Gait function and improvements across patient groups after total knee replacement: A retrospective between-cohort comparison study.

Abstract

Objective gait analysis provides insights beyond patient-reported outcomes following total knee replacement (TKR). The Cardiff Classifier has shown promise as a summary measure of gait function, but its robustness across centres is not established. This study investigated the (1) comparability of gait function pre- and post-TKR, measured via the Classifier and (2) internal responsiveness of the Classifier in two patient groups assessed at different laboratories. A British cohort (24 patients undergoing TKR, 17 subjects without pathologies) and a Swedish cohort (27 patients undergoing TKR, 25 individuals without pathologies) were included. Three-dimensional gait analysis was conducted preoperatively and one-year post-TKR in patients, and once in non-pathological references. The Classifier classified patients' gait function against their cohort-specific references and quantified the change in gait function pre- to post-TKR. Patient-reported outcomes were assessed using the Knee injury and Osteoarthritis Outcome Score. Fractional multinomial logit models were used to compare Classifier outputs between cohorts, and effect sizes quantified internal responsiveness. Gait function, assessed using the Classifier, differed between the British and Swedish cohorts pre- and one-year post-TKR, but approximately 56% of top discriminative biomechanical features were common across cohorts, mainly involving joint kinetics and ground reaction forces. The Cardiff Classifier showed large responsiveness pre- to post-TKR in both cohorts. Despite cohort-specific pre- and post-TKR gait, the Classifier was sensitive to gait changes and identified key knee osteoarthritis-related gait features across cohorts. These findings support the Classifier as a responsive measure of gait recovery, suitable for multicentre evaluation of TKR outcomes.