LAUSR

FA access was performed in three arteries because of prior common FA interposition grafts. There were 18 (6%) access site complications (8 bleeding, 9 thrombosis, 1 infection) leading to immediate (n 1⁄4 12) or delayed (n 1⁄4 5) conversion to cutdown. Factors independently associated with percutaneous access site complication were sheath outer diameter to FA diameter ratio (odds ratio [OR], 58.9; 95% confidence interval [CI], 3.5-985.8; P 1⁄4 .005), morbid obesity with a body mass index of >40 kg/ m (OR, 16.0; 95% CI, 3.0-84.5; P 1⁄4 .001), FA stenosis of >50% (OR, 52.1; 95% CI, 7.7-351.5; P # . 001), FA stenosis of >75% (OR, 271.5; 95% CI, 15.14888.6; P # .001), and postoperative anticoagulation (OR, 10.5; 95% CI, 2.0-54.0; P 1⁄4 .005). A risk prediction model based on these criteria produced a C statistic of .91, a Hosmer-Lemeshow goodness of fit of 0.99, and a Brier score of .03. A risk score of >6 of 15 corresponded with a >10% probability of groin access complication (Table). Conclusions: Percutaneous ultrasound-guided access can be safely performed in almost all patients undergoing endovascular aortic procedures; however, access site failures still occur. Application of this risk score can help identify patients at high risk for complications after initial FA ultrasound-guided percutaneous access.