Validity of the Walked Distance Estimated by Wearable Devices in Stroke Individuals

Background: Health professionals need valid devices to assess a stroke individual's ability to walk. The aim was to evaluate the validity of the estimation of the walked distance by wearable devices and the impact of the sensor's position in stroke individuals. Methods: Post-stroke patients able to walk without human assistance were equipped with several wearable devices: pedometers, Actigraph, and Sensewear Armband placed according to the manufacturers' recommendations. Participants walked for 6 min at a comfortable speed wearing all sensors at the same time. We analyzed the validity of sensor-estimated distances according to their position using Bland-Altman analysis, root-mean-square error, and coefficient of correlation. Results: In total, 35 individuals were included (mean age = 65 +/- 15 years). The best estimations were given by the Actigraph worn on the unaffected ankle (mean bias (MB) = 22.6 +/- 32.4 m; p = 0.37) and by the pedometer worn on the unaffected hip (MB = 20.5 +/- 24.6 m; p = 0.46). The other sensors and positions provided large estimation errors over 95 m (p < 0.05). Conclusion: This study led to a recommendation of a pedometer worn on the unaffected hip or an Actigraph worn on the unaffected ankle to get a valid estimation of the distance walked by stroke individuals.