Validity of the prosthetic activity monitor to assess the duration and spatio-temporal characteristics of prosthetic walking

The prosthetic activity monitor (PAM) is an instrument to assess over the long-term the duration and spatio-temporal characteristics of walking of amputees, during normal daily life. In this study, the validity of PAM-derived measurements was investigated. Twelve transtibial amputees performed an activity protocol, consisting of stationary and walking activities, and activities associated with nonlocomotor movements. The protocol also included potential sources of error and activities assumed to be prone to misdetection. Measurements consisted of the PAM and video recordings. Agreement between video analysis and PAM output was the main outcome measure. The PAM generally correctly classified stationary activities (100% inactive, 0% active, 0% locomotion), nonlocomotor activities (45% inactive, 55% active, 0% locomotion) and walking activities (0% inactive, 1.8% active, 98.2% locomotion). When walking, the number of strides taken was slightly underestimated (-1.0%). The underestimation of distance travelled (-6.2%) and walking speed (-5.8%) was greater. The agreement with video output decreased when the PAM was misaligned, when persons walked at a speed below the defined minimum speed, and when persons walked with crutches. The PAM provides valid data on activity classes and number of strides. Although the majority of the distance data was satisfactory, in some cases considerable differences were found between the PAM and the video data. The impact of alignment, walking speed, and use of assistive devices on the PAM's operation should be considered.