The effectiveness of cardiac telerehabilitation in comparison to centre-based cardiac rehabilitation programmes: A literature review

Introduction Cardiac rehabilitation (CR) is an effective, yet under-utilised, form of secondary prevention in cardiac patients. Telemedicine is one method of overcoming barriers to accessing CR. Previous systematic reviews highlight variation in the effectiveness of telerehabilitation programmes and current literature lacks identification of which telemedicine interventions are most effective, despite differences in the results of primary studies. The objectives of this literature review were to: evaluate the effectiveness of cardiac telerehabilitation compared to centre-based programmes for managing cardiac risk factors, satisfaction and adherence in cardiac patients; identify the technologies used to deliver CR; identify the key components of effective interventions. Methods A literature search was conducted using MEDLINE, EMBASE and Scopus. Randomised controlled trials (RCTs) involving an intervention group that received telerehabilitation and a control group that attended a CR centre were included. Results Twelve RCTs met the inclusion criteria. There is evidence to suggest that telerehabilitation programmes have similar effectiveness to centre-based CR. Phones were the most commonly used technology. Most studies used a combination of technologies including personal computers and self-monitoring equipment. Phase III telerehabilitation programmes using self-monitoring, motivational feedback and education were more effective than centre-based CR for increasing physical activity and functional capacity. Conclusion Cardiac telerehabilitation is delivered by a range of technologies and has a similar effectiveness to centre-based programmes. While evidence suggests that additional health benefits are seen in patients who receive a telemedicine intervention in Phase III of CR, further evidence would be required to confidently draw this conclusion.


Introduction
The global prevalence of cardiovascular disease (CVD) almost doubled between 1990 and 2019; 1 patients with these conditions are at increased risk of future cardiac events 2,3 which places a significant demand on health services. 4,5Cardiac rehabilitation (CR) is an evidence-based multiprofessional intervention used for secondary prevention in patients with cardiovascular disease, 6 playing a role in preventing further cardiac events and in the re-enablement of patients. 7[10][11] Patients who undertake CR have reduced hospitalisation and mortality, alongside improved Health Related Quality of Life (HRQoL), compared with those who do not. 12,13dditionally, there are beneficial effects on patients' cardiovascular risk factors, which satisfies the goal of cardiac rehabilitation.The UK-based National Audit of Cardiac Rehabilitation (NACR) reports that CR participation is associated with increased physical activity levels; reductions in anxiety, depression, total cholesterol (TC) and low-density lipoprotein (LDL) cholesterol levels.Furthermore, small beneficial effects are seen in the Body Mass Index (BMI), blood pressure (BP), waist circumference (WC) and alcohol consumption of patients who take part in CR. 14 Most commonly, CR programmes are centre-based whereby patients work in groups with a physiotherapist. 1416 In the UK, only 50% of eligible patients partake in CR and there are concerns that the current method of delivery makes it unrealistic for goals of higher participation to be met.14,17,18 Some of the reasons for non-participation are geographical barriers , [19][20][21][22] work and family commitments, and preferences for home-based exercise.21,23 Using telemedicine in CR, or cardiac telerehabilitation, is one way in which these barriers can be overcome.24,25 Previous research has reported that home-based CR programmes, including those delivered by means other than technology, were as effective as centre-based programmes for improving patients' functional exercise capacity and quality of life. 26,27Regarding the use of telemedicine in CR, guidelines state that it can be considered as an option for some patients.4,9 However, there are inconsistencies within the evidence that was used to create these guidelines.Two systematic reviews published in 2015 reported differing findings in the research with regards to how effective telerehabilitation was in managing cardiac risk factors, particularly pertaining to the effectiveness of telerehabilitation in improving patients' functional exercise capacity, HRQoL, anxiety and depression.28,29 The range of telemedicine interventions used in CR programmes could account for some of the variation in effectiveness. 28,29Understanding which aspects of the interventions are associated with the greatest benefit for patients could help to gain a clearer understanding of how telerehabilitation programmes could be delivered effectively.However, this has not been analysed in the current literature, despite the continued presence of discrepancies between the results of individual primary studies.30 These factors warrant a review of the more recent literature which will be representative of current attitudes to, as well as current technologies that are used within cardiac rehabilitation.Furthermore, the increase in the use of telemedicine in recent years, and especially during the Covid-19 pandemic means that confidence in its effectiveness is imperative.25 This literature review aims to appraise the most recent research that reports on the effectiveness of cardiac telerehabilitation interventions in managing the modifiable cardiac risk factors of eligible patients, as well as their satisfaction and attendance levels, compared with traditional centrebased CR.It aims to identify which technologies are used in these programmes, as well as the common features of the interventions that are most effective.This review will not include video-gaming and virtual reality (VR) as digital health interventions, as their use in CR has recently been explored.31

