Rehabilitation is an essential part of caring for people with acute or chronic health conditions, impairments, or injuries that limits functioning [1]. It is estimated that 2.5 billion people worldwide live with health conditions that benefit from rehabilitation [1]. Owing to population growth, aging, and the increasing number of people with chronic diseases and disabilities, the need for rehabilitation is steadily increasing worldwide [1].

Two of the most challenging aspects of rehabilitation are the high costs of inpatient and long-term rehabilitation programs [2] and the poor continuity of rehabilitation when patients are transferred to their homes [3,4]. To address these challenges, a shift in rehabilitation from inpatient care or rehabilitation centers to home settings is essential [5,6]. Therefore, various models have been developed to offer early home-based rehabilitation [7,8]. However, for effective home-based rehabilitation, sufficient support must be provided to clients (persons who receive health care services) and health care professionals (persons who provide health care services).

Incorporating new digital technologies into rehabilitation could help to meet these demands. Digital rehabilitation (DR) can be defined as using digital technologies as a part of the rehabilitation process [9]. DR aims to optimize functioning and reduce disability of individuals with health conditions in interaction with their environment [9]. This includes, but is not limited to, the use of tele- and remote rehabilitation applications and services, automatic services, robot-assisted technologies, wearables, emails, video, speech, and SMS text messaging solutions [9]. By using tele- and remote rehabilitation applications and services, for example, sensors and wearables, opportunities exist for monitoring clients’ health status at home [10-12]. In addition, DR improves rehabilitation outcomes in clients with heart failure, diabetes, and respiratory disease [13]. They also help clients to manage pain; increase their physical activity; and improve mental health, diet quality, and nutrition [13]. Furthermore, it appears that some parts of the DR are cost-effective [13]. With the combination of commercially available technologies, Internet of Things (IoT), and artificial intelligence (AI), there is also the possibility of remote health assessments and personalized rehabilitation interventions [14]. In addition, DR can have a positive impact on self-management [15-17]. Self-management aims to improve clients’ ability to manage their disability and improve their lifestyles. It underlines the active participation of clients, emphasizing the interactive and collaborative relationship between clients and health care professionals. Similarly, an important aspect of self-management is client responsibility, which is particularly important in home-based rehabilitation [18].

However, for the effective application of DR, it is not only important to identify digital technologies and their potential applications that can be integrated into home-based rehabilitation but also essential to equip health care professionals with the skills needed to provide high-quality rehabilitation, conferring them with the potential to develop the field multi-professionally.

The COVID-19 pandemic has exacerbated the challenges faced by individuals in need of rehabilitation [19]. In addition, there has been increased pressure on higher education institutions and health care professionals to develop DR practices that meet the needs of target populations [20].

To develop and implement DR solutions that meet client needs and ease the growing pressure on the health care system, it is necessary to provide a broad overview of existing, relevant, and future solutions shaping the constantly evolving market of technologies for home-based DR.

The scoping review aimed to identify digital technologies for home-based rehabilitation, predict new and emerging trends in DR, and report the influences of the COVID-19 pandemic on DR.

The scoping review was performed based on an adapted framework described by Arksey and O’Malley [21] by adding the improvements proposed by Levac et al [22]. This framework includes 5 phases: identifying research questions, determining relevant studies, selecting studies, charting data, and consulting with key stakeholders and experts. The scoping review was reported consistent to the PRISMA-ScR (Preferred Reporting Items for Systematic Reviews and Meta-Analyses extension for Scoping Reviews; Multimedia Appendix 1). A protocol of this scoping review does not exist.

As the first step, the following research questions were identified:

To identify references related to the research questions, a scientific database search in PubMed, Embase, CINAHL, PsycINFO, and the Cochrane Library was performed in July 2021.

The search strategy was first developed for PubMed and adapted to each database by using keywords; their synonyms; and related terms of “rehabilitation,” “home-based,” and “digital technologies.” These terms were connected with the Boolean operators “AND” and “OR” to obtain a wide spectrum of results from the various databases. The full search strategy for each database is presented in Multimedia Appendix 2. The search for relevant references was designed in 2 phases. The first phase involved searching references published from January 2015 to July 21, 2021. In the second phase, an additional search was performed in January 2022 to retrieve recent publications.

