Journal of Cardiac and Pulmonary Rehabilitation
Open Access

Coronary heart disease (CHD) is a chronic condition which often has serious physical, psychological and social consequences for the patients [1]. Coronary heart disease causes disability and it is estimated that the disability-adjusted life years (DALYs) estimate will rise from a loss of 85 million in 1990 to a loss of approximately 150 million DALYs in 2020. Coronary heart disease is the leading somatic cause of lost productivity; therefore, it is not only disabling for the patients but also a great burden to society [2]. Poor risk factor control and high prevalence of unhealthy lifestyle exist, therefore the identification of optimal patient management and CR remains a public health priority [2].

It is estimated that 42% and 38% of all deaths in European women and men, respectively that occur below 75 years of age are due to CHD [2]. The total number of patients living with CHD increases due to an ageing population in the western part of the world and a reduced case fatality of acute coronary events. The earliest and most prominent decline in death caused by CHD was seen in the more affluent countries in Europe between the 1970s and 1990s and is partly attributed to better treatments for acute coronary events, heart failure and other cardiac conditions. This development illustrates the potential for prevention of premature deaths and for prolonging healthy life expectancy [2].

In 2010, the human and economic arguments in favor of CHD prevention were estimated by the National Institute for Health and Clinical Excellence (NICE) as overwhelmingly positive regarding possible benefits and savings [3]. For secondary prevention of CHD, multimodal, behavioral interventions are recommended. A multimodal intervention may be composed by physical training and patient education delivered in weekly sessions. The interventions must include components such as nutrition advice, exercise, weight management, smoking cessation and adherence to prescribed medication. The NICE state that the benefits from such interventions are numerous and include [3].

• Narrowing of the gap in health inequalities

• Cost savings because of the number of CHD events prevented

• Prevention of co-morbidity

• Cost savings associated with CHD such as medication, primary care visits and outpatient attendances

• Cost savings to the wider economy as a result of loss of production, benefit payments and pension costs of early retirement due to CHD

• Improvement of the quality of life (QoL) and length of CHD patients’ lives.

Cardiac rehabilitation (CR) is a generic term for the multimodal and behavioral interventions in the secondary prevention of CHD, which is typically prescribed for 36 supervised exercise sessions over a 12 week period. The Scottish Intercollegiate Guidelines Network (SIGN) develops evidence-based guidelines derived from systematic reviews of the scientific literature. SIGN defines CR as: “…the process by which patients with cardiac disease in partnership with a multidisciplinary team of health professionals are encouraged and supported to achieve and maintain optimal physical and psychosocial health. The involvement of partners other family members, and careers is also important” [4].

The effect of CR has been studied in a Cochrane overview of reviews which included six systematic reviews [5]. The results are based on 148 randomized controlled trials (RCTs) and 98,093 patients. Exercisebased CR was found to be an effective and safe therapy to be used in the secondary prevention of clinically stable CHD patients in relation to reducing readmissions and improving health related QoL

A systematic review, not included in the overview, has also concluded that CR reduces patients' likelihood of depression and anxiety [6]. In addition a reduction in cholesterol and body mass index (BMI) was seen and patients tended to quit smoking and increase their exercise performance regarding intensity and duration [6]. The effect of CR was similarly studied in a large community cohort of CHD patients. The study showed that multidisciplinary CR was associated with a significant survival benefit in the first four years following an acute coronary syndrome (ACS) or cardiac intervention [7].

Patients’ lifestyles are based on long-standing behavioral patterns, which are caused by an interaction between environmental factors and personal and genetic factors [2,8]. Psychosocial risk factors such as stress, social isolation and negative emotions may act as barriers against behavioral change, and therefore should be addressed in CR [2,8].

Surveys show that a large proportion of CHD patients still do not achieve the lifestyle, the risk factor levels, and the therapeutic targets set in CR [2,8]. In the standard CR setting adherence to lifestyle recommendations and treatment regimens start to decline within six months of discharge from the hospital. Adherence to behavioral change concerning diet, exercise and smoking cessation after an ACS is associated with a significantly lower probability of re-infarctions compared with non-adherence [2].