Search strategy
A literature search was conducted using the databases Medline, EMBASE and Scopus, using the search terms shown in Table 1.Search terms related to outcomes were selected based on the risk factors that were reported in the latest version of the NACR. 14This encompassed: functional exercise capacity (FC), physical activity (PA), lipid profile, BP, BMI and WC.Search terms related to Quality of Life (QoL), satisfaction and adherence were also included as measures of acceptability of interventions to patients.Asterisks were used as truncation symbols to retrieve variations of terms.The Boolean Operators "AND" and "OR" were used to combine search terms, as shown in Figure 1.
Table 2 shows the inclusion and exclusion criteria which guided the literature search.Studies available in the English language and published between 1 st January 2016 and 21 st February 2021 were included.Studies were included if they measured one or more of the outcomes listed in the NACR. 14Studies that measured patient satisfaction and adherence were also included.Only Randomised Controlled Trials (RCTs) with a control group that received a centre-based CR service were included because of their ability to determine cause-effect relationships and lower bias compared to other study types. 32Where multiple papers reported results of the same trial at different follow-up points, the paper with the longest follow-up period was used.Studies were excluded where their intervention consisted of videogaming or VR.
OO completed the literature search independently.Search results from all three databases were exported to   EndNote-X9.Duplicates were removed and the remaining articles were screened initially by title and abstract, followed by full text in order to determine their suitability.Reference lists of included studies were examined in order to find additional relevant studies which had not been identified through the databases.Studies were assessed for quality using the Critical Appraisal Skills Programme (CASP) Checklist for Randomised Controlled Trials 33 and for bias using the Cochrane Risk of Bias 2 (RoB 2) tool. 34he included studies were categorised based on their interventions in order to allow analysis of the common features of effective interventions.The idea to categorise the interventions is from Clark et al. 28 However, the categories identified were reflective of the interventions that the included studies used, rather than the categories identified by Clark et al.

Study characteristics
As shown in Figure 2, twelve RCTs met the inclusion criteria and were included in this review.One paper displayed the results of a psychological sub-study of patients in an RCT, for which the original publication did not meet the inclusion criteria due to the date published. 356][37][38][39][40][41][42][43][44][45][46] All studies other than the CR4HER trial ,36 which was women only, included both men and women in their inclusion criteria.[40][41] Table 3 summarises the characteristics of all included RCTs.

Telerehabilitation and interventions
Seven studies 35,36,39,40,42,43,46 used telemedicine as a replacement of one or more aspects of care.Five studies 37,38,41,44,45 used telemedicine as an adjunct to usual care.Specific details of the interventions used in individual trials are summarised in Table 3. Table 4 shows the technologies that were used to deliver CR in each study.

Physical activity levels (PA)
Five studies measured PA. [39][40][41]43,46 Two reported that there were no differences between groups. 40,41Three studies reported a significant difference in favour of the intervention group.8][39]41,43,45,46 Four studies did not report time or between-group differences for these outcomes.37,41,43,45 Skobel et al. reported that BP increased in the intervention group and decreased in the control group.38 Frederix et al., reported that total cholesterol (TC) increased in both groups during the follow-up period.39 Lunde et al. found that there was an increase in systolic BP and small changes in HDL cholesterol in both groups.46

BMI and waist circumference
Four studies included BMI and waist circumference as outcomes. 37,41,43,45Additionally, Skobel et al. reported patients' BMI only as an outcome. 38our studies found no significant time or between-group differences in BMI and waist circumference. 37,38,41,42idmer et al. found that the intervention group had significantly greater decrease in BMI and waist circumference than the control group. 45

Anxiety and depression
Three studies reported anxiety and depression as outcomes 35,38,40 and Widmer et al. reported depression. 45o significant differences between groups were reported. 35,38,40,45Kraal et al. found that anxiety decreased in both groups. 40Spindler et.al found that anxiety and depression decreased in both groups 35 and Skobel et al. found no significant changes over time. 38Widmer et al. did not report significance of changes over time. 45

