Transradial Access for High-Risk Percutaneous Coronary Intervention: Implications of the Risk-Treatment Paradox

Abstract

Background: Transradial percutaneous coronary intervention (PCI; TRI) reduces adverse outcomes when compared with transfemoral PCI (TFI). However, TRI is also used less in high-risk patients. It remains unknown how baseline patient risk influences access-site choice among PCI operators and whether the absolute benefit of TRI is greater among patients at high risk for bleeding, acute kidney injury (AKI), and death. Methods: We analyzed 28 005 PCIs performed in a 7-hospital system between July, 01, 2009 and April 30, 2018, to assess the choice of access-site (TRI versus TFI) as a function of baseline risk for bleeding, AKI, and death, and examined whether the association between TRI use (versus TFI) and in-hospital outcomes is influenced by baseline risk. Results: Among 28 005 PCIs, over a 9-year period, TRI increased over time, however, a risk-treatment paradox for TRI use was observed not only for bleeding risk, but also AKI, and mortality risks, where TRI use was lower in those at highest risk. Operator variability with TRI was large. The incidences of bleeding, AKI, and death were higher with TFI versus TRI. The absolute risk difference between TRI and TFI increased with increasing baseline risk. The number needed to treat to prevent one adverse event with TRI (versus TFI) in low-, moderate- and high-risk groups, respectively, was 259, 82, and 32 for bleeding; 194, 53, and 40 for AKI; and 957, 78, and 18 for death. Conclusions: This analysis of a large multicenter cohort of patients with PCI demonstrates a risk-treatment paradox for TRI use, not only for bleeding, but also for AKI and death. Despite this, a greater absolute risk difference favoring TRI was observed among patients with the highest baseline risk. Addressing the risk-treatment paradox by preferentially selecting TRI across the spectrum of risk, but especially high-risk cases, may be an important potential strategy for improving outcomes with PCI.