Risk of Acute Kidney Injury After Percutaneous Coronary Interventions Using Radial Versus Femoral Vascular Access

Abstract

Transradial percutaneous coronary intervention (PCI [TRI]) does not involve catheter manipulation in the descending aorta, whereas transfemoral PCI (TFI) does. Therefore, the risk of acute kidney injury (AKI) after PCI might be influenced by vascular access site. We compared risks of AKI and nephropathy requiring dialysis (NRD) among patients treated with TRI and TFI. We included patients across 47 hospitals in Michigan. Primary end point was AKI (serum creatinine increase ≥0.5 mg/dL). Secondary end points were NRD and postprocedural bleeding. Odds ratios (OR) for study end points were calculated for the entire and propensity-matched population, reported as crude, and values adjusted for preprocedural calculated AKI risk. Between 2010 and 2012, a total of 82 225 PCI procedures were performed, of which 8915 were TRI. After adjustment, TRI was associated with a reduction in AKI (OR, 0.76, 95% confidence intervals [0.62-0.92]) and bleeding with a trend toward lower NRD risk. The propensity-matched population consisted of 8857 procedures per group. In this population, TRI was associated with lower adjusted odds of AKI (OR, 0.74; 95% confidence intervals [0.58-0.96]), and bleeding (OR, 0.47; 95% confidence intervals [0.36-0.63]), but no difference in NRD was observed. Although postprocedural bleeding was independently associated with AKI (OR, 2.86; 95% confidence intervals [1.75-4.66]) in the propensity-matched population, the lower odds of AKI was not mediated by a reduction in bleeding with TRI. Sensitivity analysis demonstrated that the observed association between access site and AKI could potentially be explained by a moderately strong unknown confounder. The risk of AKI was significantly lower after TRI compared with TFI. This finding needs to be evaluated in randomized controlled trials.