To complement the findings regarding the trends of home-based DR, as well as the impact of the pandemic, gray literature sources were searched for unpublished materials in the native languages by the member institutes of the project “Competences for the new era of user-driven Digital Rehabilitation (DIRENE)” (Greek, Spanish, Finnish, German, and English). The gray literature was searched using Google, Google Scholar, Yahoo, OpenGrey, specific websites from each country’s government, thesis repositories, and university library websites that offer a comprehensive list of gray literature databases.

Study selection was performed using the web-based software, Covidence (version 2021; Veritas Health Innovation) [23]. After a training period, 7 independent researchers from the DIRENE consortium performed study selection based on the eligibility criteria presented in Textbox 1. For every reference, 2 randomized researchers individually assessed the reference in terms of inclusion or exclusion. In cases of disagreement, a third researcher was included to resolve conflicts.

A bibliometric analysis was performed as the first overview of the included publications. The software Zotero (version 6.0.10; Roy Rosenzweig Center for History and New Media, George Mason University) [24] was used to indicate the number of publications per year and the journals of the included references.

To identify new and emerging technologies for home-based rehabilitation, a co-occurrence analysis was performed. The software VOSviewer (version 1.6.18; Leiden University) [25] was used for the analysis. The co-occurrence analysis was intended to map the results of the included publications, categorize the technologies mentioned in the publications, and provide a visualization of the output [26]. For this purpose, a Research Information Systems (RIS) file of all included publications was uploaded to the program, and a text-based analysis was performed using a binary counting method of the most frequently occurring keywords [26]. On the basis of the relevance score, the program automatically selected the most applicable keywords for the final analysis.

Furthermore, a full-text analysis of all included reviews was conducted to filter the trends for home-based DR technologies and to present its advantages and disadvantages. The analysis aimed to extract relevant information about the technologies in terms of the goal, specification and application, population, and advantages and disadvantages of the technologies.

The data of the gray literature search were analyzed to complement the results of the review analysis and to determine the influence of the COVID-19 pandemic on the development of DR. Each participating institution used an extraction sheet to obtain the information with respect to the research questions. All data were then summarized according to the following categories: type, aim, description of the technology, target population, and statements regarding the influence of the COVID-19 pandemic.

The aim of the stakeholder consultation was to complement the results regarding the trends of digital technologies and the impact of the pandemic on DR from different perspectives. Therefore, a meeting with key stakeholders and experts was organized by 4 of the 5 collaborating universities of the DIRENE consortium (Greece, Finland, Austria, and Germany). Participants included clients with experience or specific interest in using digital technologies at home (10/56, 18%); rehabilitation professionals (13/56, 23%); experts from companies for future trends in health care (9/56, 16%); experts in digitalization (11/56, 20%); and representatives of public health administrations (6/56, 11%), social and welfare departments (4/56, 7%), and national platforms for digitalization in rehabilitation (3/56, 5%).

The meetings consisted of a short introduction to the project and the presentation of the preliminary results of the review. Stakeholders shared their views on the identified trends in digital technologies used in rehabilitation. They discussed the potential synthesis of technologies in home-based rehabilitation and examined factors influencing DR. All the information was summarized in a final standardized report by each participating institution. The results were then analyzed in line with the Framework Approach [27]. On the basis of the research questions, a deductive approach was used to form the key themes and subthemes. After reviewing the results of the stakeholder consultation for each participating institution, the subthemes were adjusted inductively. Subsequently, the results of each stakeholder consultation were coded. Each code in a report was then summarized, abstracted, and tabulated for each subtheme. Then, the statements of all stakeholder consultations were summarized per key theme.

Bibliometric and co-occurrence analyses were performed to provide an overview of the included references and identify new or emerging technologies in home-based rehabilitation. For these analyses, 2437 records were included.

All 2437 included records for the bibliometric and co-occurrence analyses were published between January 2015 and January 2022. Over the past years, the number of publications has steadily increased annually, from 211 (8.66%) in 2015 to 544 (22.32%) in 2021. In addition, most research papers were published in 2021 (n=544, 22.32%).

In total, 53 keywords with a frequency of ≥10 were selected from 59,718 keywords, and a co-occurrence analysis was performed, as shown in Figure 2.

