OBJECTIVES:The aim of this study was to assess the validity, reliability and responsiveness of the developed Greek version of the KOOS in total knee replacement (TKR) patients.
METHODS:For the translation procedure, guidelines for cross-cultural adaptation were followed (Beaton et al, 2000). KOOS was translated into Greek independently by 3 Greek native raters (2 physiotherapists with > 5 years of clinical musculoskeletal experience and one Orthopedic surgeon).The 3 raters had a meeting where all translations were discussed and a consensus was achieved to a preliminary (synthesis) version. This version was utilized for a backtranslation (Greek to English) by another rater, an English native professional translator who had no prior knowledge of the KOOS. The synthesis and back translation were discussed in a second meeting and a consensus on a prefinal Greek version of KOOS was reached. This prefinal version was tested in a pilot study. It was administered to 10 total knee replacement (TKR) patients with variable educational levels, to confirm the comprehensibility and syntax of all questions. Following the pilot, a final meeting was organized, where all raters discussed the comments made by the patients within the pilot. There were only two points of discussion regarding a) question P4 & P5, the translation and comprehension of full flexion and extension by patients b) the translation of the word «extreme» , the 5th point of the Likertscale in questions S6-Q4. Thus, following this meeting the final version of Greek KOOS questionnaire was developed.
Validation procedure:For exploring validity, reliability and responsiveness of KOOS Greek version, 51 consecutive TKR patients (mean age 72.2±7.2 years, 39 women) within the University Hospital of Patras, Greece participated in the study. KOOS's psychometric properties were evaluated according to the COSMIN checklist.Preoperative clinical status and postoperative outcomes at 2 occasions (at discharge and 10-12 days postoperatively) were evaluated using the Greek versions of the KOOS, KOS-ADL, and SF-12 Health Survey. Psychometric properties of both Greek versions of KOS-ADL and SF-12 have been previously tested and have been found satisfactory (Roos et al, 1998, Goncalves et al, 2009). All questionnaires were administered by one orthopedic surgeon. Questionnaires were self-filled.For the analysis, criterion-related validity was evaluated by Pearson's correlation. Internal consistency by Cronbach's alpha; Reliability, by intra-class correlation coefficient (ICC).The SEM was calculated on the basis of the error variance, whereas responsiveness was calculated by effect size, standardized response mean (SRM) and by relating the smallest detectable change (SDC) with the minimal important difference (MIC).
RESULTS:Internal consistency: Internal consistency was excellent with Cronbach's a being 0.801 preoperatively, and 0.920 at discharge and at 2 weeks after TKR. Chronbach's a for the 5 subscales of KOOS was found 0.785 preoperatively, 0.822 at discharge and 0.781 at 2 wee...