ORIGINAL REPORT
Axelle GELINEAU, MSc1, Anaick PERROCHON, PhD1, Jean-Christophe DAVIET, PhD12 and Stéphane MANDIGOUT, PhD1
From the 1HAVAE Laboratory UR 20217, University of Limoges and 2Department of Physical Medicine and Rehabilitation, University Hospital Center of Limoges, Limoges, France
Background: Telerehabilitation and follow-up techniques have been developed in recent years to assess the effectiveness of diverse intervention programmes that include exergaming technologies. For patients with upper limb impairment after stroke, motion-gaming technologies can provide effective and amusing training. Beyond efficiency, professionals must analyse patient compliance with the system for self-use at home, because patients may or may not independently perform the exercises prescribed by the therapist. Questions on the sustainable use of this type of home exercise also arise.
Objective: This review examines user compliance with exercise programmes, measured according to the training rate (percentage of prescribed sessions and minutes completed) and completion rate (number of drop-outs and discontinued interventions) reported or calculable according to the data collected.
Results and discussion: Rates of compliance with training were relatively high. No group effect on compliance was found. Drop-out and discontinued intervention rates were either due to external causes or directly related to the technologies. Some studies have reported the use of supervision, most of them through home visits and remote support. Few studies performed long-term follow-up, which could provide information to help broaden practices. This narrative review considers how this field of research may evolve in the future.
The use of video games in hospitals as a rehabilitation tool in neurology is developing, particularly for stroke victims. For patients with arm problems, it can be effective and fun to use gamified systems. When the patient goes home, they must continue their rehabilitation in order to continue to progress or maintain their skills. However, performing exercises alone at home raises questions about patients’ compliance with the exercises prescribed by their therapists. Do patients complete the prescribed sessions? Are there occasional or permanent interruptions? This narrative review attempts to address these questions. The review also examines the obstacles that might hinder the use of these technologies and the facilitators that may help compliance.
Key words: stroke; compliance; exergame; motion capture; virtual reality; technology; telerehabilitation; home.
Citation: J Rehabil Med 2022; 54: jrm00325. DOI: https://dx.doi.org/10.2340/jrm.v54.2270
Copyright: © Published by Medical Journals Sweden, on behalf of the Foundation for Rehabilitation Information. This is an Open Access article distributed under the terms of the Creative Commons Attribution-NonCommercial 4.0 International License (https://creativecommons.org/licenses/by-nc/4.0/)
Accepted: Jul 21, 2022; Epub ahead of print: Aug 17, 2022; Published: Nov 24, 2022
Correspondence address: Axelle Gelineau, Faculté des Sciences et Techniques, Laboratoire Handicap, Activités Vieillissement, Autonomie, Environnement (HAVAE, EA 6310), Université de Limoges, 123 avenue Albert Thomas, 87060 Limoges, France. E-mail: axelle.gelineau@unilim.fr
Competing interests and funding: The authors have no conflicts of interest to declare.
This research received funding from the RGS@Home project H2020-EU. EIT Health, ID: 19277.
Telerehabilitation offers an alternative rehabilitation technique through the delivery of rehabilitation and habilitation services via information and communication technologies (ICTs) (1, 2). Telerehabilitation is a useful approach in outpatient rehabilitation, especially for post-stroke patients (3). Stroke was the second-leading cause of death and third-leading cause of death and disability combined worldwide in 2019 (4). Stroke is highly prevalent among older patients, many of whom have underlying health issues (e.g. diabetes, hypertension, and cardiovascular disease). Upper extremity paralysis is a predominant impairment after stroke, with a recovery rate of 10–20% (5), affecting independence and quality of life (6). This reflects the importance of post-hospital discharge support programmes. Telerehabilitation is one possible approach in outpatient stroke rehabilitation, with follow-up at home.
Exergaming, otherwise known as exercise-based games (e.g. virtual reality and interactive video-game interventions) is a relevant alternative home-based rehabilitation for persons with neurological diseases. The effectiveness of these interventions has been shown to be at least equivalent to conventional therapy or usual care (7). Most studies have focused on assessing the effectiveness of the technologies, but patient compliance with these technologies is under-explored. Even if the technology can be adapted technically or physiologically, it remains necessary for the patient to want to use it; therefore, effective implementation for self-use at home depends on patient compliance.
