Clinical Practice Guideline for Cardiac Rehabilitation in Korea

Research framework

The group that developed these CPGs was comprised of writers and consultants. Thirty-three authors, primarily rehabilitation specialists, cardiologists, and thoracic surgeons in 21 university hospitals and two general hospitals, were involved in the CPG management, development, and assessment committees. During the research and development process, opinions were garnered from 12 consultants including rehabilitation, sports, and preventive medicine specialists, CPG experts, nurses, physical therapists, clinical nutritionists, and library and information experts.

This research was based on an extensive literature review of CR, the assessment of CR outcomes, exercise therapies appropriate for CR, and educational components for secondary prevention. The key questions related to each of these aspects were selected during a process that involved all members of the CPG management and development committees, and finalized with reference to the latest foreign CPGs [4-9], in consultation with the CPG consultation committee. Individual guidelines were then developed in relation to circumstances in Korea. Key questions that had not been adequately analyzed in existing guidelines were developed de novo, and finalized in consultation with the CPG consultation committee in consideration of the circumstances in Korea.

To select foreign CPGs for adaptation, the Scottish (2017) [5], British (2013) [6], American (2011) [7], Canadian (2009) [4], and Japanese (2012) [8] CPGs for CR published within the past ten years were evaluated using accepted Appraisal Guidelines for Research and Evaluation II (AGREE II) [10]. Two investigators appraised one set of CPGs, and guidelines assigned an average ‘total score for each domain’ of 80 or greater (out of a score of 161) by two investigators, and a ‘rigor of development’ score of 70% or greater for each domain (out of a score of 56).

As summarized in Table 1, the Scottish (2017) [5] and British (2013) [6] CPGs were selected after being reviewed by the CPG management committees, and the Scottish Intercollegiate Guidelines Network (SIGN) CPGs [5], which had been published more recently, were chosen as the major guidelines for adaptation in this study. References regarding specific key questions and topics were taken from all five foreign CPGs [4-9].

Table 1.

Appraisal of foreign clinical practice guidelines for cardiac rehabilitation with AGREE II

The key questions were used to inform the CPGs were primarily selected during a process that involved all members of the CPG management and development committees. The Scottish CPGs (SIGN 2017) [5] were used as the primary reference, but the English (NICE 2013) [6], American (AHA 2011) [7], Canadian (CACR 2009) [4], and Japanese (JCS 2012) [8] CPGs were also used for reference in relation to key questions not adequately addressed in the SIGN 2017 guidelines. Those developed de novo due to the lack of available references were as follows: “Is a submaximal exercise test, such as a 6-minute walk test useful for cardiac rehabilitation?”, “Is high-intensity interval training more effective than continuous moderateintensity training as an aerobic exercise?”, and “Should policosanol be recommended as a food supplement?” (Table 2).

Table 2.

Key questions for clinical practice guidelines for cardiac rehabilitation

Data sources

To gather evidence to inform the development of the CPGs, PubMed was searched using the following key terms and search strategies: Search (‘cardiac rehabilitation’ [TIAB] OR ‘cardiac rehabilitation’ [MeSH]), Filters (Consensus Development Conference; Guideline; NIH; Practice Guidelines). In addition, PubMed was searched for guidelines using key words such as myocardial infarction (MI) as follows: Search ((rehabilitation or ‘secondary prevention’) and (‘myocardial infarction’ or ‘acute coronary syndrome’)), Filters (Consensus Development Conference; NIH; Guidelines; Practice Guidelines).

Studies published after publication of the latest foreign CPGs were primarily found in the following three databases: Cochrane Library (www.cochranelibrary.com), PubMed (www.ncbi.nlm.nih.gov/pubmed), and EMBASE (www.embase.com).

MeSH terms were used ‘cardiac rehabilitation’ in the Cochrane Library and PubMed, and Emtree terms ‘heart rehabilitation’ were used when searching EMBASE. Based on the search strategies used in SIGN 2017, CR-related search strategies for each study question were developed, and the specific search terms for each were entered using AND or OR operators. Operators and synonyms were taken from the search strategies published in Cochrane Review articles found in the Cochrane Library using the MESH term ‘cardiac rehabilitation’. After conducting a search in the Cochrane Library, the search strategy was saved, and the same search strategy used in the EMBASE. The Cochrane Library and EMBASE use the same search strategies, such as proximity operators, and the search strategy for PubMed was also nearly identical, with the exception of different proximity operators. Therefore, due to space limitations, we present only the search strategy used for the Cochrane Library in this report (refer to Appendix). When making revisions with reference to existing guidelines, we limited the searches to systematic review articles, and the year of publication was set to the year following the publication of the existing guidelines. Korean articles were searched in RISS, KMbase, and KoreaMed using terms ‘cardiac rehabilitation’ and ‘심장재활’ (cardiac rehabilitation). After the duplicate removal, the two researchers selected the first, second, and third document selection for the same documents. In case of disagreement during the article selection process, the consensus was reached after the discussion or the final decision was made by the third researcher’s arbitration.

For assessing the methodological quality of article, AMSTAR 2.0, GRADE system used for systematic review, Cochrane risk-of-bias (RoB) used for randomized controlled trials, and the RoB assessment tool for nonrandomized studies (RoBANS) used for non-randomized controlled trials. Two researchers evaluated the quality of one article. In case of disagreement during the assessing the methodological quality of article, the consensus was reached after the discussion or the final decision was made by the third researcher’s arbitration. Data obtained in the search were accepted as evidence of the studies’ populations (human cases of ACS), designs, and methods (sample size, blinding, data analysis), results (objectivity of results and consistency throughout studies), language (Korean and English), and applicability to Korea. Both adaptation and de novo methods were used. For the former, all available evidence cited in foreign guidelines was reviewed, as the level of evidence and recommendations presented by each country used different evidence appraisal methods, and an evidence chart was developed based on all the subsequent evidence published after the publication of the latest foreign CPGs. After drafting the CPGs, the levels of evidence (LOE) and strength of recommendations (SOR) were determined based on the evidence chart and with reference to the method used in the SIGN 2017 [5].

LOE ranged from 1++ (highest quality) to 4 (lowest quality) [5] (Table 3). The SOR was determined in consideration of the LOE, reliability and consistency of the LOE, impact on patients, applicability in Korea, and balance between harm and benefits, but instead of following the GRADE classification of ABCD, recommendations were classified as either strong or conditional and Good-Practice Points (GPPs) were given according to the method used in the SIGN 2017 (Table 4). The SIGN stopped using the ABCD classification in 2013 due difficulties associated with its practical application, and has been classifying recommendations into strong, conditional, and GPPs since then [5].

Table 3.

Levels of evidence [5]

Table 4.

Recommendations and Good-Practice Points [5]

Process followed to develop cPGs

Expert panels consisting of 11 rehabilitation specialists, cardiologists, and thoracic surgeons achieved formal consensus using the nominal group technique. Authors of the draft guidelines first reviewed the evidence collected with these panels, and then conducted a question and answer session, after which the final recommendations and the LOEs and SORs were determined through a vote. Participants gave their opinions using a 9-point scale (1, 2, 3 ‘inappropriate’; 4, 5, 6 ‘uncertain’; 7, 8, 9 ‘appropriate’). If the ratings given by more than 70% of the panel (8 out of 11) were within a single 3-point band, and there were no objections with the phrasing of the recommendation, a consensus was deemed to have been reached.

