Association between arterial access site and anticoagulation strategy on major bleeding and mortality: A historical cohort analysis in the Veteran population

Background: Studies have shown reduction in major bleeding with trans-radial intervention (TRI) compared with trans-femoral intervention (TFI), and with use of bivalirudin compared with heparin + glycoprotein IIb/IIIa inhibitors (GPI). We compared major bleeding, mortality and the interaction between arterial access site and the anticoagulant used for PCI in Veterans. Methods: A retrospective cohort of 1192 consecutive patients who underwent PCI at a VA hospital between 2006 and 2012 was divided into TFI-heparin (n = 192), TFI-bivalirudin (n = 272), TRI-heparin (n = 274) and TRI-bivalirudin (n = 454) groups. Primary outcomes were in-hospital major bleeding, in-hospital and 1-year all-cause mortality. Secondary outcomes were in-hospital MI, in-hospital and 1-year MACE and net adverse cardiovascular events (NACE - composite of major bleeding + MACE). Results: Femoral access was associated with a significantly increased risk of major bleeding compared with radial access (OR 11.87, p < 0.001). Correspondingly, radial access was protective against major bleeding compared with femoral access (OR 0.128, p < 0.01), but did not lower mortality or MACE by itself. Severe anemia was the only predictor of in-hospital all-cause mortality (OR = 27.62, p < 0.008). Presence of anemia and age > 70 predicted 1-year mortality, whereas major bleeding and anemia predicted 1-year MACE. An interaction was noted between anticoagulant and access site, such that heparin showed significantly greater major bleeding in the femoral group compared with the radial group. Bivalirudin resulted in similar risk of bleeding, regardless of access site. There was a synergistic interaction between radial access and heparin (HR 0.38, p < 0.05), but not radial access and bivalirudin, on reduction in 1-year NACE. Conclusion: Radial access for PCI is associated with reduction in major bleeding, but does not have an effect on in-patient or 1-year MACE and mortality. Major bleeding is associated with poor short and intermediate term outcomes. In addition, anemia is strongly associated with increased in-patient and 1-year mortality. There is a differential effect of heparin but not bivalirudin on major bleeding, depending on the access site. There is no synergism between radial access and bivalirudin in lowering the composite outcome of MACE and major bleeding at 1 year.