Monitoring activity in the acute inpatient setting

Mrs Sarah Gilmore1,2, Dr  Jodie McClelland2, Dr  Megan Davidson2

1St Vincent’s Private Hospital Melbourne, Fitzroy, Australia, 2La Trobe University, Melbourne, Australia


Monitoring activity in the acute inpatient setting: how accurate are activity monitors?


To evaluate the validity of the ActivPAL3, Fitbit Flex (FF) and Jawbone UP Move (JUM) in measuring step count immediately  following lumbar fusion.


Forty patients admitted to St Vincent’s Private Hospital Melbourne wore the ActivPAL3, FF and JUM on the thigh, and the FF and JUM on the wrist, while walking at a comfortable pace for two minutes. Twenty tests were completed with each monitor in each position. The number of steps taken were counted, and the number of steps detected by each monitor was recorded. Accuracy was determined by calculating the steps detected as a proportion of the observed step count.


The ActivPAL3 had a mean accuracy of 85% (SD 27%). The FF detected significantly more steps when worn on the thigh (mean accuracy 66%, SD 35%) than on the wrist (mean accuracy 24%, SD 34%). The JUM had a low mean accuracy on both the thigh (22%, SD 35%) and wrist (17%, SD 40%). When worn on the wrist, the FF and JUM had a lower accuracy when participants used a walking aid.

Significance of the findings to allied health:

There is increasing interest in measuring activity levels of hospital inpatients. This research indicates the ActivPAL3 provides an accurate measure of step count following lumbar fusion. While consumer grade activity monitors are a low cost alternative to research grade monitors, they have low accuracy in lumbar surgery patients who typically have a slow paced and irregular gait, and in patients using walking aids.