Occupational therapists’ perspectives on robotics integration: a scoping review informed by the consolidated framework for implementation research

With ageing populations and rising chronic conditions, two-thirds of adults over 65 are projected to have multi-morbidity by 2035 (Kingston et al., 2018), whilst healthcare faces a projected 10 million workforce shortage by 2030 (Boniol et al., 2022). Digital technologies, including robotic systems spanning rehabilitation, assistive, socially assistive, telepresence and wearable applications, are increasingly positioned to support intensive and repetitive interventions, social engagement, remote service delivery, independence in activities of daily living, and the management of therapist workload (Silvera-Tawil, 2024). The world federation of occupational therapists (WFOT, 2019) advocates for the ethical, person-centred integration of robotics into practice and calls for equitable global access to affordable, high-quality assistive technologies.

Occupational therapy (OT), originating from early 20th-century Arts and Crafts and moral treatment movements, has been fundamentally shaped by values of therapeutic relationships and individualised care. These person-centred foundations are increasingly intersecting with technological developments associated with the Fourth Industrial Revolution (Liu, 2018). As occupational therapists (OTs) continue to incorporate robotics that promise efficiency gains, a key challenge is ensuring these technologies enhance practice without compromising the relational foundations of occupational therapy.

Various robot types serve distinct purposes. For example, rehabilitation robots (e.g., MIT-Manus, Diego™) provide repetitive movement practice for upper limb function (Mackenzie et al., 2025); assistive robots (e.g., JACO arm) can support activities of daily living (Bourassa et al., 2023); and socially interactive robots (e.g., Pepper, Nao) can enhance patient engagement (Komariyah et al., 2024). To reflect this diversity, in this review the term robotics is used as an umbrella category encompassing rehabilitation, assistive, social, wearable and telepresence robotic systems. Collectively, these technologies are emerging as valuable tools for rehabilitation, offering targeted, repetitive and measurable interventions that enhance motor recovery and patient participation (Banyai and Brișan, 2024).

As robotics advances, successful integration into OT depends on both technological progress and therapists’ perceptions and experiences within practice. Robotics may enhance treatment intensity and motor recovery (Bourassa et al., 2023), with emerging evidence suggesting additional benefits, including reduced social isolation through socially assistive robots (Komariyah et al., 2024) and extended service reach via telepresence technologies (Boman and Bartfai, 2015). However, accessibility, cultural attitudes, training gaps and cost continue to limit adoption (Aguiar Noury et al., 2021), underscoring the need for implementation strategies that address these barriers.

Recent evidence indicates that research examining occupational therapists’ perspectives on robotics remains limited. Thawisuk et al. (2025), for example, identified only seven studies exploring therapists’ views on robot-assisted stroke rehabilitation and highlighted recurring barriers such as device complexity, training requirements and organisational constraints. Although robotics is becoming more prominent in rehabilitation, their review shows that understanding of how occupational therapists engage with these technologies across different populations, robot types and practice settings is still underdeveloped. This gap reinforces the rationale for the present review, which offers a broader, theory-informed synthesis of the multilevel factors influencing robotics integration in occupational therapy.

The growing importance of integration of digital health technology such as robotics into practice can be seen in policy frameworks such as the Topol Review, Topol (2019) which emphasises the need for healthcare professionals to develop digital competencies. In the UK, the royal college of occupational therapists (RCOT) (2025) 10-year workforce strategy and its research priorities highlight digital transformation and innovation as central to the profession’s future. In addition, the WFOT (2019) Minimum Standards highlight the need to embed technological literacy within professional education to ensure therapists are prepared for evolving healthcare environments.

A scoping review methodology was chosen as most appropriate for this emerging research area. Unlike systematic reviews, which focus on assessing intervention effectiveness, scoping reviews map the breadth of the literature, identify research gaps, and synthesise diverse evidence types (Munn et al., 2018). This approach is particularly suited to heterogeneous and emerging fields such as robotics in occupational therapy. This review aimed to:

• explore occupational therapists’ perspectives and experiences of using robotics across diverse technologies, populations and settings; and

• identify the individual, organisational and systemic factors that facilitate or hinder the integration of robotics into occupational therapy practice.

