Self-directed and caregiver-assisted technology-based interventions in inpatient stroke rehabilitation: A scoping review

Abstract

Objective

Technology-based interventions may enhance self-directed or caregiver-assisted activity during inpatient stroke rehabilitation, yet effective implementation strategies for technology-based tools remain unclear. This scoping review synthesises existing evidence on acceptability, feasibility and effectiveness of technology-based interventions in inpatient stroke rehabilitation, and maps barriers and enablers to implementation through the Capability, Opportunity, Motivation and Behaviour model.

Data sources

Comprehensive searches of MEDLINE, Embase, CINAHL and Google Scholar were conducted through to December 2025.

Review methods

Studies involving adult stroke survivors using self-directed technology-based interventions during inpatient rehabilitation were included. Arksey and O’Malley's scoping review framework guided data extraction. Analysis focused on acceptability, feasibility, effectiveness and behaviour change strategies.

Results

Eighteen high-moderate quality studies (20 papers) from 13 countries, involving 1081 participants, were included (10 pilot randomised controlled trials, 6 feasibility studies, 1 non-randomised and 1 observational study). Interventions were upper limb exergaming or robotics devices (n = 4) and tablet/smartphone applications with exercise videos (n = 14). Acceptability, feasibility and effectiveness outcomes were reported in 12 studies (63%), with rates of recruitment, safety and retention being reported in eight (67%), eight (67%) and six (50%) studies, respectively. All studies reported on at least one construct relating to barriers or enablers. Key enablers were caregiver involvement and additional support for more impaired individuals. Common barriers were stroke-related impairments and reduced staffing.

Conclusion

This review offers an integrated overview of the acceptability, feasibility and effectiveness of technology-based interventions in inpatient stroke rehabilitation, with identified strategies to support clinicians with implementation. Further research is needed on long-term outcomes.

Keywords

Stroke digital technology-based interventions rehabilitation review

Introduction

Stroke is a leading cause of disability worldwide, with an estimated 15 million cases annually.¹ Guidelines for stroke rehabilitation emphasise the importance of early, intensive, task-specific therapy to optimise recovery. The Australian and New Zealand Living Clinical Guidelines for Stroke Management recommend continuous engagement in rehabilitation activities through independent and semi-supervised exercises.² Despite this, stroke survivors face prolonged periods of inactivity^3,4 with limited opportunities for engagement in meaningful tasks beyond scheduled therapies.⁵ Prolonged periods of immobility can lead to a decline in physical function, increased risk of complications and delayed recovery.⁶ Importantly, early and sustained engagement in rehabilitation can lead to better outcomes in functional abilities and a reduced need for long-term care.⁷ Given the ongoing pressures on healthcare services, innovative strategies are needed to bridge the gap between recommended and current practice.

Self-directed or caregiver-assisted therapy is a promising approach to reducing inactivity outside structured therapy sessions⁸ significantly increasing overall exercise dosage in inpatient stroke rehabilitation.^9,10 Promoting self-directed therapy may support independent practice, allowing therapists to allocate resources more efficiently.¹¹ The success of self-directed or caregiver-assisted programmes depends on numerous factors, including the availability of appropriate supports for engagement.¹² Current resources to support self-directed therapy are ineffective and other approaches are needed.¹³

The growing use of technology-based tools offers clinicians working in inpatient rehabilitation an innovative way¹⁴ to support stroke survivors, and their caregivers, to better engage in self-directed activities. Technology-based interventions may be tailored to individual abilities and preferences, are able to be personalised¹⁵ and can be made to be appropriately challenging and relevant to individual users.^16,17 Whilst there is emerging evidence for the use of technology-based interventions in stroke,¹⁸ there is no synthesised evidence on feasibility, acceptability and effectiveness of using such tools in an inpatient setting after stroke.

The aim of this scoping review is to examine what is known about the use of self-directed, and caregiver-assisted, technology-based interventions to engage in self-directed, or caregiver-assisted activity in an inpatient setting after stroke. Specifically, we aim to explore:

Current approaches on research design and types of technology-based interventions which are being used in inpatient rehabilitation settings after stroke.

What is known about acceptability, feasibility and effectiveness of self-directed or caregiver-assisted technology-based interventions for stroke survivors in an inpatient setting.

What are the known barriers and enablers to the use of technology-based interventions, from a patient, caregiver and/or clinician perspective, mapped to the Capability, Opportunity and Motivation Behaviour (COM-B) change framework.

Methods and materials

We applied the five stages of Arksey and O’Malley's methodological framework in this scoping review, which include identifying the research question, identifying relevant studies, study selection, charting the data and collating and reporting the results.^19,20 We followed the Preferred Reporting Items for Systematic Reviews extension for Scoping Reviews (PRISMA-ScR) reporting guideline.²¹ Our scoping review was registered with Open Science Framework (https://osf.io/prq3e/files/osfstorage/679983be4ef5cd4daf62d164).

A preliminary search of MEDLINE, the Cochrane Database of Systematic Reviews, PROSPERO and Open Science Framework was conducted to check that no systematic or scoping reviews on our topic were published, or in progress, prior to commencing this review. We then searched MEDLINE (Ovid), Embase (Ovid) and CINAHL from inception to October 2024. This was updated in December 2025 to capture newly published studies before finalising the review. Search terms were established by members of the research team with assistance from a health research librarian. Key terms were stroke, inpatient rehabilitation, self-directed (or caregiver-assisted), digital interventions and technology. Additionally, to reduce the risk of missing any relevant literature, a basic search using the above keywords was conducted in Google Scholar, with the first 100 articles screened for relevance. An example of a full search strategy for MEDLINE (Ovid) can be found in Supplemental File S1.

To be eligible for inclusion, participants in the included studies needed to be adults receiving inpatient rehabilitation after stroke and be exposed to self-directed or caregiver-assisted technology-based interventions examining either capabilities, opportunities or motivation for engagement (Table 1). Titles and abstracts were imported into Covidence²² via EndNote²³ and duplicates were removed. The research team piloted the eligibility criteria on a sample of 100 papers and met to resolve any ambiguities. Titles and abstracts were screened independently by two authors (DP, JRAJ, SA, LJ, PYL and KJB). Similarly, all full texts were reviewed by two authors (DP and KJB). All conflicts were resolved through discussion. Reference lists of all the included studies were screened by another author (LJ).

Table 1.

Inclusion and exclusion criteria.

	Inclusion	Exclusion
Population	A diagnosis of stroke patients ≥18 years old
Context	Individuals receiving inpatient rehabilitation in a hospital setting	Individuals receiving outpatient or home-based therapy
Concept	Engagement with self-directed or caregiver-assisted resourcesANDResources offered via a technology-based tool (e.g. exergaming device, video on a tablet or an application)ANDStudies investigating either capabilities (physical and psychological), opportunities (social and physical) and/or motivation (automatic and reflective) in relation to resource engagement	Resource use is therapist-led
Sources	Peer-reviewed studies published in English with the following study designs:Randomised or non-randomised controlled trialsCohort studiesPre–post-intervention studiesQualitative or mixed-method studies	Conference abstracts and proceedingsStudy protocolsSingle case studiesCommentaries or editorialsSystematic or scoping reviews

Quality of the included randomised controlled trials was appraised using the 11-item (10-point) Physiotherapy Evidence Database (PEDro) scale,²⁴ which evaluates internal and external validity of randomised controlled trials. The scale has strong convergent and construct validity²⁵ and moderate inter-rater reliability. Because blinding of participants and therapists is often not feasible in exercise trials, a maximum PEDro score of 8/10 is considered high quality.²⁶ Studies scoring 6 to 8 on the PEDro scale were categorised as good quality, 4 to 5 as fair and 0 to 3 as poor.²⁷ Quality of observational, cohort or feasibility trials was appraised using the Critical Appraisal Skills Programme checklist.²⁸ This checklist is widely used for assessing qualitative research, with a grading system ranging from 0 to 10 based on 10 methodological questions.²⁹ Studies with a quality score of 9 to 10 were categorised as high quality, 7.5 to 8.9 as moderate, or less than 7.5 as low.³⁰

A customised data extraction form was developed and piloted. Extracted data included study design, and methodology, country of the study, participant characteristics, description of the intervention, outcome measures, key findings and future directions relevant to the review questions. Data were extracted by one reviewer (DP) and independently cross-checked by another (SA or PYL). Any discrepancies or disagreements were resolved through discussion, or by a third reviewer (KJB). Analysis involved descriptive statistics to summarise study characteristics and basic coding to categorise findings related to acceptability, adherence, feasibility and effectiveness of technology-based interventions.