Secondary prevention of CHD is considered a lifelong effort. Cardiac rehabilitation efforts help to maintain long-term adherence to the optimal treatment program by educating the patients and repeatedly emphasizing the importance of achieving and maintaining the prescribed treatments and recommended lifestyle [2,8]. Cardiac rehabilitation interventions may lead to better long-term results when it comes to behavioral change and somatic outcome [2].

Two RCTs have studied the effect of expanded CR versus standard CR [9,10]. Expanded CR consisted of everything contained in the standard CR; besides an interactive learning process, a long duration, and a mixed approach, a better collaboration with general practice, more focus on support from relatives and more intensive CR-lessons; for example with one-to-one support and hands-on practices. Patients randomized to expanded CR had significantly fewer readmissions, reinfarctions and a lower probability of cardiovascular death compared to patients randomized to standard CR [9,10]. A prospective cohort study found effect of expanded CR to socially vulnerable patients regarding significantly better adherence to prescribed medication at one-year follow-up compared to socially vulnerable patients who received standard CR. The same trend was seen in lipid profile, systolic blood pressure and BMI [11].

Based on this, it was relevant to conduct a systematic review concerning expanded CR versus standard CR. Expanded CR is defined as any interventions in addition to standard CR or usual care. It is of importance to examine the role of CR in the 21^st century given the fact that treatment options for CHD have improved rapidly during the last decades, and the fact that it therefore has been questioned whether CR has retained its efficacy [7]. An initial search in the databases PubMed, the Cochrane Library, the Joanna Briggs Institute (JBI) Database of Systematic Reviews and Implementation Reports, CINAHL and PROSPERO indicated that no systematic reviews or protocols on this topic exist.

The objectives, inclusion criteria and methods of analysis for this review were specified in advance and documented in a protocol [12]. The research question of this review was: what is the effectiveness of expanded CR compared to standard CR in adult patients diagnosed with CHD?

Studies that included adult patients (more than 18 years old) diagnosed with CHD (also known as ischemic heart disease or coronary artery disease) were considered for inclusion. The patients were included in a rehabilitation program after surviving an episode of ACS including unstable angina, non-STEMI elevation myocardial infarction (non-ST) or ST-elevation myocardial infarction (STEMI). Studies with patients who underwent percutaneous coronary intervention (PCI) or coronary bypass graft surgery were included. Exclusion criteria were studies: including CHD patients with heart failure, very old adults, focusing on co-morbidity (e.g. diabetes or depression) or on primary prevention of hypertension, hyperlipidaemia, or hypercholesterolemia in CHD patients.

This review considered only RCTs that examined expanded CR compared to standard CR. Standard CR consisted of different guideline-defined medical and behavioral interventions, typically management of medical risk factors and lifestyle risk factors (e.g. diet and smoking). CR could be delivered by anyone, e.g. by specialist nurses or physiotherapists [2,8]. There were no requirements for duration and intensity of the CR. Expanded CR was defined as any multimodal intervention/s in addition to standard CR targeting lifestyle changes and medicine management for the secondary prevention of CHD by other means than traditional CR, e.g. internetbased programs. Expanded CR was typically reinforced, multi-factorial educational and behavioral interventions delivered or coordinated by a specialist [2,8]. All kinds of settings where CR interventions are undertaken were included.

This review considered RCTs including the following primary outcomes: all-cause mortality, cardiac mortality, all cause hospital admission, readmissions due to any cardiac event.

Secondary outcomes were adherence to secondary prevention guidelines´ life-style risk factors measured as:

• Low density lipoprotein (LDL), high density lipoprotein (HDL) and total cholesterol measured as mmol/ml

• Systolic blood pressure measured as mmHg

• BMI, measured as weight (kg)/height (m) 2

• Smoking status, measured as yes/no

• Physical activity measured as hours/week of exercise, daily steps, heart rate recovery, 6-min-walk test, and the Metabolic Equivalent of Task (MET)

• Dietary habits measured as grams of fruit and vegetables per day and daily intake of saturated fat

• Self-management

RCTs included had to measure at least two outcomes. After completion of the intervention all outcomes were measured at followup with no restrictions regarding duration after CR in the included studies.

