Effectiveness of a Wearable Device-based Assisted Post-stroke Rehabilitation System

Overview

Multidisciplinary rehabilitation facilitates functional recovery after stroke, and adherence to rehabilitation guidelines is associated with improved outcomes. We assessed the efficacy of a wearable device-assisted rehabilitation system in addition to conventional therapy compared to conventional therapy alone. A total of 127 eligible acute stroke inpatients were enrolled. There were 76 males (59.8%). The mean age of the participants was 57.45 years; 63 and 64 participants were randomized to the WG and CG, respectively. Both groups showed significant improvements in mRS scores at 90 days (WG: -0.68 \[95% confidence interval (CI), -0.94 to -0.42\]; CG: -0.56 \[95% CI, -0.74 to -0.38\]; P\<0.0001), but only the WG had significant improvements between 30 and 90 days (-0.36; 95% CI, -0.72, -0.007; P=0.049). Both groups had significant improvements in secondary outcomes at 90 days, but there was no difference between groups.

Stroke remains the leading cause of chronic physical disability worldwide and the second most common cause of acute mortality despite great progress in acute stroke treatment during the past decades. Adherence to rehabilitation guidelines after acute stroke is associated with improved outcomes. Current stroke guidelines recommend early and continuous rehabilitation to facilitate functional recovery and a good prognosis, and more than two-thirds of stroke patients are encouraged to participate in outpatient rehabilitation programs after hospital discharge. The efficacy of rehabilitation might be enhanced by increasing the rehabilitation time using mobile communication devices (telerehabilitation). Moreover, the use of telerehabilitation should be emphasized during the era of coronavirus disease 2019 (COVID-19), when medical resources are limited and in-person contact is restricted. During the Wearable-Assisted Rehabilitation (WEAR) for Stroke (WEAR-Stroke) trial, we aimed to investigate the efficacy of adding a 30-minute smartphone-assisted or smartwatch-assisted session to conventional in-person rehabilitation. Several innovative interventions have been introduced beside the conventional rehabilitation because of increasing awareness of the brain's regenerative potential and advances of technology by utilizing visual feedback, transcranial magnetic stimulation and robotics. An alternative way of delivering post-stroke rehabilitation, telerehabilitation, has been proposed by information and communication technologies. The devices could detect, store and upload their physical activity recorded in the sensor-embedded mobile devices of smart watches and smartphones, with connection to the internet and get real-time feedback by the computation. The market penetration of smart mobile devices is rapidly rising in every industrial sector and professional field. The smartphone software applications (apps) are becoming increasingly popular for both health care consumers and providers. The investigators designed a customizable, sensor-driven, internet-based rehabilitation system consisting of a sensing watch, smartphone, a web server and database and users' interface for medical staff. The primary outcome was the mRS scores from baseline (D0) to 30 days (D30), from D0 to 90 days (D90), and from D30 to D90. Acute stroke therapies often use the mRS to assess the functional outcomes and activity limitations. The secondary outcomes included changes in the following measures at D30 and D90: Barthel index (BI) score for functional recovery of activities of daily living; static and dynamic balance ability (Berg Balance Scale, BBS) scores for 14 functional tasks with varying difficulty, including sitting, standing, transferring, reaching, stepping, and turning, rated using a 5-point scale (range, 0-4) based on the performance of the subject when completing the activity; the scores of the Functional Independence Measure (FIM), which is an 18-item tool that assesses motor domains, such as self-care, continence, transfers, and locomotion, and cognitive domains, such as communication and social cognition (each item was graded from 1 to 7 according to the level of assistance required, with 1 indicating total assistance and 7 indicating complete independence); and scores of the Stream Wolf Motor Function Test (SWMFT), which assesses the ability to perform simple and complex grasp tasks and global arm movements (range, 0-5 points, with 0 indicating unable and 5 indicating normal). The investigators successfully implement the system developed in the project into three hospitals.