Key features in telehealth-delivered cardiac rehabilitation required to optimize cardiovascular health in coronary heart disease: a systematic review and realist synthesis

Abstract Telehealth-delivered cardiac rehabilitation (CR) programmes can potentially increase participation rates while delivering equivalent outcomes to facility-based programmes. However, key components of these interventions that reduce cardiovascular risk factors are not yet distinguished. This study aims to identify features of telehealth-delivered CR that improve secondary prevention outcomes, exercise capacity, participation, and participant satisfaction and develop recommendations for future telehealth-delivered CR. The protocol for our review was registered with the Prospective Register of Systematic Reviews (#CRD42021236471). We systematically searched four databases (PubMed, Scopus, EMBASE, and Cochrane Database) for randomized controlled trials comparing telehealth-delivered CR programmes to facility-based interventions or usual care. Two independent reviewers screened the abstracts and then full texts. Using a qualitative review methodology (realist synthesis), included articles were evaluated to determine contextual factors and potential mechanisms that impacted cardiovascular risk factors, exercise capacity, participation in the intervention, and increased satisfaction. We included 37 reports describing 26 randomized controlled trials published from 2010 to 2022. Studies were primarily conducted in Europe and Australia/Asia. Identified contextual factors and mechanisms were synthesized into four theories required to enhance participant outcomes and participation. These theories are as follows: (i) early and regular engagement; (ii) personalized interventions and shared goals; (iii) usable, accessible, and supported interventions; and (iv) exercise that is measured and monitored. Providing a personalized approach with frequent opportunities for bi-directional interaction was a critical feature for success across telehealth-delivered CR trials. Real-world effectiveness studies are now needed to complement our findings.


Lay summary
With the rising uptake of telehealth services, evidence describing key features and mechanisms of telehealth-delivered cardiac rehabilitation (CR) influencing secondary prevention outcomes was lacking.Research was necessary to improve participation and satisfaction.Telehealth-delivered CR interventions that included personalized care with frequent opportunities for bi-directional interaction, monitoring and measuring exercise, and that engage early with participants are more likely to report improved participation in telehealth-delivered CR programmes.Further research is required to provide more qualitative data about the perception and beliefs of telehealth-delivered CR interventions across the different

Introduction
2][3] Secondary prevention programmes such as cardiac rehabilitation (CR) are a level 1A intervention for CHD management in multiple guidelines internationally. 4,5This multicomponent intervention comprises exercise training and lifestyle change education, including physical activity, nutrition, medication compliance, and stress management. 6,7Cardiac rehabilitation has been proven to reduce future cardiovascular events and disability, [8][9][10] promote healthy behaviours, and enhance health-related quality of life. 6Improved clinical outcomes underpin reductions in intervention-related costs to the health system. 2,6,8,11][14] Barriers, such as transportation issues, lack of facilities nearby, and work and/or carer commitments, have negatively impacted participation in facility-based CR programmes. 15Moreover, public health response to the COVID-19 pandemic forced the temporary closure of many facilitybased CR programmes, which exacerbated participation barriers with an ultimate reduction in service delivery. 16,17Consequently, CR providers have been strongly motivated to test out alternate CR delivery modes to enable participants to continue to access care. 18elehealth-delivered CR is a sound strategy to overcome access barriers and ultimately increase patient participation. 7,16,19This intervention uses telecommunication technologies to deliver CR services at home or close-to-home settings. 7,12,16,19,20Telehealth-delivered CR uses various technologies, including mobile apps, telephone calls, text messaging, and videoconferencing, to provide rehabilitation care. 16elehealth-delivered CR has also been proven to be safe when based at a patient's home 21 and as cost effective as hospital-based programmes. 22Multiple meta-analyses of randomized clinical trials have demonstrated that telehealth-delivered CR can result in equal or better cardiovascular outcomes when compared with facility-based programmes. 7,12,20,23,24Additionally, providing a telehealth-delivered CR option can also result in higher satisfaction 12 and increases participation rates. 2,7hile telehealth-delivered CR can provide equivalent effects on secondary prevention outcomes as compared with facility-based interventions, certain enablers are needed.For designers and implementers of telehealth-delivered CR programmes, it is currently not clear what the key 'ingredients of success' are to ensuring optimal outcomes for their telehealth-delivered option.Consequently, further analysis is required to elucidate the key features and mechanisms of successful telehealthdelivered CR programmes.Author E.E.T. had previously conducted a realist review on remote patient monitoring (RPM) interventions, 25 which included a large heart failure population (15 studies).As such to ensure a novel contribution to the literature, this review excluded heart failure studies and focused on telehealth use within the predominant CR population-CHD.
Therefore, the aim of this study was to use a realist approach 26-28 to (i) identify features of telehealth-delivered CR that improved secondary prevention outcomes, exercise capacity, participation, and satisfaction in clinical trials within the CHD population and (ii) develop recommendations for future telehealth use within this population.