The databases and trial registries searched included: PubMed, Scopus, and CINAHL for published studies, and clinical.trials.gov, and the WHO trial registry for unpublished studies. A three-step search strategy was utilized. An initial limited search of PubMed, Scopus and CINAHL was undertaken followed by an analysis of the index terms and the text words contained in the title and abstract. A second search using all identified keywords and index terms was then undertaken across all included databases and the trial registries. Third, the reference list of all identified papers was searched for additional studies.

Studies published in English, Danish, Swedish or Norwegian were considered for inclusion. Only papers published after January 1, 2000 were considered for inclusion. This was because secondary prevention underwent several changes during 1990-2000 including new recommendations in medical treatment (i.e. ACE inhibitors and statins) as well as invasive treatments (e.g. primary PCI and implantable cardioverter-defibrillators and more uniform guidelines were published and endorsed) [13,14]. Since 2000, guidelines and recommendations regarding rehabilitation and secondary prevention have been regularly updated in international guidelines and still emphasize the same core components of CR [2,8]. The search included the following keywords for the concepts population and intervention, respectively: “coronary heart disease”; “cardiac rehabilitation”; “rehabilitation”; “prevention programme”; guidelines, recommendations; “patient education”; “patient management”; “secondary prevention”; “guidelines as topic”; “patient education as topic”; “health planning guidelines”; “patient care management” (Appendix I: Search strategy).

Papers selected for retrieval were assessed by two independent reviewers (AM, KH) for methodological validity prior to inclusion in the review using the standardized critical appraisal instrument from the JBI. Any disagreements that arose between the reviewers were resolved through discussion, or with a third reviewer (ML). Papers were included if they met seven out of ten criteria: "truly randomized", "withdrawal described", "comparable control and treatment groups", "identical treatment between groups", "equal measurement of outcomes in groups", "reliable measurement" and "appropriate statistical analysis", respectively. The three criteria regarding blinding were not to be met (Appendix II: Quality assessment of included studies).

Quantitative data was extracted by two independent reviewers (AM, KH) including: specific details about the interventions, populations, study methods, time to follow-up, and outcomes of significance to the review question and specific objectives. All results were subject to double data entry. Effect sizes were planned to be calculated and be pooled in statistical meta-analysis. However, if this was not possible the findings would be presented in narrative form including tables to aid in data presentation where appropriate. In studies with missing data; i.e. standard deviations, the authors e-mailed the corresponding author in order to complete the recording. Eight of the 15 authors responded and these data were added for the analysis.

The database search identified 898 papers and two papers were identified through other resources (Figure 1). At relevance screening by reading the title and abstract by two independent reviewers, a total of 716 RCTs were excluded for not meeting the inclusion criteria due to the following reasons: reviews/meta-analysis, protocols, pilot-studies, economic evaluations, pharmacological interventions, cardiology procedures, guidelines, qualitative studies, studies on physicians’ compliance, no abstract. RCTs with CHD patients focusing on diabetes, hypertension, hyperlipidemia, hypercholesterolemia, depression, heart failure or very old adults were also excluded. Consequently, full text was sought for the 116 RCTs which were assessed for eligibility of which 54 were excluded, with reasons. Consequently 62 RCTs were included and underwent critical appraisal. The majority of exclusions were due to not fulfilling the criteria “truly randomized”.

Figure 1: Flow diagram.

After quality assessment the review included a total of 30 high quality RCTs (total N=12,206) with median follow-up 23 months (1 month-5 years). Details of included papers with characteristics of the CHD patients, settings and types of expanded CR interventions are presented in Table 1. Two papers reported on the same RCT and only reported on cholesterol level after two different follow-up periods [15,16]. Houle et al. reported on cholesterol, BMI, hypertension, and smoking [17], and in a later paper from the same RCT on physical activity only [18]. A mapping of the contents and types of intervention is presented in table 2. RCTs with shorter follow-up time than prescribed in CR [19,20], and sample sizes below 100 participants are indicated by footnotes [18,21,22]. Twenty of the expanded interventions were based on individual case management (e.g. personcentred advice coached by health professionals). Of the individualized interventions three different RCTs were based on mobile technology case-management [23-25], and in four RCTs [15,16,26,27] the intervention was delivered partly based on mobile technology. Six RCTs were home-based of which two expanded CR interventions showed significant increase in physical activity compared to standard CR [18,19]. However, the sample size was small in the study by Houle et al. [18], and the follow-up time was shorter than the usually prescribed 3 months in Peixoto et al.´s study [19].