We defined acceptability as participants’ affective attitude towards technology-based interventions, their usage intentions (e.g. willingness to engage with the tool), and their satisfaction after having engaged with the tool.³¹ Adherence was defined as actual usage (e.g. frequent interaction with the tool). Feasibility was defined as rates of recruiting users to engage with the technology-based intervention, level of assistance needed and safety considerations.³² Finally, effectiveness was defined as the extent to which engaging with a technology-based intervention produced a beneficial outcome under typical, real-world conditions.³³ Data on barriers and enablers were deductively mapped to the six sub-domains of the COM-B change framework.³⁴ The COM-B framework was chosen because it provides a clear, theory-based way to identify the capability, opportunity and motivation factors that influence uptake and ongoing use of technology-based interventions. Importantly, COM-B directly links behavioural barriers and enablers to intervention functions and behaviour change techniques, supporting both intervention design and implementation planning. This makes it a practical and actionable framework for translating research findings into real-world implementation strategies.

Results

The search strategy yielded 8797 studies and was reduced to 4750 after duplicates were removed (Figure 1). After screening titles, abstracts and reference lists, 100 full texts were retrieved. Of these, 80 were excluded based on eligibility criteria. Common reasons for exclusion were that interventions were not self-directed (n = 62), participants were not inpatients (n = 10), interventions were not technology-based (n = 5), papers were not published in English (n = 2) and participants were not stroke survivors (n = 1). Twenty papers from 18 unique studies, representing 19 datasets, were included in our review.

Figure 1.

Study selection flow chart.

Study characteristics are detailed in Table 2. The review included 18 studies (20 papers): 10 pilot randomised controlled trials,^35–44 6 feasibility studies,^8,45–49 1 non-randomised trial¹⁸ and 1 observational study.⁵⁰ All studies were published in the last 12 years with 14 out of 18 (78%) studies undertaken in the last 5 years. Studies spanned 13 high-income countries, with the United Kingdom (n = 4), Australia (n = 2), Korea (n = 2) and Netherlands (n = 2) most represented. Intervention aims included improving upper limb function, language impairment (in the presence of aphasia), functional recovery (physical and cognitive), exploring the impact of pre-commitment on levels of engagement and understanding patient preferences relating to technology-based interventions. Most studies evaluated interventions targeting stroke survivors alone (n = 16), with a smaller number looking at interventions with patient–caregiver dyads (n = 2). Intervention durations ranged from 2 to 14 weeks. Technology-based tools were varied and included upper limb exergaming or robotic devices (n = 4)^8,38,45,48 and tablet/smartphone applications (n = 14).^18,35–37^{,39–44,46,47,49,50} Data in this review represents the experience of 1081 participants. Age was reported using various summary measures, with the average of reported mean ages being 68 years (range 58–74). Sex of participants was reported in 17 (94%) studies, and most participants were male (56%). Other baseline characteristics included stroke type and severity, time since stroke, functional status and prior technology use (see Table 2).

Table 2.

Characteristics of the included studies.