Protocol and registration
A systematic literature search was conducted in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines. 29A protocol was published on the PROSPERO International Prospective Register of Systematic Reviews (#CRD42021236471).

Realist synthesis approach
We adopted a realist approach, 26 which uses published evidence reports to answer questions regarding how and why complex health interventions work (or not) in a particular context or setting. 26A realist approach [26][27][28] provides a better understanding of the relationships between context and outcomes surrounding telehealth-delivered CR implementation in clinical practice, 30 especially when comparing varied contexts, though the iterative process required can be quite time consuming and subjective.The goal of a realist synthesis is not to aggregate trial measures into a summary figure or a final judgement but, in this case, more about understanding features of CR programmes that lead them to succeed or fail.The meta-analysis by Ramachandran et al. 31 establishes the effectiveness of telehealth-delivered CR; however, further examination was required regarding how and in what contexts it leads to these successful outcomes.We explored the mechanisms by which telehealth-delivered CR impacts secondary prevention outcomes [including body mass index (BMI), lipid profile, smoking cessation, and physical activity], exercise capacity, and participation rates using a context-mechanism-outcome (CMO) configuration.Our findings were reported following the Realist And Meta-narrative Evidence Syntheses: Evolving Standards (RAMESES). 32

Search strategy
This systematic review searched for randomized controlled trials (RCTs) to provide a clear understanding of the features impacting secondary prevention outcomes among clinical interventions, which leveraged the level of evidence of this study design methodology.Due to the recent evolution of telehealth technologies, we included studies published within the last decade (January 2010 to September 2022).In February 2021, we initially searched four electronic databases: PubMed (MEDLINE), Embase (OvidSP), Scopus, and CINAHL (EBSCOhost), as well as RCTs published in the ClinicalTrial.govdatabase.To ensure the currency of information, the search was updated in September 2022.The search strategy combined terms related to telehealth, CR, telerehabilitation, and CHD.The complete PubMed search strategy is available in Supplementary material online, Table S1.

Inclusion criteria
We included peer-reviewed RCTs primarily focused on telehealthdelivered CR delivered to patients with CHD (e.g.post-myocardial infarction, post-angioplasty, and stable angina) and compared this intervention to facility-based CR or usual care.When studies targeted multiple cardiac conditions, the CHD arm was considered for inclusion.Based on the definitions described by Rawstorn et al. 24 and Hwang et al., 33 studies were included if a telehealth-delivered CR programme included at least 50% patient-provider contact using communication technologies.This information was obtained from the methods section of each selected paper.
We excluded non-randomized trials, studies primarily targeting congenital heart diseases, heart failure, arrhythmia, or valvulopathy as conditions, and study protocols or conference proceedings.Studies that did not provide a telehealth-dominant arm were excluded.

Study selection
Two reviewers (V.M.G.-R.and E.E.T.) independently screened titles and abstracts using the systematic review management tool Covidence (https://www.covidence.org/) and were blinded to each other's selections.Disagreements were discussed between these reviewers (V.M.G.-R.and E.E.T.) until a consensus was reached.The full texts of potentially relevant papers were retrieved and assessed for inclusion or exclusion by one reviewer (V.M.G.-R.), with any concerns discussed with the second reviewer (E.E.T.).

Data extraction
Data extracted from eligible articles included (i) author; (ii) country of origin; (iii) demographic characteristics of participants; (iv) details of the intervention such as exercise prescription (including duration and frequency) and the inclusion of feedback or tailoring to participant needs; (v) clinical outcomes (e.g.exercise tolerance or functional exercise capacity); and (vi) intervention outcomes (participation rates among participants enrolled, acceptability, and intervention uptake), and, based on the realist review methodology, 26 we collated authors' interpretation on reasons or mechanisms for the reported conclusions.