The SORs were indicated as strong, conditional, or GPPs, in consideration of the LOE, likely impacts on patients, their applicability in Korea, and the balance that would be achieved between benefits and harm. ‘Strong’ recommendations were used when it was deemed that the benefits of the recommendation would outweigh the harm for a vast majority of people. If recommended, it was phrased as ‘must be performed’, ‘should be performed’, or ‘should be strongly recommended’. If it was not recommended, the possible phrases used were ‘must not be performed’, ‘should not be performed’, and ‘should not be recommended’. The word ‘conditional’ was used when the benefits of the recommendation would outweigh its harms on most people. If recommended, it was phrased as ‘recommend’ or ‘is recommended’, and if it was not recommended, it was phrased as ‘should not be performed’ and ‘is not recommended’. Therefore, for interventions that were conditionally recommended, performance should be determined on the basis of an individual’s values and preferences, and after discussing the intervention with the patient and ensuring the decision is based on a thorough understanding of the intervention’s importance. Finally, GPPs were used when the SORs could not be determined due to a low LOE in the literature but the practice was deemed to be a recommended practice based on the guideline development group’s clinical experience. It was phrased as ‘is considered’ and ‘should be referred to’.

Staff from relevant academic societies and other persons concerned were invited to attend three rounds of public hearings on the draft guidelines (the Korean Academy of Rehabilitation Medicine, the Korean Society of Cardiology, and the Korean Society for Thoracic and Cardiovascular Surgery), following which they were reviewed and modified.

The CPG management committee designated one investigator responsible for each topic. These individuals then finalized the guidelines together with the CPG writing committee. The process of identifying key questions and preparing evidence summaries and meeting transcripts was used to reference the writing process. Internal revisions were made after the first review by the investigator in charge, and the CPG management committee then performed a final review. The CPG Committee at the Korean Academy of Medical Sciences was also asked to appraise the draft guidelines, and those comments will be taken into consideration in a future second revision.

Academic societies and associations for physicians caring for patients in CR will use the 2018 CPGs for CR in their professional development, as will health professionals in their consultations with patients and caregivers, and they will be used to inform various medical institutions’ decisions related to convalescence. Following their distribution, the authors will continue to respond to questions and obtain opinions from health professionals using the CPGs. A second edition will be developed once consensus has been reached on the need for further revisions based on foreign CPGs and new evidence.

Timing of cardiac rehabilitation

The Scottish CPGs (SIGN 2017) state that “all patients must be referred to CR to undergo individualized intervention and treatment plans based on individualized assessment,” [5] regardless of the type of treatment they undergo after the onset of ACS, and notes that their treatment should be comprehensive and individualized at the initiation of the rehabilitation program. There is not enough clinical evidence, however, to document the optimum timing of exercise, a critical component of CR, and early exercise is often determined passively. We therefore reviewed evidence regarding the timing of CR exercise in existing guidelines and additional studies.

US guidelines are that all eligible patients with ACS or patients’ whose status is immediately post CABG surgery or post-PCI should be referred to a comprehensive outpatient CR program either prior to hospital discharge or during the first follow-up office visit. Further, all eligible outpatients with the diagnosis of ACS, CAB surgery or PCI, chronic angina and/or peripheral artery disease in the past year should be referred to a comprehensive outpatient CR program (AHA 2011; SOR of I / LOE of A) [7]. England’s guidelines are that CR should be initiated as soon as possible after admission and before discharge from hospital. Patients should be invited to a CR session within 10 days of their discharge from hospital (NICE 2013) [6].

Haykowsky et al. [24] reported that commencing and maintaining exercise training early after the onset of MI (within 1 week of onset) leads to benefits in left ventricular (LV) remodeling (LOE of 1+). Most studies examining the timing of CR analyzed the effects of exercise training on LV remodeling, and found that exercise training in the acute phase or early phases since onset produces positive outcomes. These studies describe the timing of the exercise component of CR, as opposed to the timing of the CR per se. The risk of bias was low in clinical trials included in a 2016 meta-analysis that analyzed the impacts of the timing of exercise training on LV functions. These studies classified the timing of exercise training into an acute phase (6 hours to 7 days), a recovery phase (7 to 28 days), and a maintenance phase (≥29 days), and examined differences in the left ventricle ejection fraction (LVEF), LV end-systolic diameter, and peak VO₂ of patients with reduced LV function after MI. Although results on the changes of LVEF and end-systolic diameter varied widely, exercise training in the acute phase led to significant improvements (more than moderate improvements) in LV end-systolic diameter, and peak VO₂ that were also greater than those produced by exercise training begun in the recovery phase and the maintenance phase. In conclusion, exercise training had positive effects on LV remodeling and recovery of cardiopulmonary functions, with the greatest effects achieved when exercise training was initiated in the acute phase (LOE of 1–) [25]. In this context, we determined that it is more appropriate to provide recommendations for the timing of exercise training as opposed to the timing of CR.

The organization of cardiac rehabilitation

Because staff numbers, equipment, and facilities vary according to the size, capital, number of relevant medical professionals, and number of eligible patients in each institution, CR programs should be structured on basic concepts in consideration of organizational circumstances. The structure of a multidisciplinary team, and particularly the number of people on the team and the type of health care professionals, is directly dependent on the number of patients and contents of the program, the complexity of the patients being treated, the organization’s human resources policies, and the community resources available to deliver the program [4].

Scotland’s guidelines state that CR services should offer individualized exercise assessments, tailor the exercise components to individuals’ choices, and deliver them in a range of settings (SIGN 2017; SOR strong) [5]. Canada’s guidelines state that all CR programs must address sitespecific facilities’ equipment in conjunction with safety requirements and considerations, and all CR programs must have policies and procedures to address the management of environmental and medical emergency situations (CACR 2009) [4].

In a systematic literature review of 8 observational studies that compared patients participating in a CR program with those not participating in a CR program, CR reduced all-cause and cardiac-related mortality rates, and significantly improved patients’ health-related QOL [26]. More than half of these studies adopted a multidisciplinary team approach, and structured their teams to provide interventions regarding exercise education, healthrelated information and education, lifestyle modifications, mental health support, and social support (LOE of 1–). A meta-analysis of 71 RCTs that reported the effects of comprehensive CR in coronary artery diseases (CAD) found a reduction of all-cause and cardiac-related mortality, noncritical re-infarction and readmission rates, and also found that many types of professionals were involved in providing exercise and psychological treatments, risk factor education and counseling (LOE of 1–) [15]. In a systematic literature review analyzing 63 studies investigating the effects of exercise-based CR on CHD, CR lowered cardiovascular mortality rates and risk of hospital admission, and the rehabilitation programs included strategies for exercise and risk factor education, behavior change, psychological support, and traditional risk factor management (LOE of 1–) [14].

In 2010, the Korean Society of Cardiac Rehabilitation discussed and adapted various Korean and foreign CR guidelines, including the American Association of Cardiovascular and Pulmonary Rehabilitation (AACVPR) guidelines, and relevant studies that provided recommendations for the basic facilities, equipment, and personnel required in Korea’s CR centers [27].

Cost-effectiveness of cardiac rehabilitation

The cost-effectiveness of CR has been reported in multiple studies [14]. These authors consistently concluded that CR involves a more cost-effective use of healthcare resources relative to that of general treatments. This conclusion is based on foreign literature however, so high quality Korean studies are needed to assess the costeffectiveness of CR in Korea, based on the total cost of healthcare and the cost per quality-adjusted life year (QALY) analysis.