Medical compliance was defined in the late 1970s as “the extent to which the patient’s behavior in terms of taking medications, following diets, or executing other lifestyle changes coincides with medical or health advice” (8). Exercise compliance is defined as “a person’s compliance with a prescribed or self-prescribed fitness program” (9). The terms compliance, adherence, and concordance are often used interchangeably (10). The term compliance is used throughout this narrative review as an indication of positive patient behaviour in following an exercise programme. Many people do not feel motivated to engage in new habits, including exercise. In 2013, a meta-analysis showed that approximately 21% of people did not intend to take up physical activity, while 36% intended to, but found changing their sedentary behaviour difficult (11). Training is often limited by a lack of motivation, which can be the main reported reason for non-adherence with home exercise by individuals with chronic stroke (12). Designing optimal rehabilitation treatment programmes for stroke patients requires an understanding of “What” is the content of the treatment, “How much” treatment is required, and “When” treatment is best delivered (13).
This narrative review assesses patient compliance according to the type of upper limb technology used and the home implementation parameters.
Exergaming is defined as the integration of physical activity into a video-game environment that requires active body movements to control the game (14). The emergence of commercial video-game systems, coupled with handheld controllers and motion-capture devices, has facilitated the use of computer games for neurorehabilitation. These technologies have the potential to be effective for increasing upper limb capacities (15–18). Game-based systems may offer a motivational exercise environment that encourages continued use. Health professionals and participants have reported high levels of satisfaction and acceptance of telerehabilitation interventions (5). Piron et al. (2008) showed that post-stroke patients with arm motor impairments assigned to the home-based virtual reality group were able to engage in therapy at home through a user-friendly system, and the videoconferencing system ensured a good relationship between the patient and the remote therapist (19). Their study was based on the patient’s degree of satisfaction, an important indicator of the efficacy of the therapeutic intervention, which improves the patient’s motivation to engage in rehabilitation.
A range of commercial gaming systems are described in the literature for home use: Kinect™ (Microsoft Corporation; Redmond WA, USA) (30, 32, 34, 43), Wii™ (Nintendo, Japon) (25–28), Vive™ (HTC, Taïwan) (29), PlayStation® 3 Eye Move controller (Sony, Japon) (26), Leap Motion controller (Ultraleap, USA) (30) (Table I). These devices are mostly associated with custom games and other specific rehabilitation devices, such as robotic devices, specific controllers, and passive arm support (e.g. TheraBot system (Rehabilitation Robotics and Research and Design Lab, Milwaukee, WI, USA) (31), SaeboMAS (Saebo Inc., Charlotte NC, USA) with SCRIPT dynamic wrist and hand orthosis sensor (32), HandinMind (HiM) system (33) (Hocoma AG, Switzerland), P5 Glove (Essential Reality Inc; NY, USA) (24), Armeo®Senso (Hocoma AG, Switzerland) (34), BrightBrainer™ Grasp (Bright Cloud International, NJ, USA) (29), Myo Band (n.a) (26), MusicGlove (FlintRehab, CA, USA) (35), CyWee Z controller (CyWee Inc., Taiwan) (36) and passive arm support (30)) (Table I).
