Comparative effectiveness and safety of antiplatelet drugs in patients with diabetes mellitus and acute coronary syndrome

Abstract Purpose Comparative outcomes of treatment with antiplatelet drugs in patients with acute coronary syndrome (ACS) and co‐morbid diabetes mellitus (DM) are not well studied. Methods We performed a cohort study using US commercial claims data (2009‐2015) and conducted the following pairwise comparisons in ACS patients with DM: prasugrel vs clopidogrel, ticagrelor vs clopidogrel, and prasugrel vs ticagrelor. Outcomes of interest included (1) a composite effectiveness endpoint including myocardial infarction, ischemic stroke, or inpatient mortality; (2) a composite safety endpoint including major bleeding events requiring hospitalization; and (3) pneumonia hospitalizations as a negative control endpoint. We used calendar time‐specific propensity score matching to account for confounding and applied Cox proportional hazard models to calculate hazard ratios (HR) with 95% confidence intervals (CI). Results Comparative risk of the effectiveness endpoint was lower among prasugrel initiators compared to clopidogrel initiators (HR 0.82, 95% CI 0.68‐0.99, N = 7011 matched pairs), but no different between ticagrelor and clopidogrel (HR 1.02, 95% CI 0.76‐1.37, N = 3013 pairs) or prasugrel and ticagrelor (HR 0.83, 95% CI 0.58‐1.18, N = 2207 pairs). Bleeding risk was higher among prasugrel initiators when compared to clopidogrel initiators within the first month of treatment (HR 1.85, 95% CI 1.03‐3.35); no other comparison indicated any difference. No differences in the negative control outcomes were noted after PS matching for all comparisons, indicating adequate confounding control. Conclusions Prasugrel was associated with superior cardiovascular outcomes and a higher risk of short‐term bleeding compared to clopidogrel in patients with ACS and DM. Comparative outcomes were similar between ticagrelor and clopidogrel or prasugrel and ticagrelor.