The methodology for this scoping review was based on the stages in the methodological framework defined for scoping reviews [32], which has been further revised by Peters et al [33] into five phases: (1) identify aim and research questions, (2) search for relevant literature, (3) literature screening and selection, (4) data extraction, and (5) summarize and report the results. The plan for the study, including title, aim, research questions, screening process, and inclusion and exclusion criteria for the inclusion of literature, were specified in a protocol published on the internet [34]. This scoping review follows the PRISMA-ScR (Preferred Reporting Items for Systematic Reviews and Meta-Analyses Extension for Scoping Reviews) guidelines [35] (see Multimedia Appendix 1).

The aim of this scoping review was to gain knowledge about how ICT is used to address the collaboration among the patient, provider, and other stakeholders (eg, next-of-kin, home-care services, welfare technology personnel, or landlords) through the treatment process in rehabilitation. Studies describing a remote dialogue between patients and other stakeholders using technology aiming to address SDM or goals and means, provide decision support, and evaluate treatment were of particular interest.

A search strategy was developed in collaboration with a reference librarian. Literature search strategies were developed using medical subject headings and text words related to “ICT system,” “shared decision-making,” and “rehabilitation” processes. The search strategy for MEDLINE (OVID) can be accessed digitally.

The search was initially run on February 22, 2022, and rerun on March 22, 2022. The following databases were searched: MEDLINE (OVID), Embase (OVID), CINAHL (EBSCOhost), AMED (EBSCOhost), and Scopus (see Multimedia Appendix 2). Searches were limited to studies conducted from 2018 to March 22, 2022. With the aim to locate unpublished studies, the following sources were searched: OAIster, Bielefeld Academic Search Engine, ProQuest Dissertations and Theses, NARIC, and Google Scholar. These searches were limited to studies conducted from 2018 to July 7, 2022. We limited the searches to include studies from 2018 and newer due to the fast development in this area. Furthermore, only studies in the English language were included.

Eligible papers were those which addressed or described a remote dialogue between these parties using ICT to address goals and means, provide decision support, or evaluate certain treatment modalities within a rehabilitation context. We included papers describing the development of such digital solutions, their implementations, and evaluation of the outcomes of their usage. Studies focusing on ICT solutions only for 1-way communication between patients and health care providers or papers covering only prototypes or development of some technical features without implementation in clinical work were excluded. Digital solutions in general and their influence on organizations were not of interest either. Studies about collaboration through the rehabilitation process without digital support were not included [34].

We imported identified papers into the reference manager EndNote (version 20.4.1; Clarivate Analytics) for screening of titles and abstracts to detect and eliminate duplicates. The remaining references were uploaded to Distiller SR (version 2.38; DistillerSR Inc, 2022). For further screening, 2 authors (JIG and RJ) separately screened all paper titles and abstracts for inclusion against eligibility criteria (level 1 screening). Subsequently, all eligible papers underwent full-text review by 2 authors (JIG and RJ) to confirm whether the inclusion criteria were met (level 2 screening). All discrepancies between reviewer 1 and reviewer 2 were discussed among themselves and the third (IH) and fourth authors (TS) were consulted for making decisions in case of continued disagreement. All authors fully agreed upon which articles to include in the study.

Data extraction forms developed by the research team were used. Two of the authors extracted data independently to ensure that all relevant information was included. Data extraction was facilitated using Distiller SR. The main focus of the analysis was the collaboration among stakeholders through ICT. The information extraction included the title of the study, author, country of authors, year published, country of selected research study, the health sector studied, size and type of selection of study sample, the technology used, type of patient focus or participation, type of decision support, type of collaboration, type of digital user interface, and key findings. After level 2 screening, data extraction was completed, and 3 papers were identified. The reviewers completed the process of data synthesis, which involved identifying important findings and noting areas with gaps in knowledge. The PRISMA-ScR flowchart for the scoping review is shown in the Results section.

