Predictors of percutaneous access-related complications in aortic endovascular procedures: 'real-world' insights and a comparison to open access
George Gradinariu,Oliver Lyons,Mustafa Musajee,Trixie Yap,Oscar Johnson,Iulia Bujoreanu,Joseph Shalhoub,Jason Wilkins,Panos Gkoutzios,Mark Tyrrell,Said Abisi,Bijan Modarai,Becky Sandford,Guy's and St Thomas' Vascular Research Collaborative,Rachel Bell,Lukla Biasi,Stephen Black,Michael Dialynas,Tommaso Donati,Sanjay Patel,Ashish Patel,Prakash Saha,Morad Sallam,Hany Zayed
DOI: https://doi.org/10.23736/S0392-9590.22.04799-X
Abstract:Background: Percutaneous endovascular aneurysm repair (PEVAR) is becoming increasingly popular due to fewer access-related complications, shorter procedural times and length of stay (LOS). Our aim was to explore factors associated with access-related complications and their impact on procedural time and LOS. Methods: We retrospectively analyzed consecutive aorto-iliac endovascular procedures in a tertiary hub comprising 2 institutions and 18 consultant vascular surgeons and interventional radiologists between 2016-2017. Access-related complications were defined as: bleeding requiring cutdown or return to theatre, acute limb ischemia or common femoral artery (CFA) pseudoaneurysm requiring intervention and wound infection or dehiscence needing hospitalization. Results: Of 511 patients, 354 (69%) had a percutaneous approach via 589 CFA access sites. In this percutaneous group, access-related complications occurred in 11% of sites (65/589); Their rate varied with procedure type ranging between 3.6% to 17.6%. The most common complication was bleeding due to closure device failure in 8.5% (50/589) of access sites. When uncomplicated, percutaneous interventions were faster compared to open surgical access (P<0.0001). Operation time and median LOS (3 vs. 2 days) were longer for elective standard EVAR patients experiencing access-related complications (P=0.033). In the percutaneous group, multivariate regression analysis demonstrated significant associations between access-related complications and eGFR (odds ratio (OR) 0.984 [0.972-0.997], P=0.014), CFA depth (OR 1.026 [1.008-1.045], P=0.005), device used (Prostar vs. Proglide (OR 2.177 [1.236-3.832], P=0.007) and procedural type (complex vs. standard EVAR) (OR 2.017 [1.122-3.627], P=0.019). We developed a risk score which had reasonably good predictive power (C-statistic 0.716 [0.646-0.787], P<0.0001) for avoiding access complications. Conclusions: Physiological (low eGFR level), anatomical (increased CFA depth) and technical factors (choice of device and complex procedures) were identified as predictors of access-related complications in this large retrospective series. These are important for safe selection of patients that would benefit from percutaneous access.