Home versus hospital-based cardiac rehabilitation : a systematic review

Comprehensive cardiac rehabilitation has positive effects on many cardiac risk factors (physical activity, smoking status, cholesterol, anxiety and depression) and can lead to improvements in mortality, morbidity and quality of life. Most formal cardiac rehabilitation in the UK is offered within a hospital or centre setting, although this may not always be convenient or accessible for many cardiac patients, especially those in remote areas. The proportion of eligible patients who successfully complete a cardiac rehabilitation program remains low. There are many reasons for this but geographical isolation and transport issues are important. This systematic review examines the current evidence for homeversus hospital-based cardiac rehabilitation. Home-based cardiac rehabilitation offers greater accessibility to cardiac rehabilitation and has the potential to increase uptake. While there have been fewer studies of home-based cardiac rehabilitation, the available data suggest that it has comparable results to hospital-based programs. Many of these studies are small and heterogeneous in terms of interventions but home-based cardiac rehabilitation appears both safe and effective. Available evidence suggests that it results in longer lasting maintenance of physical activity levels compared with hospital-based rehabilitation and is equally effective in improving cardiac risk factors. Furthermore, it has the potential to be a more cost-effective intervention for patients who cannot easily access their local centre or hospital. Currently


Introduction
The benefits of cardiac rehabilitation (CR) in the treatment and prevention of cardiac disease are well established.Recent systematic reviews and meta-analyses [1][2][3] show that patients who participate in cardiac rehabilitation have significant reductions in mortality and morbidity including improvements in exercise tolerance, symptoms, blood lipid profiles, blood pressure and psychosocial wellbeing.
However, despite the evidence and inclusion of cardiac rehabilitation in national rehabilitation guidelines [4][5][6][7] , there is a lack of standardisation with respect to what a cardiac rehabilitation program should include, and participation of eligible patients in cardiac rehabilitation remains poor 8 .
In the UK, the issue of rural health has taken a back seat to ever-increasing centralist policies 9 .Therefore, there is a need to examine modes of service provision within the more remote communities.The delivery of specialist care is particularly challenging, given that many practitioners in rural areas are generalists.Rural populations may themselves introduce barriers to the uptake of specialist care.Reluctance to use services, combined with the importance of maintaining independence, a decline in community spirit and the fear of being a burden, all raise potential barriers to the delivery of high quality specialist care 10 .Conversely, evidence also suggests that rural populations tend to be accepting of the fact that living in rural areas necessitates an 'element of personal responsibility in accessing services' 11 , and that new technology is generally welcomed to improve and provide health care as close as possible to home 11 .Cardiac rehabilitation has traditionally been viewed as a hospital-based intervention.However, if rehabilitation is to be truly patient-focused, then there is a strong case for increased home-based provision, particularly for rural populations.There are several barriers to the uptake of hospital-based cardiac rehabilitation 12,13 , including distance and ease of access 14,15 , and while these are issues for those in urban environments (eg poor public transport, difficulty parking) those barriers may represent an even greater challenge for those living in remote and rural locations 16,17 .
Home-based programs, for example the Heart Manual 18 and the Angina Plan 19 , have been developed to provide a nurseled, community-based, self-help program for patients who may not be able to repeatedly attend a hospital-based program.Both the Heart Manual and Angina Plan have been shown to provide effective strategies for a self-help program for patients unable to attend a hospital-based rehabilitation program [18][19][20] .However, few data are available to assess the efficacy of such interventions in rural patients 16 .With the increasing financial burden of coronary heart disease worldwide, the development of an affordable, acceptable and appropriate method of community-based cardiac rehabilitation is of significant importance.the electronic search: involved human subjects, article published in English, involved adult subjects (>19 years).
The following MeSH terms were used: myocardial infarction/ischaemia, angioplasty, coronary artery bypass graft, heart failure, cardiac rehabilitation, exercise rehabilitation, exercise therapy, psychotherapy, community rehabilitation.Once a full list of articles was obtained, these were checked for duplication and for the inclusion and exclusion criteria (below).Initially article abstracts were reviewed to obtain relevant research articles which involved home-based rehabilitation and the outcome measures required (below).In total, 35 full-text articles were retrieved and reviewed for suitability for this review.Reference lists of appropriate studies were also hand-searched to identify further research studies for potential inclusion.This method identified 5 relevant articles which were included in the review process (Fig1).