The literature search yielded 12,203 papers. After the removal of duplicates and papers published before 2018, a total of 3206 papers remained for screening of titles and abstracts. In this level 1 screening, many studies were excluded because they did not include the use of ICT, there was no reference to a dialogue based on shared information to address goals and means, they did not provide decision support, did not evaluate treatment, or they were not carried out in a rehabilitation setting. Much of the literature was about digital 1-way communication. It could be applications for patients’ self-monitoring (and reporting) of functioning, activity, or physical parameters. Others were about applications, text messages, or websites, which health care professionals could use to send instructions, training programs, or reminders to patients. Thus, these papers did not focus on dialogue and SDM involving both patients and practitioners. After the level 1 screening, 275 papers remained for the full-text screening, of which 3 papers met the eligibility criteria and were included in qualitative analysis. The screening process is summarized in Figure 1.

Table 1 shows the characteristics of the 3 included studies. These 3 studies were highly different in design and technology or ICT used, sample characteristics, and key findings.

Two of the studies used quantitative methods [36,38], while the third had a qualitative approach [37]. Ali et al [38] had the highest number of participants (N=222) in their multicenter randomized trial exploring the use of a digital platform and phone calls to promote person-centered care in patients with chronic obstructive pulmonary disease and chronic heart failure. The quantitative study by Beit Yosef et al [36] piloted an RCT and included 16 patients with acquired brain injury. Occupational therapists remotely guided patients through video in the use of a global problem-solving strategy to focus on function and individual goal setting. Strubbia et al [37] included 8 health professionals and 8 patients in their interview study on the use of an iPad app to facilitate SDM with patients with stroke, traumatic brain injury, skin graft, or chronic leg ulcer.

Technology was used to address collaboration in different ways in the 3 studies. Beit Yosef et al [36] combined physical meetings and video sessions between occupational therapist and patients with acquired brain injury, with an initial meeting in person to establish a therapeutic relationship and complete a baseline assessment. Weekly video sessions over a period of 3 months were conducted before a second assessment was carried out [36]. Strubbia et al [37] also combined physical meetings and digital collaboration. In their study on health care professionals and rehabilitation patients, the professionals and patients chose up to 20 images on an iPad app representing goal topics from the activity and participation domain of the International Classification of Human Functioning, Disability and Health and rated them by importance for the patient. The patient and the health care professional then discussed the urgency of the chosen activities, and together they selected a maximum of 5 activities to focus on [37]. In contrast, Ali et al [38] only used digital collaboration in their study on patients with chronic obstructive pulmonary disease and those with chronic heart failure. In this study, the patient and the health care professional cocreated and followed up on a health plan through an optional number of phone calls. In the first telephone conversations, the health care professional established a partnership using communication skills such as listening to the patients’ narratives about daily life events and how they were affected by their condition. A health plan, including patient goals, resources, and needs, was then cocreated through discussion and agreement. A digital platform was used to support communication between phone calls, provide access to shared documentation (health plans and self-ratings), and access to reliable information sources. The digital platform was developed using participatory design including patients, patient partners, experts, and researchers [38].

The 3 studies used different measures to evaluate outcomes or experiences. Beit Yosef et al [36] used the Canadian Occupation and Performance Measure (COPM) to identify 5 functional goals and to measure activity performance in their study on patients with acquired brain injury. Activity performance was also measured through the Performance Quality Rating Scale (PQRS). Participation was measured using the Mayo-Portland Adaptability Inventory-Participation Index [36]. Strubbia et al [37] used semistructured interviews to collect and analyze health professionals’ and rehabilitation patients’ perspectives on using an iPad app for prioritizing goals. In Ali et al’s [38] study on patients with chronic obstructive pulmonary disease and patients with chronic heart failure, the primary end point was a composite score of general self-efficacy changes and hospitalization or death 6 months after randomization into usual care or the intervention group.

Beit Yosef et al’s [36] study with video sessions for persons with brain injuries showed significant improvements in COPM scores compared to the waitlist control group for both trained and untrained goals following the intervention. Significant improvements were also found in the PQRS and Mayo-Portland adaptability inventory-participation index scores and the patients left the house more often after the intervention. Improvements were partially maintained at follow-up. It was concluded that the intervention was feasible or effective for focusing on function and individual goal setting for adults in the chronic phase after acquired brain injury. The results gave reason to believe that strategies of problem solving learned through the intervention had a spill-over effect on other tasks. Similarly, the data from the participants’ significant other and the clinician valued the intervention as having a positive impact [36].

