Aim
The globalization of clinical trials has highlighted geographic differences in patient characteristics, treatments, and outcomes. We examined these differences in PARADISE-MI.

Methods and results
Overall, 23.0% were randomized in Eastern Europe/Russia, 17.5% in Western Europe, 12.2% in Southern Europe, 10.1% in Northern Europe, 12.0% in Latin America (LA), 9.3% in North America (NA), 10.0% in East/South-East Asia and 5.8% in South Asia (SA). Those from Asia, particularly SA, were different from patients enrolled in the other regions, being younger and thinner. They also differed in terms of comorbidities (high prevalence of diabetes and low prevalence of atrial fibrillation), type of myocardial infarction (more often ST-elevation myocardial infarction), and treatment (low rate of primary percutaneous coronary intervention). By contrast, patients from LA did not differ meaningfully from those randomized in Europe or NA. Use of angiotensin-converting enzyme inhibitors/angiotensin receptor blockers (34.8%) and beta-blockers (65.5%) was low in SA, whereas mineralocorticoid receptor antagonist use was lowest in NA (22%) and highest in Eastern Europe/Russia (53%). Rates of the primary composite outcome of cardiovascular death or incident heart failure varied two-fold among regions, with the lowest rate in SA (4.6/100 person-years) and the highest in LA (9.2/100 person-years). Rates of incident heart failure varied almost six-fold among regions, with the lowest rate in SA (1.0/100 person-years) and the highest in Northern Europe (5.9/100 person-years). The effect of sacubitril/valsartan was not modified by region.

Conclusion
In PARADISE-MI, there were substantial regional differences in patient characteristics, treatments and outcomes. Although the generalizability of these findings to a ‘real-world’ unselected population may be limited, these findings underscore the importance of considering both regional and within-region differences when designing global clinical trials.