Assessment of Arm Activity Using Triaxial Accelerometry in Patients With a Stroke

Abstract

van der Pas SC, Verbunt JA, Breukelaar DE, van Woerden R, Seelen HA. Assessment of arm activity using triaxial accelerometry in patients with a stroke. Objective To study the validity of accelerometry in the assessment of arm activity of patients with impaired arm function after stroke. Design Cross-sectional concurrent validity study. Setting Rehabilitation center. Participants Patients (N=45) at different stages after stroke. Interventions Not applicable. Main Outcome Measures All patients wore 2 triaxial accelerometers around their wrists during 3 consecutive days. Arm activity was assessed, based on unilateral (activity of the affected arm) and bilateral accelerometry (ratio between the activity of the affected and nonaffected arm). The Motor Activity Log-26 (MAL-26) Amount of Use (AOU) scale was used as the main external criterion to test the concurrent validity of arm accelerometry. In addition, the MAL-26 Quality of Movement (QOM) scale and the Stroke Impact Scale (SIS) subscale Hand Function were used. To test the divergent validity, the SIS subscale Mobility was used. Spearman correlation coefficients were calculated. In an additional regression analysis, the hypothesized confounding influence of spasm, therapy intensity, and interobserver differences was studied. Results Both unilateral (ρ=.58, P<.001) and bilateral (ρ=.60, P<.001) accelerometry were significantly related to the MAL-AOU scale. Associations of both unilateral and bilateral accelerometry with the MAL-QOM and SIS subscale Hand Function corroborated these findings. The SIS subscale Mobility was not significantly associated with unilateral accelerometry (ρ=.41, P=.01) or bilateral accelerometry (ρ=.23, P=.11). None of the hypothesized confounders influenced these associations significantly. Conclusions Based on the results, both the concurrent and divergent validity of unilateral and bilateral arm accelerometry for measuring arm activity after stroke are good.