Quality assessment
The quality of the included studies was assessed using version two of the Cochrane risk-of-bias tool (RoB 2). 34One researcher (V.M.G.-R.) completed the quality appraisal independently, which included the assessment of sequence generation, allocation concealment, blinding of outcome assessors, blinding of participants, incomplete data, selective outcome reporting, and other sources of bias.None of our articles was excluded based on the quality assessment score.

Evidence synthesis
Extracted data were recorded using a Microsoft Excel spreadsheet and uploaded to NVivo12® for thematic analysis.Initial coding was performed to identify the authors' perceptions of factors that influenced the study outcomes.Identified factors were mapped onto four outcomes: (i) secondary prevention outcomes (including BMI, blood pressure, lipid profile, smoking cessation, mood, and physical activity), (ii) exercise capacity, (iii) participation, and (iv) satisfaction.These configurations generated explanations about the contexts (C) and interactions with underlying mechanisms (M) influencing the telehealth-delivered CR outcomes (O). 26First, CMO configurations facilitated the development of theories proposing mechanisms about how specific patterns of contexts and outcomes occurred at the intervention level.Then, configurations were refined to explain how mechanisms work with the identified pattern of context and outcomes.Data synthesis was undertaken by one reviewer (V.M.G.-R.), and it was shared and discussed with a second reviewer (E.E.T.) to ensure the appropriateness of the proposed theories.Synthesis of findings was also discussed with all the co-authors to verify the consistency of findings.

Study selection
We identified 5326 records after the primary search.After duplicate removal, screening occurred on 3960 records.Among 217 full-text articles retrieved and assessed for eligibility, 37 reports describing 26 unique studies were included (Figure 1).

Study characteristics
Included studies were primarily conducted in Europe (n = 12, 46%), followed by Australia/Asia (e.g.Australia, China, and Korea; n = 11, 42%), and North America (n = 3, 12%).Sample sizes ranged from 28 35 to 710, 36 with at least 10 people in the intervention arm. 37The majority of studies reported >70% male participants (n = 16, 62%).The mean age of participants varied from 54 38 to 72 years old.Seventeen studies (65%) were conducted in metropolitan areas.Telehealth-delivered CR programmes were mainly compared with usual care (i.e.facility-based care and secondary prevention outpatient programme; n = 16, 62%), followed by only non-telehealth CR programmes (i.e.regular follow-up medical appointments; n = 9, 35%), and compared with usual care combined with non-telehealth CR 39 (n = 1, 3%).Additional characteristics of included studies are presented in Supplementary material online, Table S2.

Study quality
All 26 studies were at risk of bias due to a lack of blinding of participants or treatment delivery personnel.However, blinding participants or personnel was not possible due to the nature of the interventions.On the other hand, more than half (n = 17, 65%) reported having included blinding for outcome assessors.In one-fifth of the studies (n = 5, 19%), incomplete outcome data were a source of bias, primarily due to attrition bias among study participants.Studies consistently reported secondary prevention outcomes, exercise capacity, participation, and satisfaction rates.Risk of bias assessment is provided in Supplementary material online, Figure S1.

Modality of delivery
The most frequently used telehealth technologies were purposively designed applications (either mobile or web apps) that remotely collected and transmitted data from the patient to a health care provider 40 (n = 15, 58%), web-based applications for education support (n = 6, 23%), and either text messaging or phone call alone (n = 5, 19%).

Intervention targets
Half of the included studies (n = 13, 50%) only targeted the exercise component of CR.Five studies (19%) targeted exercise and education components, while eight (31%) provided exercise, education, and psychological support.

Intervention provider
Seven studies provided nurse-led intervention (n = 7, 27%).Three studies (n = 3, 12%) described the role of the physiotherapist as the main care provider.One study (n = 1, 8%) 41 reported the exercise physiologist leading the research team and providing care.Multidisciplinary teams, including cardiologists, general practitioners, nurses, physiotherapist, and psychologists, were reported across four studies (n = 4, 15%).