Scotland’s guidelines state that interventions are considered to be cost effective if they fall below the commonly accepted UK threshold of £20,000 per QALY. However, no evidence of the cost-effectiveness of CR was identified in the UK, and additional higher-quality trials are needed to assess the cost-effectiveness of CR (SIGN 2017; SOR conditional) [5].

According to a systematic literature review analyzing 63 RCTs involving exercise-based CR in patients with coronary syndrome, it was difficult to compare 7 studies on the cost of CR and overall health management costs due to differences in their currencies and study periods, and given that 4 studies used a QALY that ranged from $42,535 higher to $650 lower per QALY gained in the CR group compared to the control group (LOE of 1–) [14].

Use individualized assessments to develop care plans and interventions

Patients referred to CR following MI are in diverse medical, social, and financial situations. Therefore, individualized CR programs should be considered together with various factors, including differences of heart function such as cardiac output, comorbidities that hinder exercise, such as lung disease, arthritis, and low back pain, differences in lifestyle such as smoking and drinking status, and patients’ ability to visit the hospital for CR according to employment status and distance from home. This is crucial to boosting participation and adherence with the CR program and ultimately achieving its goal, namely reducing the recurrence of MI and mortality rates. For this reason, the British Association for CR describes an individualized assessment of patient needs as an important component in the early phase of CR [28].

Scotland’s guidelines state that all patients referred to CR should undergo an individualized assessment that is used to create to a care plan and identify interventions specific to their needs. In addition, CR services should offer individualized exercise assessments, tailor the exercise components of their programs to individual choice, and deliver them in a range of settings (strong) (SIGN 2017; SOR of GPP, strong / LOE of 1+) [5].

In a prospective cohort study that compared a general CR and an individualized CR program, participation was about 30% higher in the individualized CR program (p<0.0001), and the risk for admission was significantly lower with the individualized CR program (RR=0.664; 95% CI, 0.554–0.797) (LOE of 2+) [29]. A Cochrane Review of 19 RCTs (n=10,856) examined the effects of programs that consisted of goal setting and action planning based on individualized assessments of various chronic diseases, such as heart failure, kidney failure, and diabetes, and found that individualized programs were more effective than those of the control group in lowering glycated hemoglobin (-0.24%; 95% CI, -0.35 to -0.14) and systolic blood pressure (-2.64 mmHg; 95% CI, -4.47 to -0.82), and improving depression (standardized mean differences [SMD]=-0.36; 95% CI, - 0.52 to -0.20) and self-efficacy (SMD=0.25; 95% CI, 0.07–0.43) (LOE of 1+) [30]. In a RCT of 132 patients admitted for a heart disease such as CAD, a higher percentage of patients agreed to be referred to a lifestyle adjustment program in the individualized assessment group than in the conventional assessment group (27% vs. 5%) (LOE of 1+) [31]. In a RCT of 40 patients with ACS, the intervention group, members of which received educational components according to their needs at an early stage, showed significantly higher self-efficacy compared to those in the control group, and a higher percentage of the intervention group participated in more than 90% of outpatient CR sessions (47% vs. 21%) (LOE of 1+) [32]. In a RCT of patients with ACS, providing individualized goals and plans (n=78) led to a high improvement in outcome indices that included self-efficacy, return to work or the recovery of pre-onset functions, readmission, and mortality, compared to the conventional method (n=105) (21.8%, n=17 vs. 10.5%, n=11; p=0.039) [33].

Psychological evaluations and interventions

Psychological symptoms such as depression, anxiety, and stress are common among patients with MI, and these symptoms can lead to recurrence, increased mortality, and reduced QOL. At about 15% to 20%, the prevalence of depression among patients with MI is about three times higher than that of the general population [34,35], and similar prevalence rates are observed among patients who have undergone PCI or CABG [36]. Depression undermines secondary prevention by lowering adherence with treatment plans, such as drug therapies, lifestyle adjustments, and CR participation [37-39], and may increase medical costs [40] for patients with heart disease. Anxiety and stress have been associated with recurrence and mortality rates among patients with CAD [41,42]. For this reason, psychological assessments and interventions have recently been recognized as essential components of CR programs.

Scotland’s guidelines are that all patients should be offered a psychological care package, based on a cognitive behavioral model (e.g., stress management, cognitive restructuring, and communication skills) as an integral part of CR (SIGN 2017; SOR of strong / LOE of 1++) [5]. US guidelines are that patients with recent CABG surgery or MI should be screened for depression, in collaboration with their primary care physicians and a mental health specialist (AHA 2011; LOE of B) [7]. Canada’s guidelines are that all CR patients should undergo screening for active and historical depression and anxiety at their intake assessment, and a qualified professional such as a psychologist or psychiatrist should refer individuals who screen positive for depression or anxiety for assessment or treatment (CACR 2009) [4]. Japan observes that because the prevalence of depression and depressed states is high among patients with CAD, and depression has been reported to relate to increased morbidity and mortality from CD, so specialists should screen for and treat depression in this patient population. Screening for depressed moods or other psychological symptoms, as well as conducting psychological interventions for the prevention and treatment of depression, are expected to alleviate psychological symptoms, improve QOL, enhance adherence to treatment, and reduce the mortality and incidence of cardiovascular events (JCS 2012; LOE of B, C) [8].

A 2017 Cochrane meta-analysis reviewed 35 RCTs (n=10,703) with regard to the psychological interventions conducted for CAD, and found that they did not reduce all-cause mortality, revascularization, or incidence of nonfatal MI compared to conventional treatment, but they did significantly lower heart disease-related mortality rates (RR=0.79; 95% CI, 0.63–0.98), and improved perceived depression (SMD=-0.27; 95% CI, -0.39 to 0.15), anxiety (SMD=-0.24; 95% CI, -0.38 to -0.09), and stress levels (SMD=-0.56; 95% CI, -0.88 to -0.24) (LOE of 1++) [43]. In a meta-analysis of 13 studies of the outcomes of psychological interventions in CVD patients, they were effective in reducing stress (Hedges’ g effect size=0.34; 95% CI, -0.19 to 0.87), anxiety (effect size=1.04; 95% CI, 0.53–1.54), and depression (effect size=0.67; 95% CI, 0.41–0.92) (LOE of 1+) [44]. A RCT that compared early counseling focused on psychological trauma and conventional stress counseling in 190 patients with ACS, found no significant differences in psychological symptoms, mortality rates, or CVD-related readmission rates between the two groups (LOE of 1–) [45].

Symptom-limited cardiopulmonary exercise test

The symptom-limited cardiopulmonary exercise test that provides useful information related to exercise prescriptions can be obtained by monitoring heart rate (HR), blood pressure (BP), symptoms, electrocardiography (ECG), and gas analysis, while gradually increasing exercise load [46]. Assessing the risk of participation in exercise-induced CR programs based on changes in HRs, abnormal ECGs, changes in BP, and the onset of symptoms as a result of increasing exercise load can be used to develop safe and effective exercise prescriptions [47]. Aerobic exercise capacity, as measured with VO_2max, was associated with CVD patients’ prognosis, and the effectiveness of CR can also be evaluated on the basis of its improvement after CR. Thus, in the absence of contraindications, the symptom-limited cardiopulmonary test is recommended for CR [7,8,47].