| Author, year | Design Methods Data collection | Participants | Technology | Implementation parameters (number of sessions, frequency, and length) | Home training (min) | Drop out | Discontinued intervention |
| Brokaw et al. (43) | Pilot evaluation of the usability and utility in the home Qualitative Interview | 1 stroke patient | Computer screen, Kinect™ (Microsoft Corporation; Redmond WA, USA), 5 custom games | 19 sessions 30 min, 5 days/week, for 4 weeks | NR | No | 1 session hardware not used final 3 training sessions |
| Burdea et al. (29) | Feasibility study Mixed Survey + training | 7 chronic post-stroke patients 7 caregivers | BrightBrainer™ (Bright Cloud International, NJ, USA) computer, Vive™ (HTC, Taïwan) and BrightBrainer Grasp controllers, custom games, Automatic electrical meter | 20 sessions wk 1 20 min wk 2 25 min wk 3 30 min wk 4 40 min 5 days/week, 4 weeks | 605 min | 1 screening failure (low MOCA score) | NR |
| Chen et al. (26) | Qualitative study in a randomized trial Qualitative semi-structured interviews | 13 stroke patients | Computer, Myo Band, Wiimote™ (Nintendo, Japon) in a pistol-shaped holder, Power Mate, PlayStation® 3 Eye Move controller (Sony, Japon), Joystick, Logitech Trackpad, real-time video guidance | 36 sessions 70 min at a fixed time every day 6 days per week, over 6–8 weeks | NR | No | NR |
| Dodakian et al. (35) | Pilot study Quantitative Survey | 12 stroke patients | Laptop, wrist accelerometer sensor, MusicGlove (FlintRehab, CA, USA), 18 games, videoconference | 1 structured hour + 1 h of free play, for 28 days (2 × 14-day separated by a 1–3-week break) | Active time (games+exercises): 60 ± 10 min/d, including a mean of 22 ± 24 min/d of free play. Total time (active time + education questions, measuring blood pressure, reading game instructions, donning devices, and taking breaks between tasks): 182 ± 61 min/d. | No | 1 not completed 4 sessions (fatigue) 1 not completed 1 session (hardware malfunction) 1 missed 2 complete sessions (other medical appointments) |
| Fluet et al. (30) | A Feasibility and pilot study – comparison Mixed Survey + training | 11 stroke patients: 5 enhanced motivation (EM) group + 6 unenhanced control (UC) group | Leap Motion controller (Ultraleap, USA) passive arm support, custom-designed simulations | As much as possible, but at least 20 min, daily for 12 weeks | > 400 sessions. Enhanced motivation group: 95±95 min per week (range 40–276 min) Unenhanced control group: 35 ± 31 min per week (range 3–93min) | NR | NR |
| King et al. (36) | Case series (after a trial of 10 sessions of bilateral therapy using VR) (Hijmans et al.2011) Mixed Survey + Training | 3 patients with chronic stroke | Computer screen, CyWee Z controller (CyWee Inc., Taiwan) games | Chose when and for how long for in each session, for no longer than 90 min on any given day Each game at least once, after free to choose the proportion of time on any games for 8 weeks | ≥ 35 min per session ≥ 4.5 times per week ≥ 33.5 hcontinued for between 55 and 61 days at home | No | Diary: mean number total of days missed: 11.66 |
| McNulty et al. (25) | a randomized controlled trial Mixed Survey + structured interview + training | 41 stroke patients (21 + 20) | Wiimote™ (Nintendo, Japon) Wii Sports™ games | 10 sessions 60 min weekdays, progressively increasing, for 14 days follow-up 6 months | Wii-based Movement Therapy (WMT): 1,188 min modified constraint-induced movement therapy (mCIMT): 1,194 min. Completion rates 105.7 (93.6–114.7)% (WMT) and 101.0 (87.6–108.1)% (mCIMT) | 2 EG (withdrew+death) 1 CG (unrelated medical condition) | NR |
| Nijenhuis et al. (32) | Pilot randomized controlled trial Mixed Survey + semi-structured interview + training | 20 stroke patients (10 + 10) | Computer, SaeboMAS (Saebo Inc., Charlotte NC, USA) SCRIPT dynamic wrist and hand orthosis sensor, games | 36 sessions at least 30 min per day, 6 days a week, for 6 weeks follow-up 2 months | CG: 189 (143–266) min per week EG: 118 (51–176) min per week ranging from 13 to 423 min per week | 1 EG (shoulder pain due to external causes) | NR |
| Sivan et al (49) | Feasibility study Qualitative semi-structured interview | 17 (9 prototype) patients + 7 therapists | Computer screen, powered joystick, controlled assistance, 8 games | 8 weeks | NR | NR | NR |