Intervention dose
Intervention duration varied from 6 weeks (n = 3, 12%) up to 24 weeks (n = 10, 39%).Thirteen studies (n = 13, 50%) provided at least one exercise session per week up to three sessions weekly with at least one counselling or follow-up session per week.

Tailoring
Tailored feedback, defined as the information and guidance delivered by clinicians to participants promoting behavioural change and sharing their assessment of patient performance in the CR programme, 4,42 was provided in more than half the included papers (n = 18, 69%).Supplementary material online, Table S2 provides a summary of intervention characteristics.

Modifications
No studies reported on deviations from the intended intervention.

Secondary prevention outcomes
Most studies reported secondary prevention outcomes that were either non-inferior to the control group or favoured the telehealth group.Seven studies (27%) reported improved secondary prevention outcomes favouring the telehealth group compared with facility-based CR.Twelve studies (46%) reported positive secondary prevention outcomes favouring the telehealth-delivered CR group compared with usual care.Six studies (23%) reported no changes in secondary prevention outcomes, and no studies reported decreased exercise capacity in the telehealth-delivered CR group.

Participation and satisfaction
This study defined participation as the reported proportion of people in a clinical trial setting who attended at least 50% of either the education or exercise sessions while the study was conducted.Ten studies (38%) reported higher participation in the telehealth-delivered CR group than the facility-based or usual care group.Eight studies (n = 8, 31%) reported unmodified participation rates in the telehealthdelivered CR group compared with either facility-based interventions or usual care.Similarly, eight studies (n = 8, 31%) did not report modification in their participation rates when comparing the telehealthdelivered CR group vs. the control.
Regarding satisfaction rates (defined as the participant's self-reported feeling of contentment with the process of care and its outcomes), six (n = 6, 23%) studies reported higher satisfaction among participants in the telehealth-delivered CR group as compared with facility-based or usual care.In contrast, two studies 56,60 (n = 8%) reported lower satisfaction rates across their participants in the intervention group compared with those undergoing facility-based programmes.

Factors that impacted telehealth-delivered cardiac rehabilitation outcomes
Our analysis identified 29 factors that either positively (17 factors) or negatively (12 factors) affected telehealth-delivered CR outcomes in clinical trials.Factors were details of how the intervention or programme was implemented (e.g. if smartphones were provided to participants) or characteristics of the programme itself (e.g. if there was weekly contact with participants) that appeared to impact CR outcomes compared with if those factors were not present or were different.Factors were identified and then mapped onto four outcomes: (i) secondary prevention outcomes (including BMI, lipid profile, smoking cessation, and physical activity), (ii) exercise capacity, (iii) participation, and (iv) satisfaction.Due to the large number of factors, they were organized into three groups according to what aspect of the programme they related to, in order to make it easier to plan how they may be applied in the future: (i) technology related, (ii) components of care, and (iii) clinician-participant relationship (see Figure 2; Supplementary material online, Table S3).
Our analysis identified 12 factors that appear to negatively influence telehealth-delivered CR outcomes, including participation and satisfaction (see Supplementary material online, Table S4).Perceived lack of usefulness of the intervention appeared as the leading cause for low participation and satisfaction rates.Interventions lacking bi-directional interaction between participants and clinicians (passive interventions) reported low participation rates since interaction was crucial for building trust and acceptability.Additional factors limiting participation and satisfaction were lack of technological support and poor usability of the technological device. 43,56

Realist theories to improve secondary outcomes, participation, and satisfaction in telehealth-delivered cardiac rehabilitation trials
Telehealth-delivered CR outcomes and contextual factors interacted through several underlying mechanisms.To explain these interactions, we developed a CMO configuration (Figure 3) described in four theories explaining the mechanisms of action for the success of telehealth-delivered CR interventions: Theory 1-engaging early and regularly; Theory 2personalizing interventions with an opportunity to develop bi-directional interaction and shared goals; Theory 3-usable and accessible telehealth technologies; and Theory 4-measuring and monitoring exercise.

