The Dutch version of the knee injury and osteoarthritis outcome score: A validation study

BACKGROUND: The Knee Injury and Osteoarthritis Outcome Score (KOOS) was constructed in Sweden. This questionnaire has proved to be valid for several orthopedic interventions of the knee. It has been formally translated and validated in several languages, but not yet in Dutch. The purpose of the present study was to evaluate the clinimetric properties of the Dutch version of the KOOS questionnaire in knee patients with various stages of osteoarthritis (OA). METHODS: The Swedish version of the KOOS questionnaire was first translated into Dutch according to a standardized procedure and second tested for clinimetric quality. The study population consisted of patients with different stages of OA (mild, moderate and severe) and of patients after primary TKA, and after a revision of the TKA. All patients filled in the Dutch KOOS questionnaire, as well as the SF-36 and a Visual Analogue Scale for pain. The following analyses were performed to evaluate the clinimetric quality of the KOOS: Cronbach's alpha (internal consistency), principal component analyses (factor analysis), intraclass correlation coefficients (reliability), spearman's correlation coefficient (construct validity), and floor and ceiling effects. RESULTS: For all patients groups Cronbach's alpha was for all subscales above 0.70. The ICCs, assessed for the patient groups with mild and moderate OA and after revision of the TKA patients, were above 0.70 for all subscales. Of the predefined hypotheses 60% or more could be confirmed for the patients with mild and moderate OA and for the TKA patients. For the other patient groups less than 45% could be confirmed. Ceiling effects were present in the mild OA group for the subscales Pain, Symptoms and ADL and for the subscale Sport/Recreation in the severe OA group. Floor effects were found for the subscales Sport/Recreation and Qol in the severe OA and revision TKA groups. CONCLUSION: Based on these different clinimetric properties within the present study we conclude that the KOOS questionnaire seems to be suitable for patients with mild and moderate OA and for patients with a primary TKA. The Dutch version of the KOOS had a lower construct validity for patients with severe OA on a waiting list for TKA and patients after revision of a TKA. Further validation studies on the Dutch version of the KOOS should also include a knee specific questionnaire for assessing the construct validity.