Validity, reliability, and feasibility of the uSense activity monitor to register physical activity and gait performance in habitual settings of geriatric patients.

Objective: The aim of the study was to investigate the psychometric quality of a newly developed activity monitor (uSense) to document established physical activity parameters as well as innovative qualitative and quantitative gait characteristics in geriatric patients. Approach: Construct and concurrent validity, test-retest reliability, and feasibility of established as well as innovative characteristics for qualitative gait analysis were analyzed in multi-morbid, geriatric patients with cognitive impairment (CI) (n = 110), recently discharged from geriatric rehabilitation. Main results: Spearman correlations of established and innovative uSense parameters reflecting active behavior with clinically relevant construct parameters were on average moderate to high for motor performance and life-space and low to moderate for other parameters, while correlations with uSense parameters reflecting inactive behavior were predominantly low. Concurrent validity of established physical activity parameters showed consistently high correlations between the uSense and an established comparator system (PAMSysT), but the absolute agreement between both sensor systems was low. On average excellent test-retest reliability for all uSense parameters and good feasibility could be documented. Significance: The uSense monitor allows the assessment of established and-for the first time-a semi-qualitative gait assessment of habitual activity behavior in older persons most affected by motor and CI and activity restrictions. On average moderate to good construct validity, high test-retest reliability, and good feasibility indicated a sound psychometric quality of most measures, while the results of concurrent validity as measured by a comparable system indicated high correlation but low absolute agreement based on different algorithms used.