Reliability of a trunk mounted accelerometer when determining gait parameters in people with stroke

Wearable sensors and accelerometers can objectively and reliably assess gait parameters in both healthy individuals and stroke patients. The purpose of this study was to determine whether a wireless tri-axial accelerometer is reliable when measuring spatio-temporal gait parameters in patients with stroke. Thirty-one chronic stroke patients (age: 59.5±13.6 years; time since stroke: 28.1±17.8 months) completed three repeated walks along a 10m flat walkway whilst wearing a trunk mounted accelerometer (BTS G-Walk) secured around the waist of the participant over the L5 vertebrae. Outcome measures included cadence, speed, stride length, %stride length/height, gait cycle duration, step length, stance and swing phase duration, single and double support duration for both symptomatic and asymptomatic lower limbs where relevant. Reliability was assessed via intraclass correlation coefficient (ICC), standard error of the mean (SEM) and smallest detectable change (SDC) values. ICCs were > 0.75 for all parameters, excluding step length on the asymptomatic side (ICC = 0.70). SEM and the SDC were marginally larger for the symptomatic limb than the asymptomatic limb for gait cycle duration and step length, but smaller for all other outcomes. The study showed that the BTS G-Walk is a reliable tool for measuring spatio-temporal parameters in patients with stroke. Physiotherapists and clinicians often prefer detailed information on gait ability. As advanced technologies could help with specific goals relating to gait performance, such devices could be reliably implemented as an alternative to the gold standard in clinical and community settings to monitor patients outside of a lab-based environment.