In the study by Strubbia et al [37] on health professionals and rehabilitation patients, the aid for decision-making in occupational choice (ADOC) app was seen as a valuable addition to the rehabilitation process by both professionals and patients because it facilitated conversations around personally meaningful goals and person-centered goal setting. The application enabled patients to understand what they could expect from the rehabilitation process and provided them access and a tool for involvement in decisions about their care. The professionals stated that the application promoted a more patient-centered approach compared to usual goal-setting practice as it gave them a better understanding of their patients’ preferences and priorities [37].

In Ali et al’s [38] study on follow-up of patients with chronic obstructive pulmonary disease and patients with the use of phone calls and a digital platform to focus on patient-centered care, no differences in composite scores were found between usual care and the intervention groups 3 and 6 months after the intervention. However, when analyzing data from participants who actually used the digital platform and the structured telephone support, there was a significant difference between groups in composite scores 3 months after the intervention but not at the 6-month follow-up [38].

According to Beit Yosef et al [36], there were several methodological weaknesses in the video study with persons with brain injuries, small sample size, partial randomization, no active treatment control group, heterogeneity in goal-setting complexity between groups, wrong use of the PQRS, and nonblinded second and third assessment. The lack of improvement in general self-efficacy was assumed to be caused by the intervention’s focus on specific goals, which leads to effects on self-efficacy improvements specific to each goal. The lack of effect of the intervention on executive function in daily activities and caregiver burden was explained by the mentioned sample size and relatively high baseline scores. The authors discuss a potential ceiling effect with independent participants in need of only mild assistance in basic activities of daily living [36].

Strubbia et al [37] discuss several methodological weaknesses in their study of the ADOC application. Few participants, an exploratory study design, possible selection bias, and mandatory access to iPad to use the application were emphasized. Furthermore, the authors also discussed challenges related to implicit demands on users and the user interface with the ADOC app. Although the ADOC application was reported to be intuitive and instruction was given at the beginning of the study, both professionals and patients expressed the need for a user manual to keep up its use in clinical practice. Moreover, several logistical and organizational barriers were uncovered such as the availability of iPads in the clinic, challenges in incorporating a complex application in a pressured timetable, and lack of integration with the health care system. There were also app-related problems and technical issues with the ADOC app; a lack of possibility to create personalized goals and images in the application, no way to access a PDF treatment plan, incompatible email systems, and no print options within the organization [37].

In the study by Ali et al [38] on phone calls and a digital platform for follow-up at home, an explanation for the lack of results was a ceiling effect. The study included participants with a high initial score on general self-efficacy and feeling of disease stability, which might have reduced their need for the intervention. The effect of the intervention after 3 months was explained by the participants’ initial high degree of communication with the health care personnel and that the increase in global self-efficacy would attenuate over time independent of the intervention. The authors acknowledge that there could have been richer insights if they had explored the motivation and development of disease or rehabilitation needs through the project period. The authors also consider the need for tailoring interventions to the wants and needs of the user, to identify those persons who would benefit the most from it. The study concluded that person-centered care implies tailoring digital interventions to each patient’s unique needs [38].

The aim of this scoping review has been to gain knowledge about how ICT is used to support collaboration between the patient, provider, and other stakeholders (eg, next-of-kin, home-care services, welfare technology personnel, or landlords) through a rehabilitation process. Our review process suggests four different strands for discussion: (1) A low number of papers that matched the inclusion criteria; (2) the studies presented in the included papers differ in research design, sample sizes, type of technology used, and how they frame and address collaboration, effects, and limitations; (3) the complexity of ICT design and implementation in health and social care is striking; (4) there is an unaddressed implicit demand for eHealth literacy and access to health and social care.