Author	Country	Study design and purpose	Inclusion/exclusion criteria	Sample size	Technology- based intervention	Participant characteristics (intervention group)	Intervention/comparison	Outcomes
Bassindale, 2025	United States	Prospective feasibility and user experience study looking at progressive cued activity to promote functional use of the UL.	Inclusion: Unilateral stroke within the last 30 days, >18 years old and English speaking.Exclusion: Inability to give informed consent or follow 2 stage instructions, severe cognitive impairment or concurrent illness limiting participation.	30	Two wearable activity monitors, a smartphone with Bluetooth capability and a custom smartphone application which manages activity scheduling and stores data.	Age, (years), mean (SD): 61.1 (14.8)Total admission days, mean (SD): 13.3 (5.2)FMA-UE Motor, mean (SD): 45.6 (25.1)MoCA, mean (SD): 25.9 (4.2)	Intervention: Activity monitors were worn on both wrists for up to 8 h/day. Each participant was delivered 3×30-min scheduled cued activity sessions (activated every 30 s).Cues prompted participants to complete 1 of 3 activities: tap the device on the other hand, assist the arm to move and independent movement of the arm.	Usability: Systems Usability ScaleSatisfaction; Quebec User Experience and Satisfaction with TechnologyMotivation: Intrinsic Motivation Inventory
Broderick, 2021	United Kingdom	Prospective feasibility study evaluating the dose achieved using an exergaming device for the UL.	Inclusion: Adults with stroke within 4 weeks, UL weakness and capacity to consent.Exclusion: Medical instability, unremitting UL pain, visual impairment, significant language or communication barrier, photosensitive epilepsy and participation in a concurrent research trial.	30-device and standard care20-retrospective audit of standard care	An UL handheld exergaming device (non-immersive virtual reality) that senses grip force and tracks finger, wrist and arm movements.	Age, years (IQR): 72.5 (61–79)Sex (female), n (%): 14 (47)Haemorrhagic stroke, n (%): 5 (17%)NIHSS (IQR): 6 (2–10)Total admission, days (IQR): 18.5 (14–38)mRS, (IQR): 3 (3–4)Enrolment, days (IQR): 8 (5 -14)	Intervention: Research therapist taught participants how to use device.After first session, daily use with affected UL encouraged.Caregivers welcome to engage and assist participant.	Feasibility and FidelityAcceptabilityTreatment efficacy
Broderick, 2023	United Kingdom	Prospective, non-randomised feasibility study evaluating the dose achieved using an exergaming device for the UL.	Inclusion: Adults with acute or hyper-acute stroke, objective new UL weakness and capacity to consent.Exclusion: Uncompensated visual deficits, unremitting UL pain or significant language or communication difficulties.	30	An UL handheld exergaming device (non-immersive virtual reality) that senses grip force and tracks finger, wrist and arm movements.	Age (years), mean (SD): 70.3 (11.9)Sex (female), n (%): 16 (53)Haemorrhagic stroke, n (%): 8 (27)NIHSS, mean (SD): 8 (4.4)Time since stroke (days), mean (SD): 11.1 (8.1)	Intervention: Research therapist taught participants how to use device.After first session, daily use with affected UL encouraged.Caregivers welcome to engage and assist participant.	UsabilityTechnology acceptabilityTechnology adoption
De Cock, 2021	Belgium/ Netherlands	Prospective observation study of an aphasia exercise programme.	Inclusion: 18 years or older with the presence of aphasia following acute stroke with minimum proficient language in Dutch.Exclusion: Severe psychiatric, cognitive and/or visual disorders that hindered technology-delivered aphasia therapy.	25	An online therapeutic tablet application (STAPP).	Age (years), mean (SD): 65 (17)Sex (female), n (%): 14 (56)Haemorrhagic stroke, n (%): 4 (16)NIHSS admission, median (IQR): 7 (3− 15)Time since stroke, until study enrolment in days, median (IQR): 5 (3− 6)	Intervention: Programme individually tailored by the speech pathologist (type and number of exercises, number of items per exercise and difficulty level).Participants encouraged to practice for a minimum of 30 min/day, with no upper limit.	FeasibilityUsabilityAcceptability
Devittori, 2024	Switzerland	Pilot feasibility study of an UL rehabilitation robot in a clinical setting.	Inclusion: Aged (18–90), within 6 weeks from stroke, pre-stroke mRS ≤1, NIHSS ≥1 in at least one item regarding motor or sensory function and ataxia, able to consent.Exclusion: Moderate to severe aphasia and cognitive deficits, UL impairment due to other pathologies or severe pain in affected arm, Modified Ashworth Scale >2 for at least one arm muscle, pacemakers and other active implants.	13	Rehabilitation robot for the ULEnrolment (weeks): 4.	Age (years), mean (range): 65.9 (49–77)Sex (female), n (%): 3 (23)Haemorrhagic stroke, n (%): 3 (23)Time since stroke (days), mean (range): 14.7 (7–33)	Intervention: Week 1 (training), week 2 supervised by therapist with no assistance. In the unsupervised phase (week 3 and 4), the device was turned on in a freely accessible room, participants could train during a 45 min timeslot indicated on their daily schedule (business days only), as well as in their free time.	FeasibilityUsabilityTreatment efficacy
Fonte, 2025	Italy	A single-blind, randomised controlled trial looking at safety, feasibility and preliminary efficacy of cognitive training using an electronic device.	Inclusion: Ischaemic or haemorrhagic stroke within 2 weeks of onset, under 85 years of age, presence of attentional deficits as measured by the Trails Task and/or the Broken Hearts subtest of the Oxford Cognitive Screen.Exclusion: Documented visual impairments (e.g. cataracts, diabetic retinopathy, glaucoma, hemianopia), psychiatric disorders not pharmacologically stabilised; drug or alcohol abuse, history of dementia, severe comprehension deficits and spatial neglect.	23	A web-based platform, installed on a Samsung Galaxy tablet.Enrolment (weeks): 2 with daily 60-min sessions.	Age (years), mean (SD): 67 (9.64)Sex (female), n (%): 7 (30%)Haemorrhagic stroke, n (%): 4 (36%)Time from onset, days (SD):20.9 (13.7)	Intervention: Self-guided cognitive training each afternoon using an electronic device. Exercises prescribed based on individual abilities.Control: Self-administered cognitive training each afternoon using conventional paper-and-pencil materials.	Visual attention: Trail Making Test-Part AInhibitory control: Stroop test.Planning and problem-solving: Elithorn's Perceptual Maze Test
Fusari, 2022	United Kingdom	A non-randomised feasibility trial of a digital intervention aimed at supporting arm and hand rehabilitation.	Inclusion: Adults diagnosed with stroke less than 6 months ago, requiring arm and hand rehabilitation, medically stable, capacity to consent.Exclusion: Presenting without arm pain or oedema.	24	Smartphone applicationEnrolment (weeks): 14.	Age (years), mean (SD);61.1 (12.5)Sex (female), n (%): 11 (54)Haemorrhagic stroke, n (%): 2 (10)Stroke onset (days), mean (SD): 42.2 (33.04)	Baseline assessments in week 1. OnTrack intervention weeks 2–13. Follow-up assessments in week 14.Intervention: 3 main components: activity tracking, motivational content and self-management coaching.	FeasibilityAcceptabilityUsabilityAdoption/adherence
Jung, 2021	Korea	A non-randomised trial of a caregiver-supervised self-exercise programme with videos.	Inclusion: Adults with first-ever stroke within 4–6 weeks, medically stable, can participate in rehabilitation, able to communicate and have a willing caregiver.Exclusion: Previous head trauma- or brain-related illness, impaired cognition, vision loss or aphasia, non-ambulatory before stroke, diagnosed with osteoporosis or another severe musculoskeletal problem.	88/96 (historic)	Video exercise programmes on smartphoneEnrolment (weeks): 4 weeks of exercise therapy, executed with a caregiver.	Age (years), mean (SD): 60.8 (14.8)Sex (female), n (%): 39 (44%)Modified Barthel Index, mean (SD): 50.46 (24.15)Duration from stroke onset (days), mean (SD): 46.40 (11.10)	Intervention: Patient and caregiver dyads were trained. Dyads were encouraged to perform their prescribed programme 5 times/week for 60 min. Level of difficulty was progressed when patients were ready.Control: PT and OT sessions, 2 times/day, for 30 min/session. Therapy focused on UL and LL strength, balance, mobility and function.	Treatment efficacy
Kenny, 2022	United Kingdom	A randomised-controlled feasibility study of video-guided exercise for facilitating UL movement.	Inclusion: Adults diagnosed with first time stroke, mild or moderate UL paresis, no previous/existing UL pathology, not enrolled in another trial, and a predicted 4-week hospital stay with active rehabilitation goals.Exclusion: Severe aphasia and unable to consent.	5/5	Video exercises on tablet.4 weeks from first therapy session after recruitment, or until discharge if less than 4 weeks.	Age range: 41–84Sex (female), n (%): 2 (40%)	Intervention: Participants were provided with individualised videos to complete outside of therapy. Exercises were filmed during the participants’ normal therapy sessions. Participants were free to undertake exercises as often as they wished.Control: Standard physiotherapy input (repetitive UL task training and functional task training), with out-of-therapy-time exercise instructions presented verbally or in writing.	FeasibilityAcceptability
Kim, 2025	Korea	A randomised controlled trial looking at non-inferiority of an AI-driven cognitive tele-rehabilitation programme.	Inclusion: Sub-acute stroke within 6 months and cognitive impairment with a score of 26 or lower on the Korean version of the Mini Mental State Examination 2.Exclusion: Significant visual or auditory impairment, insufficient digital literacy and severe psychiatric disorders.	27/28	Self-guided, AI-driven tele-rehabilitation using Zenicog delivered via a tablet with a cloud-based monitoring system.24 sessions within 6 weeks.	Age (years), mean (SD): 66.8 (11.5)Sex (female), n (%): 23 (42)Haemorrhagic stroke, n (%): 20 (36)Since onset: (days), 1st and 3rd quartiles: 54 (30, 87)Modified Barthel Index, mean (SD): 47.67 (22.40)	Intervention: Once per day, lasting approximately 30 min, participants engaged with an AI-powered engine which dynamically generated individualised training plans based on real-time performance. The AI system comprised 4 key training modules: record collection, question analysis, performance prediction and subsequent recommendations.Control: Therapist selected training tasks manually and supervised cognitive rehabilitation using the same system. The AI recommendations were deactivated.	Treatment effectivenessUsability
Kotov, 2020	Russia	A randomised controlled trial looking at the efficacy of cognitive stimulation through PC technology.	Inclusion: Patients with ischaemic stroke admitted to hospital in the acute period.Exclusion: None.	50/50	MOTO-med and Tablet PCFrom day 5, participants received daily robot mechanotherapy.	Age (years), mean (SD): 65.02 (1.5)Sex (female), n (%): 20 (40)mRS, mean (SD): 3.44 (0.09)Rivermead Mobility Index, mean, (SD): 1.48 (0.11)	Intervention: First session was no more than 7 min, increasing session duration to 30–40 min, 3x/day. The rehabilitation programme included tablet PC technology with game apps for independent exercises by the patient with the aim of developing memory, perception, reactions and countingControl: Standard therapy.	Treatment efficacy
Mallet, 2016	Canada	Pilot feasibility study to assess the feasibility of a tablet-based speech-language therapy.	Inclusion: Mild-to-moderate communication deficits scoring 1 on the Best Language and/or dysarthria parameters of the NIHSS.Exclusion: Pre-existing speech, language, cognitive disorders or severe debilitating disorders, unable to follow simple one-step commands in English and/or unable to respond to yes/no questions reliably.	30	Tablet – iPad Air	Age (years), median (IQR): 62 (37–87)Sex (female), n (%): 8 (27)Number of days from admission to enrolment, median (IQR): 5 (6)	Intervention: An iPad with pre-loaded applications targeting communication deficits. Each participant was prescribed a programme deemed appropriate for impairment-based therapy by their speech language therapist. Introductory session of 15 min with a follow-up of 10 min.	FeasibilityAdherenceAcceptability
Pogrebnoy, 2025	Australia	Exploration of patient preferences in relation to paper or video-based resources to support independent activity on a rehabilitation ward.	Inclusion: Adult undergoing inpatient rehabilitation after stroke (>1 week prior) and able to consent.Exclusion: Not admitted for stroke rehabilitation, medically unstable/unwell or moderate to severe cognitive or communication impairment limiting ability to consent and/or participate in an interview.	9	Activities presented in a video format on a digital device.	Age, (years), range (%): 40–59: 4 (56)60–81+: 5 (44)Sex (female), n (%): 3 (33)Haemorrhagic stroke, n (%): 1 (11)Length of time on rehabilitation ward, <14 days, n (%): 5 (56)FIM score, 80 or below, n (%): 5 (56)	Intervention: Participants were shown a paper-based independent programme as well as two video prototypes (one with clinician performing the activity and the second with a stroke survivor performing the activity) and asked what they liked and disliked about each option.Participants were then asked what supports they would need to engage with each type of resource and whether they anticipated any difficulties.	Patient preferences
Rajaratnam, 2013	Singapore	A randomised controlled pilot study evaluating the efficacy of an interactive virtual reality balance retraining programme.	Inclusion: Recently experienced a first onset of stroke with moderate disability (mRS grade 3/4), medically stable and a Mini-Mental State Examination score >23.Exclusion: Terminal diseases, uncontrolled hypertension, angina, severe spatial neglect or visual impairments.	10/9	Virtual reality balance games on TV via a console.All subjects received 15 sessions of inpatient rehabilitation during their hospital stay.	Age, (years), mean (SD): 58.7 (8.6)Sex (female), n (%): 6 (60)Haemorrhagic stroke, n (%): 2 (20)mRS (grade 3/4): 6/4Days after stroke, mean (SD): 14.7 (7.5)	Intervention: Participants were introduced to either a Nintendo Wii-Fit or Microsoft Kinect game console. Each participant received 40 min conventional rehabilitation with 20 min of interactive virtual reality balance-related games.Control: 60 min of conventional rehabilitation	Treatment efficacy
Rémy-Néris, 2021	France	Phase III, randomised controlled trial, looking at the effects of self-rehabilitation using a mechanised device on upper extremity impairment.	Inclusion: 18 to 81 years old, diagnosed with haemorrhagic or ischaemic stroke 3 weeks to 3 months previously with FMA-UE score between 10 and 40 pointsExclusion: Pain in the affected shoulder >3/10, Boston Diagnostic Aphasia Examination score ≤3 points, fatigue or visual impairment and inability to sit independently.	107/108	Exoskeleton for ULUsual rehabilitation was followed by two 30-min self-rehabilitation sessions/day, 5 days/week for 4 weeks.	Age (years), mean (SD): 58.1 (14.1)Sex (female), n (%): 40 (37.4)Haemorrhagic stroke, n (%): 30 (28)NIHSS, mean (SD): 5 (2.4)FIM, mean (SD): 98.4 (17.7)	Intervention: The ArmeoSpring exoskeleton device was used for the gravity-supported, games-based self-rehabilitation.Control: Participants were instructed to perform 10 min of stretching (5 s stretches and 20 min of active exercises (10 repetitions of each exercise) that involved simple movements of the UL joints through range.	Treatment efficacyAcceptability
Studer, 2021	Germany	Randomised controlled study to assess the benefit of pre-commitment to increasing engagement with a cognitive training programme.	Inclusion: Adults with ischaemic or haemorrhagic stroke with visuo-spatial memory impairments.Exclusion: Moderate or severe aphasia, a diagnosis of dementia, severe deficits in multiple cognitive domains, inability to provide consent and presence of multi-resistant bacteria.	33/31	Tablet training for visuo-spatial working memory30 min of self-directed training with the Wizard memory game for 2 weeks.	Age (years), mean (SE): 73.7 (1.3)Days after stroke, mean (SE): 39.6 (2.8)Barthel Index, mean (SE): 60.1 (11.1)	Intervention: The pre-commitment group could choose to restrict visitors and a social punishment comprising physician surveillance.Control: No pre-commitment offered. Usual management.	Adherence
Vandenberg, 2016	Australia	Pragmatic pilot randomised trial investigated the effects of a caregiver-mediated exercise programme.	Inclusion: Understand English, in early rehabilitation phase, have a willing caregiver, able to follow instructions, mobility deficit following stroke and no depression with mini mental score of at least 18.Exclusion: Patient and caregivers with serious co-morbidity impacting participation.	31/32	Exercise app on a tablet30 min, 5 times a week for 8 weeks.	Age (years), mean (SD): 64.5 (18.5)Sex (female), n (%): 12 (39)Haemorrhagic stroke, n (%): 4 (13)	Intervention: A caregiver-mediated training programme with using a customised exercise app which was loaded with 37 standardised exercises aimed to improve gait and gait-related mobility. Participants also wore an activity monitor.Control: Usual care.	Treatment efficacyAcceptability
Vloothuis, 2019	Netherlands	Observer-blinded, randomised controlled trial evaluating whether a caregiver mediated exercise programme with e-health support can improve functional outcomes and facilitate early supported discharge home.	Inclusion: Diagnosed with stroke, lived independently prior to stroke and is planned to be discharged home, able to follow instructions, have mobility deficit and can appoint a carer. Exclusion: Patient and caregivers with serious co-morbidity which impacts participation.	32/34	Tablet/smartphone appPatient–caregiver dyads aimed for 5×/week, 30 min/day of structured exercises (total 20 h over 8 weeks).	Patient: Age (years), mean (SD): 60.5 (14.8)Sex (female), n (%): 11 (34)Time since stroke (days), median (IQR): 36 (28–57)Haemorrhagic stroke, n (%): 10 (31)Caregiver: Age (years), mean (SD): 53.9 (14.9)Sex (female), n (%): 23 (72%)	Intervention: A caregiver-mediated exercise programme (CARE4STROKE) + e-health support + usual care. The exercise programme was composed by a trained physical therapist. The therapist could choose from 37 standardised exercises. For each patient, exercises were combined into a patient-tailored, progressive training regimen, related to the patient's goals.Control: Usual care according to Dutch stroke rehabilitation guidelines.	Treatment efficacy for patient and caregiver
Vloothuis, 2018	Netherlands	Descriptive methodology paper.					Development of 8-week CARE4STROKE programme used in subsequent study as intervention.
Wischmann, 2025	Germany	A randomised controlled trial looking at whether tablet-assisted speech and language therapy using the Neolexon application is superior to standard speech language therapy.	Inclusion: Adults with aphasia and a language screening test score of <13 points, German as a native language, no known pre-existing dementia or aphasia due to other neurological disorders, life expectancy of >1 year and provision of consent.Exclusion: Not stated.	53/51	Tablet-assisted speech language therapy through a Neolexon application30 min daily.	Age, (years), mean (SD): 74.4 (11.14)Sex (female), n (%): 54 (52)Haemorrhagic stroke, n (%): 10 (9.6)NIHSS, median (IQR): 6 (3–12)mRS, median (IQR): 4 (3–5)	Intervention: Individualised digital exercises via the app were provided by a qualified speech and language therapist.Control: Speech Language Therapy provided according to local clinical standards supported printed handouts.	Treatment effectiveness

