BackgroundThe use of new oral anticoagulants (NOACs) in patients with impaired renal function has raised major concerns, in particular the possibility of an increased risk of bleeding due to accumulation. The aims of this work were to assess the safety of NOACs in patients with renal failure and describe the relationship between clinical events and drug renal excretion magnitude. MethodsAll phase III trials comparing NOACs with vitamin K antagonists (VKAs) in patients with estimated glomerular filtration (eGFR) rate <50mLmin(-1) were eligible. The main safety and efficacy outcomes were major bleeding and thrombosis. A meta-regression was performed to estimate the correlation between the treatment effect estimate and the percentage of renal excretion. ResultsNine studies (12272 patients) were included. A significantly greater relative reduction in major bleeding was seen for NOACs with renal excretion <50% (RR, 0.61; CI, 0.51-0.74) than for those with high renal excretion (RR, 0.96; CI, 0.85-1.07) (interaction test, P<0.0001). A linear relationship between the relative risk of major bleeding and the magnitude of renal excretion was found by meta-regression (R-2=0.66, P=0.03). For thrombosis, a greater treatment effect of NOA vs. INR-adjusted VKA was observed in patients with eGFR <50mLmin(-1) (RR 0.78, CI 0.67-0.92), but no correlation between treatment effect and renal excretion was found. ConclusionsNew oral anticoagulants were at least as effective as VKAs, with reduced risks of major bleeding and thrombosis in patients with eGFR <50mLmin(-1). The renal excretion of these new drugs seemed to modify the safety profile, contrary to the efficacy.