Sensor-derived TKA gait metrics correlate more strongly with in-clinic performance tests than PROs

The purpose of this study is to determine the correlation among gait metrics from a sensor integrated into a total knee arthroplasty (TKA) implant, performance tests and Patient Reported Outcome (PROs) after TKA. Fifty-six patients underwent TKA with integrated sensors and completed the following PROs tests: Oxford Knee Score (OKS), Knee Injury and Osteoarthritis Outcome Score for Joint Replacement (KOOS JR) and the following in-clinic performance tests: Time Up-and-Go (TUG), and fast 4-meter walk test (4mWT) at four to eight weeks. Sensor-derived gait metrics included six metrics: walking speed while performing activities of daily living in the community, or “Community Walking Speed” (CWS), cadence, knee range-of-motion (Knee ROM), tibial range-of-motion (TROM), stride length and step count. Average sensor-derived metrics from the week of the clinic visit were compared to performance tests and PROs using Pearson correlation coefficients and 95% confidence intervals. The correlation between CWS and TUG was the strongest, rho=-0.49 (p=0.0001), followed by CWS and 4mWT, rho=-0.37 (p=0.005). Correlations between the sensor-derived gait metrics and PROs were not statistically significant except for cadence to OKS (rho=0.36, p=0.008) and stride length to OKS (Corr: -0.43 p=0.0009). The correlation between TUG and 4mWT was strong (rho=0.51, p<0.0001) while the correlation between the performance tests (4mWT and TUG) and PROs (OKS, KOOS JR) were not statistically significant. Gait metrics derived from a sensor embedded in a TKA tibial extension correlate more strongly with performance tests (TUG and 4mWT) than with PROs (OKS and KOOS JR). While further study against controls is needed to confirm, early research indicates objective gait metrics better reflect physical function than subjective PROs. II Prospective Cohort Study