American Heart Association statistics indicate that every 25 seconds an American has a coronary event, every 40 seconds someone has a stroke; every 38 seconds someone dies of cardiovascular disease (CVD), and every minute someone dies of a coronary event. These events are often without warning, with the first manifestation being catastrophic (ie, sudden death). Survivors are left with considerable morbidity. These stark numbers emphasize the importance and urgency of preventing CVD.
One component of preventing CVD is to identify and target those at highest risk of disease. Thus, “risk prediction” is one of the central tenets of preventing CVD. The recent decade has seen major advances in our understanding of the pathophysiology of CVD, and a plethora of biomarkers (both circulating and genetic ones; proteomic, lipidomic, metabolomic biomarkers, and small molecules/cells) have emerged as playing a key role in determining the onset of CVD. These novel biomarkers have become candidates for CVD prediction. In parallel, there has been a burgeoning growth in the number of CVD prediction algorithms, accompanied by important advances in assessing which metrics to use to evaluate increments to contemporary CVD risk prediction tools.
The purpose of this series of articles in Circulation is to elucidate these advances in CVD risk prediction systematically. The series begins with an elaboration of the basic concepts of risk prediction, including the different time horizons of risk prediction (short-term, long-term, lifetime), and an interpretation of relative and absolute risks of disease. We also explore how the explosion of genomics and systems biology (OMICS) has led to the inundation of the literature with so many biomarkers, and how best to synthesize a parsimonious set of predictors from the numerous eligible and worthy candidates. We explore separately the expectations from and the incremental contribution of novel biomarkers to risk prediction in the primary and secondary prevention settings, where the goals of prediction may be distinct. One of the articles directly discusses how different the various contemporary risk prediction algorithms are from each other, and evaluates if these risk assessment scores truly make a difference in practical terms of patient management and disease outcomes. Last but not least, in this series we explore the standards for evaluating the clinical utility of new biomarkers (including an assessment of their cost-effectiveness) with a focus on optimal study designs to address this important issue.
In summary, this series of articles on risk prediction presents the state-of-the-current-art of CVD risk prediction in primary and secondary prevention settings, outlines the future ahead of us in terms of personalized risk prediction, and clarifies the desirable designs and standards for measuring clinical utility of new biomarkers of CVD risk.