Inclusion criteria
The inclusion criteria for the present review were as follows: • Human subjects • Adults (>19 years old) • English language text • Cardiac rehabilitation study in a home or community setting • Patient had been discharged from hospital and article focused on post-discharge care

Definition of home-based cardiac rehabilitation
In many studies the definition of 'home', 'community' and 'hospital' CR was not clear.For the purposes of this review a pragmatic approach was taken; 'home based' was defined as CR which was delivered either in the patients' home or in a local, non-hospital location while 'hospital based' was defined as CR delivered in a hospital or medical centre setting.

Results
In total, 131 articles were identified by electronic search and 17 of these met the inclusion criteria and were deemed suitable for review, with a further 5 articles sourced from a manual search of reference lists (Fig1).There were 8 studies that directly compared home-based with hospital-based cardiac rehabilitation participants, and the remaining studies compared home rehabilitation with a control group (which varied from hospital-based cardiac rehabilitation to 'usual' or 'standard' care), although this was often poorly defined (Table 1).

Mortality
In general, participation in CR is associated with a relative reduction in mortality of approximately 25%.In terms of absolute risk reduction there is great heterogeneity between studies due to different study populations with mortality rates varying from 3.8% in the Ontario Exercise Heart Study 21 , to 26.4% in a Helsinki-based study 22 .Nevertheless, there appears to be little difference between hospital and home-based cardiac rehabilitation in terms of reduced mortality or cardiovascular event rates 23,24 .

Cardiovascular risk factors (cholesterol, high blood pressure, smoking)
Telephone follow up and home-based cardiac rehabilitation can produce a greater reduction in serum cholesterol when compared with usual care 25,26 .Similar reductions in cholesterol levels are also observed following hospital-based rehabilitation 23,27 .A key study by Jolly et al 23 directly compared home-with hospital-based programs.The results showed a reduction in blood pressure following both forms of intervention, with no difference in blood pressure reduction at 6 months, suggesting home rehabilitation is as effective at reducing blood pressure as a hospital-based intervention.A reduction in smoking habit can be seen in both home-and hospital-based groups with a similar improvement in both groups when compared with baseline levels.However, there was no improvement observed in patients who received a GP-based form of rehabilitation, suggesting a more focused or intense approach is required 28 .

Prevalence of angina
A significant reduction in frequency of angina was reported by an early community-based study 29 .A more recent study by Jolly et al 23 also reported an improvement in the frequency of angina, along with self reported chest pain on movement and shortness of breath.Improvements in all these factors help to improve quality of life of the cardiac patient.There were no clinically significant differences when comparing patients who completed their rehabilitation at home with those who did so in hospital.

Hospital re-admissions
Hospital re-admissions are often measured as a primary outcome 24,[30][31][32] 1).There were 25% (n=35) re-admitted in the treatment group compared with 41% (n=51) from the usual care group within the first 100 days following discharge, suggesting that home visits reduce subsequent admission to hospital 24 .One Australian community-based study found that patients who attended a rehabilitation program were admitted less frequently and spent less time in hospital than those receiving usual care 30 .Over a 12 month period, less than 1% of patients from the intervention group and 4% from the control group were re-admitted.This impressive reduction in hospital re-admissions was important when assessing the efficacy and cost-effectiveness of home-or community-based rehabilitation.However, it should be noted that, of the 954 patients registered in the program, 621 attended fewer than four sessions so were excluded from the final study analysis.

Anxiety, depression and quality of life
Both home and hospital cardiac rehabilitation reduce anxiety and depression 23,27 with no difference in effect between modalities.However, there also appears to be a natural temporal improvement in anxiety and depression in cardiac patients following an event 33 .Similar improvements in quality of life are observed when comparing home-with hospital-based rehabilitation 23,27,34,35 Marchionni et al 35 also found improvements in both home and hospital groups' quality of life, and noted an improvement in younger patients who had received no formal rehabilitation.This reinforces similar findings in post-CABG patients working with the Heart Manual 36 .A significant improvement in quality of life can be seen with heart failure patients receiving home-based physical activity intervention when compared with an education-only group 37 .These data would appear to support the inclusion of exercise in home rehabilitation programs.

