HealthTech draft guidance

Cardiac rehabilitation is an established option for secondary prevention of CVD. A meta-analysis comparing cardiac rehabilitation programmes with at least 6 months of follow-up data compared with no exercise reported statistically significant risk reduction in cardiovascular mortality and hospitalisations (Dibben et al. [2023]). Conventional cardiac rehabilitation may consist of face-to-face sessions, or a hybrid programme of in-person group-based and home-based programmes (including paper manuals, live online classes, home visits, or telehealth).

The core components of cardiac rehabilitation according to the British Association for Cardiovascular Prevention and Rehabilitation (BACPR) (PDF only) are health behaviour change and education, lifestyle risk factor management, psychosocial health, medical risk management and long-term strategies. The BACPR recommends that cardiac rehabilitation is offered before discharge from hospital for people who have had an eligible cardiovascular condition.

Not all people with CVD are currently offered cardiac rehabilitation, and the provision of rehabilitation services varies by region (NHS commissioning standards for cardiovascular rehabilitation). It is beyond the remit of this evaluation to determine whether cardiac rehabilitation should be offered to a wider population than in current practice. But the committee noted that increasing the number of options for cardiac rehabilitation could reduce regional variation.

Patient experts explained that cardiovascular conditions are lifelong conditions that affect all aspects of their life. They said that all the core components of cardiac rehabilitation are important (see section 3.3), not just exercise, and that digital tools should not exclude the human and social aspects of cardiac rehabilitation. An expert stated that having access to peer support was valuable to reduce feelings of isolation. They emphasised the importance of personalisation to ensure that digitally supported programmes meet the needs of the individual with CVD. One patient expert suggested that digital platforms could be made more engaging by gamification. The committee members agreed that the features, potential for personalisation and accessibility of a digital platform were important considerations when choosing a suitable programme.

The patient experts highlighted that current cardiac rehabilitation is typically focused on older people with prioritised conditions such as myocardial infarction. This does not meet the needs of all people with CVD, especially those who are younger or were previously fit and exercising frequently. There is limited evidence on the views of people with CVD who feel that the current offer of conventional cardiac rehabilitation does not work for them. The patient experts suggested that offering digital options could help avoid feelings of alienation and exclusion, which reduce uptake and completion in these groups. The committee concluded that further research was needed on patient experiences of the current offer of cardiac rehabilitation, and on reasons for declining offers of cardiac rehabilitation, to better understand where different modes might be suitable.

The EAG prioritised 15 clinical-effectiveness studies for this evaluation, of which 5 were randomised controlled trials (RCTs).Eligible evidence was available for 8 of the 13 technologies:

• 3 studies on Activate Your Heart (2 RCTs)

• 1 study on Datos Health

• 2 studies on D REACH-HF

• 1 study on Digital Heart Manual

• 1 study on Gro Health HeartBuddy

• 2 studies on KiActiv (1 RCT)

• 2 studies on myHeart (1 RCT)

• 3 studies on R Plus Health (1 RCT).
  
  For the other 5 technologies, no relevant evidence on clinical effectiveness was available.

The committee thought that the study populations were generally representative of the populations that access cardiac rehabilitation in the NHS. The studies variously included people with stable angina, myocardial infarction, heart failure, coronary artery disease, and revascularisation, as well as broad CVD populations. The populations typically included older adults (mean or median ages between 50 and 66 years) with a high proportion of white male participants. People who lacked digital literacy, access to the internet or smart devices were commonly excluded across studies. In some studies, people with high depression or anxiety scores, or cognitive impairment, were also excluded. The EAG did not identify any other evidence addressing the scoped subgroups to determine whether the interventions had different effects in groups already underserved such as women, younger people and ethnic minority groups. So further research is needed that includes underserved populations, and that analyses outcomes by subgroup.

The committee concluded that evidence from outside the UK was unlikely to be generalisable to clinical practice in the NHS. Studies providing evidence for Datos Health took place in Israel, and those for R Plus Health were in the USA and China. The EAG noted that both population and healthcare system can affect generalisability. It stated that the population and healthcare systems in Israel and China were quite different to that in the UK, and that the study done in the USA was likely to have a more similar population, but the healthcare setting was very different. A clinical expert noted that the design of cardiac rehabilitation programmes is also very different between countries, and that outcomes collected in these studies such as VO₂ peak were not typically collected in NHS practice. The committee concluded that technologies should be designed to support cardiac rehabilitation programmes similar to those currently used in the NHS, and that evidence based in the UK is important to demonstrate the feasibility and effectiveness of digital platforms in the NHS.

