Measurement properties of the Knee Injury and Osteoarthritis Outcome Score (KOOS) for patella fractures

BACKGROUND: The study aimed to investigate relevance, construct validity, reliability, and responsiveness and estimate the minimal clinically important difference of the Knee Injury and Osteoarthritis Outcome Score (KOOS) to adult patients with patella fractures. MATERIALS AND METHODS: The study design was a prospective cohort study including patients treated conservatively and surgically following a patella fracture (AO-34). The primary outcome measure was the KOOS. The KOOS was repeated at 14 days, 15 days, six weeks, and finally at six and 12 months. Relevance was evaluated by patients and surgeons ranking of relevance of the 42 items in the KOOS, construct validity by correlation to Eq5d, test-retest reliability by an interclass correlation coefficient, and responsiveness by a priori hypothesis and supplemented by effect size. Furthermore, estimation of minimal clinically important difference (MCID) based on longitudinal confirmatory factor analysis. RESULTS: Included were 65 patients with a mean age of 63.3 years (range 24 to 89) with 74 % female gender. Results showed an acceptable relevance and construct validity. The test-retest reliability was high for all five subscales, with an interclass correlation coefficient ranging from 0.8 to 0.9. Evaluation of responsiveness showed high correlation to the predefined hypothesis (r < 0.6) and moderate to large effect sizes. The MCID of the KOOS subscales were Pain 7.0 (1.6 - 9.8), Symptoms 7.3 (2.0 - 12.6), ADL11.8 (3.2 - 16.0), Sport/Rec 9.3 (3.4 - 15.3) and QOL 8.8 (6.1 - 17.8). CONCLUSION: The Knee Injury and Osteoarthritis Outcome Score (KOOS) showed acceptable relevance, construct validity and responsiveness and high reliability to adult patients with patella fractures.