First, since information exchange and data access for all stakeholders are prerequisites for patient-centered care, 1 potential challenge is the lack of technological solutions to support data access and exchange when needed. Lack of access complicates remote dialogue and may result in a fragmented information flow between stakeholders [13]. This was reported by Strubbia et al [37] in their study of the ADOC app on health professionals and rehabilitation patients. In their study, the lack of integration with the health care system, for example, electronic health records, logistics, problems with personalization of the application, and the lack of ability to access and distribute results, were major drawbacks with the solution [37]. These challenges might have been experienced differently by the stakeholders, even though they were unanimous in their critique.

Second, even without restrictions on data and information flow, there are numerous possible pitfalls concerning digitalizing dialogue between stakeholders in rehabilitation. As illustrated by Levesque et al [9], there are many different processes or situations a potential patient has to navigate to be able to engage in a health care encounter, for example, perceive, seek, reach, pay, engage, and interact with on the institutional or professional side to gain access. For instance, the capacity to seek health care services depends on the patient’s personal and social values, culture, gender, autonomy, socioeconomic position, and living conditions. On the societal side, cultural and social factors influence the possibility for people to accept the aspects of the service (eg, the gender or social group of providers and the beliefs associated with systems of medicine) [9]. The engagement of the patient in health care may depend on the fit between services and the patient’s needs, its timeliness, the amount of care spent in assessing health problems and determining the correct treatment, and the technical and interpersonal quality of the services provided. ICT designers are embedded in their own context and have their prejudices and knowledge gaps, like all people. To avoid what has been called “Script by design” [39-41], that is, cultural stereotypes and prejudices reiterated in the technology, ICT must incorporate a vast number of personal factors concerning the patients’ access and equivalent dimensions on the provider side through health care. This complexity can be a possible explanation for the paucity of research found in this study. The 3 included studies also report several methodological challenges. In the study by Beit Yosef et al [36] on persons with brain injuries where OTs used video to remotely guide patients in the use of a global problem-solving strategy, an initially high baseline score probably explained the lack of effect in part of the intervention. The same was the case in the study by Ali et al [38], where people with chronic obstructive pulmonary disease and patients with used a digital platform and phone calls. Ali et al [38] suggest that a high initial score on general self-efficacy and feeling of disease stability reduced the need for the intervention. This highlights the complexity described above, and as Ali et al [38] concluded, “person-centered care implies tailoring digital interventions to each patient’s unique needs.” Based on the authors’ methodological critique, we would argue that there is no best fit between the participants’ needs and wants, the baseline tests, and the outcome. Given a different methodological approach, with a systematic inclusion of participants in planning and carrying through of the studies and an explicit use of patient-reported outcome measures or experiences (where COPM sits), different outcomes might have been produced.

Third, there is a need to design digital health solutions that meet people’s needs and wants, take the users’ contexts into consideration, as well as embedding a range of possibilities for adaption to impairments or disabilities [42]. User-friendly and adaptable design is important for all stakeholders, whether it be patients, providers, and other stakeholders’ needs in different contexts. Technology development has been driven by technical possibilities to a greater extent than the needs of the different stakeholders [40,41,43]. Technology can act as a barrier against access if the design does not fit the context of use [11]. Traditional design science has not recognized the role of the organizational context in the development and implementation of technology, where a suitable demonstration context is selected after building the artifact [44]. A consequence may be that the shaping of the IT artifact condones the interests, values, and assumptions of the user end of the artifact. One can assume this is another factor influencing the paucity of research on digital collaboration in rehabilitation. The 3 included studies in this review point to this. The ceiling effect, mentioned by both Beit Yosef et al [36] and Ali et al [38], indicates that interventions are not tailored to the patient’s wants and needs. Both studies explained the lack of effects caused by initial high scores on several measures. In addition, in the study by Strubbia et al [37], several conditions were reported that indicate a lack of adaptability of the technology to individual needs. Both the need for user manuals, challenges in incorporating a complex application in a pressured timetable, and lack of personalization possibilities highlights unmet individual requirements for technology [37]. The authors do not address the most obvious lack in their methodology, including those concerned to a greater extent, probably at odds with what is recognized as the best and promising practice [30,31].