AI: artificial intelligence; FIM: functional independence measure; FMA-UE: Fugl-Meyer Assessment-upper extremity; IQR: inter-quartile range; LL; lower limb; MoCA: Montreal Cognitive Assessment; mRS: modified Rankin Scale; NIHSS: National Institutes of Health Stroke Scale; OT: occupational therapy; PC: personal computer; PT; physiotherapy; STAPP: speech therapy app; TV: television; UL: upper limb.

Quality appraisal

Using the PEDro Scale, most randomised controlled trials were rated good (n = 6)^38,40–44 or fair (n = 3)^35,37,39 in quality, with one assessed as poor quality³⁶ (Supplemental File S2). The non-randomised, feasibility and observational studies were either rated as high (n = 3)^8,36,49 or moderate (n = 6)^{8,18,42,43,45,48} in quality, using the Critical Appraisal Skills Programme checklist (Supplemental File S3).

Outcome measures

Acceptability outcomes were reported in 12 (63%) studies^{8,36,38,40,43,45–51} (Table 3). Measures for acceptability varied across the included studies, with the most common concepts pertaining to participants’ and caregivers’ perceived usefulness, ease of use and satisfaction. Most participants perceived technology-based tools to be convenient, usable and acceptable. In one study,⁴⁹ participants preferred a technology-based format over paper alone, as it was perceived to be more engaging and meaningful. Adherence to the intervention protocol was reported in eight (67%) studies^43–48^,50,51 and varied from 42% to 83%. Feasibility outcomes were reported in 12 (63%) studies.^{8,36,38,40,43–48}^,50,51 Recruitment rates, reported in eight studies (67%), ranged from 38% to 81%. Retention rates, reported in six studies (50%), were generally very high and commonly attributed to daily contact and encouragement from treating therapists. Levels of assistance required for participants to be able to interact with their technology-based tool were reported in five (42%) studies. In three of those studies,^36,38,45 only minimal assistance was needed by enrolled participants, whereas in the other two studies,^8,50 half of their participants required at least some level of assistance to engage with their technology-based tool. One study⁵⁰ reported that limited prior experience with the use of computers or tablets, as well as moderate aphasia, was negatively correlated with participants’ level of independence. However, in another study⁴⁵ participants felt well supported in using technology-based tools even without prior experience. Safety was reported in eight (67%) studies,^{3,8,40,43–46}^,48 with seven (88%) reporting no adverse events. In one study,⁴⁴ there were a significant number of participants who withdrew their consent, but this did not appear to be related to safety concerns pertaining to the intervention. Three episodes of device failure were reported in another study,⁸ but these were rectified in a timely manner by a member of the research team.