The Digital Heart Manual and D REACH-HF are adapted from paper manuals used to support home-based cardiac rehabilitation. The EAG did not think the evidence for paper manuals was relevant for this assessment, because it had received clinical advice that paper manuals and digital versions do not perform the exact same role in cardiac rehabilitation. The committee agreed that evidence for a paper manual is not necessarily generalisable to a digital tool, because the change in format may affect how people use it. The company representatives mentioned that the content and healthcare professional facilitation for the Digital Heart Manual was identical between the paper and digital versions, but that D REACH-HF had additional functionality over the paper REACH-HF manual. Home-based cardiac rehabilitation is recommended as an option in the NICE guideline on acute coronary syndromes and the NICE guideline on chronic heart failure in adults, which would typically involve using a paper manual. The clinical experts stated that the risk that the digital adaptations would result in significantly worse outcomes than the paper manuals was low, but that this was uncertain. The committee concluded that further data would be needed to determine any difference in effectiveness between the paper formats and digital adaptations of Digital Heart Manual and D REACH-HF. But, they could still be used while the evidence is generated.

Uptake, adherence and completion of digital cardiac rehabilitation programmes were reported in a large number of studies, but the definitions of the outcomes varied. Uptake ranged from 7.51% to 100%, and completion ranged from 82% to 92.4% across the studies. There was limited data comparing uptake and completion between digital and conventional programmes. Because of the lack of subgroup analysis, there is no evidence that digital tools would increase uptake in underserved populations. But 1 study on Activate Your Heart reported that 54% of users would not have attended conventional outpatient cardiac rehabilitation. Users generally reported positive usability experiences across the studies, but the scales used were not validated. The clinical and patient experts stated that it was likely that digital technologies would increase uptake in certain groups who cannot or do not want to access in-person cardiac rehabilitation, such as people with full-time work or caring commitments, younger people or those who feel the in-person offering is not suited to their needs (see sections 3.6 and 3.9). The committee concluded that further evidence is needed on uptake, adherence and completion in subgroups that are likely to benefit from using digital technologies.

The committee noted that the evidence supporting the clinical benefit of digital cardiac rehabilitation was limited, but the available evidence suggested that the digital technologies could improve clinical outcomes. Definitions of clinical outcomes varied across the studies. Overall, the limited evidence available showed improvements in exercise capacity, cardiovascular risk profile and health-related quality of life for people using digital technologies. The maximum length of follow up in the included studies was 6 months. So, the effectiveness of the digital tools beyond 6 months is uncertain. The committee noted that long-term data was needed to evaluate the true effectiveness of these technologies and highlighted the importance of consistency in the outcome measures in future research and evidence generation.

The committee recalled that the cost effectiveness was driven by the cost differences resulting from the reduced number of face-to-face appointments (see section 3.15). For the base case, the cost for each cardiac rehabilitation session was based on NHS reference costs, which resulted in a total cost of £862.17 for 8 sessions including consultant-led initial and final assessments. The model was sensitive to the cost of an in-person session, which was explored in the sensitivity analyses using alternative cost inputs. Where lower costs for cardiac rehabilitation sessions were used, digital tools were dominated by conventional cardiac rehabilitation. The clinical experts questioned the use of consultant-led reference costs for the initial and final assessments and stated that that these appointments were usually held by a cardiac rehabilitation nurse or physiotherapist. They stated that it is unlikely that a consultant would lead this task, and that multiple allied healthcare professionals may be involved. The EAG suggested that the consultant cost could be seen as representing the time of 1 or more cardiac rehabilitation specialists. The EAG also noted that this cost was applied for both digital and conventional cardiac rehabilitation arms and so had little impact on the results. The cost difference between arms was largely because of the avoidance of 6 non-consultant-led sessions when using the digital technologies. The committee concluded that further data is needed to determine the true cost of conventional cardiac rehabilitation.