US guidelines recommend a risk assessment with a physical activity history and/or an exercise test for all patients, to guide their prognosis and prescription (AHA 2011; SOR of I / LOE of B) [7]. Canada recommends that a directly supervised graded exercise test (GXT) be part of the initial CR assessment prior to the initiation of therapy (CACR 2009) [4]. Japan recommends using a symptom-limited exercise stress test on day 14–21 after onset to predict prognosis, prescribe exercise training programs, and assess the treatment efficacy of CR (JCS 2012; LOE of B) [8].

Symptom-limited cardiopulmonary exercise testing is currently accepted as a standard in CR programs for patients after ACS, so it is difficult to conduct a new RCT to directly compare the clinical outcomes of patients who undergo this testing and those who do not. Consequently, we could not find such clinical trials in our search. However, large cohort studies on CVD patients found that VO_2max measured with cardiopulmonary exercise testing or exercise testing-based risk score are strongly correlated with all-cause mortality or heart disease-related mortality with consistent dose-response correlations. This suggests that maximal cardiopulmonary exercise testing is a useful tool for predicting the outcomes and risks during exercise for patients with ACS (LOE of 2++) [44]. In a study of 100 patients who underwent coronary angioplasty due to MI, patients were randomly divided into an exercise testing group (n=50) and a non-exercise testing group (n=50); physicians permitted seven out of 11 activities, including the return to work, earlier for the exercise testing group than for the non-exercise testing group (LOE of 1–) [48].

Submaximal exercise test (6-minute walk test)

The symptom-limited exercise test is recommended to assess exercise capacity and risk, prescribe exercise, and predict outcomes for patients undergoing CR [7,8,46]. However, exercise, ECG, gas analysis equipment, and adequate space are needed to conduct this standard cardiopulmonary exercise testing, and the associated costs hinder the universal application of this testing across all medical institutions [49]. Where maximal exercise testing is difficult, a submaximal exercise test, such as a 6-minute walk test, is commonly recommended [50,51], and a 6-minute walk distance is used to set exercise intensity and evaluate the effectiveness of CR [52,53]. In Korea, the 6-minute walk test is generally used as a submaximal exercise test, so recommendations in consideration of the validity, reliability, and responsiveness of the 6-minute walk test are needed.

Japan’s guidelines state that a submaximal exercise test conducted on day 4 after onset or thereafter should be used to predict prognosis, prescribe exercise-training programs, and assess treatment efficacy (JCS 2012; LOE of B) [8].

A 2012 systematic review of studies of the 6-minute walk test included six studies on validity, one study on reliability, and 11 studies on responsiveness [54]. The maximum metabolic equivalents (METs) obtained from the 6-minute walk test and symptom-limited exercise test were more than moderately correlated, and the maximum heart rate obtained from the 6-minute walk test and symptom-limited exercise test were also more than moderately correlated. Changes in reliability ranged from 2% to 8%, but the intraclass correlation coefficient (ICC) was 0.97, based on which level of evidence was evaluated to be moderate. In a meta-analysis of studies on the responsiveness of the 6-minute walk test in CR, the mean difference of a 6-minute walk distance before and after treatment was 60.43 m (95% CI, 54.57–66.30; p<0.001), indicating a high responsiveness (LOE of 2+). In a subsequent study, a repeated 6-minute walk test within 1 week after onset of MI had high reliability (ICC=0.879; 95% CI, 0.785–0.939), and the ICC for HR, rate of perceived exertion (RPE), measured using the Borg scale, and BP ranged from 0.880 to 0.934, also showing high reliability (LOE of 2+) [55]. Hanson et al. [56] found a high reliability for the 6-minute walk test (ICC=0.94), but the 6-minute walk distance tended to increase with subsequent tests, suggesting the need to consider this fact when assessing treatment responsiveness (LOE of 2+). A moderate correlation between 6-minute walk distance and VO_2max was found in a recent study of 49 patients referred to CR (r=0.56, p<0.01) (LOE of 2++) [57].

Strategies to improve cardiac rehabilitation participation

Despite the effectiveness of CR in patients with ACS, the problem of low CR participation lingers. In the US, CR participation among patients with MI is about 35%, and 20%–30% in Europe, while that in Japan is low, at about 4%–8% [58]. A recent Korean study also found a hospital-based CR participation rate of about 30% among patients with ACS in three university hospitals, and various factors that included distance, cost, time, comorbidity, and perceptions of CR were identified as hindering factors [59]. Therefore, effective strategies that could increase early uptake or participation in and adherence to CR are essential. In the United States, the Million Heart Project has been undertaken to lower recurrence, readmission, mortality, and medical costs among patients with ACS by increasing their CR participation based on grounds for effective strategies [60].

Scotland’s guidelines are that interventions to improve self-efficacy should be considered for inclusion in a CR program. Further, the strategies associated with increased uptake in at least one trial were: (1) structured telephone calls/visits by a nurse/therapist after hospital discharge; (2) early appointments to CR; (3) motivational letters based on the theory of planned behavior; (4) CR programs tailored for women; and (5) intermediate phase programs (self-management, instruction, and exercise monitoring) for older patients (SIGN 2017; SOR of conditional / LOE of 1++) [5]. Canada recommends maximizing the number of patients who may benefit from CR programs by developing systematized, preferably automated, referral mechanisms that are sensitive to socioeconomic and ethno-cultural diversity with key patient care partners and other CVD stakeholders. An automated referral process should be used to significantly increase referrals from acute care to CR and subsequent enrollment (CACR 2009) [4].

A 2014 Cochrane Review of RCTs on CR participation among CVD patients included 10 studies on early CR participation (n=1,338) and eight studies on CR maintenance (n=1,167), but a meta-analysis could not be performed due to the varying types of treatment involved [61]. Eight out of 10 studies and 3 out of 8 studies, however, found a significant increase in CR participation and maintenance, respectively. Strategies that increased CR participation included structured phone interviews or house visits after discharge [62-65], early visits to rehabilitation programs, [66] and letters that recommended CR participation [67]. Three studies that observed an increase in CR maintenance rates used physical monitoring based on daily journals, goal setting, and execution plans as strategies (LOE of 1++) [68-70]. In a RCT of 141 patients with ACS, physical activity rates increased in the intervention group that received motivating counseling compared to the control group (LOE of 1+) [71]. In a RCT that compared a female-only CR program and a mixed-gender CR program, there were no significant differences in participation rates between the two groups (LOE of 1++) [72]. In a RCT of 825 patients with ischemic heart disease or heart failure, participants were divided into two groups, one of which received instructions on learning and coping strategies and another that participated in conventional CR program strategies. The results showed that a higher percentage of the former group participated in more than 75% of the CR sessions (80% vs. 73%; OR=1.48; 95% CI, 1.07–2.05) (LOE of 1+) [73].

In a systematic literature review of 14 observational or intervention studies on the use of systematized automatic referrals and a liaison system to boost CR participation, meta-analysis revealed that the CR participation rate was highest when both systematized automatic referrals and a liaison system were applied (66%; 95% CI, 54%–77%), as opposed to using only the conventional CR referral system (24%; 95% CI, 18%–32%), only the systematized automatic referral system (45%; 95% CI, 33%–57%), or only the liaison system (44%; 95% CI, 35%–53%) (LOE of 2–) [74].