| Standen et al. (27) | Prospective cohort study plus qualitative analysis Mixed Semi-structured interview + training | 29 stroke patients, including 17 in the qualitative study | PC, Virtual glove, Wiimote™ (Nintendo, Japon), 3 custom games | 3 times a day for periods of no more than 20 min, for 8 weeks | Percentage of the duration of use ranged from 1.46 to 70.6 percentage of days used ranged from 10% to 100% | Allocation: 4 EG (family issues, not interested, arm pain, severe aphasia) Follow-up: 1 CG (measures onerous) + 3 EG (illness, ill family member, going on holiday) | NR |
| Thielbar et al. (21) | Randomized trial - crossover, interventional study Mixed Survey + Training | 20 chronic post-stroke patients (10 in the FSU group and 10 in the FMU group) | Computer, Kinect™ Microsoft Corporation; Redmond WA, USA) wireless mouse. 3 custom exercises, voice communication, remote central server, custom Google Cloud Virtual Machine | 8 sessions 4 sessions of 1 h each week, for 4 weeks: 2 weeks with multi-user (MU) + 2 single-user (SU) version | For the SU mode, FSU group: 30.2 ± 10.8 min and FMU: 39.9 ± 12.8 min For the MU mode, the FSU and FMU groups had similar times | 1 (medical reasons) 1 FMU group (declined participation) | MU mode: 18 of 20 subjects participated in all 8 sessions, and all 20 subjects participated in at least 7 sessions SU mode: only 15 of 20 subjects participated in all 8 sessions, and 16 of 20 participated in at least 7 sessions |
| Wingham et al. (28) | Pragmatic multi-centre RCT with a qualitative study and health economics analysis Qualitative semi-structured interview | 19 stroke patients 10 caregivers | Television, Wii console, Wii games | For up to 45 min daily for 6 weeks followed up at 6 weeks and 6 months | EG: 37 min (range 4–108 min) CG: 32 min (range 4–63 min) | 1 (timetable) | NR |
| Wittmann et al. (34) | Feasibility study Mixed Structured patient interview + training | 11 stroke patients | Armeo®Senso (Hocoma, Switzerland) = wearable inertial measurement units (IMU) | As often as possible for 6 weeks | 26.5 ± 11.5 (min 8, max 41) days out of 42 days duration per week: 137 ± 120 min (min 15, max 357) training sessions on 4.4 days per week training (gaming) duration per session: 30 ± 16 min (min 11, max 56) | NR | NR |
| NR: non-recorded; EG: experimental group; CG: control group; EM: enhanced motivation group; UC: unenhanced control group; WMT: Wii-based movement therapy; mCIMT: modified constraint-induced movement therapy; SU: single user; MU: multi-user, wk: weeks; min: minimum; max: maximum; MoCA: montreal cognitive assessment. | |||||||
Some studies included real-time audio or video systems (30, 31, 39, 43, 47). The exergames used in the studies were specific or non-specific systems for upper-limb stroke rehabilitation. Devices can be used alone or in combination with other technologies (e.g. exoskeleton, robotics, exogenous stimulation, virtual reality headsets, and augmented reality), potentially further improving recovery. It is of interest to determine whether combinations can have a positive impact on compliance. For example, in the REINVENT platform (37), researchers combined the principles of action observation in head-mounted virtual reality with brain-computer interface (BCI) neurofeedback for stroke rehabilitation to try to elicit optimal rehabilitation gains. It would be of interest to assess the relevance of this on functional activities and in different settings (e.g. laboratory, clinic, home). As reported by Perrochon et al. (2019), future studies in the home should include the multiple observations reported in the literature to ensure an optimal exergaming design (7, 38).
Several systematic reviews have indicated that intensive treatment is favoured, but there is no consensus on the optimum amount, intensity, distribution, or duration of therapy (39). Across a large number of studies, the key elements of task-specific training are repeated, challenging the practice of functional, goal-oriented activities (40–42). Motion-tracking systems provide real-time feedback regarding execution of task-specific exercises, but it is unclear which is the most efficient training. To our knowledge, there are no guidelines regarding the appropriate maximum duration of research and clinical sessions to maintain compliance. Dose, frequency, and duration scheduling of home-based therapy are shown in Table I.
