Cardiac rehabilitation (CR) is a meticulously structured multidimensional intervention designed to support individuals who have experienced cardiovascular events or are at high cardiovascular risk. CR has demonstrated a multitude of benefits, including improved quality of life, increased functional capacity, enhanced adherence to pharmacological therapy, mitigation of anxiety and depression levels, reduced hospitalization rates, and even mortality reduction.1–5 These benefits mean CR is one of the most cost-effective interventions for managing cardiovascular disease.6,7

Cardiac rehabilitation typically comprises three phases. Phase 1 begins in the hospital, ideally 24–48 hours after an uncomplicated acute cardiac event. Phase 2 (exercise and education phase) ideally starts within the first two weeks after hospital discharge and lasts for three to six months.8 It can be conducted within the hospital setting, specialized CR centers, or in some cases, at the patient's home. This phase should include an individualized physical exercise program, educational sessions regarding lifestyle modification and fostering autonomy and psychological monitoring.8 Phase 3, also known as the maintenance phase, starts after the second phase and is intended to last throughout the patient's lifetime.

The impact of phase 2 of CR on patient prognosis has been extensively studied over the years.9,10 Physical exercise is currently recommended for heart failure (HF) patients, not only for its benefits in improving functional capacity and symptoms, but also for reducing the risk of hospital readmissions.11–13 The ExTraMATCH II meta-analysis demonstrated that participation in exercise-based CR programs was associated with improved quality of life and exercise tolerance in HF patients. However, it did not show a significant impact on mortality or hospitalizations.14

In 2019, the updated Cochrane review showed that CR appeared to have limited impact on short-term mortality (<12 months),15 but it was associated with reduced hospitalizations and seemed to be linked to a slight reduction in long-term all-cause mortality (>12 months).

For patients with coronary artery disease (CAD), the mortality benefit of CR is better established. The CROS II meta-analysis indicated a significant reduction in mortality after acute coronary syndrome or coronary artery bypass grafting (CABG).5

Most studies primarily focus on phase 2 CR programs, thus creating uncertainty regarding phase 3 CR programs, including its prognosis, ideal exercise protocols, training frequency, monitoring, components. Only two meta-analyses have exclusively addressed phase 3 studies, concluding that maintenance programs are associated with improved quality of life, lipid profiles, and functional capacity.16,17 However, due to insufficient studies, the impact on mortality could not be assessed.

The main objective of this study was to evaluate the impact of the maintenance phase of a CR program on all-cause mortality. Additionally, the study aimed to analyze phase 3 CR program types, exercise protocols utilized, frequency, duration, and included components beyond physical exercise training.

This systematic review with meta-analysis was conducted in accordance with the Preferred Reporting Items for Systematic Review and Meta-Analysis (PRISMA) guidelines and did not require ethical approval.

In total, 1867 articles were identified, of which eight met the eligibility criteria (Figure 1) and were included in the review.20–27

Figure 1.

Systematic reviews and meta-analyses flow diagram of study selection process.

Meta-analysis

Phase 3 of a CR program was associated with significantly lower mortality, with an HR of 0.56 (95% CI 0.49–0.65), and no substantial evidence of statistical heterogeneity (I2=5.3%), as presented in Figure 2. Despite the work by Nakayama et al.21 and Gabrys et al.,20 which were the studies that contributed the most to the final result, sensitivity analysis by individually removing studies demonstrated that the outcome in terms of direction, significance, and magnitude remains consistent, regardless of the study removed. In an analysis excluding each study individually, the statistical significance persisted, and the magnitude of the effect also remained consistent.

Figure 2.

Forest plot showing the results of individual studies and meta-analysis of the impact of phase 3 of CR on global mortality.

In a cumulative evaluation based on the temporal publication of data, a statistically significant impact of phase 3 of RC on mortality has been observed since 2012. This is reflected in the cumulative results of the studies by Giannuzzi et al. (2008),26 Onishi et al. (2010),27 and Belardinelli et al. (2010).25

In a subgroup analysis based on the included population (comprised of five studies with CAD patients, one with just HF patients, one including also MINOCA and another one with an unspecified etiology), no statistically significant differences were observed between the groups (p=0.59).

