Comparison of percutaneous decannulation and open surgical repair for large-bore arterial access sites of extracorporeal membrane oxygenation.

Abstract

Following weaning of peripheral venoarterial extracorporeal membrane oxygenation (ECMO) support, removal of cannulas and repair of large-bore arterial sites is traditionally by open surgical repair (OSR). The use of a percutaneous vascular closure device (VCD) offers a minimally invasive alternative to OSR with potential for reduced operative complications, reduced hospital length of stay and in-hospital mortality. A systematic review of Medline and Embase databases was conducted for studies comparing VCD-assisted decannulation with OSR following decannulation of peripheral ECMO. The primary endpoint was rate of post-procedural complications, namely wound infection and limb ischaemia. The secondary endpoint was in-hospital mortality. Eight studies, with a total of 685 patients, met inclusion criteria. Forty-eight percent (n = 328) of patients underwent VCD-assisted decannulation with the remainder decannulated by OSR. Pooled analysis demonstrated that patients undergoing VCD-assisted decannulation are at significantly lower risk of overall complications than surgical repair (OR 4.34; 95% CI 2.19-8.57; p < .001). Specifically, there was a substantial decrease in groin infections (OR 6.77; 95% CI 3.07-14.97; p < .001) without notable differences in limb ischaemia or pseudoaneurysm occurrence between the VCD and OSR cohorts. There was a lower incidence of intra-operative bleeding and haematoma in the VCD cohort (p < .001 and p = .002, respectively). Vascular closure device-assisted decannulation of peripheral ECMO offers a significantly reduced risk of complications, particularly groin infections and bleeding. Future research should encompass larger cohorts, randomised controlled trials, cost-benefit analyses, and the training of surgeons, cardiologists and intensivists in VCD-assisted decannulation, potentially through the integration of simulation-based training.