The Consolidated Framework for Implementation Research (CFIR) was applied as an analytical lens to structure interpretation of findings in relation to both aims.

This scoping review followed Arksey and O’Malley (2005) five-stage framework, refined by Levac et al. (2010), with reporting adhering to PRISMA-ScR guidelines (Tricco et al., 2018). The five stages are:

1. identifying the research question;

2. identifying relevant studies;

3. study selection;

4. charting the data; and

5. collating, summarising and reporting findings.

Each stage is described in detail below.

This review used the Population, Concept, Context (PCC) framework to guide the development of research questions, structure the inquiry, and inform the search strategy. This review aimed to understand:

A comprehensive search strategy, developed with a research librarian, was conducted on 18 April 2025 across four databases: CINAHL Complete, MEDLINE, OTSeeker and IEEE Xplore. Database selection captured health sciences (CINAHL, MEDLINE), occupational therapy-specific (OTSeeker) and technical literature (IEEE Xplore). The search combined two concept groups: Population (“occupational therapy” OR “occupational therapist” OR “occupational therapists” OR “OT”) AND Technology (“robot” OR “robots” OR “robotic” OR “robotics” OR “assistive robotics” OR “rehabilitation robotics” OR “robotic systems” OR “robot-assisted therapy” OR “robotic device*”). Subject headings (MeSH, CINAHL headings) were combined with free-text searching, with truncation applied to capture word variations.

Searches were limited to English-language publications from 2010–April 2025. The 2010 start date was chosen because preliminary scoping indicated occupational therapy-specific literature on rehabilitation robotics emerged following this period, coinciding with increased commercial availability and clinical implementation (Kyrarini et al., 2021).

Grey literature searching was conducted on 18 April 2025 to capture practice-based and policy documents not indexed in academic databases. Google Scholar was searched using the primary search terms, with the first 100 results screened for relevance. Reference lists of all included studies were also screened to identify additional relevant sources.

Endnote software was used for citation management and removal of duplicate articles. Studies were included if they involved qualified, registered occupational therapists with clinical experience using robotic technologies and if they examined therapists’ perspectives, attitudes, experiences or implementation-related factors concerning robotics. Eligible studies were required to be peer-reviewed primary research, using qualitative, quantitative or mixed-methods designs, conducted within clinical, rehabilitation or community-based therapeutic settings, and published in English between January 2010 and April 2025. Studies were excluded if they did not report occupational therapy data separately from other professional groups, if they focused exclusively on technical or engineering specifications without therapist-perspective data, or if they were conducted in non-clinical or laboratory-only environments. Conference abstracts, dissertations, study protocols and other non–peer-reviewed outputs were also excluded from the review.

The screening process involved two stages conducted via the Rayyan QCRI platform (rayyanLink to the cited article). Firstly, titles and abstracts were reviewed, followed by full-text assessment. Two reviewers (CD and LD) independently evaluated all records at both stages using pre-established inclusion and exclusion criteria. Any discrepancies were addressed through discussion, and a third reviewer (CR) was consulted when agreement could not be reached. Study selection is reported using PRISMA-ScR flowchart (see Figure 1).

A structured data extraction form was developed, informed by review questions, CFIR domains (Damschroder et al., 2022) and previous implementation reviews (Mitchell et al., 2023). Extraction categories included publication details, study characteristics, participant demographics, robotic technology specifications, implementation factors and key findings. Recognising that therapist perspectives in qualitative studies are often interwoven with author interpretation, we extracted entire results and discussion sections (Thomas and Harden, 2008), carefully distinguishing primary data (participant quotations, quantitative findings) from secondary interpretation (author conclusions) during analysis.

The data extraction form was piloted on two randomly selected studies. Two reviewers (LD, CD) independently extracted data, met to compare results and identify ambiguities, then incorporated third reviewer (CR) feedback. Modifications included creating separate fields for device characteristics and clarifying ‘implementation factors’ using CFIR domains. Following refinement, the lead reviewer (LD) extracted data from all eight studies. As specified in the protocol, a second reviewer (CD) independently verified data extraction for two studies (representing the planned 20% sample), purposively selected to reflect variation in design, geographic context and robot type. Discrepancies, primarily related to categorising implementation factors, were resolved through discussion, and no systematic patterns of disagreement were identified.

