Characterization of Step Count Accuracy of Actigraph Activity Monitor in Persons With Lower Limb Amputation

In Brief Activity monitors can be used to objectively assess patient progress and to quantitatively measure the impact of new technologies on function. The purpose of this study was to characterize the accuracy of the step counter, GT1M, when worn on the waist and ankle location in persons using a prosthetic limb to ambulate. Potential subjects were given the opportunity to enroll while at their usual clinical prosthetic checkup. The subject population consisted of nine males (two transfemoral and seven transtibial amputations). Accuracy of GT1M was determined by measuring the percent difference between visually observed steps and the steps recorded by the GT1M. Statistical significance was achieved if the p value was ≤0.05. The accuracy of the GT1M when worn on the ankle (90.2% mean ± 16.3% SD) was greater than when worn on the waist (63.5% mean ± 39.2% SD) as determined with a paired t-test (p = 0.028). Linear regression demonstrated that inaccuracies of the GT1M highly correlated to use of a walking aid (r2=0.789, p = 0.001). Therefore, the GT1M demonstrated the highest accuracy when worn on the prosthetic ankle and when ambulating over level surfaces without the use of a walking aid (97.8% ± 4.0%). Percent accuracy was as low as 10% in a participant who did not meet these criteria. If planning to use GT1M to quantify walking in persons with lower limb amputation, validation of the GT1M on the subjects and walking terrain to be studied is recommended. Activity monitors can be used to objectively assess patient progress and to quantitatively measure the impact of new technologies on function. This study aimed to characterize the accuracy of the step counter, GT1M, when worn on the waist and ankle location in persons using a prosthetic limb to ambulate. The GT1M demonstrated the highest accuracy when worn on the prosthetic ankle and when ambulating over level surfaces without the use of a walking aid (97.8 ± 4.0%). Percent of accuracy was as low as 10% in a participant that did not meet these criteria. If planning to use the GT1M to quantify walking in persons with lower limb amputation, validation of the GT1M on the subjects and walking terrain to be studied is recommended.