Table 1: Included studies. AMI: Acute Myocardial Infarction; CABG: Coronary Bypass Graft Surgery; CHD: Coronary Heart Disorder; CR: Cardiac Rehabilitation, GP: General Practitioner; MI: Myocardial Infarct; PCI: Percutaneous Coronary Intervention.

Table 2: Contents and types of interventions. *Significant effect of expanded versus standard cardiac rehabilitation, p<0.05. a, follow-up time<3 months b, sample size<100 participants.

Due to heterogeneity among the health professionals involved in interventions, the settings, and time to follow-up statistical pooling was not possible, and a meta-analysis was not performed. Therefore, the findings are presented in narrative form. Table 3 presents available details of the data (mean values of outcomes, standard deviations) on the following outcomes: mortality, readmission, cholesterol, BMI, hypertension, smoking status, physical activity, dietary habits and measures of self-management, e.g. self-efficacy and risk score.

Table 3: Results of the interventions. HDL: High-Density Lipoprotein; LDL: Low-density Lipoprotein; MET: Metabolic Equivalent of Task; NS: Non-Significant.

Results on mortality (due to all causes and cardiac events) were included in six RCTs, of which only one attained significant results on mortality after the expanded CR intervention [28] and Gullikson et al reported significantly 41% fewer CVD events compared to the control group (hazard ratio 0.59, 95% CI 0.42-0.83), p=0.002 [29].

Regarding readmission (due to all causes and cardiac events) 12 RCTs in total reported this outcome, of which eight RCTs reported significant effect on reduction of readmission after the expanded CR intervention [9,10,26,28-32]; i.e. Gullikson et al reported 38% fewer recurrent AMI events, p=0.02 [29].

At time for follow-up cholesterol level was the most included outcome. Cholesterol (level of total, HDL, and LDL) was reported in 17 RCTs of which nine of the expanded CR interventions reported significant decreases in total cholesterol, HDL and/or LDL-level, respectively [15,16,21,23,25,28,33-35]. However, in two studies the sample size was small; therefore the effect estimation is less robust [21,22].

BMI was reported with favourable effect in four of 10 RCTs of the expanded CR intervention compared to standard CR [9,21,23,34] of which two RCTs included a small sample size [21,22].

Hypertension measured by systolic blood pressure was included in 13 RCTs of which five reported significant differences between the expanded CR and standard CR [21,23,34-36]. Besides Jorstad et al. reported the percentage of participants with blood pressure>140 mm Hg, which in the groups compared were 264% vs. 211%, p<0.001 [36].

In total 14 RCTs reported on physical activity/exercising by a number of different outcome measures (hours/week of exercise, daily steps, heart rate recovery, 6-min-walk test, and the Metabolic Equivalent of Task (MET). In nine of these the expanded CR resulted in significantly better outcomes compared to standard CR [9,18,19,23,24,33,34,36,37]. Five RCTs were exercise-only programs focusing on physical activity promotion [18,19,24,37] and only the study by Johnson et al. resulted in a non-significant difference, between groups [38]. One RCT found that the number of participants defined as inactive (by <600 MET minutes/week) was significantly less in the expanded CR compared to the standard CR group, 37.4% vs . 68.8%, p<0.001 respectively [23]. Another RCT found that participants receiving standard CR were 76% more likely to stop exercising during the year following CR than participants receiving expanded CR, p=0.02 [37].