The Mixed Methods Appraisal Tool (MMAT) was used by the lead author (LD) to assess the methodological quality of all included studies across diverse designs, with each criterion rated as “Yes”, “No” or “Can’t tell” (Hong et al., 2018). Consistent with scoping review guidance, this appraisal was undertaken to enhance transparency rather than to influence study selection or weighting. The PRISMA-ScR checklist explicitly designates critical appraisal as optional (Item 12) and states that, where conducted, authors should simply report the rationale and describe how (if at all) the information was used in the synthesis (Tricco et al., 2018). In keeping with this guidance, quality appraisal was applied descriptively, and no studies were excluded based on MMAT scores.

Study characteristics (publication years, geographic distribution, study designs, sample sizes and participant demographics), robotics applications (technology types, clinical applications, settings and implementation contexts), and temporal trends were summarised and presented in tables and figures.

This review adopted the CFIR to provide a comprehensive structure for identifying multilevel factors that influence implementation (Damschroder et al., 2022). CFIR (Damschroder et al., 2022) is a comprehensive determinant framework organising implementation factors into five domains: (1) innovation characteristics (device features such as complexity and adaptability), (2) outer setting (external policies, economic context), (3) inner setting (organisational culture, resources, readiness), (4) characteristics of individuals (therapist knowledge, beliefs, self-efficacy) and (5) process (implementation planning and execution). CFIR enables systematic examination of multilevel factors and their interactions, moving beyond simple barrier lists to understand how contextual elements shape outcomes. Whilst widely used in health services research, its application to occupational therapy robotics represents a novel contribution.

In this review, CFIR was used specifically as an analytic rather than descriptive tool. Its purpose was to guide the deductive component of data analysis and provide a structured lens for interpreting implementation-related findings, rather than to determine the extent of robotics use in practice. CFIR was incorporated only during Stage 5 of the review, during coding and synthesis, and not during study selection or data extraction.

Qualitative data from the results and discussion sections, comprising therapist quotes, author interpretations, narrative findings, reported benefits, challenges and implementation factors, were compiled into a single data set and thematically synthesised (Thomas and Harden, 2008). A hybrid coding strategy was adopted: inductive coding captured emergent experiences and perceptions, while deductive coding mapped implementation-related data to CFIR domains (Nevedal et al., 2021).

The lead reviewer (LD) engaged in iterative reading and line-by-line coding using NVivo 14 to identify patterns. Initial codes reflected specific concepts (e.g. “setup time as a barrier”) without reference to existing frameworks. Concurrently, deductive coding applied CFIR constructs flexibly, recognising that some data might not align perfectly with predefined categories (e.g. device complexity mapped to Innovation Characteristics – Complexity, funding constraints to Inner Setting – Available Resources).

Following both inductive and deductive coding, CFIR domains were used to classify, organise and compare implementation determinants across studies. This step enabled multilevel factors (e.g. organisational readiness, individual confidence, innovation complexity) to be systematically synthesised and linked to broader contextual influences. Preliminary themes were reviewed and refined through iterative team discussions to ensure internal consistency and external distinctiveness. The final thematic structure integrated inductive themes emerging from therapist experiences with deductive themes aligned to CFIR domains. This hybrid approach facilitated both exploratory syntheses, identifying what perspectives exist, and theoretically informed interpretation, examining how these perspectives relate to implementation success.

The initial database search retrieved 669 articles; four additional records were identified through grey literature and reference list searching. Following duplicate removal, 377 articles were screened, with 17 full-text articles reviewed. Eight studies met the inclusion criteria (see Figure 1), published between 2015 and 2025 across diverse international contexts.

Figure 1: Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) – flow diagram of study selection process (Tricco et al., 2018).

The review includes four qualitative studies (Boman and Bartfai, 2015; Flynn et al., 2019; Mashizume et al., 2021; Flynn et al., 2025), three mixed-methods (Komariyah et al., 2024; Mackenzie et al., 2025; Proulx et al., 2023) and one quantitative survey (Bourassa et al., 2023). Notably, Flynn et al. (2019) captured therapist perspectives prospectively, prior to clinical use of the InMotion device, meaning findings represent anticipated rather than practice-based experience.