Strategies to maintain long-term physical activities

Maintaining physical activities is important to reducing CVD recurrence rates and improving patients’ QOL in the long-term, after undergoing exercise-based CR. Although fitness is improved and maintained for a certain amount of time after participating in a CR program [75], these effects are often not translated into maintaining physical activities in the long-term [76,77]. In a Korean study administering physical activity surveys before and after the onset of MI, 37% of the patients who were active prior to MI became inactive after onset, and 11% of the patients who were inactive prior to MI maintained physical activities after onset, thereby suggesting the need for strategies to promote the maintenance of physical activities in the long-term [78].

Scotland’s guidelines are that psychoeducation (goal setting, self-monitoring) should be considered for patients in CR to facilitate adherence to physical activities, and noted that the interventions reported to be of benefit were: (1) goal setting, (2) action planning, and (3) using daily diary entries to monitor activities (SIGN 2017; SOR of conditional / LOE of 1++) [5].

In a meta-analysis of 209 patients in four studies, psychological education (using strategies such as goal setting, problem solving, and self-monitoring) led to greater long-term physical activities compared to those promoted by simple education about exercise and risk factors (SMD=0.62; 95% CI, 0.30–0.94) (LOE of 1++) [79]. In a systematic literature review of strategies used to maintain physical activities in the long-term, one study examined the effects of a problem-based learning strategy, but the study failed to find any significant effects (LOE of 1+) [80]. Further, a study on improving self-efficacy also did not find significant differences [81], while one RCT that used self-monitoring observed significant effects [82]. In a recent RCT of 47 patients with CVD, no significant differences in the maintenance of physical activities 1 year after the program were found between a voluntary, incremental physical activity program and the existing supervised exercise program (LOE of 1+) [83]. However, the addition of group education regarding exercise based on the use of a pedometer and face-to-face counseling to the conventional CR program significantly increased the duration of patients’ continuous moderate to vigorous exercise 1 year after the program (LOE of 1+) [84].

Timing of cardiac rehabilitation in the case of coronary artery bypass surgery

Early CR is usually recommended if PCI is performed for ACS. In cases of CABG, CR may be delayed due to concerns about general deconditioning and the recovery of the sutured site after sternotomy. Thus, we propose appropriate timing for CR in cases involving CABG for ACS.

US guidelines recommend CR for all patients after CABG, with the referral ideally performed early post-operatively during the surgical hospital stay (AHA 2015; SOR of I / LOE of A) [85], and all eligible patients with ACS, or those whose status is immediate post-CAB surgery or post-PCI should be referred to a comprehensive outpatient CR program either prior to hospital discharge or during the first follow-up office visit (AHA 2011; SOR of I / LOE of A) [7].

Patients with ACS who underwent CABG and participated in CR showed significantly reduced mortality rates, CVD incidence, readmission rates, and length of hospital stays, compared to those in the control group [86,87]. According to AHA recommendations in 2015, CR is recommended for all patients who undergo CABG, and being referred to CR as an inpatient early after surgery is ideal [85]. Furthermore, according to AHA recommendations made in 2011, all patients who undergo CABG should begin CR while hospitalized, and must be referred immediately after discharge or at the first outpatient visit for a more comprehensive outpatient-based CR program [7].

A systematic review of nine randomized controlled studies on the effects of CR after CABG on patients with ACS found that the CR group showed improved quality of life and reduced cardiovascular risk factors [88]. However, this study could not systematically analyze the effects of specific rehabilitation regimens, since the type, intensity, and frequency of exercise varied in different CR programs across the studies, and diverse assessment parameters were used to analyze the effects of CR (LOE of 1–). It has been reported in Korea that beginning CR in the hospital after CABG leads to lower resting heart rates and increased 6-minute walking distances at the time of discharge and 1 month after surgery [89].

Aerobic exercise program

CR programs encompass the management of various risk factors that increase the risk of CVD. In particular, interventions for physical activities are a major component of CR programs [90]. Among various interventions for physical activities, we presented published evidence from aerobic exercise programs, and reviewed the different effects of CR according to the types of aerobic exercise involved.

Scotland’s guidelines are that aerobic and resistance exercises should be considered part of the exercise prescription for patients attending CR (SIGN 2017; SOR of conditional / LOE of 1++) [5]. Canada recommends that patients in CR engage in aerobic and resistance exercise 3–5 times per week, at 40%–85% of their HR reserve, for 20–40 minutes per session. Each exercise session should include an appropriate warm-up and cool-down period. These guidelines further note that regardless of the form of interval training, there is compelling evidence that interval training can lead to improvements in peak VO₂, functional status, and overall QOL. Accordingly, CR programs have increasingly used interval training as an alternative to traditional continuous aerobic training (CACR 2009) [4].

The exercise component of CR lowers cardiovascular mortality and readmission rates while increasing QOL [91]. Though the methods, frequencies, durations, and intensities of exercise vary widely across the studies conducted, nonetheless, studies have reported that CR is effective, regardless of specific frequency, duration, intensity, and place of exercise (LOE of 1++) [18,19,92]. The SIGN 2017 recommends that aerobic exercise should be included in exercise prescriptions for patients participating in CR [5].

In 18 RCTs that compared aerobic exercise-based CR and control groups among patients who underwent PCI or CABG for ACS [93], the aerobic exercise group showed reduced systolic BP (95% CI, -6.67 to -0.91), increased high-density lipoprotein (HDL; 95% CI, 1.24–6.43), and reduced low-density lipoprotein (LDL; 95% CI, -10.35 to -0.73) compared to the control group. Furthermore, peak VO₂ (95% CI, 2.41–4.53) and LVEF (95% CI, 0.26–4.93) were improved when aerobic exercise was performed. However, there were no significant differences in diastolic BP, total cholesterol, and triglycerides between the two groups. Regarding the effects of aerobic exercise according to duration, exercise for 8 to 12 weeks led to lower systolic BP at rest and significantly higher peak VO₂ compared to the conventional treatment group, and exercise for more than 12 weeks led to positive effects on HDL, LDL, triglycerides, peak VO₂, and LVEF. Regarding the amount of exercise per week, 60–90 minutes of exercise per week had positive effects on systolic BP, HDL, triglycerides, peak VO₂, and LVEF, and there were no additional effects, even when exercise was performed more than 90 minutes per week (LOE of 1–).

In a systematic literature review comparing the effects of CR with high-intensity interval training (HIT) and that with moderate-intensity continuous training (MCT) [94], HIT led to greater changes of peak VO₂ than did MCT (mean difference [MD]=1.78; 95% CI, 0.45–3.11), while MCT led to lower resting HRs (MD=-1.80/min; 95% CI, 0.71–2.89) and body weight (MD=-0.48 kg; 95% CI, 0.15–0.81) than did HIT. There were no differences in changes to blood glucose, triglycerides, and HDL after exercise between the two exercise groups (LOE of 1+). A Korean study also reported that HIT (1–2 times a week for 9–10 weeks) led to a greater increase in peak VO₂, and a 6-minute walking distance led to a greater reduction in perceived depression and fatigue in patients with low to moderate-risk stratification for exercise-related cardiovascular adverse events [95,96].

Resistance exercise program

CR programs encompass the management of various risk factors that increase the risk of CVD. In particular, interventions for physical activities are a key component of CR [90]. Among different types of physical activities, we presented evidence of resistance training for CR, and reviewed the different effects of CR according to types of resistance exercise.

