Purpose: To define Minimal Important Change (MIC) values for the Knee injury and Osteoarthritis Outcome Score (KOOS) in patients with knee osteoarthritis (OA) receiving physical therapy (PT) or undergoing total knee replacement (TKR). Methods: Two cohorts of patients with knee OA were included: 195 patients receiving PT in Portugal and 102 patients undergoing TKR at LundUniversity Hospital inSweden.KOOSwas administered with a set of anchorquestions at 4weeks following PTand at 6 months,12 months and 5 years post-TKR. KOOS baseline values were obtained pre-surgery and before PTwas initiated. The KOOS is a 42-item patient reported outcome measure where 5 subscales are scored separately on a 0 (worst) to 100 (best) scale; Pain, Symptoms, Activities of Daily Living (ADL), Sport and Recreational activities (Sport/Rec) and Quality of Life (QOL). The anchor questions asked the patients to rate their perceived change on a scale ranging from “much improved” too “much worse”. The PT cohort used a 15-point global rating of change scale. The TKR cohort used 5-point scales relating to each KOOS domains. MIC values were calculated for each cohort, time point andKOOS subscale with the MeanChange method and the Visual anchor-based approach. With the Mean Change method, the anchor points 2: “a little bit better” and 3: “somewhat better” were combined in the PT cohort, and the anchor point 2: “better” in the TKR cohortwas chosen to represent those reporting a relevant improvement. The mean change KOOS scores for these patients represent the Mean Change MIC values. With the Visual Anchor-based approach a Receiver Operating Curve-Statistics was performed and the cut-off points associated with the least amount of misclassification: the sum of 1-sensitivity and 1- specificity were chosen as the ROC MIC values. Results: Mean Change MIC values for the five KOOS subscales increased with the length of follow-up, and higher Mean Change MIC values were found for patients treated with TKR (range 20.6-38.2 and 27.9-48.5 at 6 and 12 months, respectively) than patients treated with PT (range 10.9- 15.3 at 4 weeks). The ROC MIC values were based on a limited number of unchanged patients (13 and 3-12 for PT and TKR, respectively), indicating results being unreliable. Conclusions: For the KOOS, Mean Change MIC values in patients with knee OA varied with intervention and increased with length of followup, indicating MIC being context dependent. To adequately determine the MIC in different study contexts, adequate anchor questions should supplement administration of the KOOS score in future studies. (Table presented).
Ingelsrud, L. H., Terwee, C. B., Gonçalves, R. S., & Roos, E. M. (2014). Minimal important change for the knee injury and osteoarthritis outcome score (KOOS) in patients with knee osteoarthritis. Osteoarthritis and Cartilage, 22, S179–S180. https://doi.org/10.1016/j.joca.2014.02.338