Sample sizes ranged from 3 to 113 occupational therapists, predominantly with 2–20 years’ experience in adult rehabilitation. One study focused on paediatric therapists working with autism spectrum disorder (Komariyah et al., 2024). Settings included acute hospitals, inpatient/outpatient rehabilitation and community services. Table 1 summarises key characteristics of the included studies and associated robotic technologies.

MMAT appraisal revealed that six studies met all quality criteria, demonstrating rigorous methodology across data collection, analysis and interpretation (Boman and Bartfai, 2015; Flynn et al., 2019; Flynn et al., 2025; Komariyah et al., 2024; Mackenzie et al., 2025; Mashizume et al., 2021). Two studies had unclear elements, particularly regarding sampling representativeness and data integration in mixed-methods designs (Bourassa et al., 2023; Proulx et al., 2023). One study (Bourassa et al., 2023) showed methodological limitations related to nonresponse bias, with a large number of non-respondents potentially affecting generalisability. The results of the full quality assessments are outlined in supplementary material (Appendix).

The included studies examined five categories of robotic systems, each serving distinct therapeutic purposes. Rehabilitation robots (e.g. InMotion2, Diego™, ReoGo-J) provided repetitive, task-oriented upper limb training with adaptive assistance, visual feedback and gamified interfaces (Flynn et al., 2019; Flynn et al., 2025; Mackenzie et al., 2025; Mashizume et al., 2021). Wearable devices such as the ExoGlove offered motorised assistance for targeted hand rehabilitation (Proulx et al., 2023). Social robots, including humanoid forms (Nao, KASPAR, Zeno) and animal-inspired designs (Paro, AIBO, CuDDler), facilitated interaction and skill development for children with autism (Komariyah et al., 2024). Telepresence robots like Giraff enabled remote monitoring and environmental assessment in brain injury rehabilitation (Boman and Bartfai, 2015). Finally, wheelchair-mounted robotic arms (JACO, iARM) supported independence in daily tasks through adaptive control interfaces (Bourassa et al., 2023).

Thematic synthesis integrated the CFIR with inductive insights to interpret how individual, organisational and contextual factors shaped occupational therapists’ engagement with robotics. Rather than serving solely as a categorisation tool, CFIR constructs deepened understanding of emerging themes, particularly the interplay between perceived benefits, practical constraints and evolving professional identities in the adoption process. Four overarching themes emerged:

1. Cautious Optimism.

2. Barriers to Implementation.

3. Contextual and Environmental Influences.

4. The Evolving Role of Occupational Therapists in Technology Integration.

The relationship between CFIR domains and themes is presented in Table 2.

Mapped primarily to the CFIR constructs of Relative Advantage (Innovation Characteristics) and Knowledge and Beliefs about the Innovation (Characteristics of Individuals), this theme captures how therapists weighed perceived therapeutic value against practical uncertainties, a balance that shaped an overarching sense of cautious optimism regarding the integration of robotics into practice. While OTs acknowledged the therapeutic potential of these technologies, they remained mindful of practical limitations. Optimism was largely driven by perceived benefits such as improved patient engagement, opportunities for intensive and repetitive practice, customisation options and the ability to achieve meaningful functional outcomes:

Examples such as the ExoGlove (Proulx et al., 2023) and the JACO arm (Bourassa et al., 2023) were perceived by therapists as enabling clients to regain previously lost abilities, reshaping therapists’ perceptions of recovery potential. Similarly, the implementation of robotics was described as providing patients with “a sense of accomplishment” (Mashizume et al., 2021, p. 4):

However, this enthusiasm was tempered by concerns about environmental constraints, resource limitations, technical complexity and the need for specialised training.