Scotland recommends that aerobic and resistance exercises be considered part of the exercise prescriptions for patients attending CR (SIGN 2017; SOR of conditional / LOE of 1++) [5]. US guidelines are that it is reasonable for a clinician to recommend complementary resistance training at least 2 days per week [5] (AHA 2011; SOR of IIa / LOE of C) [7]. Canada in turn comments that patients in CR may safely derive fitness benefits from a supervised program of resistance training, and resistance training should be offered to all eligible CR participants (CACR 2009) [4].

The exercise component of CR lowers cardiovascular mortality and readmission rates while increasing patients’ QOL [91]. The methods, frequencies, durations, and intensities of exercise vary widely across studies, however, it has been reported that CR is effective, regardless of the specific frequency, duration, intensity, and place of exercise (LOE of 1++) [18,19,92].

The SIGN 2017 recommends the inclusion of resistance training in exercise prescriptions for CR [5], and the AHA 2011 reports that resistance exercise improves muscle strength, endurance, fitness, independence, and QOL, regardless of the presence of CVD [7]. One limitation of these studies on the reported effects of resistance training, however, was that they were generally small-scale studies involving low-risk groups for exercise-related adverse cardiovascular events. Thus, the AHA (2011) presents the absolute and relative contraindications of resistance exercise [97] (Table 6).

Table 6.

Absolute and relative contraindications to resistance training

In a systematic review analyzing the effects of progressive resistance exercise in CR programs, both progressive resistance exercise and aerobic exercise led to increased peak VO₂ and fitness, with no significant differences between the two types of exercise. Although the improvement of peak VO₂ did not differ between the aerobic exercise plus resistance exercise group and the aerobic exercise only group, the aerobic exercise plus resistance exercise group had greater improvements in maximal exercise capacity. When progressive resistance exercise was performed, there was a significant increase in upper and lower limb muscle strength compared to the control group. When both resistance and aerobic exercises were performed, there was a significant improvement in lower limb muscle strength compared to aerobic exercise alone. Regarding types of exercise, the combined exercise group showed significant improvements when isotonic exercise was performed, but the combined exercise group and the aerobic exercise group did not significantly differ in muscle strength improvements when isokinetic and isometric exercises were used. For lower limb muscles, combining isotonic resistance with aerobic exercises led to significantly better improvements in lower limb muscle strength compared to the effects of using only aerobic exercise (LOE of 1–) [98].

In a systematic review of three clinical trials that compared the effects of eccentric and concentric exercises in patients with acute MI, there were no significant differences in peak VO₂ and 6-minute walk test results between the two groups (LOE of 1–) [99].

In a systematic literature review of studies analyzing the effects of CR according to the intensity of resistance exercise, the amount of VO_2max increase was greater when lowmoderate-intensity resistance exercise (<70% 1RM or ≥12 repetition/set) and aerobic exercise were combined, than when aerobic exercise was performed alone. When performed for less than 12 weeks, combined exercise led to a greater increase in VO_2max than did aerobic exercise alone, but when performed for more than 12 weeks, studies could not find significant differences in the increases in VO_2max between the two groups. One study that examined high-intensity exercise (≥70% 1RM or <12 repetition/set) did not find differences between the two exercise methods. Combined exercise led to superior enhancements in muscle strength compared to those with aerobic exercise, regardless of the intensity and duration of exercise (LOE of 1+) [100].

Safety issues of exercise training in cardiac rehabilitation

The safety of patients who have experienced ACS must be ensured during CR. To this end, it is universal practice to assess patients’ cardiac states after ACS, stratify risk for exercise-related cardiovascular adverse events through cardiopulmonary exercise testing, and have patients perform ECG-monitored exercise at the hospital for a certain period early after onset.

With regard to safety issues related to CR and exercise, US recommendations are that all patients receive a risk assessment with a physical activity history and/or an exercise test to guide prognoses and prescriptions, and the clinician should counsel patients to report and be evaluated for symptoms related to exercise (AHA 2011; SOR of I / LOE of B, C) [7]. Canada advises that risk stratification is the key to safe and effective exercise prescriptions, and all patients entering CR programs must have a medical assessment and undergo determination of their cardiometabolic fitness prior to the initiation of therapy. A directly supervised GXT is recommended as part of the initial CR assessment prior to the initiation of therapy, and the provision of continuous EKG telemetry should be at the discretion of the CR program medical director. Patients in CR are encouraged to wear a heart rate monitor during physical activity and exercise, and follow the target heart rates prescribed for them (CACR 2009) [4].

Findings such as VO_2max, duration of exercise testing, presence of angina symptoms, and ECG abnormalities from cardiopulmonary exercise testing are helpful for predicting long-term outcomes such as mortality for patients who underwent PCI or CABG for ACS [101-105]. Furthermore, demographic factors and CVD history, such as sex, age, BMI, smoking, systolic BP, total cholesterol, HDL, hypertension, diabetes, and drug history are known to be predictors of CVD recurrence and mortality [4,106].

Therefore, risk stratification for exercise-related cardiovascular adverse events based on history and cardiopulmonary exercise testing is important for patients who undergo PCI or CABG [7]. Risk stratification is critical for ensuring the effectiveness and safety of exercise prescriptions, and therefore, cardiopulmonary exercise testing is recommended prior to beginning an initial CR exercise program [4]. The incidence of heart attacks when conducting exercises such as jogging, swimming, playing tennis, and engaging in cross country skiing is reported to be higher in the heart disease population than in the population without heart disease, with one case per 12,000–121,955 hours, and one case per 375,000–888,000 hours, respectively [107]. However, appropriate exercise programs significantly lower the disease-related mortality of patients with heart disease, so recommending aggressive CR exercise training to these patients is a reasonable decision in which the benefits far outweigh the harms. The type and intensity of exercise and monitoring during exercise are known to have a serious impact on exercise-related deaths, and continuous ECG monitoring markedly lowers the incidence of heart attacks during exercise [107].

According to one Korean report, there were 17 cases of angina symptoms without ECG anomalies, 31 cases of ECG anomalies without angina symptoms, 10 cases of ECG anomalies with angina symptoms, and 10 cases of temporary blood pressure instability during 13,934 hours of CR by 975 patients over a 10-year period, but during this time there was not a single major adverse cardiac event, such as a death or AMI [108].

Home-based cardiac rehabilitation

Home-based CR does not refer to an exercise performed alone by the patient at home; it is a form of exercise performed at home after medical professionals have assessed the patient’s cardiac functions at the hospital through exercise testing, and the patient has been given an exercise prescription and education about risk factor management. Effective home-based CR is based on continuous bidirectional communication with the medical team, and continuous follow-up assessments. Owing to recent advances in information and communication technologies (ICTs) and medical technologies, studies are continuing to examine new approaches to homebased CR and their effects. Hence, we reviewed relevant evidence and prepared a recommendation tailored to the circumstances in Korea.

Scotland’s guidelines state that an exercise component in CR reduces cardiovascular mortality and hospital admissions, and improves health related QOL regardless of the type of CHD, type of CR, or setting. Benefits appear to be independent of a specific frequency, duration, or intensity of exercise, or whether it takes place in a hospital, home, or community setting. Further, in homebased CR programs, BCTs such as social supports and goal setting were shown to be effective in reducing CVD risk factors, with results comparable to those of hospitalor center-based programs. It concluded that there is little difference between home- and center-based CR in terms of the number of program withdrawals (SIGN 2017; LOE of 1++) [5]. US guidelines are that a home-based CR program can be substituted for a supervised, center-based program for low-risk patients (AHA 2011; SOR of I / LOE of A) [7]. Canada advises that CR programs should follow comprehensive program models, both home-based and institution-based, and integrate the core program components presented within its guidelines (CACR 2009) [4].

