Background: The polypill strategy could become widely accepted in cardiovascular prevention due to reduced costs and its simplicity, which promote compliance. Aspirin is often included as a component of the polypill for primary prevention, but three powerful recent trials failed to show any favorable net benefit even in high-risk subgroups. Our objective is to estimate the net benefit associated with aspirin in primary cardiovascular prevention.

Methods: We simulated the impact of different polypill compositions combining pravastatin, ramipril, hydrochlorothiazide, with or without aspirin, on a realistic French virtual population between 35 and 65 years old. We assessed how this impact on myocardial infarction and stroke varied according to gender, diabetes, and arterial hypertension. We identified the subgroup of individuals whose specific benefit from aspirin was greater than twice the risk of serious bleeding it induced.

Results: The absolute benefit associated with aspirin was reduced by co-prescriptions. No subgroup of women benefited from aspirin, and the subgroup of women with a clear net benefit represented 128 women out of 529,421. Men at high risk of cardiovascular death, or with diabetes and hypertension, had a benefit from aspirin exceeding the risk of bleeding induced, but this risk represented more than half of the benefit. No subgroup analyzed did show a benefit greater than twice the risk of bleeding. The proportion of men whose expected benefit from aspirin was greater than twice the risk of bleeding represented 3% of all men. An optimal polypill strategy in primary prevention between the ages of 35 and 65, combining three drugs but not aspirin, can hope to save two out of three strokes and more than one out of two myocardial infarctions. It would prevent a major cardiovascular accident every 16 to 193 individuals treated according to the subgroups considered.

Conclusion: Until proven otherwise, aspirin has only a limited place in individuals between 35 and 65 years without a cardiovascular history. We showed how simulating therapeutic strategies on a realistic virtual population could be used for best applying available evidence.