Responsiveness and interpretability of the Animated Activity Questionnaire for assessing activity limitations of patients with hip or knee osteoarthritis

Purpose The aim of the study was to determine the responsiveness and interpretability of the Animated Activity Questionnaire (AAQ), an online questionnaire in which osteoarthritis patients select animations that best match their performance of daily activities. Methods A longitudinal study was carried out, in which 94 patients with hip or knee osteoarthritis were assessed at baseline, and 3 and 6 months after treatment (conservative and surgical). Responsiveness was assessed by means of testing hypotheses about expected correlations between change in AAQ, a Global Rating Scale of change (GRS) and change in the Activities of Daily Living subscale of the Hip disability or Knee injury and Osteoarthritis Outcome Score (H/KOOS), and a combination of performance‐based tests (the 30 s chair‐stand test, the timed up‐and‐go test and the nine‐step stair climbing test). The minimal important change (MIC) was estimated by means of the receiving operating characteristics (ROC) method. Results The correlations of the AAQ with the H/KOOS were as expected, but other correlations were lower than anticipated. The area under the ROC curve was 0.74 at 6 months. At 3 months' follow‐up, the correlations were too low to calculate a MIC. A total of 20% of the results at 3 months and 80% of the results at 6 months were in accordance with the hypotheses. The MIC was 9 points at 6 months. Conclusions The AAQ was sufficiently responsive at the six‐months follow‐up, but not at the three‐month follow‐up. The MIC at the 6‐month follow‐up (9 points) was slightly lower than the smallest detectable change of 14 points found in a previous study.