Objective To describe specific causes of death and evaluate whether bleeding events and infection contributed to mortality in all ticagrelor-treated and clopidogrel-treated patients with acute coronary syndromes.

Methods In the PLATelet inhibition and patient Outcomes (PLATO) trial, ticagrelor significantly reduced rates of vascular and total death compared with clopidogrel. In the 905 patients who died postenrolment in the PLATO trial (n=18 624), reviewers, blinded to study treatment, subclassified direct causes of death and evaluated whether infection or bleeding events contributed to fatal events.

Results Among vascular deaths, there were significantly fewer sudden deaths (63 (0.7%) vs 98 (1.1%), p<0.01) but no significant difference in deaths caused by acute myocardial infarction (179 (1.9%) vs 194 (2.1%), p=0.43) or heart failure (31 (0.3%) vs 42 (0.5%), p=0.20) with ticagrelor compared with clopidogrel. For non-vascular deaths, there was no difference between treatments in deaths directly caused by infection. Although, patients treated with ticagrelor were at lower risk for death where infection was either a direct cause or contributed to death (51 (0.5%) vs 76 (0.8%), HR 0.67 (0.47 to 0.95), p<0.05) but not for bleeding (42 (0.5%) vs 42 (0.5%), HR 0.99 (0.65 to 1.53), p=0.98).

Conclusions In this post hoc analysis, ticagrelor compared with clopidogrel reduced total and cardiovascular mortality, which appeared to be mainly mediated by a reduction in sudden death. Importantly, bleeding causing or contributing to death did not differ between treatments.

Clinical trial registration number NCT00391872 (http://www.clinicaltrial.gov).