Fourth, the differences in design, results, and challenges in the included studies can also be attributed to the complexity of health care and the situations where ICT is assumed to help. Video, an iPad application, a digital platform, and phone calls are technologies used in the included studies in this review. However, what technologies are eligible for addressing the patients’ and other stakeholders’ wants and needs? This can be seen in the challenges reported in the studies, with a complex fit between methodological challenges and challenges in providing individual access for patients and the match with the provision of access from the supply side. None of the included studies are addressing the implicit knowledge demand put on users, whether these are professionals, patients, or other stakeholders. ICT research in health has shown that to be able to maximize the use of ICT in this sector all users need digital skills, dexterity, cognitive capacity, user interfaces, access to support (eg, introduction, guidance, maintenance), basic knowledge about health and health and social systems, and access to safe storage. ICT for health has been driven by technology architects and commercial interests, which creates unnecessary barriers to both commercial success and access.

It is important to acknowledge the limitations of this scoping review, despite following the appropriate methodology. First, the strict eligibility criteria resulted in a limited number of included studies, which may have caused the omission of important information relevant to the study's objective. In addition, restricting the search to a 5-year period starting from 2018 could have hindered the identification of relevant literature. Moreover, the decision to only include papers in English may have resulted in the exclusion of important information from non-English sources.

The search strategy for this study was developed by selecting relevant subject headings, text words, and their combinations from a larger pool of potential terms. Despite careful quality assurance measures taken during this process, there is still a chance that some crucial elements may have been inadvertently excluded from the search.

However, the low number of included papers illustrates the key finding of this study. The most salient feature of the subject matter of the study is the complexity of human and technological interfaces and collaboration, which is difficult to research.

The use of ICT has been proposed as a solution to address and to support the individual management of the rise in noncommunicable diseases. Furthermore, a lesson learned after the COVID-19 pandemic is that ICT can be a valuable tool for shifting the tables among the stakeholders in rehabilitation to better meet the needs and wants of rehabilitees and other stakeholders and to provide remote support and care. Despite the widespread optimism that ICT can better access, clinical outcomes, and cost-effectiveness, evidence supporting such effects is limited. Technology use is rising globally, but there is an urgent need to include consideration about global differences not only in health burdens but also in socioeconomic factors and living conditions [45,46]. A low-tech user-friendly technical solution might have a much larger global potential to aid the rehabilitees’ process, support those concerned and their families, and reduce the demands on professional staff both in low-income and high-income countries. Low-tech and intuitive user interfaces are also paying heed to the necessity for universal design, which increases access for all.

There is an inherent contradiction between the hallmark of rehabilitation; individually tailored and complex intervention; and the study design’s lack of thorough assessments of the rehabilitees’ wants and needs, and the project’s wish for homogeneity, simplicity, and standardized goals. The understanding of a rehabilitation process is always embedded in a treatment plan, and henceforth in a research design. If these perspectives are at odds, the likelihood of success is lower.

Based on this scoping review, we still argue that ICT holds the potential to facilitate communication between stakeholders in the complex and collaborative process of rehabilitation. However, a prerequisite for eliciting this potential is to systematically include those concerned in the design and implementation process and to consider simplicity, low-tech, and low cost to lower barriers for successful user experiences and outcomes.

Critical appreciation of the 272 papers read in full text uncovered implicit biases toward end users (rehabilitees, next-of-kin, and professionals), for example, a top-down approach, fragmented approach, or high demands on eHealth literacy. This probably creates barriers for the stakeholders’ participation in rehabilitation processes. Systematic feedback to designers seems lacking. There is an urgent need for more research on how implicit biases can be uncovered and how end users’ experiences and needs can be systematically fed back to designers. Implicit bias creates barriers to a strength-based, empowering professional relationship or needs-based inclusive design [40,41]. Critical rehabilitation studies [31] and critical disabilities studies [30,47] have repeatedly shown that if rehabilitation is to be successful, persons with disabilities must be acknowledged as competent and creative actors and included in the process of technology development and implementation from the launch of the ideas to commercialization.