As presented in Figure 2, the research themes of home-based DR can be divided into 6 clusters: sensors (blue), robotics (purple), gamification (green), virtual reality (VR) and augmented reality (AR; turquoise), mobile apps (red), and digital platforms (yellow). Each node represents a keyword. The size of the node indicates the number of occurrences of that keyword, and the link connecting the 2 nodes indicates that a keyword appears in common with another keyword. The thickness of the connection line indicates the strength of co-occurrence between the 2 keywords. The visualization indicates that the keyword “sensor” is the most frequently occurring term throughout the included publications compared with other keywords represented in this cluster.

Further analysis of keywords that occurred ≥10 times provided an overview of the mentioned populations in which the identified technologies were used in home-based rehabilitation, as shown in Table 1.

It can be seen that the cumulative sum of stroke and related synonyms are the most occurring conditions related to home-based rehabilitation among the publications. Chronic obstructive pulmonary disease (COPD) and its synonyms ranked second. The third most common disease mentioned in the included publications is multiple sclerosis (MS).

The 3-year pandemic (with May 5, 2023, as the end) has changed many rehabilitation processes for clients and health care professionals. To present the influence of the COVID-19 pandemic on DR, 40 records were included.

To continue rehabilitation during the pandemic, remote digitally driven rehabilitation approaches, such as videoconferencing, apps, and web-based programs, were frequently used in practice. However, the crisis did not lead to the development of new DR technologies but rather to an increased use of technologies that were already on the market.

This has led to more experiences with DR among clients, health care professionals, and health care providers. Clients accepted DR remotely as a substitute for face-to-face therapy during the pandemic [110,185-188]. However, they also identified barriers to the use of technologies [186,188,189]. To increase clients’ adoption, recommendations have been developed [110]. Because many health care professionals are not trained to use virtual approaches in low-income countries [190], as well as in industrialized countries [191], additional training was offered to build competences in this area. Bernocchi et al [192] further stated that not only clients but also health care professionals require adequate training to use digital approaches.

The pandemic has led to an increased awareness of the need to develop further digital health measures (eg, digital vaccination certificate, digital sick note, improvement of digital tools, and simplified access to digital processing). In many countries, efforts have been made to implement digital solutions in the health care sector or help meet the preconditions for their use. However, the COVID-19 pandemic acted as an additional driver in the implementation of these applications. This was supported by political measures such as the softening of strict guidelines, which made it possible to launch and use digital solutions even without proof of their effectiveness [193]. However, clients have expressed concerns regarding their effectiveness and safety [194].

In addition, in some countries (Austria and Germany), videoconference (therapy via videoconferencing) was legally approved, considering individual requirements, such as a specific diagnosis, the confirmation of the service provided, safe technical equipment, and a positive prognosis regarding the success of the application [195,196]. In Germany, this permission remains permanent regardless of the pandemic situation [197].

However, the rapid implementation has also led to the frequent implementation of incomplete solutions that have met the clients’ short-term needs but are not sustainable in the long term, as pointed out by stakeholders. In addition, evaluations of the measures have been less rigorous or have not been conducted, leading to the risk of perpetuating rapidly implemented solutions after the pandemic [193].

This scoping review revealed various digital technologies intended for use in home-based rehabilitation. Trends were identified, and the advantages and disadvantages of each technology were presented. In addition, the influences of the pandemic on the development of DR were shown.

The findings of this review reveal that sensors, robotic devices, gamification, VR and AR, and digital and mobile apps are already widely used in home-based rehabilitation. However, AI and machine learning, exoskeletons, and digital and mobile apps represent emerging trends in rehabilitation. Compared with the other identified technologies, VR, AR, and robotics cannot be used independently for home-based rehabilitation for usability and safety reasons. Thus, there is a need to develop sufficient and feasible DR practices that demonstrate clinically relevant effectiveness. Furthermore, we discovered that the COVID-19 pandemic has led to an increased use of digital technologies in a remote approach, especially apps, web-based programs, and videoconferencing, but not to the development of new technologies. Clients and health care professionals accepted this approach during the pandemic, but they also expressed concerns about it. The pandemic acted as a driver for implementing remote approaches for health care systems. However, sustainable solutions that can be applied even after the pandemic should be implemented.

