54.16 Accelerometers as an adjunct to mobility assessment in the intensive care unit: a feasibility study

K. Ricci1, C. Horwood1, K. Castellon-Larios1, C. Byrd1, S. Steinberg1, D. Vazquez1  1The Ohio State Wexner Medical Center,Surgery,Columbus, OHIO, USA

Introduction:  

Early mobility of patients in the intensive care unit (ICU) has been associated with decreased complications and length of stay. We currently rely on subjective documentation by staff of the degree of activity among patients.   This is a report of a feasibility study of utilizing a wrist-worn accelerometer as an adjunct to developing an automated and objective mobility scoring system in a surgical ICU.

Methods:

 An IRB approved pilot project was conducted which involved placing an off-the-shelf activity tracking device (Fitbit Flex II) on six non-intubated surgical ICU patients.  The device was left on the patients for a total of 72 hours. The patients had similar demographics and acuity scores. Data on amount of steps, calories burnt and number of minutes of sleep was collected. Subjective mobility was also documented by nursing staff.

Results:
There was a large variation in results between patients.  The range of steps taken per day was 0 to 1962 with a mean of 489. The calories burnt per day ranged from 1543 to 2378 with a mean of 1625. Finally, sleep per day ranged from 0 to 1036 minutes, with a mean of 200 minutes.  The objective actimetry data for steps and calories appeared to correlate with the subjective daily nursing documentation of patient activity (see figure 1).  In contrast, the device did not appear to accurately reflect observed sleep.  This might be due to a limitation of the intrinsic sleep algorithm within the device, which requires an uninterrupted 60 minutes of no activity before it will reflect “sleep”.

Conclusion:
It is feasible to utilize wrist-worn accelerometers to objectively quantify activity in patients in the ICU.  However, the data obtained from the Fitbit Flex II appears to be at times inaccurate and unreliable particularly when assessing sleep patterns. Further work is needed testing other available devices prior to creating and validating an objective mobility scoring system.