Patient-Reported Outcome Measures for the Knee

In the past 20 years, there has been considerable growth in the number of knee instruments and rating scales designed to measure outcomes from the perspective of the patient. Only a few of these instruments have been evaluated for reliability, validity, and responsiveness. The purpose of this systematic review was to examine the psychometric evidence of patient-reported outcome measures for the knee and identify the best scores for specific knee conditions. A literature search was performed to retrieve references relating to the development and evaluation of knee-specific instruments. Twenty-four unique instruments were identified, and most have satisfactory evidence for internal (alpha > or = 0.82) and test-retest reliability (intraclass correlation coefficient > or = 0.80). Face/content validity was typically assessed during the item selection process, and construct validity was evaluated through strict hypothesis testing or correlations with other clinical measures. For many instruments, effect sizes and standardized response means measuring responsiveness were large (> or = 0.80) within disease-specific populations. Based on the psychometric data, recommendations include the Cincinnati Knee Rating System, Knee Injury and Osteoarthritis Outcome Score (KOOS), and Lysholm Knee Score for anterior cruciate ligament (ACL) injuries, the Kujala Anterior Knee Pain Scale for anterior knee pain, the International Knee Documentation Committee (IKDC) Subjective Knee Form, KOOS, and Lysholm Knee Score for focal chondral defects, the Western Ontario Meniscal Evaluation Tool (WOMET) for meniscal injuries, and the KOOS for osteoarthritis (OA). Although the IKDC can be used as a general knee measure, no instrument is currently universally applicable across the spectrum of knee disorders and patient groups. Clinicians and researchers looking to use a patient-based score for measurement of outcomes must consider the specific patient population in which it has been evaluated. Using a diagnostic algorithm that measures the anatomic parts of the knee as separate constructs may solve this dilemma, allowing for measurement of treatment outcomes across patient groups and selection of the optimal clinical intervention.