The meta-regression based on the follow-up time showed a trend towards greater benefit from CR programs over time, although without meeting statistical significance (Figure 3). Regarding biases in the meta-analysis, Egger's test showed a significant absence of publication bias/small study effects, with a p-value of 0.57. However, upon evaluating the funnel plot that assesses the dispersion of studies, a slight asymmetry is noted due to a greater magnitude of effect in studies with smaller standard errors (Figure 4A). In this context, a “trim-and-fill” analysis was conducted with the imputation of a study to the right on the funnel plot (Figure 4B). This imputation showed a HR=0.58 (CI=0.497–0.680), and no difference was observed between the data from this assessment and the results of the primary evaluation.

Figure 3.

Meta-regression plot evaluating the effect estimate of CR phase on mortality during follow-up time.

Figure 4.

(A) Funnel plot showing the dispersion of studies according to effect and standard error around the meta-analysis effect. (B) Funnel plot with the imputation of a study on the right side of the plot.

The bias risk assessments for each non-randomized study were evaluated according to the NOS. The majority of studies have good quality (4 out of 6 studies),21–23,27 with only 2 of reasonable quality (NOS=5 and 6).20,24

Regarding the bias risk of the two included randomized studies,25,26 both studies are classified as “some concern” in domain 2 of Rob 2, which relates to bias due to deviations from the intended intervention, as neither study is blinded, and patients and professionals are aware of each patient's study arm. In an analysis by study design type, randomized and non-randomized, there were no statistically significant differences between the groups (p=0.23).

This meta-analysis included over 60000 patients and demonstrates, for the first time, the benefit of phase 3 in reducing all-cause mortality. The positive impact of phase 2 CR is already well established in various prospective studies and meta-analyses, particularly in terms of quality of life, symptomatic improvement, reduction in re-hospitalizations, and mortality.1,2,4,5 However, the prognostic impact of phase 3 CR had not been established until now. This is the first meta-analysis designed to specifically assess the benefit of attending a phase 3 CR program on mortality.

Incorporating the meta-analysis data, phase 3 CR program was associated with a significant reduction in all-cause mortality, with a risk reduction of 44%. The statistical analysis demonstrated consistent results, with minimal evidence of heterogeneity among the included studies. Remarkably, the impact on mortality remained robust even after sensitivity analysis, which involved the individual removal of studies. This suggests the stability of the observed effect regardless of the inclusion or exclusion of specific studies.

This work highlights the scarcity of randomized and/or large-scale studies aimed at evaluating the prognostic impact of phase 3, as it only included eight studies, and of these, only two were prospective randomized studies. The bias risk analysis carried out through NOS showed that the included cohort studies were mostly of good quality. However, it is important to note that one of the studies with the largest sample size, although of moderate quality, had a substantial influence on the overall results, underscoring the need for cautious interpretation of the findings considering potential methodological limitations.

Consistent with the literature, a significant proportion of patients included in this meta-analysis had CAD (98.1%). Additionally, there was representation of studies focused on specific populations with HF, MINOCA, and mixed populations. This diversity in patient profiles emphasizes the relevance of phase 3 CR across a wide spectrum of cardiovascular conditions.

Also, in line with the majority of CR studies, our meta-analysis demonstrates a much higher prevalence of male patients compared to female patients (82% vs. 18%). Despite the similar benefit of CR between sexes, even for phase 2, the participation of women is generally significantly lower.28,29 Studies show that women are 12% less likely to be referred for CR, 9% less likely to initiate a program, and 13% less likely to complete it.30,31 These findings highlight the urgent need to develop targeted strategies to increase female participation in CR programs. Several barriers may contribute to this underrepresentation, including social, cultural, caregiving responsibilities, financial issues, and perceived risks. Therefore, exploring approaches that can overcome these obstacles is essential. Promising strategies include the implementation of women-only programs or tailored interventions that address their specific needs, as well as awareness campaigns directed at women and healthcare providers. Other actions might involve flexible scheduling, offering psychological support, support groups, and interventions that address social and cultural factors. Further research into the unique needs and preferences of women can guide the design of more inclusive and accessible CR services.

