The Accuracy of Variously Positioned Pedometers for Lower-Limb Prosthesis Users

• Main Authors: Gary Guerra, John D. Smith, Paula Gomez, Juthamas Siriwatsopon
• Other Authors: Texas A&M University
• Format: Article
• Published: 2020
• Subjects: Engineering; Medicine
• Online Access: https://repository.li.mahidol.ac.th/handle/123456789/50833
• Institution: Mahidol University

© Lippincott Williams & Wilkins. Introduction Tracking the physical activity of persons with amputation is a valuable outcome measure. Identifying affordable and accurate ways of measuring activity for individuals with amputation in developed and resource-limited environments is equally valuable. Limited walking data and methods to reliably measure walking of persons with amputation exist. The purpose of this study was to evaluate accuracy of the Omron HJ-329 worn by persons with lower-limb amputations at different locations. Methods Nineteen lower-limb prosthesis wearers wore pedometers around the neck, in both affected and nonaffected side pockets, and on the hip of both the affected side and the contralateral side while walking at a self-selected pace for 200 m. Actual step counts (ACs) were recorded by a hand tally counter. Repeated-measures analysis of variance was used to determine differences in pedometer counts between affected and nonaffected sides and between ACs. Agreement was assessed with intraclass correlation coefficients (ICCs) and Bland-Altman plots, and absolute percent error was calculated using ([steps detected by pedometer - AC]/AC) × 100. Alpha was set at 0.05 for all tests. Results No significant differences existed in pedometer counts between affected and nonaffected sides (P > 0.05), nor were any significant differences evident between any of the pedometers and ACs (P > 0.05). In our sample, there were more left affected participants; among all participants, ICCs were highest on the left hip followed by the right hip, and least in the left pocket and around the neck. Limits of agreement in the Bland-Altman plots reflected this trend as well. Absolute percentage error of pedometers was greatest in the left pocket (6.8%) and lowest for the pedometer located at the left hip (2.2%) and right hip (2.6%). Conclusions The individual with amputation can wear the activity monitor at any of the locations and on either the affected or nonaffected side, and it seems wearing the devices on the hip will provide the most accurate results. Among the individuals with unilateral amputation, there were no differences in metrics between affected and nonaffected limbs. For our bilateral participants, the same nondifferences were observed. The Omron HJ-329 is thus an affordable, viable activity monitor for the population of persons with lower-limb amputation.