Technological monitoring of motor parameters to assess multidimensional frailty of older people in the PRO-HOME project

Abstract An interconnected system employing Kinect Azure and Fitbit Sense for continuous and non-intrusive data collection was used in the PRO-HOME protected discharge program, aiming at monitoring functional and clinical parameters in hospitalized older patients at different risks of frailty. The present study shows the findings on 30 older patients included in the PRO-HOME project. The Fitbit Sense recorded the mean daily and hourly number of steps, mean daily walked distance, and time spent inactive. Moreover, Kinect infrared camera captured gait speed and daily mean latero-lateral (body sway) and antero-posterior oscillations (lean-in). Patients underwent a standard Comprehensive Geriatric Assessment (CGA) to compute the Multidimensional Prognostic Index (MPI), including basic and instrumental activities of daily living (ADL, IADL), cognition (Short Portable Mental Status Questionnaire, SPMSQ) and nutrition, risk of pressure sores (Exton-Smith Scale, ESS), comorbidity, number of drugs and cohabitation status. Significant correlations between the mean hourly number of steps and MPI (p = 0.022), IADL (p = 0.013), SPMSQ (p = 0.006), ESS (p = 0.009), and both mean and maximum automated gait speed (p = 0.046 and p = 0.048) were found. Automated gait speed was also correlated with mean walked distance per day (p = 0.007) and lean-in (p = 0.047). Domotic technological monitoring through Fitbit Sense and Kinect Azure provides information on multidimensional frailty, including mobility and cognitive and functional status, in older people.