Among the five RCTs including dietary habits all reported significant intervention effect of the expanded CR compared to standard CR; e.g. decreased intake of saturated fat [33,39], healthier diet [9,34], and recommended daily intake of >200 g vegetables and >2 fruits, was reported among 286 vs . 244, 332 vs . 292 participants, respectively, p<0.01 [36].

A few RCTs reported significant effects on self-management (e.g. stress or self-efficacy in in a risk score) [9,25,40].

Cardiac rehabilitation is the standard of care for CHD patients. This review demonstrated that expanded CR showed positive treatment effect compared to standard CR although narrow beneficial effects. The majority of the expanded CR was individual case management of which seven were based on mobile technology for coaching, feed-back and surveillance. The interventions were delivered in different settings (in patients´ homes, rehabilitation centers, or hospital-based).

This review found among the 30 included RCTs several significant reductions regarding clinical primary outcomes (e.g. death and nonfatal AMI), and favorable effect on several CHD lifestyle risk factors. The effects found were irrespective of patient case mix (i.e. % of AMI patients), characteristics of RCTs (i.e. sample size, location, length of intervention and follow-up), or year of publication. However, compared to an updated Cochrane review including 23 RCTs [41], this review only found effect on physical activity in two RCTs.

Two thirds of the RCTs did not report on readmission; mortality was even less often reported, and only one RCT found a significant difference between the groups. This finding is in accordance with a recent Cochrane review, which found only few RCTs reporting clinical effects and no effect on all-cause mortality and non-fatal AMI from interventions based on internet [42]. A new Cochrane review on education-based CR found some evidence of a reduction in events, but based on only two studies [43]. However, reductions on these clinical outcomes are important to evaluate in future studies on effect of CR; considering their significant impact on the patients´ QoL and on health care cost.

The effect on adherence to recommendations in secondary prevention guidelines could be explained by some of the multiple risk factors, CR usually target. For behavior changes regarding lifestyle risk factors, it was the intervention groups´ actual level at the time for follow-up which was compared to standard CR. However, management of adherence to healthy lifestyle recommended in secondary prevention such as dietary habits and exercise attendance was not reported except in Moore et al.´s study which focused on participants´ self-efficacy enhancement, problem-solving skills, and relapse prevention strategies addressed exercise maintenance, and found favourable intervention effect [37].

A majority of the expanded CR interventions were focusing on multiple life style risk factors, as seen in an overview of reviews [5]. However, cholesterol was the most common reported outcome, and most RCTs reported only few outcomes. Our findings on cholesterol, BMI, hypertension and smoking status are similar to those in a recent Cochrane review, which found some evidence for internet-based interventions on lifestyle behaviors; however no firm conclusions could be made [42]. Like in the present review the outcomes measured differed, there was incomplete reporting, studies reported at different follow-up times, therefore the interpretations of findings are very complex [42].

Although exercise therapy has been identified as a central element in guidelines for CR comprehensive programs, and are recommended by both the American and the European societies of cardiology [44], only some of the RCTs measured physical activity outcomes. Thus, the expanded CR often included exercise therapy but reported other outcome measures (readmissions, lifestyle habits). A recent metaanalysis on effect from exercise-based CR showed reductions from cardiovascular mortality, readmissions and improvements in health related QoL across patients and interventions types [44]. Only one RCT measured heart rate recovery, although it is more highly associated with overall risk than are physical activity and exercise therapy. Cardiorespiratory fitness is directly influenced by the maintenance of regular training, and a faster heart rate recovery after training has been shown to be associated with a decrease in cardiac events and mortality [20]. It seems that a home-based intervention is sufficiently effective regarding physical activity [18,19]. However, the adherence to daily activity was not reported on.

Our findings indicated that expanded CR may provide treatment benefit compared to standard CR. However, to determine whether the effect was sustained after the interventions´ follow-up, measures of adherence to lifestyle behavior is needed. A recent intervention with mobile technology-delivered CR program showed a lifestyle behavior change after 3 months compared to standard CR [45]. The adherence to healthy lifestyle behaviors was measured by a composite behavior score.

This review showed that an individual case management intervention based on mobile technology was seen to have significant impact on several lifestyle factors [15,16,23-25]. In one RCT text messages were tailored to the individual needs of the patient according to baseline lifestyle characteristics [23].