Theory 1: early and regular engagement
Early engagement is defined as clinician-patient contact during an inpatient stay or within 1 week after hospital discharge to initiate or introduce the option of telehealth-delivered CR. 45 Participation tended to be higher when participants received early engagement, and it was reported to promote higher participation rates and improved measurements of depression and anxiety among participants in telehealth-delivered CR. 45,51 Lee et al. 45 reported that providing a personalized CR programme from early stages was safe and effective.This strategy appeared to reduce anxiety and increase exercise capacity, improving health-related quality of life, participation, and satisfaction rates. 45Similarly, Yudi et al. 51 highlighted the significant contribution  of early engagement to goal achievements in exercise capacity and participation.Their results were influenced by early familiarization with telehealth technologies (e.g.initiating telehealth-delivered CR programme on hospital discharge), which allowed participants to develop confidence in using the intervention. 51elehealth-delivered CR trials providing repetitive contact with participants were more likely to report improved secondary prevention outcomes, including increased exercise capacity and physical activity levels.These trials provided at least one contact session per fortnight 52 to deliver feedback to support lifestyle modification and improve secondary prevention outcomes.Session delivery occurred through different telehealth modalities, including videoconferencing, 61 telephone, and instant messaging 51 platforms.Maintaining repetitive contact between participants and clinicians appeared to increase programmes' perceived usefulness, leading to increased participation and acceptability in telehealth-delivered CR interventions.
The repetitive contact between participants and clinicians was vital for improving participation rates and self-management.Repetitive contact facilitated intervention engagement and self-management skills provision to enhance participants' satisfaction rates, participation, and perceived usefulness.At the same time, repetitive contact was also critical to reducing anxiety and depression among participants in telehealth-delivered CR care.Similarly to early engagement, repetitive contact could also benefit participation and secondary prevention outcomes when provided during CR Phase 1.

Theory 2: personalize interventions and develop shared goals
Personalized telehealth-delivered CR interventions were more likely to report improved secondary prevention outcomes and increased participation and satisfaction by including strategies adjusted to participants' needs and goals.These strategies included providing feedback embedded in behavioural change theories, ensuring the intervention is available in multiple modalities to suit the participant's needs, and providing bi-directional opportunities 47,59 for contact.Consequently, these personalization strategies included in telehealth-delivered CR interventions were reported to improve acceptability and feasibility of goals, resulting in increased motivation and satisfaction rates among patients, which potentially led to increased secondary prevention outcomes.For example, as Sankaran et al. 39 reported, a personalized intervention (including personalized feedback delivered by the clinician) could increase participants' comprehension of their own progress and ultimately achieve physical activity goals effectively.Neubeck et al. also recommended a participant-specific approach since actively engaging patients in decision-making may have increased participation rates.Additionally, bi-directional interaction can allow greater personalization due to the increased opportunity for feedback and support. 49On the other hand, lack of personalization may result in loss of interest in the intervention and contribute to null effects on secondary prevention outcomes.Most of our included studies were more likely to report that telehealth-delivered CR interventions could improve results when a personalized approach is included within service delivery.
Engaging participants in decision-making appeared to increase participation rates.In order to develop shared goals, interventions included behaviour change theories, such as goal setting and motivational interviewing.This feature required repetitive contact between participants and clinicians, where feedback provision was possible.Feedback aimed to increase self-efficacy by promoting competence among participants to support the adoption of lifestyle changes, which could lead to improved exercise capacity. 62Additionally, established bi-directional communication with repeated contact sessions (e.g.telecoaching) supported patients to achieve pre-defined goals in telehealth-delivered CR interventions.As a result, increased self-efficacy contributed to boosting confidence and adherence to the interventions. 62,63As reported by Kraal et al., 64 by providing feedback through telecoaching, participants were empowered to execute an exercise routine at home independently and, at the same time, achieve the outlined goals from the exercise prescription. 64,65eedback timeliness did not appear to influence goal achievement.Asynchronous feedback, 59 such as in chat platforms and texting, worked as a medium to provide emotional support and targeted education that supported lifestyle changes.Real-time feedback provision increased participation and improved secondary outcomes by maintaining participants' confidence in the intervention. 19,41,55Authors primarily used RPM technologies to provide real-time support about exercise execution and reduce exercise-related anxiety. 38