Table 3.

Acceptability and feasibility.

	Acceptability		Feasibility
Study	Outcome measure	Results	Recruitment/retention	Adherence	Assistance	Safety
Bassindale, 2025	System Usability Scale	Mean participant SUS scores: 80.1 (SD 13.8)	12 participants withdrew (attrition rate: 28.6%)	Participants wore the devices; days, mean (SD) 5.6 (2.6)		There were no failures due to battery loss or other causes.
	Quebec User Experience and Satisfaction with Technology	Mean QUEST score (4.52 ± 0.46)
	Intrinsic Motivation Inventory	Mean (SD) IMI scores: value/usefulness 6.2 (0.9), perceived choice 6.0 (0.9), perceived competence 5.8 (1.0), effort/importance 5.4 (1.4) and interest/enjoyment 4.9 (1.3)

Broderick, 2021	Technology Acceptance Model	TAM rating: 77% Perceived usefulness: 63% Intent to use: 73% Ease of use: 57%	40% recruitment (30/75 eligible)	Not reported	62% required some assistance to use device.	No adverse events reported.
						Three episodes of device failure occurred
		Stroke severity (NIHSS) correlated negatively with technology acceptance.				Three episodes of device failure occurred
Broderick, 2023	Technology Acceptance Model (TAM)	TAM rating: 68% Perceived usefulness: 77% Intent to use: 86% Ease of use: 58%	40% recruitment (30/75 eligible)	Median 26 (SD 12) min/day.	Technology usable for 26/30 (87%) participants.4 participants unable to use the device due to weakness.	Not reported
De Cock, 2021	Usability questionnaire5 statements on 5-point scale	Median satisfaction rate of 91% (IQR = 75–100)	Recruitment 81% (25/31 eligible).Retention 96% (23/24).	Adherence 42% (10/24).Practice median 23 min/day.	58% (14/24) were able to use the app independently. Aphasia severity and lack of experience with technology use was significantly associated with the inability to work independently with the app (p = 0.011 and 0.009 respectively)	Not reported
Devittori, 2024	User experience questionnaire	Customer satisfaction score was 81.8% for Usability 1 and 80.0% for Usability 2.	Not reported	Adherence 77%	12/13 able to progress to unsupervised training.	No adverse events or devise deficiencies.
Fusari, 2022	System Usability Scale	84.6 (13.1)	Recruitment: 65% (24/37 eligible)Retention: 50% (12/24)	Adherence 71% (17/24)A 63% mean increase in activity	Not reported	No adverse events reported
Kenny, 2022	Qualitative questions	Intervention valued as a visual guide for error correction, environment set-up and motivation. Cognitively impaired patients less likely to engage.	Recruitment 44%	Not reported	None of the participants had difficulty handling the tablet, despite fears that finger dexterity would limit capacity	Not reported
Kim, 2005	Usability across eight domains: Equitable and flexible Simple and intuitive Information is perceptible Tolerance for error Low physical effort Size and space for use Overall product quality and Overall satisfaction	Both the intervention and control groups showed similarly favourable responses across all domains, with no statistically significant differences.Median scores for overall product quality: 3.6 [3.2, 4.0] andoverall satisfaction: 4.0 [3.3, 4.0]	76 potential participants were screened. 63 of those met inclusion criteria (83%).55 participants completing the intervention (87% of eligible participants).	Some participants reported mild inconveniences such as poor visibility of on-screen text or difficulty using the touch interface; however, these were minor and did not affect overall adherence.	Unable to measure	No serious adverse effects related to the intervention were identified throughout the study period.
Mallet, 2016	Convenience survey	92% rated 3/5 (moderately convenient) or better	Recruitment 68% (30/44 eligible)Retention 97% (29/30)	Adherence 83% complete the recommended 1 h/day	Not reported	Not reported
Pogrebnoy, 2025	Semi-structured interviews to understand patient preferences.	Four key themes identified: Role of the clinician in setting up an independent practice programme (including opportunity to practice, provision of resources and environmental set-up). Factors that motivate survivors of stroke (type of content, family and loneliness). Drivers for patient preferences (preferences changing over time). Key accessibility needs for survivors of stroke (paper vs video and assistance needed).
Rémy-Néris, 2021	Perception questionnaire, 5 statements on 5-point scale	Ease of learning: 4Ease of performing: 4Satisfaction: 5	Recruitment 38% (215/568 eligible)None discontinued intervention		Sense of independence: 4	No serious adverse events related to the intervention were reported.
Vandenberg, 2016	Satisfaction questionnaire	Caregivers reported a higher self-efficacy (p = 0.0078) and less fatigue (p = 0.0369)	Recruitment 72% (66/92 eligible)	Not reported	Caregiver-mediated intervention	No adverse events reportedLower readmission rates in intervention group.
Wischmann, 2025			Recruitment: 2.5% (104/4097)Retention: 54% (56/104)	A numerically higher number of participants in the tablet-assisted SLT group conducted self-training compared to the control group. Amongst those who performed self-training, participants using the Neolexon app trained for longer total durations than those using analogue training materials.		Eleven participants (10.6%) died before completion of the 90-day follow-up, 28 (26.9%) withdrew consent, 4 (3.8%) were not able to participate in study visits due to medical reasons and 2 (1.9%) were lost to follow-up.

IMI: Intrinsic Motivation Inventory; IQR: inter-quartile range; NIHSS: National Institutes of Health Stroke Scale; QUEST: Quebec User Experience and Satisfaction with Technology; SLT: speech language therapy; SUS: System Usability Scale; TAM: Technology Acceptance Model.

Eleven studies (58%) evaluated the effectiveness of a technology-based intervention (Table 4). The most common outcome measuring effectiveness was the completion of self-directed activity outside of therapy (measured in minutes). Other outcomes included stroke-related disability (modified Rankin Scale, Functional Independence Measure, National Institutes of Health Stroke Scale and Stroke Impact Scale), mobility (Timed Up and Go Test), activities of daily living (modified Barthel index), upper limb function (Fugl-Meyer Assessment–Upper Extremity, Action Research Arm Test), self-efficacy (Self-Efficacy Scale), aphasia severity (Bielefelder Aphasia Screening-Rehabilitation Version) and balance (Berg Balance Scale, Functional Reach Test). Effectiveness for caregivers was measured using the Caregiver Strain Index and Caregiver Quality of Life Scale.⁴¹ Seven (64%) studies^{8,18,35,37,40,42,45} reported significant differences in favour of the intervention groups whilst four (36%) studies^38,41,43,44 found no significant between-group differences.

Table 4.

Effectiveness.