A new review including seven interventions delivered in part by mobile phone text messaging found that simple text appeared to most effective. Three of the interventions improved adherence to medication, and two studies significantly increased physical activity behavior [46]. In our review one of the expanded CR interventions present significant favourable results on a composite score based on physical activity, perceived self-efficacy and rate of return to work [40].

Our review included 30 RCTs fulfilling a strict quality assessment, thus the findings may be evident and not violated by bias. However, in some of the RCTs the sample size was small. Due to more variation than expected a meta-analysis could not be performed, which is a limitation.

The comparison groups in all RCTs were to be described as standard or usual care following guidelines. Thus, we excluded a study which compared the intervention with no intervention [47]. However, it was a limitation that the content of the usual care/standard CR were not thoroughly described, and that the studies were heterogeneous in terms of follow-up time.

Effect sizes were relatively modest, and due to variation of expanded CR components performing meta-analysis was impossible. In general, the lack of blinding resulted in a risk of performance bias. However, blinding of participants in CR or non-medical interventions in general is not possible.

Outcomes such as self-management (e.g. self-efficacy enhancement, problem-solving skills, and a composite score of self-efficacy, physical activity, and return to work were seldom reported. Stress management is not routinely included as a component of standard CR, despite considerable evidence that high stress is associated with health outcomes. In the present review one study reported effect on stress management after 3 years CR intervention [9], and a recent study showed that this component may benefit CHD participants in regard to heart disease biomarkers [48].

Although health-related QoL is another relevant factor in all patients´ lives, there is no universally accepted single measure used. Therefore the concept was not included. However, standard CR has been shown to improve QoL in CHD patients [48], and an updated Cochrane review on education-based CR also found evidence of improved QoL [43].

The review did not include risk scores, which is a limitation. Although these categorizations are arbitrary, they may be indicator of the likely benefits of prevention of future cardiac events and useful for both the individual patient and for the clinician in helping decide at which risk the patient belong and whether lifestyle modifications are needed. Only a few RCTs included risk scores such as the Framingham risk score [25].

Most often the RCTs did not include a detailed description of the content of the interventions in standard CR. Future RCTs ought to improve their quality of reporting, particularly in terms of risk of bias [8], details of the intervention and also content of standard CR. Furthermore, expanded CR should be tailored to sustainable lifestyle behavior changes measured by adherence to healthy lifestyle, and include outcomes of functioning not solely mortality and readmission [49,50].

The population included may be regarded as representative of CHD patients, and the results therefore used for minor groups of patients. Future RCTs may recruit patients who are less representative of the broader CHD population, both including those with stable angina and those at higher risk, e.g. with major comorbidities and disabilities.

Furthermore, the RCTs included relatively small study groups as well as varied outcome measures and follow-up times. It is a limitation that the findings of the RCTs are reported at different follow-up times. RCTs were generally of short duration (beyond 12 months) and the duration of interventions varied between 2 months and up to 5 years.

For future practice it is recommended that CR interventions are delivered by individualized mobile technology and tailored to the individual patient´s needs. Before this, there is a need for development of a standardized risk score instruments to help the clinician in decision of treatment.

In summary, this review found effect primarily on lifestyle risk factors of expanded CR compared to standard CR. Only few RCTs found effect on mortality and readmission outcomes. Thus, there was some evidence to suggest that expanded CR (e.g. person-centred mobile based interventions) have potential for use in CHD population and may be more beneficial than only standard CR following guidelines.

However, expanded CR should be tailored to sustainable lifestyle behavior changes measured by adherence to healthy lifestyle and outcomes of functioning. There is need for large scale and longer term studies of lifestyle behavior changes and complete reporting of management and adherence to lifestyle behaviour.

The review confirmed that expanded CR showed positive treatment effect compared to standard CR on lifestyle risk factors, whereas only few RCTs reported the mortality and readmission outcomes.

There are no conflicts of interest to declare.

We thank the librarian from the Aarhus University Library, Hanne Caspersen for assistance in the search of literature.