Previous studies have proposed several methods of home-based CR, including one in which patients perform CR at home by themselves after undergoing a cardiac function assessment at the hospital and being given an exercise prescription, with periodic visits from a therapist to provide further instructions on CR. Another approach involves teaching patients to exercise using a brochure or phone, and monitoring their adherence [109].

A systematic literature review that analyzed the effects of home-based CR and hospital-based CR found no differences between the two groups in total mortality at 12 months after onset (RR=1.19; 95% CI, 0.65–2.16), fitness at 12 months after onset (MD=-0.13; 95% CI, -0.28 to 0.02), fitness at 12 to 24 months after onset, quality of life parameters at 24 months, CR completion rates, and total cholesterol, LDL, triglycerides, systolic BP, diastolic BP, and smoking cessation rates at 3 months and 12 months (LOE of 1+) [110,111]. Furthermore, home-based CR and hospital-based CR had equal effects on maximum heart rate and weight loss (LOE of 1–) [111]. Similarly, both CR programs had equal effects on physical and emotional components of QOL as well as on depression (LOE of 1–) [111]. Regarding anxiety, one study reported that homebased CR and hospital-based CR had similar levels of effectiveness [111], and another study reported that homebased CR was slightly more effective (LOE of 1–) [112]. Regarding blood HDL, one study found increased blood HDL after hospital-based CR, [110] while another study found hospital-based CR to be more effective, though the statistical significance of the difference was unclear [111], thus calling for additional analysis (LOE of 1+). Behavioral change techniques such as social supports and goal setting during home-based CR programs are effective in decreasing cardiovascular risk factors, and the levels of effects were similar to those of hospital-based CR programs (LOE of 1++) [113].

A report conducted in Korea suggested that homebased CR programs using a mobile ECG data transmission device (HeartCall) led to improvements in the participants’ fitness and QOL [114]. Owing to recent advances in ICTs, researchers in Korea are attempting to study CR equipment based on smart devices and sensors, and additional studies are needed to investigate the clinical effects of home-based CR programs using these technologies [115,116].

Cardiac rehabilitation for elderly patients

The need for CR is also growing due to aging and the subsequent rise of coronary syndromes among those aged 65 years and older. In many cases, however, the elderly engage in fewer daily physical activities, and have a low level of participation in CR exercise programs due to other comorbidities. Based on our review of the relevant literature, we developed recommendations for CR for the elderly.

Canada’s guidelines state that elderly patients with CAD will benefit from and should participate in CR programs that include both aerobic and resistance training (CACR 2009) [4].

One-year survival rates for acute MIs is known to decline with advancing age [117]. There is also a report that those aged 65 years or older who have had MIs show lower VO_2max, exercise testing durations, depression indexes, and grip strength compared to those with other heart diseases [118]. According to an article that reviewed various studies observing the effects of CR by age, CR improves aerobic fitness and physical functioning scores equally, and lowers body fat percentages and LDL/HDL ratios in patients 65 years and older similarly to its effects on patients under the age of 65 [119]. Even among the superaged over 75, CR can lead to functional improvements, and for this group it is important to initiate CR aggressively in the acute phase, immediately after being admitted for heart disease, and continuing into hospitalcentered outpatient CR after discharge [120]. Therefore, elderly patients with CAD must also be provided with and be motivated to participate in CR that is comprised of aerobic and resistance exercise [4]. A systematic literature review that analyzed the effects of resistance training on elderly (≥65 years) patients and mid-aged patients with CAD also found that resistance training improves upper and lower limb muscle strength, exercise capacity, and mobility equally in both patient groups (LOE of 1++) [121].

Although no Korean study has analyzed the effects of CR in those aged 65 years and older, it has been reported that CR improves VO_2max and anaerobic thresholds in MI in patients aged 60 years and older [122].

Education for secondary prevention

Omega-3 fatty acids

England advises physicians not to offer or advise people to use the following to prevent another MI: omega-3 fatty acid capsules and/or foods supplemented with omega-3 fatty acid. There is, however, no evidence of harm if people choose to take these capsules or foods. It further advises people not to take supplements containing betacarotene, and does not recommend antioxidant supplements (vitamins E and/or C) or folic acid to reduce cardiovascular risk (NICE 2013) [6].

Patients who had MIs and take omega-3 supplements or increase the α-Linolenic acid (ALA) content in their diet for secondary prevention will not experience negative consequences. Further, omega-3 or ALA does not have a harmful effect on lipid metabolism, and may also reduce triglycerides [148]. However, there is no clear evidence suggesting that they influence cardiovascular events or all-cause mortality (LOE of 1++) [149].

There have been reports of relatively mild adverse events, such as diarrhea, nausea, delayed hemostasis, and allergies after taking omega-3 as food supplements, but omega-3 lowers triglycerides and increases HDL-cholesterol. Several clinical trials and meta-analyses, however, found little evidence suggesting that taking omega-3 lowers all-cause mortality, cardiovascular events, or cardiovascular mortality, the incidence of CAD, cerebrovascular disease, or arrhythmia [149-152]. An ALA-rich diet did not lower all-cause mortality or cardiovascular mortality, and its effects on the incidence of stroke were not clear. It did, however, slightly reduce deaths from CAD and the incidence of arrhythmia [149].

Polycosanol

In a study that assessed the effects of policosanol in patients with hyperlipidemia, polycosanol lowered total cholesterol and LDL-cholesterol and increased HDL-cholesterol with no significant effects on triglycerides or body weight [153]. Patients with a history of CAD must take high-dose statins for secondary prevention purposes, however, in cases where side effects included poor drug adherence and necessitated the use of low-dose statins, combining low-dose statins with Armolipid Plus (red yeast rice, policosanol, berberine, folic acid, coenzyme Q10, astaxanthin complex) led to a greater reduction of LDL-cholesterol and higher achievement of targeted LDL-cholesterol rates without differences in adverse events (LOE of 1–) [154].

Another study that assessed the safety and effects of adding polycosanol for patients receiving antiplatelets after drug eluting stent insertions found that adding polycosanol did not significantly elevate cardiovascular events, deaths, or bleeding tendencies (LOE of 1–) [155]. However, findings on the association of polycosanol and improvements in dyslipidemia are inconsistent [156], and there have been no clinical trials on the long-term clinical outcomes or secondary prevention outcomes related to polycosanol intake in patients with CAD.

Antioxidants

Vitamin E, beta-carotene, and vitamin C are antioxidants widely used as food supplements. Vitamin C has been found to delay the progression of early atherosclerosis, with no benefits for atherosclerosis in later stages in vitro. A meta-analysis of the effects of these antioxidants on major cardiovascular events found that there were no significant differences between these antioxidants and placebos in major cardiovascular events, that is, MI, stroke, all-cause mortality, heart disease-related mortality, revascularization rates, CAD, and heart failure [157]. Further, antioxidants did not prevent cardiovascular events in women (LOE of 1++) [158,159].

