ABSTRACT
OBJECTIVES: To estimate the frequency and critically appraise the use and reporting of composite outcomes in randomized clinical trials on pharmacological interventions for coronary artery disease. STUDY DESIGN AND SETTING: A metaresearch study. A search strategy was developed to retrieve references from MEDLINE. We considered articles, published from 1st January 2020, to December 31, 2021, reporting results of clinical primary outcomes from randomized clinical trials which assessed pharmacological interventions, used alone or in combination, for the treatment or secondary prevention (previous coronary event) of coronary artery disease. RESULTS: From the 34 included studies, 28 (82.35%) had a primary composite outcome. Thirteen unique composite primary outcomes were used with the most frequent being "cardiovascular death, myocardial infarction, stroke" (12/28, 42.86%). The term major adverse cardiac events was used for five distinct composite primary outcomes. A combination of 12 different components resulted in the 28 primary composite outcomes, with stroke being the most frequent component present in 96.43% (27/28) of the primary composite outcomes. From the included studies, 60.71% (17/28) reported the estimates for each individual component and the direction of the effect was consistent between all components and the composite outcomes in 58.82% (10/17) of them. Additionally, no included study discussed potential limitations and/or related advantages of the composite outcomes. CONCLUSION: In randomized clinical trials on pharmacological interventions for coronary artery disease, composite outcomes are frequently used, but the definition of their components is very heterogeneous. The estimate for individual components within the composite outcome is often not fully reported, which prevents a complete analysis of their adequacy for clinical practice. The term major adverse cardiac events was used inconsistently and to refer to different set of components, which can also be misleading and confusing.
