Feasibility of Unsupervised Therapy After Neurological or Orthopaedical Injury in the Home Setting With a Hand Rehabilitation Device (ReHandyBot) and an Assistive Smartphone Application (RehabCoach)

Many neurological (e.g., stroke, traumatic brain injury) and orthopedic patients do not fully recover hand function, with a consequent decrease in independence and quality of life. Growing evidence shows that increasing the dose of upper limb therapy for in- and outpatients could improve functional outcomes. However, due to limited resources such as the low number of therapists, providing a higher dose of therapy to patients after discharge is highly challenging.

Unsupervised technology-assisted therapy, i.e., patients training with rehabilitation technologies without external supervision or intervention, may allow an increase in therapy dose (in clinical settings and at home) with little impact on the additional resources needed.

At the Rehabilitation Engineering Laboratory (RELab, ETH Zurich) we developed ReHandyBot, a portable device for hand rehabilitation. ReHandyBot was developed for unsupervised application in the home setting and preliminary results of an ongoing study show that stroke patients can learn how to use the device independently and bring it home to keep training with it after discharge from the clinic, increasing their therapy dose. However, during the unsupervised phase at home, patients' adherence to the recommended dose of therapy with ReHandyBot (i.e., at least one therapy session per day) was not always optimal.

At home, without the supervision of healthcare providers, patients often experience a decline in motivation and engagement over time. To address this, we developed RehabCoach, a smartphone application designed to support unsupervised therapy with ReHandyBot. RehabCoach incorporates chatbots (i.e., conversations with virtual assistants) and gamification elements designed to enhance therapy adherence. Until now, RehabCoach has been tested in some pilot usability studies, but a more extended evaluation has not been performed yet.

This study, therefore, investigates the use of ReHandyBot supported by RehabCoach at the homes of patients with stroke, traumatic brain injury, or orthopedic issues. The main objectives of this study are therefore to investigate (i) the feasibility of unsupervised therapy with ReHandyBot and RehabCoach within patients' home after discharge from the clinic, (ii) compliance to the therapy schedule after discharge, (iii) the usability of ReHandyBot and RehabCoach and user experience when using them, and (iv) the feasibility of extending therapy with ReHandyBot to populations other than stroke (i.e., patients with orthopedic problems or traumatic brain injury).

Sex and gender are not relevant for this study, as we do not expect the feasibility of using the device without supervision to be dependent on them.

The primary objective of this study is to investigate feasibility of unsupervised home therapy with the ReHandyBot supported by RehabCoach and to determine subjects' compliance to unsupervised robot-assisted therapy in the home setting as measured by the amount of self-administered dose of therapy with ReHandyBot.

Main secondary objectives comprise the evaluation of usability of ReHandyBot and RehabCoach and of user experience during robot-assisted therapy and when using the app, the feasibility of extending therapy with ReHandyBot to individuals with traumatic brain injury or orthopedic injury, functional recovery of the upper limb, parameters possibly influencing the use of ReHandyBot and RehabCoach, the impact of the goal and reward systems implemented on RehabCoach on motivation to train with ReHandyBot, as well as the feasibility and usability of the educational program delivered by the app.

The primary outcome of this study is the dose (e.g., therapy time, number of repetitions) of self-administered robot-assisted therapy performed by subjects in the home setting without supervision.

Main secondary outcomes are the number of subjects who can use ReHandyBot and RehabCoach unsupervised at home out of the total number of tested subjects, attendance to the therapy schedule during the unsupervised phase at home, adherence to the interactions required by RehabCoach, the number of times and reasons why the investigators need to intervene to help participants with eventual issues during the phase at home, safety, usability of ReHandyBot and RehabCoach and user experience (System Usability Scale, mHealth App Usability Questionnaire for RehabCoach, Customer Satisfaction Score, Net Promoter Score, open interview, etc.), number of additional therapy sessions initiated by the patients compared to the minimum recommendations, functional recovery.

