Cross-cultural adaptation and validation of the French version of the Knee injury and Osteoarthritis Outcome Score (KOOS) in knee osteoarthritis patients

Objective: To adapt the Knee injury and Osteoarthritis Outcome Score (KOOS) into French and to evaluate the psychometric properties of this new version. Methods: The French version of the KOOS was developed according to cross-cultural guidelines by using the "translation-back translation" method to ensure content validity. KOOS data were then obtained in patients with symptomatic knee osteoarthritis (OA). The translated questionnaire was evaluated in two knee OA population groups, one with no indication for joint replacement (medicine), and the other waiting for joint replacement (surgery). The psychometric properties evaluated were feasibility: percentage of responses, floor and ceiling effects; construct validity: internal consistency using Cronbach's alpha, correlations with osteoarthritis knee and hip quality of life domains using Spearman's rank test, and known group comparison between medicine and surgery groups; reliability: intra-class correlation coefficient ([CC), Bland and Altman representation; responsiveness using data obtained prior to and 3 months after surgery: standardized response mean (SRM), and effect size. Results: Thirty-seven patients were included in the medicine group (68% women, mean age =70 +/- 10 years) and 30 in the surgery group (73% women, mean age = 71 +/- 10 years). The percentage of responses was excellent. Neither a floor nor a ceiling effect was observed, except for the sport and recreation subscale (20.6% of patients with the worst possible score in the medicine group, 40 and 0% in the surgery group prior to and after surgery, respectively). Results for internal consistency (Cronbach's alpha ranging from 0.76 to 0.93), and convergent and divergent construct validity were satisfactory. The patients waiting for knee surgery presented with significantly lower scores in all KOOS domains. The reproducibility of measurements of all KOOS subscales was good to excellent, with ICC ranging from 0.755 to 0.914. The responsiveness was high, with SRM ranging from 0.89 to 1.93, and effect size from 1.31 to 2.8. Conclusion: The French version of KOOS is a valid, reliable, and responsive instrument to capture specific aspects of functional disability affecting quality of life of knee OA patients. (C) 2007 Osteoarthritis Research Society International. Published by Elsevier Ltd. All rights reserved.