Contribution of individual components to composite end points in contemporary cardiovascular randomized controlled trials - 26/11/20
![](/templates/common/images/mail.png)
Background |
Cardiovascular randomized controlled trials (RCTs) typically set composite end points as the primary outcome to enhance statistical power. However, influence of individual component end points on overall composite outcomes remains understudied.
Methods |
We searched MEDLINE for RCTs published in 6 high-impact journals (The Lancet, the New England Journal of Medicine, Journal of the American Medical Association, Circulation, Journal of the American College of Cardiology and the European Heart Journal) from 2011 to 2017. Two-armed, parallel-design cardiovascular RCTs which reported composite outcomes were included. All-cause or cardiovascular mortality, myocardial infarction, heart failure, and stroke were deemed “hard” end points, whereas hospitalization, angina, and revascularization were identified as “soft” end points. Type of outcome (primary or secondary), event rates in treatment and control groups for the composite outcome and of its components according to predefined criteria.
Results |
Of the 45.8% (316/689) cardiovascular RCTs which used a composite outcome, 79.4% set the composite as the primary outcome. Death was the most common component (89.8%) followed by myocardial infarction (66.1%). About 80% of the trials reported complete data for each component. One hundred forty-seven trials (46.5%) incorporated a “soft” end point as part of their composite. Death contributed the least to the estimate of effects (R2 change = 0.005) of the composite, whereas revascularization contributed the most (R2 change = 0.423).
Conclusions |
Cardiovascular RCTs frequently use composite end points, which include “soft” end points, as components in nearly 50% of studies. Higher event rates in composite end points may create a misleading interpretation of treatment impact due to large contributions from end points with less clinical significance.
Le texte complet de cet article est disponible en PDF.Plan
What is already known about this subject?
What does this study add?
How might this impact on clinical practice?
|
|||||||||||||
Wilbert S. Aronow, MD, served as guest editor for this article. |
Vol 230
P. 71-81 - décembre 2020 Retour au numéroBienvenue sur EM-consulte, la référence des professionnels de santé.
L’accès au texte intégral de cet article nécessite un abonnement.
Déjà abonné à cette revue ?