Study	Outcome measure/s	Time points of outcome measurement	Unsupervised therapy time	Physical function	Psychological/cognitive function
Broderick, 2021	Duration of exergaming device use Self-report by patients/caregivers and Recorded by the device	N/R as patient commitments were different.	Median 31 min	Increased daily UL exercise duration 2-fold (on average), and repetitions 8-fold, compared to conventional supervised therapy.Nil significant difference between conventional (supervised) therapy times between groups.	Half of the participants had cognitive scores within the impaired range (MoCA 22/30) and there was no evidence for less device use with reduced cognitive ability.
Devittori, 2024	Dose of unsupervised robot-assisted therapy measured as: Therapy duration (min/day) Total minutes over 2 weeks Number of task repetitions and % increase in physical therapy time	Baseline and 2 weeks	Mean 360.0 minRange from 197.2 to 608.7 min)	12 of 13 participants did not require a therapist to use the device and were able to complete therapy unsupervised.Participants did not tolerate 45 min scheduled sessions. No significant difference recorded between therapists' presence or absence.	Participant with the lowest cognitive function score could not train unsupervised.
Fonte, 2025	TMT-A (s)Dual task (comparing performance on a single task with performance under dual task conditions)Stroop-time (s)	Baseline and 2 weeks	60 min every day for 2 weeks	Within-group comparison showed changes between pre- and post-treatment scores on the 3 outcome measures for the experimental group.	All participants who completed the protocol were able to carry out the sessions with minimal supervision.The intervention was well tolerated, easily integrated into standard inpatient rehabilitation schedules and perceived by participants as engaging and intuitive.
Jung, 2021	Berg Balance ScaleModified Barthel IndexSF-36	Baseline and 4 weeks	60 min every day for 4 weeks	Significant improvement in the intervention group to control:BBS: 33.56 ± 20.04Improved balanceMBI: 66.45 ± 23.03Improved functional independence level	Significant improvement in the intervention compared to control with the SF-36 scores
Kim, 2025	K-MMSE-2Trail Making Test A and BDigit Span Test5-level EQ-5D versionSelf-Efficacy ScaleModified Barthel Index	Baseline and 6 weeks	Once per day for 30 min	The AI-based intervention was determined to be non-inferior to the conventional therapist-supervised programme. Overall, the study provides clinically relevant evidence supporting the use of self-guided digital platforms as a complementary approach to traditional cognitive rehabilitation.	Both groups achieved significant gains across functional and cognitive domains, including MBI and MMSE, DSF, DSB, TMT-A, TMT-B.
Kotov, 2020	Modified Rankin ScaleBeck Depression and Anxiety Scale	21 days,3 months and 6 months.	First session is no more than 7 min, followed by increasing sessions to 30–40 min three times/day.	Significant improvements were shown in the mean values of the modified Rankin Scale that were reduced for the study group in comparison to the control for 21 days, 3 months and 6 months.	Statistically significant decrease in the Beck Scale at day 21 days as well as 6-month time points.
Rajaratnam, 2013	Modified Rankin ScaleReach TestTimed Up and GoBerg Balance ScaleCentre of PressureModified Barthel Index	Baseline and after 15 sessions of rehabilitation	60 min over 15 sessions	Significant difference in the FRT scores between the two groups after 15 sessions. Experimental group also had significant improvements in TUG and MBI.	N/R
Rémy-Néris, 2021	Fugl Meyer Upper ExtremityARATAshworth scorePain VASFIMSIS hand functionEQ- 5D	Baseline, 4 weeks, 3 months, 6 months and 12 months	30 min session/day, 5 days for 4 weeks	No significant difference recorded for any of the outcomes for both groups	N/R
Vandenberg, 2016	The amount of (additional) practice performed by the participating dyads in both the intervention and control group was self-reported with a diary.Mobility part of the stroke-specific self-reported health status measure of the Stroke Impact Scale	Baseline, 8 and 12 weeks	30 min session 5 days/week for 8 weeks	Less fatigue reported by caregivers and higher self-efficacy reported at 12 weeks for the intervention groupThere were also significant between-group differences in favour of the intervention group for the Nottingham Extended ADL index at week 8The intervention group also reported 2 h a week of extra therapy over the 8 weeks.	Improved caregiver well-being at 12 weeks.
Vloothuis, 2019	Self-reported mobility domain of the Stroke Impact ScaleLength of stay from stroke onset until discharge from inpatient rehabilitationHospital Anxiety and Depression Scale (HADS)Caregiver Strain Index, Caregiver Quality of Life Scale	Baseline, 8 weeks and 12 weeks	30 min over 5 days for 8 weeks (150 min/week) and a total 20 h over 8 weeks in addition to usual care.	No significant difference in SIS or LOS	Patient HADS Anxiety and Caregiver HADS Depression both improved in intervention group.
Wischmann, 2025	Change in BIAS-R percentile rankSelf-training frequencyCumulative self-training durationNational Institutes of Health Stroke ScaleModified Rankin ScaleBarthel Index	Baseline and at 90 days	30 min/day	After adjustment for age, baseline NIHSS and LAST Tablet-assisted SLT with Neolexon was associated with an improvement in BIAS-R percentile ranks

ADL: activities of daily living; BBS: Berg Balance Scale; ARAT: Action Reach Arm Test; BIAS -R: Bielefelder Aphasia Screening Rehabilitation Version; DSB: Digit Span Backward; DSF: Digit Span Forward; EQ-5D-5L: 5-level EQ-5D; FIM; Functional Independence Measure; FRT: Functional Reach Test; HADS: Hospital Anxiety and Depression Scale; K-MMSE-2: Korean Mini-Mental State Examination; LAST: Language Screening Test; LOS; length of stay; MoCA: Montreal Cognitive Assessment; MBI; Modified Barthel Index; NIHSS; National Institutes of Health Stroke Scale; N/R: not reported; SES: Self-Efficacy Scale; SF-36: 36-Item Short Form Health Survey; SIS; Stroke Impact Scale; SLT: speech language therapy; SUS: System Usability Scale; TMT-A/B: Trail Marking Test-A/B; TUG: Timed Up and Go; UL: upper limb; VAS: Visual Analogue Scale.

Barriers and enablers

Barriers and enablers were mapped to the three domains of the COM-B framework (Figure 2).

Barriers to physical capability were identified as severity of stroke, poor literacy in technology use as well as significant cognitive, motor and visual deficits. Reported enablers were optimising the technology-based intervention to the intended user, family involvement, education and patients having an opportunity to familiarise themselves with the technology-based tool before being expected to manage the tool independently. Barriers to psychological capability were participant low mood and not understanding the purpose of self-directed practice. Enablers were regular contact with prescribing clinicians, education for caregivers and patients and availability of caregivers.

Barriers to physical opportunity included the inability of patients to manage navigation of the technology-based tool due to stroke-related impairments. Enablers were a well set-up room, availability of required equipment and reliable technical support. Barriers to social opportunity were set-up costs, a lack of multi-lingual resources and a lack of external encouragement whilst enablers were structured clinician and family support and shorter sessions.

Barriers to reflective motivation were fatigue and not having clear targets for expected activity outside of therapy time, whilst enablers were having access to motivating content, providing patients with choice, customisation of the technology-based tool and offering training opportunities to caregivers and patients. Barriers to automatic motivation were being prescribed repetitive tasks and not understanding the purpose of self-directed therapy, and enablers were scheduling self-management on the timetable, clarity in instructions and group activity.

Figure 2.

Data mapped to the COM-B framework.³⁴ Capability refers to whether an individual has the knowledge, skills and abilities to engage in a behaviour. Domain of capability is separated into physical and psychological. Psychological capability includes mental state, knowledge and skill, whereas physical capability is when an individual feels physically capable of performing a behaviour or achieving an outcome. Opportunity refers to external factors that make the execution of a behaviour possible. Domain of opportunity is broken up into social and physical. Physical opportunity describes opportunities provided by the physical environment, whereas social opportunity is opportunities provided through a social setting. Motivation refers to internal processes that influence decision making and behaviour. Domain of motivation is broken up into automatic and reflective. Reflective motivation describes the reflective process involved in making plans, whereas automatic motivation more describes impulses and inhibitions. COM-B: Capability, Opportunity, Motivation and Behaviour.

Discussion

Our findings, derived from 18 studies (19 datasets), suggest that it is feasible, acceptable and safe to use technology-based interventions for most individuals undergoing inpatient stroke rehabilitation. It appears that technology-based interventions have the capacity to successfully promote increased patient activity outside of structured therapy time. Key enablers were clinician support during set-up, opportunities for practice and structured scheduling. Barriers included participant cognitive or physical limitations, fatigue and low mood.

Engaging in self-directed programmes outside of structured therapy time can address the known gap that exists between recommended levels of activity and the known inactivity amongst survivors of stroke undergoing inpatient rehabilitation.⁸ Literature on existing resources, aimed at supporting patients to increase their activity levels during inpatient rehabilitation, often provided in the form of paper-based prompts or verbal instructions, suggests that current practice is ineffective.⁵² Our review suggests that technology-based interventions have the potential to more effectively support increased activity outside of therapy time when compared to usual care, with 7 out of 11 included studies reporting superior outcomes for participants who engaged with technology-based interventions. These findings align with emerging literature supporting the role of technology in increasing rehabilitation engagement.^53–55

Clinicians who wish to prescribe the use of technology-based tools during inpatient rehabilitation need to consider patient, prescribing clinicians and environmental barriers as part of their prescription. Findings from our review highlight that technology-based interventions, which are intended for use by survivors of stroke, need to be designed to consider stroke-specific impairments which are known to influence engagement. Accessible design needs to not only consider physical impairments but also take into account the impact of visual and memory deficits as well as the impact of altered cognition and fatigue. Content needs to be motivating, and user feedback needs to be provided regularly to promote sustained engagement. Finally, environmental factors, technical support requirements and safety considerations need to be carefully planned for during programme set-up. Clinicians play a critical role in helping patients overcome barriers to engagement when they initially prescribe technology-based interventions to their patients. Clinicians are encouraged to reflect on their own practice and consider if they are appropriately supporting their patients in setting up effective self-directed programmes when they are initially prescribed and subsequently reviewed during the rehabilitation episode. Findings from this review should give clinicians the confidence that prescribing technology-based interventions to survivors of stroke undergoing inpatient rehabilitation can effectively promote increased self-directed (or caregiver-assisted) activity.