Physical activity
When comparing home rehabilitation with comprehensive hospital-based care there appears to be little or no difference in physical activity outcomes between the two 23,[34][35][36]38,39 suggesting that there is no difference between these two approaches to cardiac rehabilitation and that both are effective. Home diac rehabilitation interventions are certainly associated with improvements in physical activity levels, from an improved 6 min walk test 28 , increase in estimated VO 2 25,36 increased daily physical activity index 40 , and improved functional capacity in heart failure patients 37 after completing a rehabilitation program.Patients involved in a focus group after completing the 12 week Heart Manual program found the exercises to be well planned but were worried about exercising on their own, especially in the early days 41 .However, home-based exercise may have longer lasting effects.While Marchionni et al 35 observed an improved total work capacity in both hospital and homebased groups, at 12 months post-discharge, total work capacity had reverted to baseline levels in the hospital group, but not in the home-based patients.These data suggest that home-based rehabilitation exercise may have longer lasting effects than hospital-based rehabilitation in terms of activity levels.It has been suggested by patients that home rehabilitation is seen as 'more of a lifestyle change…rather than treatment' 41 . Patients feel  onus of control themselves during home rehabilitation; whereas, in hospital others are 'in control' 41 .

Cost of rehabilitation
Research comparing home with hospital rehabilitation 23 has determined the average cost per patient to be £198 and £157, respectively.When costs for patient travel and time were included, the cost for hospital rehabilitation rose close to that of the home program (£157-£181).When comparing 6 weeks' provision of the Heart Manual with 8-10 weeks of comprehensive hospital cardiac rehabilitation, Taylor 44 found the home-based approach to be, on average, £30 cheaper per patient.The reasons for this were attributed largely to the reduction in personnel costs for this particular program.Over a 9 month period, there was no significant difference in healthcare costs between the two patient groups.

Participation and concordance
Less than 50% of eligible cardiovascular patients benefit from cardiac rehabilitation in most European countries 45 and participation rates remain low in those who are referred.
Reasons for non-attendance vary from patients being 'not interested', illness, need to work, re-admission to hospital and transport issues 23,46 .When comparing adherence rates between home and hospital-based interventions, the Birmingham Rehabilitation Uptake Maximisation (BRUM) study found that 96.1% of home participants received 5 contacts with a rehabilitation nurse, whereas only 56.1% of centre-based participants attended this number of classes 23 .
For patients living in remote and rural areas, the most prominent barriers are accessibility and distance 15 .