This theme draws on constructs spanning Innovation Characteristics, particularly Complexity and Cost, and the Inner Setting domain, including Access to Knowledge and Information and Available Resources. Together, these constructs reflect how barriers to robotics integration operated simultaneously at the level of the device itself and within the organisational environments in which therapists practise. Robotics integration in occupational therapy was constrained by barriers spanning financial, technical and educational domains. Financial barriers were prominently reported, with device acquisition costs identified as a significant obstacle to implementation. These financial barriers relate specifically to the cost of acquiring and sustaining robotic devices in clinical settings and are distinct from broader systemic funding and resource allocation issues, which are addressed in the following section:

Technical complexity, including device setup, calibration and interface usability, complicated clinical workflows and limited patient accessibility (Proulx et al., 2023; Boman and Bartfai, 2015). Design limitations particularly affected patients with severe impairments:

Beyond hardware challenges, limited knowledge and insufficient training undermined confidence in implementation. Many therapists reported minimal awareness of the evidence base or practical functionality of robotic devices. In addition, infrequent use emerged as a major barrier, with therapists reporting that sporadic exposure led to declining skills and confidence, making earlier training difficult to apply:

These findings reveal a complex interplay of financial, technical and educational challenges requiring comprehensive strategies to foster effective robotic integration in clinical practice:

Anchored in the CFIR Outer Setting domain, specifically External Policy and Incentives and Patient Needs and Resources, and the Inner Setting constructs of Available Resources and Implementation Climate, this theme reflects how factors beyond individual control fundamentally shaped whether and how robotics were adopted in practice. Broader contextual and environmental factors significantly shaped robotics adoption in occupational therapy practice. At a systemic level, policy-driven funding structures and institutional resource allocation emerged as primary contextual constraints across multiple studies (Bourassa et al., 2023; Komariyah et al., 2024). Rather than reflecting the direct cost of individual devices, these constraints operated at an organisational and policy level, determining whether institutions prioritised, procured and sustained robotic programmes over time. External policies, institutional incentives and resource availability created foundational environments that either supported or hindered implementation. Komariyah et al. (2024) demonstrated how environmental and institutional constraints influenced therapists’ engagement with social robots, pointing to systemic barriers extending beyond individual willingness or device functionality. This aligned with findings from Bourassa et al. (2023), where despite high interest in robotics, actual clinical recommendation remained low, suggesting that system-level factors such as funding priorities, procurement policies and organisational readiness played decisive roles in bridging the gap between enthusiasm and practical implementation.

Physical location also acted as an under recognised barrier, with devices positioned outside occupational therapy spaces limiting access and routine use. These environmental challenges were compounded by a lack of structured implementation planning, with therapists describing that sometimes, efforts to embed robotics into practice were ad hoc:

Socio-cultural factors further complicated adoption. In regions where robotics represented unfamiliar technology, particularly rural areas, novelty generated hesitation among patients and providers:

The implementation climate, encompassing organisational culture, leadership support and interprofessional collaboration, emerged repeatedly as a critical mediator of adoption. Supportive institutions that prioritise innovation and provide training create a fertile environment for robotics integration. Conversely, limited resources, lack of support and unclear role definitions undermine these efforts (Flynn et al., 2019; Flynn et al., 2025; Mackenzie et al., 2025):

This theme is grounded in the CFIR domain of Characteristics of Individuals, drawing on the constructs of Knowledge and Beliefs about the Innovation, Self-Efficacy and Individual Stage of Change, together illuminating how therapists’ readiness to engage with robotics was shaped not only by confidence and competence but by evolving professional identity. Occupational therapists demonstrated increasing engagement with robotics integration, influenced by individual knowledge, confidence and perceived readiness (Mashizume et al., 2021; Bourassa et al., 2023). Therapists emphasised their essential roles in assessment, customisation and implementation of robotic interventions while maintaining strong professional identity. They articulated that robotics enhanced rather than replaced the therapeutic relationship:

Therapists showed openness to adopting technologies that aligned with therapeutic goals and integrated into functional activities (Flynn et al., 2019; Proulx et al., 2023). However, readiness varied, with some requiring evidence of tangible benefits before full commitment (Flynn et al., 2019). Negotiation of professional roles around robotics was fluid and collaborative, with occupational therapists emphasising patient benefit over “disciplinary ownership”:

Personal motivation, curiosity and professional development aspirations drove engagement, with therapists acknowledging the importance of skill development in an evolving clinical landscape:

This scoping review synthesised evidence from eight studies spanning diverse designs, countries, settings and robotic technologies, providing an overview of how occupational therapists appear to engage with robotics in practice. The discussion addresses both aims of the review: firstly, to explore occupational therapists’ perspectives and experiences of robotics across varied contexts, reflected in the themes of cautious optimism and the evolving professional role; and secondly, to examine the multilevel facilitators and barriers influencing implementation, captured in the themes relating to barriers to adoption and contextual and environmental factors. Overall, the findings reveal a measured optimism, with therapists in the included studies indicating the potential of robotics to enhance patient engagement, motivation and functional outcomes, while also acknowledging practical constraints such as device complexity, training and confidence gaps, cost and unclear implementation pathways.

