In Canada, stroke survivors have difficulty accessing community-based rehabilitation services because of a lack of resources [1]. Evidence indicates that there is potential for recovery, even several years after stroke [1,2]. However, rehabilitation services are generally provided in the acute and postacute stages [1]. A common long-term consequence of stroke is hemiparesis, or weakness on one side of the body, leading to loss of upper extremity (UE) motor function with a significant long-term impact on everyday activities [3]. Given the chronic nature of stroke, it is essential to develop interventions that provide community-dwelling stroke survivors opportunities for further personalized training.

Telerehabilitation and virtual reality technologies could play an important role in providing novel rehabilitation approaches to optimize stroke recovery in the chronic phases, as suggested by Canadian stroke guidelines [1]. More specifically, telerehabilitation can be used to increase accessibility to rehabilitation programs and follow-up for persons no longer receiving rehabilitation (or discharged from intensive rehabilitation), whereas virtual reality technologies, which involve engaging activities for practice, can provide the intensity needed for optimal recovery. Moreover, behavioral and motivational techniques [4] could be used with these technologies to empower stroke survivors to continue exercising and using their affected UE in everyday activities (eg, brushing their hair, getting dressed, and eating). A few studies have examined the combined use of telerehabilitation and virtual reality technologies in stroke survivors [5-7]. These studies reported an improvement in UE motor function and high adherence to the treatment plan, which suggests that adding a motivational component to the technology may foster gains and changes in behavior in the long term.

VirTele, a personalized remote rehabilitation program combining virtual reality exergames and telerehabilitation, was developed to provide stroke survivors an opportunity to pursue rehabilitation of their chronic UE deficits at home while receiving ongoing follow-up from a clinician [8]. More specifically, VirTele used Jintronix exergames [9] and the Reacts platform (Koninklijke Philips NV) [10] to provide personalized training for the UE and enable videoconference sessions with a clinician, respectively. At the time of the intervention, the Jintronix exergames included 5 types of UE games (Space Race, Fish Frenzy, Pop Clap, Apple Picking, and Kitchen Cleanup) performed in a sitting position. The performance of the affected UE (score, percentage of compensation, and number of repetitions) and the duration and number of sessions played can be accessed on the web through a clinician portal within the Jintronix system. Reacts is an internet-based audiovisual platform that can be used through a computer or mobile phone to conduct secure videoconferences and share content (images, videos, messages, etc). It enables screen sharing (viewing the participant’s computer screen) to supervise in real time the stroke survivors’ performance, provide direct feedback, and adjust difficulty level in collaboration with the stroke survivors, taking into account their preferences and capacities (as observed during real-time Reacts sessions and through the data available on the Jintronix web portal).

An initial study was conducted with a stroke survivor to test the VirTele technology and study protocol during the development phase of the intervention [11]. The results showed that it was feasible to use the VirTele program for remote UE rehabilitation [11]. Meaningful determinants of technology use were identified, including performance (perceived improvement in UE use during daily activities and unlimited time of exercises), effort (feeling comfortable using VirTele and experiencing only minor technical issues, which the stroke survivor could easily resolve), and social influence (positive feedback from family and friends) [11]. Preliminary efficacy results showed improvement in UE motor function, UE quality and quantity of use in activities of daily living, and quality of life [11]. Hence, there is interest in studying this technology further to explore varied experiences among more participants, including clinicians and stroke survivors.

Behavior-modification strategies (eg, patient-centered counseling, action planning, and self-monitoring) have been implemented in exercise promotion interventions to enhance motivation, exercise participation, and maintenance [12-14]. As gains achieved during rehabilitation are not always maintained in the long term [15], chronic-stroke survivors may benefit from such behavior-modification strategies when they are integrated into postrehabilitation programs. These strategies could be used to empower stroke survivors to continue exercising and using their UE in everyday activities (eg, brushing their hair, getting dressed, and eating). Furthermore, there is increased recognition that programs aimed at changing behaviors should have a strong theoretical basis [16]. Self-determination theory (SDT) [17] states that human beings have a natural tendency to autonomously pursue goals or achieve healthy changes when 3 of their psychological needs are satisfied, namely autonomy (a person’s ability to act according to their own values and aspirations), competence (a person’s belief in their ability to achieve changes), and connectivity (a feeling of belonging) [17]. Therefore, social environments where the clinician engages in a partner relationship with the stroke survivor while supporting their autonomy (shared decision-making, choice of exergames, etc), competence (the stroke survivor’s belief in their capacity to achieve their goals, etc), and connectivity (a nonjudgmental interaction) may result in greater autonomous motivation [18]. Previous studies [19,20] demonstrated that support of the 3 psychological needs predicted greater autonomous motivation, which resulted in better adherence to exercises. A recent meta-analysis of SDT-informed interventions [21] found small-to-medium effects of physical health outcomes (physical fitness and function, weight-related outcomes, blood pressure, etc) at the end of the interventions and during the follow-up period (ranging from 1 week to 30 months after the interventions). As autonomous motivation is a key element for developing maintained change, a supportive psychological needs environment should be integrated into the VirTele intervention. Thus, motivational interviewing [22], consistent with SDT, was incorporated into the VirTele program to ensure that shared decision-making and empowerment were consistently integrated into the intervention. The behavioral and motivational techniques incorporated into motivational interviewing may enhance autonomous motivation to adhere to the treatment plan and change behavior regarding UE use in activities of daily living. In addition, combining real-time videoconferencing (telerehabilitation) with virtual reality technology could allow for adequately monitored and engaging theory-based UE rehabilitation programs, which may enhance stroke survivors’ empowerment and sustain gains in the long term.