Discussion
There are limited data regarding 'home versus hospital' cardiac rehabilitation but both appear to be effective at improving clinical parameters and fitness.Home rehabilitation may prove more successful in maintenance of physical fitness in cardiac patients 35 .The variation in mortality between studies is likely to reflect the lack of standardisation of entry criteria to cardiac rehabilitation programs even within clinical trials.The risk of further cardiac events and death are clearly related to patient characteristics but also temporal distance from the index event; thus, if there is a delay in recruitment of patients into cardiac rehabilitation, death rates within the program will tend to be lower because a proportion of higher risk patients may have died before commencing cardiac rehabilitation.
Similarly, measurements of the prevalence of angina should be interpreted with caution.While frequency of angina in cardiac patients may be a major factor in their quality of life, angina may increase following successful cardiac rehabilitation because patients are exercising more, or being more socially active.This is particularly true in patients with chronic stable angina, because an increase in the frequency of angina may not actually result in a reduction in quality of life.Frequency of angina should be examined in the context of activity and other quality of life measures.Hospital readmission is an important outcome in terms of costeffectiveness and has arguably greater implications for those living in remote and rural areas.Re-hospitalisation rates following initial recovery from MI range from 5-41% 24,32 .While it is impossible to prevent all hospital readmissions through the use of cardiac rehabilitation programs, the current evidence suggests that communitybased programs, in a variety of forms, are effective in reducing re-admissions to hospital, and patients who are admitted have a shorter stay.The implications for providing community or home-based services for those living in remote and rural areas are therefore promising.
There is a well established relationship between anxiety and depression and patients with coronary heart disease 45 , which may be due in part to the poor risk factor profiles, including diet, smoking and exercise.Previous studies have shown that cardiac rehabilitation improves symptoms of anxiety and depression in patients 47,48 .The prevalence of depression in CHD patients ranges from 16% to 25%, and from 10% to 29% for anxiety disorders 49 .Anxiety and depression can affect heart rhythms and blood pressure, and can elevate insulin and cholesterol levels and increase smoking, with highly anxious patients at 3-6 times greater risk of MI and sudden death 49 .In the two main home-based rehabilitation programs under review, namely the Heart Manual and the Angina Plan, psychological components are central to the successful rehabilitation of patients.
A recent Cochrane Review 50 carried out a full meta-analysis on research of home versus hospital studies.This review found no statistically significant difference between home and hospital rehabilitation for the following outcomes: mortality, cardiac events, exercise capacity, modifiable risk factors, blood pressure, total cholesterol and health related quality of life.This supports findings from the BRUM 23 and Cornwall Heart Attack Rehabilitation Management Study (CHARMS) 27 studies and shows that patients receiving both home-and hospital-based rehabilitation benefit in many ways by completing their rehabilitation.Flint et al 51 have stated that 29% of networks had increased uptake of home options (during the period 2007-2008) within the English Cardiac Networks group.This evidence suggests an improvement in the provision of home rehabilitation as an option.This is promising data because attendance at cardiac rehabilitation remains poor.
At present, home-based cardiac rehabilitation may be offered in some areas as an alternative to hospital-based programs.The lack of standardisation of rehabilitation programs within hospital, community and home settings makes direct comparisons difficult.It is possible where there is a small treatment effect that patients receiving usual care or that have returned to the care of their family physician have attended rehabilitation and exercise classes elsewhere, but in the majority of cases this is not identified.This might explain the variety in outcomes seen when comparing home rehabilitation with usual care and its lack of impact in some studies 26,28,43 .Nevertheless, when home rehabilitation in its many forms has been directly compared with hospital-based programs, there appears to be a consensus that there is no significant difference in outcomes for patients between these two approaches, with both approaches being effective at improving a number of clinical and psychological parameters.However, there is some evidence to suggest that patients receiving home rehabilitation may maintain greater levels of physical activity than those completing hospital-based programs 34,35 .Thus, it appears that home-based rehabilitation can be both safe and effective for those unable to attend a secondary or tertiary care centre.
Therefore those receiving home-based care should be at no disadvantage compared with their hospital counterparts.The review of the data suggests that home programs can and should be offered alongside hospital intervention, instead of as a secondary option.This would support an approach based on patient preference and, indeed, such may help increase the uptake of cardiac rehabilitation which remains dismally poor, especially in those who see distance or lack of time as a major barrier to attendance.There are several studies into the reasons for poor participation rates in CR 12,15,46,50 but there is clearly a need for more investigation.Although the distance a patient has to travel to attend cardiac rehabilitation is a well known barrier to attendance, there are other issues relevant to remote and rural patients which need to be further investigated.A high drop-out rate is not unusual for cardiac rehabilitation programs but may limit the generalisability of these results.The interaction between rurality and participation rates has not been established and there is a need for further research in this area.

Study limitations
One major limitation with these data is the inconsistency between studies in terms of what was being provided as 'home' or 'community based' cardiac rehabilitation.In the late 1980s and early 1990s the position was similar with respect to hospital-based cardiac rehabilitation programs, where the benefits were known but there were no standard guidelines 52

Implications
The implications of the systematic review key points (Fig2), merit special consideration.The evidence suggests that home-based rehabilitation using resources such as the Heart Manual 18 , can be an acceptable and appropriate alternative to the more traditional hospital-based setting.However, it must be pointed out that not all rural communities are homogenous, and each will have separate needs and requirements.The UK National Health Service (NHS) is moving toward encouraging self care 53 in patients, and this review supports the need and potential benefit to offering patients a choice in their mode of rehabilitation.The role of telehealth in rural communities has an ever-increasing presence and can also be used to expand on the home rehabilitation the patient receives, with potential for specialist input via video conference, or one-to-one discussions with an exercise specialist with regard to progress or similar uses.The increasing use of home rehabilitation as an option 51 protects patients from missing out on vital information and education to help aid recovery from a cardiac event.

