Background
Outcome measure choice and definition can determine the result of the study. We describe outcome measures and their definitions for cardiovascular studies in highly cited medical journals.
Methods
Cardiovascular phase III or IV randomised clinical trials (RCTs) or multicentre observational studies published in the New England Journal of Medicine, Lancet or Journal of the American Medical Association between 1 January 2013 and 6 June 2024 from Embase and Ovid Medline were included. Two independent reviewers selected the studies and extracted the primary and secondary outcome measures from each publication.
Results
386 studies (83% RCTs; 17% observational) representing 10 699 147 participants were included. Studies investigated coronary heart disease (51%), cardiomyopathy/heart failure (22%), heart rhythm disease (15%), valvular heart disease (11%) and ‘other’ cardiovascular diseases (1%), with 45% investigating a device and 48% funded by industry. The most frequently reported primary outcome measure was a composite (63%), the most frequent component of which was myocardial infarction (58%). The use of a composite for the primary outcome measure increased from 49% of studies in 2013 to a peak of 85% in 2018. From 2013 to 2023, the median number of secondary outcome measures per study increased for RCTs (3–8) and observational studies (0–7). Definitions for cardiovascular mortality, myocardial infarction and stroke varied across the studies.
Conclusions
For cardiovascular studies published in highly cited journals, there has been an expansion in the use of primary composite outcome measures and secondary outcome measures, with heterogeneity in the definition of primary outcome measures. A standardised approach to the use of cardiovascular outcomes measures is required.