The balance between enthusiasm and hesitation echoes broader findings in rehabilitation technology research, where uptake lags behind innovation. While external studies suggest potential therapeutic benefits of robotics, the findings in our review reflect therapists’ perceptions of potential benefit, rather than empirical evaluations conducted within the included studies. Distinguishing therapist perceptions from the wider evidence base is therefore important. Although OTs in the included studies viewed robotics as potentially beneficial for engagement and functional progress, these views do not always align with evidence. For example, Thawisuk et al. (2025) noted that while therapists recognised the potential of robotics, evidence for effectiveness over conventional therapy on activity and participation outcomes remains inconsistent. More broadly, many studies report feasibility and usability rather than functional outcomes, suggesting that adoption may be shaped more by perceived benefit and practical advantages than by robust outcome data. This reflects Rogers’ Diffusion of Innovations theory, Rogers (2003), which suggests that while perceived relative advantage drives interest, adoption is often slowed by complexity, limited trialability and organisational or contextual barriers (Ward, 2013).

This review adds to existing literature by drawing attention to the centrality of professional identity in shaping therapists’ responses. Earlier discussions have raised concerns that automation could diminish therapist roles and lead to less personalised care (Stahl and Coeckelbergh, 2016). However, findings from this review indicate that some occupational therapists perceive robotics as complementing, rather than replacing, therapeutic relationships. This interpretation aligns conceptually with the Beyond Adoption Framework by Greenhalgh et al. (2017), which offers an evidence-based, theory-informed and pragmatic approach for evaluating the success of technology-supported health interventions. The framework emphasises that clinicians actively shape how technologies are integrated, embedding them into practice in ways that maintain core professional values.

Findings support a context-sensitive approach to implementation, showing that therapists’ readiness depends on organisational culture, training access and alignment with patient-centred goals. Successful integration may therefore be influenced by systemic, rather than solely technical or individual, factors, reinforcing calls in implementation science to consider both organisational and external factors (Dryden-Palmer et al., 2020).

This review underscores the need for targeted strategies to support robotics adoption in occupational therapy. Contextual factors, such as cultural attitudes and healthcare system differences, shape how robotics are received, calling for locally tailored implementation. Integrating robotics and digital health into pre-qualification curricula and professional development may help to build confidence and competence, aligning with WFOT (2019) standards on technological literacy in professional education.

Greater access to hands-on training, technical support and interdisciplinary collaboration in practice can further enhance practitioner readiness. This aligns with the priorities outlined in RCOT’s (2025) 10-year workforce strategy, which identifies digital transformation and innovation as central to the future of occupational therapy.

The Topol Review, Topol (2019) calls for investment in digital capability and cultural change to prepare the healthcare workforce for technological advances. Sustainable funding models should cover device, training and support costs, while organisations foster supportive climates with adequate time, infrastructure and resources, particularly in resource-limited settings. Clear national guidelines and evidence-informed best-practice frameworks may support efforts to ensure that robotics integration supports, rather than undermines, the core values of occupational therapy.

To advance integration in this field, occupational therapy professional bodies may consider promoting the adoption of robotics in ways that uphold core professional values. Coordinated attention to educational preparation, policy development and practice frameworks may be important for supporting effective and sustainable implementation:

• For education: Integrating hands-on robotics teaching within pre-qualification occupational therapy curricula, supported by structured placement opportunities in settings where such technologies are actively used, may help build foundational competence and readiness for the use of robotics in practice.

• For practice: Developing sustained continuing professional development pathways for robotics, incorporating mentorship between experienced and novice users and ensuring protected time for ongoing skill development, may reduce reliance on one-off training and support more consistent implementation in clinical settings.

• For policy: Professional organisations, such as RCOT and WFOT, may play a role in developing evidence-informed guidance for robotics integration, including considerations for procurement processes, funding mechanisms and ethical frameworks to promote equitable access.