Eventually, the SDT-informed VirTele intervention may not only help patients and clinicians decide together on the best treatment options but also allow clinicians to identify potential problems once the patient has reintegrated into the community. Thus, this study will also document the experiences of the stroke survivor as well as the clinician when using the VirTele program, which are key aspects for the successful eventual implementation of such interventions.

The objectives of this study were as follows:

This study used a multiple case design, which allows extensive data collection with varied methods across different cases [23]. This design enables the exploration of the studied phenomenon across a more varied range of characteristics compared with a single-case model [23]. The unit of analysis in this multiple case study is each stroke survivor and their respective clinician (physiotherapist) participating in the VirTele intervention. A range of experiences in terms of age, sex, familiarity with technology, and living arrangements were sought.

This multiple case study is embedded into a 2-armed randomized clinical trial comparing an experimental group (receiving the VirTele intervention) with a control group (receiving standard care) in Montreal, Quebec, Canada, and registered with ClinicalTrials.gov (NCT03759106) [8].

The qualitative data were collected between June 2019 and August 2020 by the first author (DRA; who was not involved in the VirTele intervention), a PhD student under the supervision of DK and JH who had previous experience in qualitative research and stroke rehabilitation research. This multiple case study was reported according to the Standards for Reporting Qualitative Research [24].

This study targeted the stroke survivors who were assigned to the experimental group receiving the VirTele intervention in the context of the 2-armed randomized clinical trial and who had completed the 2-month program. This group of stroke survivors was screened for eligibility before enrollment, and participants were selected based on the inclusion and exclusion criteria described in the published protocol (refer to the Participant Selection and Recruitment Strategy section) [8]. The clinicians included in the main study were physiotherapists who had experience with stroke rehabilitation. All participants had to speak French or English.

This study received ethics approval from the research ethics board of the Centre for Interdisciplinary Research in Rehabilitation of Greater Montreal (review number CRIR-1319-0218; June 28, 2018) [8]. All participants provided informed written consent before starting the VirTele intervention.

The VirTele intervention protocol is illustrated in Figure 1. The stroke survivors were invited to use Jintronix exergames for 30 minutes at least 5 times a week and conduct 1-hour videoconference sessions with a clinician, using Reacts, for a period of 2-months. The videoconference sessions took place 3 times a week for the first 2 weeks, twice a week for the next 2 weeks, and then once a week for the remaining 4 weeks [8]. The clinicians only received training to familiarize themselves with the VirTele technology (Jintronix exergames and Reacts platform) because they were already trained in motivational interviewing (including SDT concepts), an approach they had been using in their practice for >2 years before the study began. Although SDT and motivational interviewing were developed independently, a resemblance exists between them [25]. In fact, motivational interviewing techniques [4] are consistent with the 3 psychological needs of SDT [25]; for example, motivational interviewing promotes shared decision-making (selection of treatment goals, exercises, etc), behavior change techniques (such as express advantages and disadvantages of change, goal setting, and review of goals) [26], and reflective listening (shows empathy), which emphasize autonomy, competence, and connectivity, respectively [4]. Thus, motivational interviewing, including motivational and behavioral techniques, was achieved through videoconferencing sessions conducted using the Reacts platform (Figure 1).

The Unified Theory of Acceptance and Use of Technology (UTAUT) [27] was used to explore the factors that may influence VirTele adoption intention and use behavior among stroke survivors and clinicians, to fulfill the third objective of this study. The UTAUT states that the intention to adopt a new technology is determined by 3 main factors: expected performance (the degree to which the technology is perceived as helpful and useful), expected effort (ease of use and complexity associated with using the technology), and social influence (positive or negative feedback that family and friends may have regarding the technology) [27]. In addition to the intention to adopt a new technology, contextual conditions (such as the ability and knowledge to use a new technology and interoperability) may facilitate the use behavior regarding the technology [27].

