36. However, the Knee Society functional score displayed negligible to low correlation with its WOMAC functional counterpart preoperatively. The Knee Society pain and functional scores, respectively, showed marked and moderate association with satisfaction. The change in the Knee Society pain and functional scores had moderate association with the SF-36 and WOMAC counterparts, except low correlation was displayed between the pain scores for the Knee Society rating system and the SF-36. The Knee Society rating system pain score was found to be the most responsive of the measures with a standardized response mean of 1.6, whereas the Knee Society rating system functional score was found to be the least responsive at 0.7. CONCLUSIONS Currently, there is no so-called gold standard that optimally reflects the status of the knee, as well as the patient, prior to and following revision total knee arthroplasty. Ideally, numerous assessment scales should be administered to the patient in order to accurately reflect the patient characteristics for the purpose of academic study, but from a practical standpoint, this may not be feasible. We encourage further research and development of a simple and concise standardized questionnaire for use before and after revision total knee arthroplasty.