External applicability of the COMPASS trial: the Western Denmark Heart Registry

Abstract

Abstract Aims In the COMPASS trial, combined aspirin and rivaroxaban treatment reduced ischaemic events in patients with stable coronary artery disease (CAD) or peripheral artery disease (PAD). We estimated the proportion of COMPASS eligible patients among unselected patients undergoing coronary angiography (CAG) and compared outcome rates among COMPASS eligible and non-eligible patients. Methods and results We applied the COMPASS study criteria on patients undergoing CAG in Western Denmark (2004–11). Both COMPASS eligible and non-eligible patients had CAD/PAD and met no exclusion criteria, but only COMPASS eligible patients met the inclusion criteria. We assessed the COMPASS primary endpoint of cardiovascular death, ischaemic stroke, haemorrhagic stroke, or myocardial infarction (MI). We computed event rates and adjusted incidence rate ratios (aIRRs). Of 80 071 patients undergoing CAG, 27 939 did not have CAD or PAD and were not considered. Of the 52 132 patients remaining, 11 930 were COMPASS eligible. Rates of the primary endpoint were 4.8 (95% confidence interval 4.6–5.0) events per 100 person-years among COMPASS eligible patients and 2.3 (2.2–2.4) among COMPASS non-eligible patients [aIRR 1.7 (1.6–1.9)]. COMPASS eligible patients also had higher risks of cardiovascular death [aIRR 2.5 (2.1–3.0)], ischaemic stroke [aIRR 1.4 (1.2–1.6)], and MI [aIRR 1.9 (1.7–2.1)]. Conclusion In this all-comers CAG cohort, 15% were eligible for combined aspirin and rivaroxaban treatment. COMPASS eligible patients had up to 2.5-fold higher rates of cardiovascular events than non-eligible patients. The higher incidence of ischaemic events in COMPASS eligible patients highlights an unmet need for additional preventive measures.