The UTAUT also incorporates 4 moderators—age, sex, experience, and willingness to use—that can influence technology adoption intention and use behavior [27]. In the context of this study, it is interesting to capture the expectations of users as well as their actual experiences to see whether the technology meets the needs of end users. Operational definitions of the UTAUT and concrete examples are provided in Multimedia Appendix 1.

Triangulation was used for this multiple case study. This involved the use of various methods to collect qualitative data [28]. First, semistructured interviews (lasting from 30 minutes to 1 hour) were conducted 4 to 5 weeks after the end of the VirTele intervention with the stroke survivors and clinicians. Two interview guides were developed and tailored to the clinicians (eg, What was your role or responsibility during VirTele? Did you have any concerns when you first started using the technology?) and the stroke survivors (eg, Did you perceive any change in your arm function? Can you describe this change?) to facilitate the interview administration while allowing new ideas to emerge. Questions were structured to target the key concepts of each theory. Key UTAUT concepts were used to identify major factors that influenced the VirTele experience in the stroke survivors as well as the clinicians. For the stroke survivors, SDT concepts were used to explore the indicators of empowerment in terms of autonomy, competence, and connectivity. For the clinicians, SDT was used to identify which motivational interviewing technique was used and which need was supported (autonomy, competence, or connectivity) when interacting with the stroke survivors. The interview took place either face to face at the research center or remotely through the Reacts platform.

Second, logbooks were used by the clinicians to collect data related to technical difficulties, number of videoconference meetings, complementary activities suggested in addition to the exergames, and motivational strategies used. Third, reflexive notes were used by the researchers to collect VirTele intervention–context-related data (technical difficulties, adverse events, etc). Demographic information about the stroke survivors was also collected. A sample size of 10 stroke survivors and 4 clinicians was targeted to diversify the experiences and enrich the data. However, only data saturation can predict the final sample size [29].

Abductive thematic analysis was conducted. This type of analysis seeks to go beyond inductive and deductive reasoning [30]. By adopting this approach, researchers can generate new ideas through a dynamic interaction between data and theory [30]. First, a predetermined coding scheme was developed based on UTAUT and SDT constructs. Next, the transcript text was examined to identify which meaning unit reflected one of the predetermined codes. Codes and assemblies were frequently revised, and relevant new codes were assigned to meaning units that could not be coded or categorized within the initial scheme codes. Finally, the new codes were examined and either represented as subcategories (reflecting border concepts related to the UTAUT or SDT) or new categories of codes (enriching the corpus of existing theories).

QDA Miner (Provalis Research) [31] was used to enter the list of predetermined scheme codes and retrieve the highlighted text into meaning units, which were condensed and then coded and categorized using the scheme codes.

In each case, the stroke survivors’ and clinicians’ experiences with the VirTele intervention and indicators of SDT variables were developed and examined independently and across the duos (stroke survivors and clinicians) for a within-case comparison. Next, experiences were examined among cases, using a cross-case analysis, to explore differences and similarities regarding the determinants of VirTele use and indicators of SDT variables. Underlying similarities and constant associations were then developed to form more general explanations. The analysis was conducted by 3 members of the research team (DRA, DK, and JH). The verbatim transcripts were translated from French into English for publication and verified by bilingual team members (DRA and DK).

The principles of Lincoln and Guba [32], including confirmability, credibility, reliability, and transferability, were applied to ensure study rigor. Audit trails and verification were conducted to ensure confirmability. An external verification by members was carried out for credibility. Reliability was confirmed through verification of a portion of the data by 3 coders (DRA, DK, and JH). For transferability, reflexive notes and a detailed description of the context of the intervention were compiled. The variation in the cases may increase the robustness of the qualitative data [29].

Five stroke survivors were assigned to the intervention group and completed the VirTele intervention (Table 1).

However, recruitment was halted in mid-March 2020 at the onset of the COVID-19 pandemic in Canada, and all research activities were suspended from March 2020 to October 2020. Of the 5 stroke survivors allocated to the VirTele group, 1 (20%) could not be reached to conduct the interview, and 1 (20%) was excluded because he did not speak French or English fluently. Of the 3 remaining stroke survivors, 2 (67%) were women (participant ID1 and participant ID5), and 1 (33%) was a man (participant ID11); their mean age was 58.8 (SD 19.4) years, and they varied in terms of computer familiarity, Chedoke-McMaster stroke assessment score, time since stroke, and dominance of UE. Two physiotherapists participated in administering the VirTele intervention. Participant ID11 received a 3-month VirTele intervention instead of 2 months, as was the case for participant ID1 and participant ID5, given that it was impossible to retrieve the technology material during the COVID-19 pandemic period. We decided to give this participant the opportunity to benefit from the services offered by this technology for an additional month. For readability, each participant was given a pseudonym: participant ID1 identified as Carolina, participant ID5 identified as Helene, and participant ID11 identified as Jack.