Early initiation of self-directed rehabilitation is known to be beneficial for patients who require inpatient rehabilitation after stroke.⁵⁶ Findings from our review suggest that technology-based interventions are acceptable and feasible for most survivors of stroke undergoing inpatient rehabilitation. This is in keeping with prior work in relation to self-directed interventions in an inpatient rehabilitation setting but existing literature is not specific to stroke or to technology-based tools which can be used in a self-directed manner.¹² Most participants in studies included in our review found technology-based tools easy to use, they perceived the interventions to be useful, and reported feeling satisfied with their experience of interacting with their technology-based tool. Findings from this review suggest that patients with aphasia or with limited technology exposure prior to their stroke are still able to benefit from being prescribed a technology-based intervention but may require additional education and support from their clinician, or caregiver, to help build their capability. It should be noted that in two studies,^8,44 stroke severity correlated negatively with technology acceptance and in another,⁵⁰ aphasia severity and lack of experience with technology use was significantly, and negatively, associated with self-directed engagement. From this, a reasonable hypothesis is that whilst technology-based tools appear to be acceptable to most survivors of stroke undergoing inpatient rehabilitation, they may only be a feasible option, for self-directed use, by patients with mild-to-moderate physical impairment, mild-to-moderate aphasia and for patients who have a caregiver available for support.

We believe future research should focus on evaluating the long-term impact of technology-based interventions on the physical, cognitive and social recovery of survivors of stroke including health-related quality of life and participation in activities of daily living. It is important to further explore which types of technology-based tools and approaches are most acceptable and effective in the inpatient rehabilitation setting for survivors of stroke and their caregivers. Finally, literature on applicability and acceptability of technology-based interventions in diverse healthcare settings and with culturally and linguistically diverse patient cohorts remains sparse and should be a focus of future research.

A key strength of this review is our comprehensive search strategy, increasing the likelihood that all relevant published work, in this emerging field, was captured. The inclusion of studies using a variety of technology-based interventions and involving participants with a range of functional abilities increases the generalisability of our findings. However, the total number of studies included was relatively small (n = 18), and variation in study design as well as the types of technology-based tools used limited the ability to draw strong conclusions. Only studies published in English were included, which may limit the international relevance of this review. Furthermore, all studies took place in high-income countries; therefore, findings from this review may not be translatable to other settings where access to technology and literacy in technology use may differ.

Clinical messages

Technology-based interventions are acceptable, feasible and safe during inpatient rehabilitation after stroke.

Some barriers to engagement include cognitive impairment, severe aphasia, low mood, not understanding the purpose and insufficient support.

Enablers include appropriate training, targeted scheduling and motivating content with real-time feedback.

Further research must investigate their effects.

Supplemental Material

sj-docx-1-cre-10.1177_02692155261428703 - Supplemental material for Self-directed and caregiver-assisted technology-based interventions in inpatient stroke rehabilitation: A scoping review

Supplemental material, sj-docx-1-cre-10.1177_02692155261428703 for Self-directed and caregiver-assisted technology-based interventions in inpatient stroke rehabilitation: A scoping review by Dina Pogrebnoy, Jennifer RA Jones, Shahad Alsawaf, Lourdes Joseph, Pak Yin Lau and Kelly J Bower in Clinical Rehabilitation

Supplemental Material

sj-docx-2-cre-10.1177_02692155261428703 - Supplemental material for Self-directed and caregiver-assisted technology-based interventions in inpatient stroke rehabilitation: A scoping review

Supplemental material, sj-docx-2-cre-10.1177_02692155261428703 for Self-directed and caregiver-assisted technology-based interventions in inpatient stroke rehabilitation: A scoping review by Dina Pogrebnoy, Jennifer RA Jones, Shahad Alsawaf, Lourdes Joseph, Pak Yin Lau and Kelly J Bower in Clinical Rehabilitation

Supplemental Material

sj-docx-3-cre-10.1177_02692155261428703 - Supplemental material for Self-directed and caregiver-assisted technology-based interventions in inpatient stroke rehabilitation: A scoping review

Supplemental material, sj-docx-3-cre-10.1177_02692155261428703 for Self-directed and caregiver-assisted technology-based interventions in inpatient stroke rehabilitation: A scoping review by Dina Pogrebnoy, Jennifer RA Jones, Shahad Alsawaf, Lourdes Joseph, Pak Yin Lau and Kelly J Bower in Clinical Rehabilitation

Footnotes

Acknowledgements

We would like to acknowledge the University of Melbourne Doctor of Physiotherapy students, Zac McCosh, Amen Kassa and Chelsea Lee, who contributed to the early stages of this review as part of their Research Capstone subject.

ORCID iDs

Dina Pogrebnoy

Kelly J Bower

Funding

The authors received no financial support for the research, authorship and/or publication of this article.

Declaration of conflicting interests

The authors declared the following potential conflicts of interest with respect to the research, authorship and/or publication of this article: We acknowledge that a qualitative study included in this scoping review is co-authored by an author of this review. Data extraction from this qualitative study was completed by two independent authors to reduce the risk of bias.

Data availability statement

All data relevant to this review is included in the manuscript or as a . All reasonable requests for additional information can be made to the corresponding author.

Supplemental material

Supplemental material for this article is available online.

References

World Health Statistics 2021: monitoring health for the SDGs, sustainable development goals. Geneva: World Health Organization 2021. Licence: CC BY-NC-SA 3.0 IGO.

Stroke Foundation. Living Clinical Guidelines for Stroke Management. Stroke Foundation. 2023. Available at: https://informme.org.au/guidelines/living-clinical-guidelines-for-stroke-management.

Foley

McClure

Meyer

, et al. Inpatient rehabilitation following stroke: amount of therapy received and associations with functional recovery. Disabil Rehabil 2012; 34: 2132–2138.

McPake

Mahal

. Addressing the needs of an aging population in the health system: the Australian case. Health Syst Reform 2017; 3: 236–247.

Eng

Brauer

Kuys

, et al. Factors affecting the ability of the stroke survivor to drive their own recovery outside of therapy during inpatient stroke rehabilitation. Stroke Res Treat 2014; 2014: 1–8.

Bernhardt

Dewey

Thrift

, et al. Inactive and alone. Stroke 2004; 35: 1005–1009.

Bernhardt

Godecke

Johnson

, et al. Early rehabilitation after stroke. Curr Opin Neurol 2017; 30: 48–54.

Broderick

Almedom

Burdet

, et al. Self-directed exergaming for stroke upper limb impairment increases exercise dose compared to standard care. Neurorehabil Neural Repair 2021; 35: 974–985.

Dorward

Devlin

Brusco

, et al. Patients’ perceptions of participating in self-directed activities outside supervised occupational and physiotherapy within inpatient and home-based rehabilitation settings: a qualitative study. Disabil Rehabil 2025; 47. DOI: 10.1080/09638288.2024.2341872.

10.

Whittaker

Brusco

Hill

, et al. A self-management program increases the dosage of inpatient rehabilitation by 26 min per day: a process evaluation. Disabil Rehabil 2025; 47. DOI: 10.1080/09638288.2024.2339533.

11.

Schneider

Lannin

Ada

, et al. Increasing the amount of usual rehabilitation improves activity after stroke: a systematic review. J Physiother 2016; 62: 182–187.

12.

Brusco

Walpole

Kugler

, et al. Barriers and facilitators to implementing self-directed therapy activities in inpatient rehabilitation settings. Aust Occup Ther J 2023; 70: 617–626.

13.

Janssen

Ada

Bernhardt

, et al. Physical, cognitive and social activity levels of stroke patients undergoing rehabilitation within a mixed rehabilitation unit. Clin Rehabil 2012; 28: 91–101.

14.

Chin

Rosbergen

ICM

Hayward

, et al. A self-directed upper limb program during early post-stroke rehabilitation: a qualitative study of the perspective of nurses, therapists and stroke survivors. PLoS ONE 2022; 17: e0263413.

15.

Alhusayni

Cowey

Coulter

, et al. Personalised online upper-limb physiotherapy for stroke survivors during the inpatient phase: a feasibility study. Healthcare 2023; 11: 2582–2582.