The committee recalled that the available clinical evidence suggested that digital technologies for supporting cardiac rehabilitation may be clinically effective (see section 3.8), but that it was uncertain whether digital and conventional cardiac rehabilitation could be considered equivalent (see section 3.18). It noted that there was limited evidence on the impact on overall uptake of cardiac rehabilitation of offering digital tools, but thought it was likely that introducing them would increase uptake. The committee concluded that offering digital technologies was likely to improve health by providing an option for people who currently have nothing at all. But there needs to be an initial assessment to determine suitability of the technology for the person.

There is limited UK evidence on some digital technologies to support cardiac rehabilitation. The committee noted that the health systems and structure of cardiac rehabilitation programmes in China, Israel and the US are not comparable to those in the UK (see section 3.10), so UK evidence is needed before Datos Health and R Plus Health can be widely used. It recalled that the Digital Heart Manual and D REACH-HF did not have clinical evidence, but that there was a substantial evidence base for the paper predecessor manuals. The committee concluded that it is unlikely that the change in format would add much clinical risk, especially for people who would otherwise not do cardiac rehabilitation (see section 3.11). So, these technologies can be used while more evidence is generated on the impact of changing the mode of delivery. Beat Better, Luscii vitals, Pumping Marvellous Cardiac Rehab Platform, Get Ready and Sword Move had no relevant clinical evidence, so the committee concluded that they should be used in research only.

Some committee members noted that younger people with CVD may prefer digital tools, because they allow more independence while the rehabilitation programme is still facilitated by healthcare professionals. They stated that people with work or caring commitments are less likely to do in-person cardiac rehabilitation (see section 3.12) and noted that most unpaid carers are women. So, introducing digital tools could reduce inequalities in uptake of and adherence to cardiac rehabilitation by age and sex.

The committee recognised that using digital tools may be challenging for some people such as:

• older people

• people with dexterity issues

• people who do not have regular access to smart devices or the internet

• people who do not have English as a first language

• people experiencing homelessness, living in homes of multiple occupancy, or in residential care.
  
  Additional support and resources may be needed for these groups. The clinical experts also noted that some of these groups are already less likely to do cardiac rehabilitation, and so introducing digital tools could widen existing equality gaps. The EAG's economic model assumed that a tablet computer and monthly internet access would be provided to reduce the risk of digital exclusion. The committee also recalled that the evidence base was limited and there was no analysis on subgroups (see section 3.9). The committee concluded that more data is needed on the usability and acceptability of digital tools in different groups.
  
  The committee recalled that many studies excluded people with high depression or anxiety scores, or cognitive impairment (see section 3.9). The clinical experts stated that these people would need further clinical support outside of the cardiac rehabilitation programme, and that this should be recognised at the initial assessment. The committee concluded that further research is needed to collect evidence of uptake and effectiveness in these subgroups, but that the suitability of the technology should be considered in the initial assessment.

The clinical experts noted that people's ethnic, religious or cultural background may affect how cardiac rehabilitation should be delivered. For example, dietary advice may need to be tailored to the cultural background of the person with CVD. The committee stated that healthcare professionals should discuss the language and cultural content of digital technologies as part of the initial assessment.

The clinical experts highlighted that the study populations did not entirely represent the cardiac rehabilitation population. The studies generally excluded people with low digital literacy, limited access to the internet or smart devices or with significant comorbidities, which are factors that may affect people who currently benefit less from cardiac rehabilitation. These people could be disadvantaged by using digital technologies. The committee agreed that further research is needed on the benefits and risks of using digital cardiac rehabilitation in populations commonly excluded from cardiac rehabilitation.

There was no eligible evidence for some of the technologies. Evidence comparing the digital technologies with the scoped comparator of conventional cardiac rehabilitation was limited. The committee agreed that further research is needed on the offer of both digital and conventional cardiac rehabilitation compared with conventional cardiac rehabilitation alone. There is also a need for research on any difference in effectiveness between the paper and digital formats.

The reported outcomes varied across the studies. Length of follow up was relatively short, with no studies having follow up of longer than 6 months. There was no evidence for outcomes in the subgroups defined in the scope, especially for uptake and completion of the programmes. No evidence on adverse events was reported. The committee noted that long-term data is needed to evaluate the true effectiveness of these technologies. Also, more research is needed to determine if digital technologies improve uptake and completion of cardiac rehabilitation programmes in certain subgroups (see section 3.9). More research is also needed on adverse events.

The cost of conventional cardiac rehabilitation was a key area of uncertainty, which had a substantial effect on the model results. So, more evidence is needed on the cost of delivering conventional cardiac rehabilitation.