Health related quality of life
][39][40][41][42][43][44][45][46] Three studies reported significant differences between groups.Bravo-Escobar et al. reported improvements in the control group (p = 0.007), and non-significant deterioration in the intervention group. 37Frederix et.al and Widmer et.al found that there was a greater improvement in quality of life in the intervention group than the control group (p = 0.005 and 0.003, respectively). 39,45f the eight studies that reported no significant differences between groups, four reported that QoL improved 35,41,42,44 and two reported that it did not. 38,40Lunde et al. reported that QoL improved in the intervention group, but between-group differences were insignificant. 46Widmer et al. did not report significance of changes over time. 43

Adherence
3][44] Two studies reported that the higher value in the range of number of sessions attended was greater than the total number of sessions prescribed in the intervention group. 40,42Lunde et al. measured adherence in the intervention group only, finding that it was 'high', with 71% of patients completing all tasks. 46

Satisfaction
Two studies measured satisfaction as outcomes. 40,44Hwang et al. reported equally high satisfaction in both treatment groups. 44Kraal et al. reported higher satisfaction in the intervention group. 40

Cigarette and alcohol consumption
No included studies measured cigarette and alcohol consumption as outcomes.

Assessment of quality and bias
Table 5 shows the results of individual studies from the CASP RCT checklist Table 6 presents a summary of risk of bias in each study.

Technologies used to deliver cardiac telerehabilitation
This review highlighted that CR can be delivered through a range of technologies; a finding that is consistent with a recent systematic review. 30Phones were the most commonly implemented technology which were used for mobile applications, phone-calls or text messages.Most studies also used remote self-monitoring equipment, 35,[37][38][39]40,[42][43][44][45] which has previously been reported to enhance patients' motivation for behaviour change. 47,48 PC for web-applications, videoconferencing and e-mailing were other examples of technologies that were also used.6][37][38][39][40][41][42][43][44][45][46] In relation to the technologies used, these findings differ from the findings of a systematic review conducted by Huang et al. in 2015.29 This previous review reported that out of nine studies, seven used phone calls as their only form of technology.The remaining two studies were reported to have used a computer application alongside phone calls with remote monitoring.This may be more reflective of the technologies that were used over 10 years ago, as the review by Huang et al.only considered studies that had been published prior to 2011.29 The effectiveness of cardiac telerehabilitation Overall, the results showed that telerehabilitation had a similar overall effectiveness as centre-based CR in managing the outcomes that this review aimed to evaluate.[35][36][37][38][39][40][41][42][43][44][45][46] The results also showed that CR attendance is slightly higher when patients used telerehabilitation services compared to centre-based CR. 36,40,[42][43][44] This suggests that telemedicine can be used to effectively deliver CR services, either to complement centre-based services or, for some patients, replace them.[37][38][39][40][41][42] This is in keeping with the World Health Organization recommendation that telemedicine is used in order to complement existing services, rather than to replace them.25 For BP, lipids, BMI, waist circumference, anxiety, depression and quality of life, the studies included in this review produced varying results with regards to the overall effectiveness of CR in general in improving these outcomes.35,[37][38][39][40][41][42][43][44][45][46] This contradicts previous findings where CR was reported to be an effective form of secondary prevention and positively impacted these outcomes.13,14 A reason for this difference in findings could be due to the small number of studies measuring these outcomes in this review.

Common features of the most effective telerehabilitation programmes
6][37][38][39][40][41][42][43][44][45][46] As such, identifying common components of interventions that were most effective in managing these outcomes was not possible.A larger number of studies reported physical activity, functional capacity, The intervention group had a significantly higher aerobic capacity (MD: 2.139 mL/min/kg, p = 0.032) at follow-up.PA IG had a significantly higher moderate or vigorous physical activity levels (p = 0.01) at follow up.