The overall dose ranged from 20 min (29) to 120 min per day (1 h structured activity and 1 h free play) (35). The recommended time increased stepwise each week for 1 study (29), whereas another progressively increased the time over 10 consecutive weekdays (25). Three studies did not give the number of expected min per day (30, 34, 36). Three studies indicated a time limit not to be exceeded; users played for as long as they wanted in a session and whenever they wanted, but for no longer than 90 min on any given day (36); 3 times per day for periods of no more than 20 min (27); or for up to 45 min daily (28). Conversely, another study indicated a minimum time limit requirement of 20 min daily (30). Time of day can be defined to accommodate the necessary appointments with therapists (21, 26). The current review observed variability in the prescriptions: studies proposing 30 min per day were considered as sub-therapeutic. In contrast, planning 1–2 h per day of task time and assessing repetitions and intensity accurately were recommended (13).
The number of sessions ranged from 8 (21) to 36 (26, 32). Six studies did not specify the number of expected sessions ((27, 28, 30, 34–36). The number of days per week were set at 5 (29, 43), 6 (26, 32) or 7 (28, 35). Hence, the sessions were discontinuous or continuous, depending on the study.
The most often used training duration in the studies was 6 weeks (26, 28, 32, 34, 36). Other studies had a total intervention duration ranging from 14 days (25) to 12 weeks (30). Only 1 study planned an interruption after 14 days of 1–3 weeks halfway through the intervention (35). They explained that 14 days was selected based on the success of the EXCITE trial of constraint-induced therapy, which showed benefits when making daily demands on patients for 14 days (44). Only 3 studies proposed follow-up to assess the persistence of improvements in motor function. They did not indicate any information on long-term engagement.
Some authors referred to the training rate to assess compliance quantitatively. The ratio of the quantity of training performed by the patient to the quantity prescribed by the therapist was reported or calculated. Some authors considered the total training time, in minutes, performed by the study subject, the mean number of sessions per week, and the mean minutes per session for home-based interventions. These data can be compared with the authors’ recommendations.
Certain studies calculated high compliance rate of approximately 96% (29). Other authors compared groups, for example, compliance was very high (99%) for the multi-user mode and 89% for the simple-user mode, suggesting that the training mode can affect the compliance (21). More, the change in the difficulty and complexity showed a greater increase in training time per session than the control group (30). Some other studies have a compliance rate of over 100, taking 100% as the time prescribed by the therapist. Indeed, some studies allow patients to do more exercises if they wish following their prescribed time (30, 32). For example, the completion rate was 105.7% (93.6–114.7%) for the Wii-based movement therapy and 101.0% (87.6–108.1%) for the control group (median, interquartile range) (25). Conversely, a high compliance rate was found for the control group (105%), but not for the experimental group using computerized gaming exercises (65.6%) (32).
A few studies reported the number of repetitions performed during the intervention (29, 34, 35). However, with few comparative data, we chose not to report the data in this review.
Participants in the study by Nijenhuis et al. (32) recorded the frequency and duration of training in a diary to assess user compliance with training duration and motivation. King et al. (36) used the same methods to assess patient engagement quantitatively, using diaries to record the occurrence and duration of the intervention.
The attrition rate is calculated as the number of drop-outs and discontinued interventions. Within eHealth interventions, the exponential decrease in adherence has been described as the “Law of Attrition” (45). It is essential to consider the main causes of the rate and to set up tools to reduce the level of attrition as much as possible. Dropping out of therapy is an implicit marker of dissatisfaction or unacceptability. Variability in acceptable discontinued intervention/drop-out rates was found in the literature, ranging from 5%, to 20%, and 30% for follow-up of more than 1 year (46).