Theory 3: ensure interventions are usable, accessible, and supported
Technological interventions need to be easy to use and adjusted to the participants' digital health literacy.Across the telehealth interventions, the authors highlighted the need to ensure adequate internet access and technological provisions for participants.For example, Guiraud et al. 59 relied on the widespread knowledge of telephone use to provide counselling sessions that included support and feedback to increase physical activity levels.Similarly, Leemrijse et al., 44 who used telephone calls, and Dorje et al., 55 who used a smartphone application, highlighted the impact of telehealth-delivered CR interventions on improving secondary outcomes when telehealth delivery CR interventions are accessible and easy to use by participants.Consequently, accessing these types of interventions yielded additional benefits from the programmes, such as awareness and self-efficacy, vital for goal achievement on secondary prevention outcomes, participation, and participants' satisfaction.
Easy-to-use technologies were required to effectively establish bidirectional communication between participants and clinicians through different platforms such as text messaging, 66 RPM, 38,67 and web platforms. 38,67,68These interventions should include user-friendly features to prevent participants' frustration and delays in feedback delivery.Familiarization and demonstration of the technologies was important to increase user confidence.For some (e.g.Kraal et al. 64 ), this involved supervised training with instructions on how to use devices and upload data.Many studies reported high education levels of their participants 69 and exclusion of participants that did not have access to the internet or computer at home. 64Therefore, supporting digital health literacy may be even more important in the general population.Skobel et al. 56 and Kayser et al. 60 developed systems that resulted in poor usability and connectivity of the technological device and negatively influenced participation outcomes.Moreover, strict safety protocols preventing app use 57 and lack of technical support 43 also impact the ease of use, limiting participation.

Theory 4: measure and monitor exercise
Measuring exercise appears to contribute to maintaining participants' motivation to continue with the intervention, optimize satisfaction rates, and improve secondary prevention outcomes.Across telehealth interventions, authors agreed that measuring exercise achievements and improving physical activity levels (objectively or self-reported) motivated participants to achieve pre-set CR goals.They increased the perceived usefulness of the intervention. 55,59,64,70Remote patient monitoring technologies (i.e.wearable devices and/or smartphones that collected biometric information) were the preferred modality to measure and monitor exercise.Synchronous and asynchronous exercise measurements both benefited exercise outcomes 47,71 and supported the adoption of physical activity habits. 51Similar to our realist Theory 2, feedback embedded in behavioural change theories, such as motivational interviewing and goal setting, was the proposed mechanism. 65,70,71In this sense, measuring and monitoring exercises required accessible and easy-to-use telehealth-delivered CR technologies (Theory 3) to achieve effective care delivery. 41However, a lack of exercise measurement and monitoring could reduce participation. 69

Recommendations for future telehealth use within cardiac rehabilitation
Recommendations for optimizing outcomes for patients participating in telehealth-delivered CR include the following: having low-cost technology or providing devices, offering a variety of platforms and modalities of care, tailoring exercise prescriptions, objectively measuring exercise, and having regular two-way contact between the patient and provider.Patients should also be engaged early with the intervention (e.g.before being discharged from the hospital) and should have a choice on their preferred modality of care.

Discussion
We found personalized approaches in telehealth-delivered CR interventions with repetitive bi-directional interaction between the health provider and participants are key success factors for improving secondary prevention outcomes (i.e.BMI, lipid profile, smoking cessation, and physical activity), exercise capacity, participation, and satisfaction in participants.Repetitive interaction that leverages behavioural change theories promoted the adoption of lifestyles favouring cardiovascular health.Participants need to be engaged early and often and access telehealthdelivered care through accessible, usable, and supported interventions.
To the best of our knowledge, this is the first review to analyse key features of telehealth-delivered CR using a realist synthesis approach.Gass et al. 72 narratively reviewed papers across preventative cardiology populations to investigate the impact of telemedicine design features on adherence.Further, Ramachandran et al. 31 have previously reported equivalent effects on secondary prevention outcomes when telehealthdelivered CR was compared with facility-based interventions and increased effectiveness compared with usual care.
Aligning to Theory 1, Gass et al. 72 also highlighted the need for recurrent personal contact (either in-person or via phone or video) to promote adherence and participation in telehealth-delivered programmes.The broader (non-telehealth specific) literature also supports regular clinician-patient engagement as a key factor in promoting participation, 73 with evidence to show earlier enrollment in CR after a patient is discharged leads to improved attendance. 74This appears to be more successful when delivered by either nurses or allied health professionals. 14Potentially, this early engagement facilitates positive clinicianparticipant relationships and rapport building previously described in the literature to facilitate successful CR outcomes. 75imilar to our results (Theory 2), Ramachandran et al. 31 acknowledged the effect of providing personalized interventions that target selfefficacy and are underpinned by behavioural change theories.Personalization and bi-directional feedback have previously been reported as crucial factors influencing the effectiveness of home-based CR programmes. 31Here, we provide additional evidence to be considered regarding the importance of targeted feedback in regard to exercise monitoring (Theory 4)-this is an area where RPM devices can really enhance the information received by both the clinician and patient.
To increase telehealth-delivered CR implementation effectiveness, Batalik et al. 76 recommended a comprehensive approach including increasing health e-literacy support and social support in the intervention.Our findings support the necessity to provide support for technology adoption and to increase trust and acceptability among participants toward the intervention (Theory 3).The best way to do this requires further investigation.Moreover, the Ramachandran et al. 31 review added a recommendation to consider the place of data privacy on participant willingness to engage with telehealth-delivered CR programmes.Although data privacy is an important consideration, this did not appear among perceived factors affecting intervention outcomes in our review.To examine this aspect, specific assessments of acceptability may be needed.