It has been reported that vitamin E lowers the incidence of MI and stroke in patients aged 65 years or older [160], but this was only observed in a RCT that involved supplements manufactured by a pharmaceutical company, and there is a lack of evidence suggesting that it has preventive effects against CVD (LOE of 1++) [157,161]. Given published findings and the results of the analyses conducted in developing these CPGs, there is not sufficient evidence to conclude definitively that omega-3, polycosanol, and antioxidants have benefits for the secondary prevention of CAD.

It would clearly be harmful if patients overestimated the effects of these food supplements and deluded themselves into thinking they were undergoing adequate treatment, and so paid less attention to adhering to other established measures for secondary prevention, that is, taking prescribed medicines, exercising, and giving up smoking. Therefore, healthcare professionals must intervene aggressively in this matter (LOE of 4).

ICT-based modality for patient education

ICTs (e.g., phones, cell phones/smartphones, mobile apps, tablet computers, the Internet, biosensors) are now widely used in healthcare. The World Health Organization defines medical and public health practice using mobile devices (e.g., cell phones, patient monitoring devices, personal digital devices, other wireless devices) as mobile health (mHealth) [162], though its definition is still vague, and it is often confused with electronic health (eHealth) and telehealth. This technology can, however, help overcome factors that limit participation in CR, such as distance, traveling, scheduling, access to caregivers, and costs. These technologies can increase access to home-based CR, enable two-way communications with experts and monitoring by experts, and is also helpful for hospital-based programs. In particular, mHealth is very useful for managing chronic diseases such as CVD, by increasing a patient’s knowledge and awareness of their disease, and providing appropriate education to induce and maintain good self-care.

As of 2015, Internet usage rates and smartphone penetration rates among adults in advanced countries worldwide were 87% and 68%, respectively. At 94% and 88%, respectively, these rates are highest in South Korea. Although there is a great gap in developing countries, which have rates of about 54% and 37% respectively, rates in these countries are rising sharply [163]. As of 2017, 50% of global Internet traffic was on mobile devices (smartphones or tablet computers) [164], and mHeatlh apps were estimated to have been downloaded 3.7 billion times [165]. In other words, the Internet and mobile digital devices based on ICTs that are already a routine part of our daily lives are widely used for health-related purposes.

Scotland’s guidelines are that technology-based interventions should be considered for patients participating in CR, and psychoeducation (goal setting, self-monitoring) should be considered for patients in CR to facilitate adherence to engaging in physical activities (SIGN 2017; SOR of conditional / LOE of 1++) [5].

The results of ICT-based interventions intended to promote exercise are not consistent [111,166]. An Australian patient-centered program that consisted of a 1-hour initial consultation followed by 3 months of repeated phone follow-ups significantly reduced total cholesterol (mean 154.7±3.9 vs. 181.8±3.9 mg/dL; p<0.001), systolic BP (mean 131.6±1.8 vs. 143.9±2.3 mmHg; p<0.001), and BMI (mean 28.9±0.7 vs. 31.2±0.7 kg/m²; p=0.025) [166]. In another multicenter RCT in Australia, goal setting with 6 months of phone follow-ups led to a greater reduction in total cholesterol (MD=21 mg/dL and 95% CI, 16–25 mg/dL vs. MD=7 mg/dL and 95% CI, 3–11 mg/dL), body weight, BMI, dietary fat intake, saturated fat intake, and anxiety [166]. These interventions, which had efficacies similar to those of traditional CR, may serve as alternatives for patients who cannot participate in hospital-based CR, or those who simply prefer this method (LOE of 2+) [111,166]. In a RCT that used exercise prescriptions and behavior support over the Internet and text messages (4–6 per week for 24 weeks) for patients with ischemic heart disease (mean age, 60 years), this intervention did not improve exercise capacity (VO_2max), but significantly increased leisure activities (MD=110.2 min/week; 95% CI, -0.8 to 221.3; p=0.05) and walking time (MD=151.4 min/week; 95% CI, 27.6–275.2; p=0.02). In addition, patients’ self-efficacy and QOL were significantly improved (LOE of 1++) [167]. The use of digital pedometers in a small-scale RCT significantly increased the total number of physical activities, walking time, and walking frequency of patients undergoing CR (LOE of 1+) [168].

Combining physical activities with video games increased involvement in physical activities and improved QOL, well-being, and symptoms of depression in patients with stroke or heart failure aged 50 years or older. These attempts were evaluated to be safe and fun (LOE of 2+) [169]. Internet-based interventions, such as websites that provided individual instructions, also improved the number of steps and exercise capacity of patients in distant locations, or those who had decided not to participate in hospital-based CR (LOE of 1++) [71,170,171]. In a home-based RCT that provided a web-based rehabilitation program for patients with angina, at the 6-week follow-up the intervention group showed an average increase of 497 steps daily, while the control group showed an average decrease of 861 steps daily (LOE of 1+) [170]. Patients who were provided with 6 months of online personal instruction called CardioFit, an Internet-based system developed to promote physical activities in patients with CAD, increased the amount of physical activities they engaged in and their QOL at 6 months and 12 months (LOE of 1++) [172]. In addition, the intervention group, which used an Internet-based virtual CR program for 4 months, showed a significant increase in maximum treadmill duration (45.7 seconds; 95% CI, 1.04–90.48) at 16 months, compared to that of the control group. The intervention group also had decreased total cholesterol, LDL-cholesterol, and dietary saturated fat intake (LOE of 1++) [171]. Home-based multilateral CR models using ICT were associated with improvements in CVD risk factors comparable to those of traditional hospital-based CR programs [166]. Thus, these CR models may be integrated into traditional treatment models, and increase the number of options available to patients (LOE of 2+).

A recent study reviewed and meta-analyzed 10 clinical trials of 607 patients in 5 countries about the utility, acceptance, and usefulness of mobile apps for self-care of CVD and management of risk factors [173]. The mobile app group showed improvements in readmission rates, disease knowledge, psychosocial well-being, BMI, waist circumference, cholesterol and fitness, physical activities, drug adherence, and smoking cessation rates [173]. However, many of these studies were small-scale, had a short study period, had no blinding due to the nature of the intervention, and used biased data, thereby necessitating additional studies to address these shortcomings. Nevertheless, these studies confirmed the potential mobile apps had for inducing behavioral changes and ameliorating CVD risk factors. These findings allude to the usefulness of mHealth as a healthcare delivery system that improves access to CR for patients who have difficulty participating in traditional hospital-based CR (LOE of 1–).

Maintaining medication adherence is critical for the secondary prevention of CVD, and is associated with readmission, disease onset, mortality, and increased medical costs [174]. Hence, behavioral interventions such as education and counseling intended to enhance medication adherence are essential components of CR. In this context, study findings that ICT-based interventions are useful in improving medication adherence for patients with CVD, as with other chronic diseases are mounting. A recent meta-analysis of RCTs used mHealth devices for CVD patients had medication adherence as its primary outcome. According to this meta-analysis of 10 clinical trials (1–18 months of follow-up), using text messages, Bluetooth-electronic pillboxes, online messaging platforms, and conversational voice calls all enhanced drug adherence, though the degrees of improvement varied [130]. This meta-analysis was also limited because it analyzed mostly small-scale studies with short-term follow-ups and self-reported results, but the results nevertheless suggest that mHealth tools may be useful for improving drug adherence among CVD patients (LOE of 1–).