Literature results showed a high attrition rate for the home-based setting, although the drop-out rate was higher for the experimental group than for the control group (25, 27, 32). The completion rate was not described in all studies. However, among the studies that evaluated the completion rate, variable results were found. For example, a 94.7% rate for intervention completion was reported by Brokaw et al. (43), with only 1 missed session and a 97.9% rate was described by Dodakian et al. (35) with 3 patients not completing or missing sessions. King et al. (36) provided the number of days of intervention over the total intervention period for each subject, recorded by the participants in diaries, and reported a rate of 79.9%. Thielbar et al. (21) compared the training attrition rate between groups: in the multi-user (MU) mode, 90% of subjects participated in all 8 sessions, and 100% participated in at least 7 sessions. However, in the single-user (SU) mode, only 75% of subjects participated in all 8 sessions and 80% participated in at least 7 sessions (21).
Non-compliance was intentional or non-intentional. Some factors were due to internal factors (no interest, pain, fatigue, severe aphasia, seizure, illness, death, etc.) and others to external factors (family issues, ill family member, holidays, schedule). This led to limited compliance, and patients either dropped out or discontinued the interventions. Some non-compliance factors were correlated with technical issues: Brokaw et al. (43) reported that participants completed 18 of the planned 19 sessions, but hardware issues prevented participants from using the system for the final 3 training sessions; Some studies reported technical issues with the hardware (26, 29, 35, 43, 49), the software (35), communication reliability causing storage data problems (29), the home setting (26, 29, 49), unsuitability for use with spastic hands (29), and time constraints (26). These aspects did not necessarily stop the patients from participating in the study. As reported by Perrochon et al. (2019), the rate of drop-out and discontinued interventions were due either to external causes or directly related to the technologies (7). Dodakian et al. (35) produced a summary document of the issues, solutions, and lessons learned for future telerehabilitation studies. This initiative is relevant for future perspectives. As Ong et al. (2018) stated, identifying manipulable aspects of treatment that reduce the probability of drop-out can guide the development of more acceptable interventions (47).
Some authors proposed a familiarization stage with the device before home use. Some studies provided laboratory familiarization before home use: 1 session (43), a brief 2- or 3-day orientation (21), and 10 sessions (36). Familiarization directly at home was also prescribed: 1 session (29) and 2 sessions (43). Other studies established an initial visit for training purposes (27, 30). These sessions equipped participants with the skills to continue the programme safely and independently at home.
Other means were established to carry out the intervention. One study supplied an instruction manual at the beginning (frequently asked questions and tips) (27), whereas another used a transfer package in each session including different forms to emphasize patient safety, monitor home practice, improve handling and compliance, and provide a forum for patient-centred goal-setting and problem-solving (25). This allows for self-care before calling in the therapists.
However, remote assistance was necessary in most of the studies for which it was made available. Technical issues were expected during studies with this type of technology. Some systems can be used as remotely monitored telerehabilitation; for example, some studies used only telephone calls while others used videoconferencing, education videos, web-based chats, or virtual reality systems (48). Clinicians can propose appropriate responses by adapting and personalizing the planned rehabilitation activities and offering technical assistance when needed. It is important to verify whether the use of telerehabilitation systems can improve patient engagement by conducting their rehabilitation training at home in a safe environment.
Four studies supervised the interventions by combining home visits and remote support via phone calls (27, 29, 30, 43), whereas 1 study relied only on home visits (32). Two studies used videoconferencing and remote support (26, 35) and 1 study scheduled remote appointments using the remote therapist-patient systems and telephone support (21). Remote access to data on participants' playing performance can be used to verify compliance with the protocol (29). It can be a way to automatically inform the researcher if the data has not been uploaded for three consecutive days (29). The first home visits included a time dedicated to installation and explanations with patients and possibly caregivers. Some studies did not offer supervision (34, 36, 49).
Furthermore, some authors requested patient feedback. Participants in the study by Chen et al. (26) specified having received support following minor technical issues. Participants and caregivers in the study by Wingham et al. (28) said they felt reassured and supported by the weekly phone call, but also appreciated the social connection.
Dodakian et al. (35) measured the assistance given by phone during the study. The authors supposed that the decrease in assistance required over time might be due to patient familiarity with the system and improved performance of the study team.
