Exercise-based cardiac rehabilitation: different angles to grasp its beauty

Abstract

Given the ever-improving quality of invasive and pharmacological therapy for coronary artery disease, one might wonder if the role of an older treatment modality, cardiac rehabilitation (CR), persists. To combat any doubts, many have recently sought out to answer this question. All have chosen a different methodology and all have arrived at different, albeit mainly positive, conclusions. A recent Cochrane review chose to include only randomized controlled trials (RCTs) into their analyses. 1 Specifically, RCTs were chosen in which exercise-based interventions were compared to no exercise. As a main result, the study found that exercise-based CR reduced the risk of cardiovascular mortality but not total mortality. The CROS study 2 and its successor the CROS-II study 3 looked at CR from a different perspective. With the aim in mind to reflect current clinical reality as closely as possible, strict criteria were set to only include recent studies about multi-component CR. Also, and interestingly , retrospective and prospective controlled cohort studies are included in the analysis. The authors chose to do so to expand the study population, but also to better reflect real-life clinical reality, as RCTs tend to select for highly motivated people eager to participate in clinical trials. Both studies show effectiveness of CR in selected populations (acute coronary syndrome and coronary artery bypass grafting) by reducing total mortality. Twice however it is underscored that large heterogeneity persists in CR delivery throughout Europe and that international standards for CR delivery, but also for scientific evaluation are highly needed. The study from Ekblom et al. 4 shines yet another light on the problem. This study nicely demonstrates how a thoroughly kept large registry can offer very interesting and applicable results for clinical practice. In a large study population of over 20 000 patients, it was shown that exercise-based CR was associated with reduced total mortality, and for the first time this was shown independently in both men and women. The methodology using the nationwide cohort from the SWEDEHEART registry offers real-world data that is highly applicable to clinical practice. As the authors mention, limitations of the registry-based study are that cause-effect relationship cannot be confirmed, detailed data about dose and duration of each modality in CR is not known and cause-specific mortality data is lacking. All these studies shine a different light on CR and, while using different methodologies, none of these methods should be considered superior to one another. In their perspective paper, Murad et al. 5 argue that the classical depiction of the evidence pyramid, in which RCTs, systematic reviews, and meta-analyses are at the top, falls short as a classification of levels of evidence and as a surrogate for risk of bias. Depending on the research question other, non-random-ized study designs can be considered equal or superior to RCTs as long as the methodologies are designed to maximally reduce the risk of bias. The variety in methods and results of the aforementioned studies only illustrates the difficulty in studying CR.