Date Presented Accepted for AOTA INSPIRE 2021 but unable to be presented due to online event limitations.
This poster will discuss the use of commercially available accelerometers to monitor poststroke home rehabilitation of the upper extremity during repetitive, occupation-based exercises. Secondary data analysis of outputs includes (1) comparison of self-reported practice sessions with sessions captured via the accelerometer, (2) comparison of self-reported practice time versus time captured via accelerometer, and (3) patterns in angular velocity based on data processing.
Primary Author and Speaker: Kate De Castro
Additional Authors and Speakers: Elena Donoso Brown
Contributing Authors: Sarah Wallace, Rachael Neilan
PURPOSE: Accelerometers have been used in the literature to measure outcomes related to hand function for persons post-stroke (Bailey et al., 2015; Reiterer et al., 2008; Uswatte et al., 2006). While research previously found accelerometers to be valid and reliable real-time monitors for home rehabilitation programs, challenges regarding data management and translation of accelerometer output information into clinically interpretable results continue to exist (Noorkõiv et al., 2014). The purpose of this secondary data analysis was to explore the feasibility of using a commercially available accelerometer with custom software to collect data corroborating self-reported practice in a repetitive task practice home program.
DESIGN: The original study was a single-subject design intervention study that tracked home practice for upper extremity rehabilitation and the effects of adherence with and without music. Participants were adults with chronic stroke resulting in aphasia and hemi-paresis in one upper extremity.
METHOD: Primary data collection (n = 7) involved wearing commercially available accelerometers on both upper extremities while completing several sets of two to three pre-determined repetitive, occupation-based exercises. Participants were asked to practice twice per day for four weeks. Accelerometer data were downloaded and imported into Excel and exercises were cut into individual practice sessions according to participant reported practice times. Angular velocity and acceleration data were graphed in a custom Matlab program and the start/stop of each set was visually identified. Key outputs included duration, average angular velocity, and average acceleration for total practice and individual sets of practice. Comparisons were made between the following: 1) the number of self-reported sessions and the number of sessions captured via accelerometry, 2) self-reported duration of practice and duration captured via accelerometry, and 3) average angular velocity captured from total practice sessions and individual sets within a practice session. Descriptive statistics have been completed as an initial analysis.
RESULTS: Preliminary data analysis from four participants indicated key findings for data capture, accuracy of self-reported time, and trends in angular velocity. Across practice weeks, the average percent of self-reported data captured via the accelerometer ranged from 37.39% to 71.04% (Median = 49.68%). The average percent of active time during practices, calculated by accelerometer individual set time over the total practice time, ranged from 63.34% to 91.01% (Median = 80.73%). Additionally, three out of four participants on average had higher self-reported total practice times, with the difference between self-reported and accelerometer captured ranging from 26 to 107 seconds. In all four participants, individual set average angular velocity was higher than total practice angular velocity.
CONCLUSION: Only 50% of self-reported data were accounted for via accelerometers, despite trainings and weekly check-ins to facilitate data capture. In future work, the use of cloud-based data monitoring of accelerometer data will be essential to increase amount of data captured and support real-time technology related problem solving. When accelerometer data was captured, it was reflective of self-reported data for duration and the differences observed were likely due to elapsed time between writing the start/stop time and when practice began/ended.
IMPACT STATEMENT: This project highlighted the challenges and the potential parameters for clinical interpretation of accelerometer data for use in home programs which will support the extension of occupational therapy practice into the community.
References
Bailey, R. R., Birkenmeier, R. L., & Lang, C. E. (2015). Real-world affected upper limb activity in chronic stroke: An examination of potential modifying factors. Topics in Stroke Rehabilitation, 22(1), 26-33. https://doi.org/10.1179/1074935714Z.0000000040
Noorkõiv, M., Rodgers, H., & Price, C. I. (2014). Accelerometer measurement of upper extremity movement after stroke: A systematic review of clinical studies. Journal of Neuroengineering and Rehabilitation, 11(1), 144. https://doi.org/10.1186/1743-0003-11-144
Reiterer, V., Sauter, C., Klösch, G., Lalouschek, W., & Zeitlhofer, J. (2008). Actigraphy–a useful tool for motor activity monitoring in stroke patients. European Neurology, 60(6), 285-291. https://doi.org/10.1159/000157882
Uswatte, G., Giuliani, C., Winstein, C., Zeringue, A., Hobbs, L., & Wolf, S. L. (2006). Validity of accelerometry for monitoring real-world arm activity in patients with subacute stroke: Evidence from the extremity constraint-induced therapy evaluation trial. Archives of Physical Medicine and Rehabilitation, 87(10), 1340-1345. https://doi.org/10.1016/j.apmr.2006.06.006