Participants will receive robot-assisted therapy sessions with the ReHandyBot in parallel to their conventional therapy program for a period of around nine weeks. Additionally, they will interact with an app, RehabCoach, which will motivate patients to maximize the amount of training they perform with ReHandyBot.

This study consists of the following phases:

* Preparation phase: This phase includes a screening visit and an evaluation session with a therapist to perform baseline clinical assessments. * Supervised phase: During the first week, subjects perform 3 sessions of 30 minutes of therapy with ReHandyBot supervised by a therapist or researcher. This phase is performed at the clinic and is used to familiarize the subject with the rehabilitation device. Additionally, the participants are also introduced to RehabCoach. * Minimally supervised phase: During the second week, subjects interact with RehabCoach daily and perform 3-7 sessions of 30 minutes of minimally supervised therapy with the device at the clinic. The supervisor is still present but helps only in case patients encounter problems or if they have any questions. At the end of this phase, the supervisor uses a checklist to determine if the subject can proceed to the unsupervised phase. If not, the participant stops the study at the end of the minimally supervised phase. At the end of the minimally supervised phase, subjects need to fill in some questionnaires to rate the usability of ReHandyBot and RehabCoach. * Unsupervised phase: If patients are ready for unsupervised training, during the unsupervised phase they take the ReHandyBot home and have the opportunity to independently exercise with the robot, with the recommendation of training at least 30 minutes per day (i.e., one complete therapy session per day) but also informing them that the more they train, the better. Patients who can use RehabCoach at home will also continue with the daily interactions with the app. * Final phase: At the end of the protocol, subjects repeat the clinical and robotic assessments and the usability evaluation.

• Feasibility - Adverse events [Time frame: This is monitored over the whole protocol duration (i.e., from beginning of the first week to the end of the six weeks of unsupervised therapy at home).]
• Feasibility - Device deficiencies [Time frame: This is monitored over the whole protocol duration (i.e., from beginning of the first week to the end of the six weeks of unsupervised therapy at home).]
• Feasibility - App deficiencies [Time frame: This is monitored over the whole protocol duration (i.e., from beginning of the first week to the end of the six weeks of unsupervised therapy at home).]
• Feasibility - Number of subjects performing unsupervised robot-assisted training [Time frame: This is calculated at the end of the study, after performing the final measurements (i.e., at the end of the six weeks of unsupervised therapy at home).]
• Feasibility - Attendance [Time frame: This is calculated at the end of the study, after performing the final measurements (i.e., at the end of the six weeks of unsupervised therapy at home).]
• Usability as assessed by the System Usability Scale (1) [Time frame: This is measured during the intervention (at the end of the familiarization phase, before the 6 weeks unsupervised at home start).]
• Usability as assessed by the System Usability Scale (2) [Time frame: Usability is measured again at the end of the study protocol (i.e., at the end of the six weeks of unsupervised therapy at home).]
• Usability as assessed by the raw NASA Task Load Index (1) [Time frame: This is measured during the intervention (at the end of the familiarization phase, before the 6 weeks unsupervised at home start).]
• Usability as assessed by the raw NASA Task Load Index (2) [Time frame: Usability is measured again at the end of the study protocol (i.e., at the end of the six weeks of unsupervised therapy at home).]
• Usability as assessed by the mHealth App Usability Questionnaire (MAUQ) for RehabCoach (1) [Time frame: This is measured during the intervention (at the end of the familiarization phase, before the start of the 6 weeks unsupervised at home).]
• Usability as assessed by the mHealth App Usability Questionnaire (MAUQ) for RehabCoach (2) [Time frame: This is measured again at the end of the study protocol (i.e., at the end of the six weeks of unsupervised therapy at home).]
• Usability as assessed by the Chatbot Usability Questionnaire (1) [Time frame: This is measured during the intervention (at the end of the familiarization phase, before the start of the 6 weeks unsupervised at home).]