One major finding was that the relevance of AI and machine learning will increase in the field of rehabilitation in the future, especially for diagnostic procedures, decision-making, and the development of client-centered care, even though they are not yet broadly applied in rehabilitation. Furthermore, it was found that AI, machine learning, and digital and mobile apps would be essential to process and optimize resources in the health care sector. This finding is consistent with that in the current literature. For example, Hamet and Trembley [198] stated that electronic medical or health records are essential tools for use in personalized medicine, early detection, and targeted prevention, with the aim of increasing their clinical value and decreasing health costs. Moreover, Róman-Belmonte et al [199], who investigated the influence of AI on musculoskeletal disorders, highlighted that AI can produce a paradigm shift in musculoskeletal health, a move from descriptive to predictive medicine.

Nevertheless, there are challenges that need to be overcome when applying AI and machine learning widely in the health care sector. Jiang et al [200] report a lack of ethical and legal supervision. No globally unified laws or regulations regarding the application of AI in medicine are currently in place to standardize the behavior of practitioners [200]. In addition to ethical and legal issues, one major challenge is the clear need for a standardized, comparative evaluation of the effects of AI on health indicators and measures of changes in psychological and physical status, side effects, and outcomes [198]. The findings regarding acceptance were incoherent. Health care professionals support the application of AI in rehabilitation and desire training for its application. However, people who are less well informed about AI fear being replaced by this technology, often because of a lack of knowledge about AI [200]. Therefore, it is important to outline the benefits and barriers of using AI for rehabilitation within society.

One further result was that technologies such as gamification, VR and AR, and digital and mobile apps that are already used in home-based rehabilitation have the potential to improve client adherence and motivation. Similarly, approaches based on IoT can increase client participation in the rehabilitation process because they allow self-monitoring; thus, client self-management can be increased. This represents an important factor for successful therapy of chronic diseases to improve quality of life and reduce the use of health care resources [201].

This study revealed that most technologies could be used in home-based rehabilitation without the presence of health care professionals. The emerging developments in the design and possibility of asynchronous or synchronous monitoring via apps, web-based programs, or videoconferencing make the remote rehabilitation approach possible. In other studies [54,56,202], the results showed that the independent use of VR and AR technologies and robotics in a home setting is limited owing to low usability and safety concerns. Clients with disabilities may require assistance in attaching or using the device or face the risk of harming themselves when excessive force is transferred to the body. However, to provide independent, usable, and safe rehabilitation at home, trained caregivers supporting the clients [76,203] or the construction of the rehabilitation device have the ability to overcome this challenge. For example, attachment mechanisms could be designed to enable the client to don and doff the device [159]. Furthermore, Kim et al [204] used emergency stop buttons and added safety features limiting ROM and joint velocity and stopping the robot in case of excessive force and torque interaction.

In order to offer client-friendly rehabilitation in the home setting, sufficient space for the equipment must also be ensured. Furthermore, the technical equipment needs to meet requirements such as access to the internet or a digital and mobile device, and the costs of purchasing, repairing, and maintaining the technology need to be low [62] because such costs are not always covered by health insurance funds.

In addition to the safety and usability factors, the use of DR should be based on an individual rehabilitation goal, considering the motivation of clients and health care professionals to use technologies and the possibility of receiving support when using such technologies [205]. Furthermore, individual physical conditions should be considered, which may affect the use of the selected therapy (grade of dementia, disability of vision or hearing, and degree of impairment). It must be noted that Cottrell and Russell [205] limited their recommendations to videoconferencing in physiotherapy practice. Given that the technology must be customized to the client’s body, these recommendations can also be considered when using other technologies.

Another major finding is that the effectiveness of many technologies has not yet been confirmed scientifically. Several studies have described projects in which the technology is currently under development or tested in laboratory settings. However, it can also be seen that 221 RCTs referring to DR are already registered via ClinicalTrials.gov for 2021 and 2022 (September 15, 2022). Because many technologies have only been used under controlled conditions, their application in daily life cannot be assessed.

In addition, we found that feasible approaches to the implementation and integration of digital technologies for the rehabilitation process are lacking. Accordingly, it is essential to consider not only the effectiveness of the intervention in terms of dedicated outcomes, the design of the intervention itself, and the characteristics of individuals but also the setting in which the approach will be applied, the process of implementation, and factors influenced by policies and government regulations [206].