The prevalence of cardiovascular risk factors, such as dyslipidemia, hypertension, diabetes, and smoking, was substantial, emphasizing the relevance of CR in controlling these risk factors. Compared with the EURORASPIRE IV study,32 which included patients exclusively with CAD, the average prevalence of risk factors in our study was lower (hypertension: 44% vs. 78%; dyslipidemia: 68% vs. 74%; smoking: 11% vs. 26%). This fact may be justified by the greater variety of populations included in our study and the scarcity of data from some of the studies.

In contrast to the vast majority of literature regarding phase 2, which shows that center-based programs are most common, in this meta-analysis, most programs are exclusively or partially home-based. A meta-analysis published in 2018 showed that 57.5% of programs were center-based,33 and the UK National Audit of Cardiac Rehabilitation (UK NACR) reported that 75.4% of patients received this type of CR, with only 8.8% being home-based.34 This variation might reflect the evolving landscape of CR and the need to adapt programs to individual patient needs, especially for longer programs as anticipated in phase 3.

Regarding types of CR programs, there is a wide variety in terms of program duration and components. Program duration varied widely, ranging from six months to a decade, and this discrepancy may affect estimates of the mortality impact, particularly with longer follow-up. Despite the absence of guidelines regarding ideal components for phase 3, the majority of studies in our meta-analysis offered a multidimensional program. While the specific form or frequency of these components was not extensively detailed, half of the authors mentioned that programs also included nutritional counseling, smoking cessation, educational sessions, stress management, and psychological counseling, highlighting the holistic nature of a phase 3 CR.

European recommendations suggest that a CR exercise session should include aerobic training and dynamic muscular strength training, which should be individually prescribed.35 Among the five studies that provided sufficient information, 60% (3/5) included strength training in addition to aerobic training. The discrepancy in rehabilitation programs among different studies makes it difficult to reliably assess the best type of training and underscores the need for further research to establish guidelines or standards, similar to those in existence for phase 2.

All the programs of the analyzed studies were moderate continuous intensity programs. Although high-intensity interval training (HIIT) has been studied in phase 2 of CR, there are currently no studies specifically designed to evaluate its safety/impact in maintenance programs. Exercise stress testing was used in all studies to prescribe training intensity, with the exception of one that did not mention it. Most studies are in line with the European recommendations and use CPET to determine peak VO2 and ventilatory thresholds to guide training intensity prescription.36 Only a minority (two out of eight studies) used conventional stress testing, which is another way to assess patients’ functional capacity and a useful test to prescribe exercise according to the European Guidelines.36 Regarding the repetition of this test in follow-up, only two studies mentioned repeating it annually to adjust training intensity parameters.

This meta-analysis highlights the importance of attending a phase 3 CR program, especially of continuing a phase 2 program to improve patient outcomes and justify further research to optimize CR program implementation. Probably tele-rehabilitation and hybrid programs with more personalized prevention, adapted to patients’ needs and preferences will increase the implementation of long-term CR programs with its positive impact on patients’ lives.

Although our study aimed to provide innovative information on the impact of phase 3 CR program on mortality and program characteristics, it is important to acknowledge some limitations that may affect the interpretation of our results.

This systematic review demonstrates the significant benefit of long-term phase 3 cardiac rehabilitation (CR) programs in reducing all-cause mortality among cardiovascular patients, particularly those with coronary artery disease (CAD), which was the primary focus of the included studies.

This highlights the importance of the existence of specialized Phase 3 CR centers.

Our comprehensive analysis also highlights the heterogeneity of phase 3 CR programs, which vary in structure, duration, and components. This variability underscores the need for further research to optimize exercise protocols, standardize CR components, and develop guidelines, with particular emphasis on strategies to increase referral and participation among women.

This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.

None declared.