Need for further research
There is lack of research covering longer term follow up of patients completing home-based rehabilitation.This information would provide vital knowledge on outcomes such as mortality and physical activity maintenance for those receiving this mode of service delivery.More work is needed on the long-term effectiveness and safety of different modes of home service delivery.This includes home rehabilitation and also for more modern approaches such as 'tele-rehab' or interactive internet-based, self-help programs, which could provide a more flexible option for some patients.These approaches are potentially accessible to many and would offer further choice and flexibility previously not available, particularly for remote and rural residents.Research investigating effectiveness and cost-analysis for these modern approaches to rural care would offer patients the chance to make an informed choice on their mode of rehabilitation.Few data are currently available to inform healthcare providers on attendance issues for rural patients.Expansion of knowledge in this area would allow an insight into the service redesign that may be beneficial in rural areas.

Conclusion
Patient groups who are most likely to benefit from the provision of home rehabilitation services vary, and those living in remote and rural locations are likely to be one of those groups.In geographical regions where a considerable proportion of patients live in remote communities, and where the current provision and accessibility of cardiac rehabilitation is inadequate, home-based intervention appears a safe, viable and effective option and offers a convenient means of delivering the information that would be missed by not being able to attend a hospital-based program.Self-help manuals such as the Heart Manual and Angina Plan are not new to cardiac rehabilitation, and minimise the cost to the patient and are accessible to friends and family supporting the patient.The evidence shows that this is an effective method of rehabilitation, yet few NHS providers offer this type of intervention to their patients as a standard adjunct, or as an alternative to hospital-based care.Home rehabilitation is a safe and effective therapy which could, and possibly should, be offered to all eligible cardiac patients.

•Figure 2 :
Figure 2: Home cardiac rehabilitation systematic review key points.

Table 1 : Summary of studies from systematic review (n=22) 17
,23-43Home based via the Heart Manual plus home visits and telephone contact.Commenced on hospital discharge.Heart Manual provides a self help program for 6 weeks post MI, including education material, home-based exercise program and stress and relaxation management, including a relaxation tape.Common themes between modes of CR included loss of confidence, realising importance of exercise and awareness of the benefits of CR.Themes exclusive to hospital CR included ease of access, enjoyment from group sessions, motivation to attend regular sessions and reluctance to exercise outside hospital environment.Home program patients experiences included unanimous praise for the Heart Manual, helpfulness of relaxation tapes, positive views on nurses support provided and thought the exercises well planned but expressed anxiety when beginning to exercise alone.Using METs to establish if activity was light or moderate.Intervention group significantly improved PA index (278.2to525.5 counts/min/day) and the time spent in moderate intensity PA (from 16.8 to 63.7 min/day).No changes were observed in the control group.At 12 months follow up (12 months post completion of program, 18 months after baseline measurements), peak VO 2 sustained in home group but declined in hospital CR patients.Physical HRQoL higher in home group at follow up, with mental HRQoL showing minor deterioration over time in both groups.Home patients had higher habitual PA levels than hospital at 12 months.
1 session per week for 7 weeks.Including 1 hour exercise and 1 hour education per session.Commenced 3-6 weeks after registration to program.19% of control group re-admitted, 6% of intervention group re-admitted (within 2-14 months after initial registration).Patients contacted 1 week post discharge, then every month for 1 year.Patients identified as being at risk of psychological distress following completion of general health questionnaire were visited and followed up by nurse as required.No overall survival impact.Higher cardiac and all cause mortality among women in the intervention group.No evidence of benefit or harm to male participants.The program's impact on anxiety and depression was small ADLs, Activities of daily living; AF, atrial fibrillation; BP, blood pressure; CABG, coronary artery bypass graft; CCS, Canadian Cardiovascular Society scale; CR, cardiac rehabilitation; ECG, electrocardiograph; HAD, hospital anxiety and depression; HDL, high density lipoprotein; HR, heart rate; HRQoL, health-related quality of life; IHD, ischaemic heart disease; LDL, low density lipoprotein; METs, metabolic equivalent of task; MI, myocardial infarction; NYHA, New York Heart Association; PA, physical activity; PCI, percutaneous coronary intervention; QoL, quality of life; TWC, total work capacity, VO2, oxygen consumption.