Therefore, studies are required that present an effective and valid concept that can enable clients and health care professionals to apply digital technologies for (home-based) rehabilitation. This includes the presentation of plans for the modus of DR (blended therapy or replacing face-to-face therapy), the reimbursement of health care professionals, and cost coverage for the technologies needed by clients and rehabilitation units. Other issues must be addressed for a feasible implementation, such as privacy compliance, data protection, insurance coverage, and the assumption of responsibility and liability if harm occurs during remote or unsupervised rehabilitation.

The social distancing regulations enacted during the COVID-19 pandemic caused many clients and health care professionals to use DR, especially videoconferencing, apps, and web-based programs. Clients and professionals described their relatively high level of DR positivity along with improvements over the course of the COVID-19 lockdown. However, the question of whether this therapy model will continue to be accepted without exceptional circumstances remains.

The literature shows contrary results regarding the acceptance of digital interventions by clients and health care professionals [13,207]. However, it is noteworthy that many terminologies exist for Digital Rehabilitation that are associated with various technologies. It is likely that the acceptance rates of different technologies vary, which does not allow for a general statement about their acceptance among clients and health care professionals.

Through the COVID-19 pandemic, more training opportunities for health care professionals were offered to allow them to gain competence in the field of DR. To be adequately prepared for working life and systematize the acquisition of competencies, health care professionals should already have acquired the necessary skills in higher education. Because DR technologies have already been implemented in rehabilitation, it is crucial that health care professionals are sensitized to the possibility of integrating digital options, acquire useful competencies, and have the ability to recognize which clients will benefit from this approach and then provide adequate assistance to them.

Although the pandemic was, in many ways, a driver of the use of DR, it also led to the rapid introduction of solutions that were not based on an efficient concept and left many questions unanswered with regard to realistic implementation. Therefore, careful planning with a phased, linear implementation approach is crucial for establishing sustainable DR practices that will last beyond the pandemic and have the potential to meet the challenges of rehabilitation in the future [208].

The results of this study paint a very broad picture of existing and emerging technologies in rehabilitation because a large number of records were included in the analyses, which were not limited to a specific target group or rehabilitation outcomes. The results derived from the scientific databases were complemented by a search of the gray literature. Through a meeting with key stakeholders, insights were gained beyond those reported in the literature. In addition to health care professionals, politicians, experts in the field of digitalization in rehabilitation, and clients with experience or a specific interest in DR were included.

Bibliometric and co-occurrence analyses were performed with a minimum of 10 occurrences per keyword. This may have influenced the completeness of the representation of technologies if newer technologies occurred fewer than 10 times throughout the data set. However, the stakeholder meeting served to close this gap. Moreover, we did not weigh the strength of the evidence from all papers or appraise the efficacy of the approaches, which would have been beyond the scope of this review. In addition, new technologies may have been developed in the meantime and were not presented here because of the rapid development of the market.

In the search for relevant literature, we used the word “rehabilitation” and not other terms commonly used in the context, such as “therapy,” “training,” “intervention,” or “treatment,” because we wanted to stick to the concept of rehabilitation, which implies a multimodal, collaborative, and patient-centered process rather than the stand-alone interventions suggested by those terms [209]. However, it resulted in few articles being included related to mental and psychosocial disorders, despite the Cochrane definition of rehabilitation states that rehabilitation also includes people with mental problems [209]. Thus, rehabilitation seems to be still more closely associated with the improvement or optimization of physical functions.

These findings reflect the growing interest in the use of digital technologies in rehabilitation. Multiple tools are already available and implemented for home-based rehabilitation; however, there are some limitations to their use, such as low usability, safety concerns, ethical challenges, and a lack of efficacy and legal frameworks. DR implementation should be based on the clients’ goals and motivation. AI and machine learning could be of particular interest in redesigning rehabilitation to address future challenges in the rehabilitation sector.

On the one hand, the pandemic acted as a driver for the application and acceptance of existing digital solutions in rehabilitation. On the other hand, digital solutions that only met the requirements of the clients during the pandemic were implemented. The results of this research reflect the need for feasible and effective approaches to implement DR sufficiently to meet clients’ needs and adhere to framework conditions to be sustained apart from exceptional situations.