Strengths and limitations
Our review identified limitations in the design and reporting of the included studies.In most studies, blinding of participants was impossible because of the studies' settings (telehealth-delivered interventions vs. facility-based or usual care).Included study populations lacked diversity with most studies excluding people with disability, sensory impairments, cognitive impairments, and English as a second language.Included population groups were largely from metropolitan areas, under 75 years old, spoke English, and had good digital literacy.Additionally, all studies had a higher proportion of males to females (the majority of studies reported >70% of included participants were male).Research has shown an underrepresentation of women in cardiovascular trials relative to proportion of women with cardiovascular disease.Consequently, our findings may not adequately represent the needs of women.Authors' reports about clinicians' and other care providers' contributions to the intervention were limited.In this sense, there needs to be improved quality of reporting of clinical trials by promoting clinical trial reporting to adhere to current guidelines (e.g.CONSORT).
This study has many strengths.We included a wide range of databases in our search, studies from different parts of the world, and large sample sizes that increase this review's external validity.We also explored multiple technologies with varied familiarity and complexity (e.g.telephone or short-text messages and virtual reality or remote participant monitoring).While interventions included features described in the four realist theories, these seemed effective irrespective of their intervention delivery technologies.

Future directions
As participants' acceptability of programmes had an impact on clinical outcomes, pragmatic RCTs should be considered to enable participant choice in their model of care.Additionally, large-scale, 'realworld' effectiveness studies are needed to provide evidence of how telehealth-delivered CR programmes perform when implemented in complex clinical settings.We note some clinical trials currently underway (e.g. the RESTORE trial being conducted in Poland 77 and the LeIKD trial in Germany 78,79 ) may provide new lessons in this area.Such studies need to incorporate data on patient choice to use or not use the programme and perception and beliefs of telehealthdelivered CR, and this will support learnings in the patient groups most likely to benefit from telehealth options.Studies should also properly report clinician involvement in the programme delivery as this has implications for workforce use and planning.Our review did not find relevant information about how best to engage clinicians in telehealth-delivered CR interventions.We also agree with other recently articulated CR research priorities including that future work must address how telehealth-delivered CR can be adapted to support more diverse populations. 80

Conclusions
Our realist review provides a practical resource that contributes to telehealth-delivered CR service implementation.A personalized approach with opportunities for early and regular bi-directional interaction and inclusion of exercise monitoring appears to be key ingredients in engaging and effective programmes.Additionally, it cannot be understated the importance of ensuring these programmes are easy to use with appropriate technical supports.

Figure 1
Figure 1 Preferred Reporting Items for Systematic Reviews and Meta-Analyses flow diagram representing the number of search records screened, included, or excluded at each stage of study selection.CHD, coronary heart disease; CR, cardiac rehabilitation; RCT, randomized controlled trial; TH, telehealth; UC, usual care.

Figure 2
Figure 2 Identified factors that either positively or negatively impacted telehealth-delivered cardiac rehabilitation outcomes.

Figure 3
Figure 3 Context-mechanism-outcome theories required to optimize telehealth-delivered cardiac rehabilitation outcomes.