16.

Cikajlo

Rudolf

Mainetti

, et al. Multi-exergames to set targets and supplement the intensified conventional balance training in patients with stroke: a randomized pilot trial. Front Psychol 2020; 11: 72.

17.

Peiris

Shields

Brusco

, et al. Additional physical therapy services reduce length of stay and improve health outcomes in people with acute and subacute conditions: an updated systematic review and meta-analysis. Arch Phys Med Rehabil 2018; 99: 2299–2312.

18.

Jung

S-H

Park

Kim

J-H

, et al. Effects of Self RehAbilitation Video Exercises (SAVE) on functional restorations in patients with subacute stroke. Healthcare 2021; 9: 65.

19.

Arksey

O’Malley

. Scoping studies: towards a methodological framework. Int J Soc Res Methodol 2005; 8: 19–32.

20.

Levac

Colquhoun

O'Brien

. Scoping studies: advancing the methodology. Implement Sci 2010; 5: 69.

21.

Tricco

Lillie

Zarin

, et al. PRISMA extension for scoping reviews (PRISMA-ScR): checklist and explanation. Ann Intern Med 2018; 169: 467–473.

22.

Veritas Health Innovation. Covidence [Systematic review software]. 2024. https://www.covidence.org/ .

23.

Clarivate. EndNote 21 [Reference management software]. 2023. https://endnote.com/ .

24.

PEDro. PEDro scale [Data collection tool]. Physiotherapy Evidence Database. 1999. https://pedro.org.au/english/resources/pedro-scale/ .

25.

Macedo

Elkins

Maher

, et al. There was evidence of convergent and construct validity of physiotherapy evidence database quality scale for physiotherapy trials. J Clin Epidemiol 2010; 63: 920–925.

26.

Pinheiro

Reis

AHS

Baldwin

, et al. Quantity and quality are increasing but there's room for improvement: a scoping review of physical activity intervention trials. Braz J Phys Ther 2024; 28: 101051. PMID: 38574557; PMCID: PMC10999812.

27.

Cashin

, et al. Clinimetrics: physiotherapy evidence database (PEDro) scale. J Physiother 2020; 66: 59.

28.

Critical Appraisal Skills Programme. CASP systematic review checklist . 2018. https://casp-uk.net/casp-tools-checklists/ .

29.

Long

French

Brooks

. Optimising the value of the critical appraisal skills programme (CASP) tool for quality appraisal in qualitative evidence synthesis. Res Methods Med Health Sci 2020; 1: 31–42.

30.

Butler

Hall

Copnell

. A guide to writing a qualitative systematic review protocol to enhance evidence-based practice in nursing and health care. Worldviews Evidence-Based Nurs 2016; 13: 241–249.

31.

Perski

Short

. Acceptability of digital health interventions: embracing the complexity. Transl Behav Med 2021; 11: 1473–1480. PMID: 33963864; PMCID: PMC8320880.

32.

Bowen

Kreuter

Spring

, et al. How we design feasibility studies. Am J Prev Med 2009; 36: 452–457. PMID: 19362699; PMCID: PMC2859314.

33.

Kim

. Efficacy versus effectiveness. Korean J Fam Med 2013; 34: 27. PMID: 23904951; PMCID: PMC3726789.

34.

Michie

van Stralen

West

. The behaviour change wheel: a new method for characterising and designing behaviour change interventions. Implement Sci 2011; 6: 42.

35.

Kotov

Isakova

Sheregeshev

. Possibilities for correcting emotional and behavioral impairments in stroke patients during rehabilitation therapy. Neurosci Behav Physiol 2020; 50: 156–161.

36.

Kenny

Gilmartin

Thompson

. Video-guided exercise after stroke: a feasibility randomised controlled trial. Physiother Theory Pract 2022; 38: 609–620.

37.

Rajaratnam

Gui Kaien

Lee Jialin

, et al.

Does the inclusion of virtual reality games within conventional rehabilitation enhance balance retraining after a recent episode of stroke?

Rehabil Res Pract 2013; 2013: 649561.

38.

Rémy-Néris

Le Jeannic

Dion

, et al. Additional, mechanized upper limb self-rehabilitation in patients with subacute stroke: the REM-AVC randomized trial. Stroke 2021; 52: 1938–1947.

39.

Studer

Timm

Sahakian

, et al. A decision-neuroscientific intervention to improve cognitive recovery after stroke. Brain 2021; 144: 1764–1773.

40.

Van den Berg

Crotty

Liu

, et al. Early supported discharge by caregiver-mediated exercises and e-health support after stroke: a proof-of-concept trial. Stroke 2016; 47: 1885–1892.

41.

Vloothuis

JDM

Mulder

Nijland

RHM

, et al. Caregiver-mediated exercises with e-health support for early supported discharge after stroke (CARE4STROKE): a randomized controlled trial. PLoS ONE 2019; 14: e0214241.

42.

Fonte

Damora

Abbruzzese

, et al. Self-administered cognitive rehabilitation using an electronic device in subacute stroke patients: a proof-of-concept study on safety, feasibility, and preliminary efficacy. NeuroSci 2025; 6: 09.

43.

Kim

Park

S-W

Jeong

, et al. AI-driven cognitive telerehabilitation for stroke: a randomized controlled trial. Front Neurol 2025; 16: 1636017.

44.

Wischmann

Brasch

Franzen

, et al. Tablet-assisted speech and language therapy for acute post-stroke aphasia: a randomized clinical trial (LEXI study). Eur J Neurol 2025; 32: e70443.

45.

Devittori

Dinacci

Romiti

, et al. Unsupervised robot-assisted rehabilitation after stroke: feasibility, effect on therapy dose, and user experience. J Neuroeng Rehabil 2024; 21: 52.

46.

Fusari

Gibbs

Hoskin

, et al. What is the feasibility and patient acceptability of a digital system for arm and hand rehabilitation after stroke? A mixed-methods, single-arm feasibility study of the ‘OnTrack’ intervention for hospital and home use. BMJ Open 2022; 12: e062042.

47.

Mallet

Shamloul

Corbett

, et al. RecoverNow: feasibility of a mobile tablet-based rehabilitation intervention to treat post-stroke communication deficits in the acute care setting. PLoS ONE 2016; 11: e0167950.

48.

Bassindale

Golus

Horder

, et al. The feasibility and user experience of a program of progressive cued activity to promote functional upper limb activity in the inpatient rehabilitation setting with follow-up at home. Appl Sci 2025; 15: 3010.

49.

Pogrebnoy

Neale

Tillyard

, et al. Exploration of patient preferences in relation to paper or video-based resources to support independent activity on an inpatient rehabilitation ward after stroke. Disabil Rehabil 2025; 1. DOI: 10.1080/09638288.2025.2546075.

50.

De Cock

Batens

Feiken

, et al. The feasibility, usability and acceptability of a tablet-based aphasia therapy in the acute phase following stroke. J Commun Disord 2021; 89: 106070.

51.

Broderick

O'Shea

Burridge

, et al. Examining usability, acceptability, and adoption of a self-directed, technology-based intervention for upper limb rehabilitation after stroke: cohort study. JMIR Rehabil Assist Technol 2023; 10: e45993.

52.

Bhattad

Pacifico

. Empowering patients: promoting patient education and health literacy. Cureus 2022; 14: e27336. PMID: 36043002; PMCID: PMC9411825.

53.

Mitchell

Shirota

Clanchy

. Factors that influence the adoption of rehabilitation technologies: a multi-disciplinary qualitative exploration. J NeuroEng Rehabil 2023; 20: 80.

54.

Pearce

LMN

Pryor

Redhead

, et al. Advanced technology in a real-world rehabilitation setting: longitudinal observational study on clinician adoption and implementation. J Med Internet Res 2024; 26: e60374.

55.

Roberts

Chaboyer

Gonzalez

, et al. Using technology to engage hospitalised patients in their care: a realist review. BMC Health Serv Res 2017; 17: 88. Erratum in: BMC Health Serv Res. 2017 Sep 13;17(1):650. doi: 10.1186/s12913-017-2577-5. PMID: 28587640; PMCID: PMC5461760.

56.

Coleman

Moudgal

Lang

, et al. Early rehabilitation after stroke: a narrative review. Curr Atheroscler Rep 2017; 19: 59.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.02 MB

0.03 MB

0.02 MB