Videoconferencing allowed therapists to observe patient performance, answer questions, review the treatment plan, and, on select days, perform brief study assessments (26). In the same study, daily 30-min half-therapy sessions were supervised by videoconference, followed by 40 min during which the patient was alone. Dodakian et al. (2017) proposed 3 videoconferences per week, focusing on the prior week’s activities, followed by a structured interview including specific inquiries on pain, adverse events, and the telerehabilitation system performance. The authors explained that providing feedback and encouragement could increase motivation, according to the standard of care (35). Johnson et al. (2011) described simple encouragement provided through video and audio feedback via a robot assistant (31). In addition to the aspects of communication with the user and monitoring of progress, it could also allow for remote monitoring of the the user's position when using the device (49).
Fluet et al. (30) described modest motivational enhancements and discussed areas for increasing the training time per session. One suggestion was that supervision in the form of email and text message reminders of weekly tasks could increase the number of training sessions.
It is important to make the games interesting and stimulating to increase the use of the device (49). Integration of the multi-player mode within a game-based task could also be a factor in increasing compliance (21, 49).
Compliance behaviour is complex, requiring an analysis based on theoretical models. Over the years, specific theoretical models for studying compliance have been described. In the Health Belief Model (HBM), compliance is determined by the knowledge and attitudes of the patient (50). There are 4 essential areas in the development of compliant behaviour: perceived susceptibility; perceived severity; perceived benefits; and perceived barriers. More recent forms of the HBM emphasized 2 other factors in the decision to engage in a behaviour: (i) self-efficacy, which is the patient’s belief that he or she is capable of taking the recommended action; and (ii) action cues, which are aids that teach or remind the patient of the recommended action (51). Designing a novel technology for stroke patients involves a deep understanding of the persons who use the system and perform the activity, and the context in which that activity takes place. User expectations and needs must be understood, as must their motivations, and previous experiences (52). The context of use is also a key element to explore at this pre-implementation stage and to evaluate at the post-implementation stage. The architectural, material, and human environment must be considered with regard to use of the system in the patient’s home. Evaluating user experience with interactions and the use of the system is relevant to obtain patient reactions in post-task interview questions (53), and in particular, in terms of compliance with use and engagement.
Most researchers used an off-the-shelf system (i.e. a commercial system) as a primary tool and combined it with elements or games adapted to neuro-rehabilitation. According to Tamayo-Serrano et al. (2018), low-cost solutions are expected to promote the adoption of in-home rehabilitation systems by patients and health organizations (54). To be able to assess cost-effectiveness, quantified cost and time data must be considered. For example, Lloréns et al. (55) reported the cost of telerehabilitation and in-clinic programmes. Costs were reported in terms of human resources (time spent on assistance and guidance during the intervention, progress monitoring, and troubleshooting), round trips to the neurorehabilitation unit, and instrumentation (laptop, Kinect™, and internet access) (55). The cost of the clinical instrumentation was not considered, but this would have provided additional information for comparison. More transparency on the costs incurred would be useful for future studies. Table II summarizes the current literature on compliance.
This narrative review identified current practices for user involvement in the development of stroke-patient decisions on home-based gaming technology. This review studied compliance with technologies adapted to post-stroke rehabilitation. The literature on the subject is recent. Heterogeneity is found on the type of technologies, the intervention regime, and the supervision, making it difficult to draw conclusions on compliance. More robust studies are needed to provide additional data.
In general, home-based gaming therapies are well received by patients and no significant problems occur. Additional experimental studies are required to understand which determinants, intrinsic to devices, impact the user’s compliance. Further research into the conditions for the personalized specification of the technological devices, developed in co-construction with the user, would be useful. Current studies focus on short-term evaluations; however, a long-term view is necessary to assess user compliance for high-ecological validity. Future studies should investigate how emerging technologies enable long-term use and the transfer of the acquired knowledge to everyday life, and should identify resources and assess the suitability of the system’s settings based on the best situation for home use.
The authors thank Hugo Landais and “La Fondation de l’Avenir pour la recherche médicale appliquée” (Paris, France) and the European Institute of Technology (EIT Health) for their scientific insights and unyielding support.
All those designated as authors meet the International Committee of Medical Journal Editors (ICMJE) requirements for authorship.