"Greek KOOS-Child: a valid, disease specific, diagnostically accurate and responsive PROM in children with knee-related pathology".

Abstract

The Knee Injury Osteoarthritis Outcome Score (KOOS)-Child questionnaire is one of the frequently child-friendly measures used in pediatric studies. The aim of this study was to transculturally adapt the Greek version of KOOS-Child and evaluate its clinimetric properties in children with knee disorders. Children visiting the Outpatients Orthopaedic Clinic of a Greek Paediatric General Hospital were considered eligible if they were aged 8-14years, had a knee soft tissue injury and associated physical limitations. The transcultural adaptation was based on a multistage backward translation approach. Participants completed the KOOS-Child at their first visit to the orthopedic specialist (baseline), 2weeks and 3months after baseline. Content validity of the KOOS-Child was evaluated using general QoL measures (KIDSCREEN and Kid-KINDL) and construct validity was explored by correlating relevant items. Responsiveness was evaluated according to the children's response on the given orthopeadic treatment. Sample consisted of 59 children (30 males), aged: 11 ± 1.8years. The KOOS-Child showed high internal consistency (Cronbach's a: 0.80-0.96). Adequate convergent validity with > 75% relevant a priori hypotheses was confirmed. Construct validity was moderate to strong (Pearson's r correlations between related KOOS and Kid-KINDL subdimensions: 0.54-0.62). KOOS and KIDSCREEN subdimensions correlations were fair (Pearson's r correlations: 0.32-0.65). KOOS-Child's diagnostic accuracy was high. Factor analysis extracted height factors accounting for 76.15% of the total variance, confirmed by the scree plot. Responsiveness was moderate to high with Cohen's d from 0.6 to 1.4. The Greek version of the KOOS-Child demonstrated excellent internal consistency, good construct validity, diagnostic accuracy and interpretability as well as good responsiveness. The measure could be used across Greek children with orthopaedic knee problems. Generalisability of findings is limited due to the relatively limited cohort. II.