A study investigating the validity of an accelerometer in quantification of step count in adult hospital inpatients recovering from critical illness

Objective
To determine the validity of the ActiGraph GT3X accelerometer in step count quantification when compared to observed step count in hospitalised adults recovering from critical illness.

Setting
Large National Health Service (NHS) Hospitals Trust.

Subjects
In total, 20 hospital ward-based adults (age: mean 62.3, SD 11.5) who had required greater than 48 hours of mechanical ventilation in the intensive care unit.

Main measures
Participants walked self-selected distances and speeds as part of a semi-structured movement protocol not exceeding 3 hours. Two ActiGraph GT3X accelerometers were worn, one on the thigh and one on the ankle of the non-dominant leg. Accelerometer-recorded step counts were compared against observed step counts.

Results
In total, 31 separate walking episodes were analysed. A mean (SD) of 45.87 (±19.72) steps was calculated for observed step count (range 15–90). Mean differences (95% limits of agreement) of −0.84 steps (−3.88 to 2.2) for the ankle placement and −17.7 steps (−40.63 to 5.25) for the thigh were calculated. Intraclass correlation coefficients (95% confidence intervals) of 0.99 (0.99 to 1.0) and 0.46 (−0.1 to 0.78) were determined for the ankle and thigh, respectively. Placement sites were well tolerated by 95% of participants.

Conclusion
An ankle-mounted ActiGraph GT3X accelerometer demonstrates validity in quantification of step count in hospitalised adults recovering from critical illness. A thigh placement was not considered valid.