A detailed case description of the 3 duos (stroke survivor and respective clinician), collected from the interviews, logbooks, and exergame portal, is provided in Multimedia Appendix 2. A summary of the techniques used by the clinicians during motivational interviewing and their impact on stroke survivor empowerment, collected from logbooks and interviews, is provided in Table 2. The differences among the 3 cases are illustrated in Multimedia Appendix 3. The determinants of VirTele use, as expressed by the stroke survivors and clinicians during the interviews, are presented in Table 3. Although we did not reach our target sample size because of the COVID-19 pandemic, the data collected from the 5 participants allowed us to achieve a certain level of data saturation because many of the reported experiences were repeated across cases.

Differences and similarities have emerged regarding the determinants of VirTele use between the duos (stroke survivor and clinician).

In the clinicians’ interviews, some meaning units, reflecting different recommendations related to the use of the VirTele program, were assembled. They are presented as a bulleted list in Multimedia Appendix 4. Lessons can be learned from these recommendations (Textbox 1) regarding the use of telerehabilitation combined with exergames. In fact, these lessons provide relevant instructions for the use of exergames via telerehabilitation and suggest useful strategies to optimize the potential of this technology for the rehabilitation of the affected UE.

This study’s results corroborate the findings in the study by Caughlin et al [34], which confirmed the supportive role of caregivers during telerehabilitation interventions (facilitating the use of the technology). The high level of adherence to the exergames and the increased use of UE in stroke survivors echo the findings of a previous systematic review [21], which found that interventions involving tailored counseling strategies such as goal setting and monitoring, motivational interviewing, and follow-up seem to be effective at promoting long-term physical activity participation after stroke. Furthermore, the use of the affected UE in self-directed exercises may result in improved motor function. At this stage of the study only the evaluations of the first few participants of the randomized clinical trial [35] were performed, and firm conclusions cannot be drawn regarding the results obtained on the sensorimotor measures. However, a trend in improvement was observed regarding motor function measured using the Fugl-Meyer Assessment [35] in Carolina and Jack as well as UE activity measured using the Motor Activity Log [35] (quality and quantity of use) in all participants. These gains were maintained 2 months after completion of the VirTele intervention [35]. The high adherence to the exercise program demonstrated by the participants could optimize the motor gains. In addition, the change in behavior with respect to the use of the UE in daily activities, as observed in Carolina and Helene, could justify the maintenance of the gains in the long term. Furthermore, it is important to note that Jack, who did not intend to use the affected UE after the end of the VirTele program, still managed to maintain long-term gains (improvements noted in the Fugl-Meyer Assessment and Motor Activity Log scores [35]), highlighting the importance of adhering to the VirTele program and its potential to maximize gains.

In a previous study, Sit et al [36] found that stroke survivors (n=105) receiving motivational techniques similar to those in our study (encouragement, verbal persuasion, goal setting, partner relationship between the clinician and the patient, action plan, and self-management steps) significantly improved functional indices (Barthel and Lawton indices, which are scales used to assess activities of daily life performance on independent living) and self-management outcomes (medication adherence, self-monitoring of blood pressure, communication with physician, etc) compared with a control group receiving standard care. Hence, further research is needed to explore the correlation between motivational interviewing and UE motor function outcomes among stroke survivors.

Moreover, the determinants of VirTele use, identified through this study, are in part consistent with the determinants reported by other studies deploying telerehabilitation [37] and virtual reality exergames [38]. Despite the technical issues, the 3 participants were receptive to the VirTele program and continued using the system, which echoes the findings of previous studies among stroke patients and clinicians using telerehabilitation [34].

Participants’ expectations regarding the VirTele intervention before the start of the study were not documented. Therefore, it was not clear whether the intervention met the participants’ expectations, which could affect technology acceptability and their compliance to the rehabilitation program. Future studies could investigate expectations before the start of the intervention to better capture end user expectations of similar interventions. Furthermore, given the small sample size, the results of this study should be interpreted with caution.

In conclusion, the factors predicting intention to use the VirTele intervention and use behavior among stroke survivors and clinicians include technology performance, effort, social influence, contextual facilitators, and challenges. The empowerment attained by stroke survivors is promising for the future deployment of such an intervention to encourage the use of the affected UE in activities of daily living and achieve impactful long-term improvement. The lessons learned from this study regarding the resilience of stroke survivors and adaptability of clinicians with respect to technology limitations, role of the caregiver, new responsibilities of clinicians during the VirTele intervention, impact of aphasia diagnosis, and empowerment of stroke survivors may help to guide the implementation of similar interventions. However, further studies in different contexts are needed to better understand the factors affecting intention to use such technologies and use behavior.