Quality of life:
6][37][38][39][40][41][42][43][44][45][46] The studies reporting physical activity and functional capacity [35][36][37][38][39][40][41][42][43][44][45][46] were consistent in showing that the most effective interventions were delivered in Phase III of CR and included strategies which target behaviour change.0][41][42][43][44][45][46] This is contrary to the results of a RCT by Snoek et al. 51 that was not included in this review, where the intervention group similarly received a telemedicine intervention in phase III of CR, whereby healthcare providers were also trained in motivational interviewing.Snoek et al. found no significant differences between groups for changes in functional capacity and physical activity levels, highlighting that additional evidence may be necessary to draw this conclusion confidently. 51 In 2015, Clark et al. reported that telerehabilitation programmes that were most effective were those that were individualised. 28As reported in Table 3, individualised programmes were used in several studies across all categories 35,36,39,41,43,46 and were common in Phase III programmes, which could be another reason for their effectiveness.][38][39][40][41][42][43] Guidelines state that all CR programmes should be individualised 4,7 and this may have been a feature of more studies without it being stated.Clark et al. 28 also found that exercisebased telerehabilitation programmes were not an effective alternative to usual care.[42][43][44][45] Patients in these studies received an educational component of CR delivered through means other than technology, as per guidance, which could be one reason for the difference in results. 37,40,42,43

Summary of limitations of reviewed studies
6][37][38][39][40][41][42][43][44][45][46] In studies by Grace et al. and Skobel et al., concerns arose when patients listed the telerehabilitation intervention itself as a reason for leaving the trial, highlighting the importance of accessibility in the technologies used in healthcare. 36,38one of the included studies measured all the outcomes specified by the NACR. 14In particular, smoking and alcohol consumption were not measured as outcomes by any of the studies despite their role in cardiovascular disease [35][36][37][38][39][40][41][42][43][44][45][46]52 and the recommendations in guidance for these behaviours to be addressed in CR programmes. 4 Moreover, since the follow up period of studies was mostly under one year, [36][37][38][42][43][44][45] additional evidence would be required to confidently conclude that these interventions are effective in managing all cardiac risk factors and that the results remain present long-term.

Limitations of the review methodology
This literature review was undertaken as part of a programme of academic study by the main author.The selection of studies, data extraction and appraisal of studies were undertaken independently by the main author (OO).
Although RCTs are suitable for minimising bias and determining a cause-effect relationship, many otherwise-eligible studies may have been excluded from this review due to their study design.For example, one non-randomsied study had similar fidings to that of this review, reporting no significant differences between the telemedicine and control groups in the change in patients' cardiovascular risk factors over time. 53ost and adverse events were not considered as outcomes in this review.However, results from an economic analysis of a RCT show that a phase II telerehabiliation programme was associated with a small reduction in cost for both the patient and healthcare provider compared to a centre based programme, where patients in both groups saw similar improvements in health outcomes (this included BMI, WC, waist circumference depression, anxiety and functional capacity).Researchers also identified the potential for further cost savings if the telerehabiliation programme was to be implemented on a larger scale. 54

Conclusion
The main findings of this literature review are that cardiac telerehabilitation can be delivered through a range of technologies and that it has a similar effectiveness to traditional centre-based programmes in patients for whom the use of technology is appropriate.][37][38][39][40][41][42][43][44][45][46] The limitations highlighted in this review suggest that future primary research in this area should consider measuring the effectiveness of cardiac telerehabilitation on all of the relevant cardiac risk factors that CR seeks to address, especially patients' smoking and alcohol consumption habits.

Statement of originality
The authors confirm that this work has not been previously published in any other format, in any other publication, in whole or in part.

Declaration of conflicting interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Figure 1 .
Figure 1.The use of boolean operators in the literature search.

-
Available in the English language.-Published in the past 5 years (Between 1st January 2016 and 21st February 2021).-Published in a peer reviewed journal -Population is eligible adults participating in cardiac rehabilitation.-The telemedicine intervention fits the definition of 'ICT used to deliver a healthcare service'.-Must have measured the change in one or more traditional modifiable cardiac risk factor, quality of life, satisfaction, or adherence.-RCTs with a control group receiving centre-based CR. -Studies measuring the impact of VR or video games in cardiac rehabilitation.-Studies where telemedicine is not used in the intervention group e.g.home-based phase II cardiac rehabilitation using leaflets and booklets.-Studies where the only technology used is telephone calls with a frequency of less than once per week.-Studies where the comparison group received no form of cardiac rehabilitation.-No cardiac risk factors or adherence or satisfaction are measured.-Study types which are not RCTs including systematic reviews and pilot trials.-Control group also does home-based CR rather than centre based CR.

Table 2 .
Inclusion and exclusion criteria.

Table 3 .
Characteristics of included RCTs.

Table 4 .
Technologies used for cardiac rehabilitation in included studies.

Table 5 .
Results from the CASP RCT checklist for included studies.

Table 6 .
Overall risk of bias in included studies.