This study has several key strengths. It followed Arksey and O’Malley framework with PRISMA-ScR reporting, ensuring transparency and reproducibility. Dual independent screening and MMAT quality assessment strengthened methodological rigour. Using the CFIR framework added theoretical depth, highlighting individual, organisational and systemic factors that an atheoretical approach might overlook.

Several limitations should be acknowledged. The small sample size (n = 8) reflects the emerging nature of research in this area. Requiring separate reporting of occupational therapist perspectives may have excluded insights from mixed-profession studies. Evidence was primarily drawn from adult physical rehabilitation contexts, limiting generalisability to other populations and settings. Variation in therapist experience, organisational context and prior exposure to robotics may also have influenced findings. In addition, quality appraisal using the MMAT was conducted by a single reviewer, which may introduce bias. Restricting the search to English-language publications could have excluded relevant international literature. Grey literature searching was limited to the first 100 results in Google Scholar, potentially omitting additional sources. Finally, while CFIR provided a useful structure, reliance on predefined constructs may have constrained identification of novel themes, indicating the value of combining deductive and exploratory approaches in future work.

Three research priorities emerged. Firstly, comparative effectiveness studies examining robotic versus conventional interventions across diverse populations, conditions and settings would provide essential evidence for clinical decision-making.

Secondly, exploration of underrepresented contexts including paediatric services, mental health, community rehabilitation, would clarify whether implementation factors identified in adult physical rehabilitation transfer to other practice areas or whether context-specific barriers require tailored strategies.

Finally, and perhaps most importantly, patient perspectives on the use of robotics in therapy, particularly regarding therapeutic benefit, autonomy and satisfaction, represent a critical gap. Implementation frameworks appropriately focus on clinician and organisational factors, yet patient experience is also a key determinant of intervention success.

Beyond empirical gaps, an important question remains: to what extent will robotics enhance or reshape occupational therapy practice? Across the included studies, participants often viewed robotics as tools that may support rather than replace clinical expertise. Yet advances in AI and machine learning may demand new competencies in data interpretation, algorithm oversight and ethical governance. The profession’s role in this transition will depend on strategic action from professional bodies, curriculum adaptation by educational institutions, and therapists’ willingness to integrate digital innovation into practice.

This review intentionally focused on occupational therapy perspectives to foreground the profession’s distinctive philosophical orientation and its core emphasis on occupation, functional performance, client-centred practice and the therapeutic relationship. Occupational therapists evaluate robotics not only in terms of motor improvement but in relation to meaningful activity, role participation and the broader occupational goals of the individuals they support. While perspectives from other professions are valuable, they reflect different disciplinary priorities and therefore fall outside the scope of this review. We acknowledge, however, that this professional focus necessarily excludes client and patient perspectives, an important gap for future research. Integrating patient experience data will be essential to ensure that implementation frameworks reflect the needs and expectations of those receiving robotic interventions, not solely those delivering them. A further consideration for the profession is whether occupational therapy’s commitment to occupation-centred practice is adequately reflected in how robotic technologies are selected, configured and evaluated in real-world settings.

This scoping review synthesised occupational therapists’ perspectives on robotics across eight international studies, revealing an overarching sense of cautious optimism. While therapists recognised benefits for engagement and outcomes, they also cited barriers such as design complexity, limited training and high costs, underscoring a gap between technological potential and clinical readiness. Using CFIR, the review highlighted that successful adoption depends on organisational culture, system support, training access and alignment with therapeutic goals rather than individual motivation alone. Therapists upheld their professional identity while integrating technology, countering assumptions that robotics dehumanise care. Advancing integration will require coordinated efforts across education, policy and practice to ensure ethical, effective and sustainable implementation.

• Occupational therapists view robotics with cautious optimism due to barriers in training, cost and design complexity.

• The adoption and implementation of robotics in occupational therapy are influenced by a range of factors, including professional identity, organisational readiness and broader contextual conditions.

This research received no specific grant from any funding agency in the public, commercial or not-for-profit sectors. This study was completed as part of the requirements for an MSc in Occupational Therapy.

The authors thank Professor Emerita Elizabeth McKay for her support and advice, particularly during manuscript preparation.

No ethical approval was required for this study.

The supplementary material for this article can be found online.