Comparison of vascular closure devices vs manual compression after femoral artery puncture: the ISAR-CLOSURE randomized clinical trial.

<bold>Importance: </bold>The role of vascular closure devices (VCD) for the achievement of hemostasis in patients undergoing transfemoral coronary angiography remains controversial.<bold>Objective: </bold>To compare outcomes with the use of 2 hemostasis strategies after diagnostic coronary angiography performed via transfemoral access-a VCD-based strategy with 2 types of devices, an intravascular device and an extravascular device, vs standard manual compression. The primary hypothesis to be tested was that femoral hemostasis achieved through VCD is noninferior to manual compression in terms of vascular access-site complications. A secondary objective was the comparison of the 2 types of VCD.<bold>Design, Setting, and Participants: </bold>Randomized, large-scale, multicenter, open-label clinical trial. We enrolled 4524 patients undergoing coronary angiography with a 6 French sheath via the common femoral artery from April 2011 through May 2014 in 4 centers in Germany. Last 30-day follow-up was performed in July 2014.<bold>Interventions: </bold>After angiography of the access site, patients were randomized to hemostasis with an intravascular VCD, extravascular VCD, or manual compression in a 1:1:1 ratio.<bold>Main Outcomes and Measures: </bold>Primary end point: the composite of access site-related vascular complications at 30 days after randomization with a 2% noninferiority margin. Secondary end points: time to hemostasis, repeat manual compression, and VCD failure. An α-level of .025 was chosen for primary and secondary comparisons.<bold>Results: </bold>Of the 4524 enrolled patients, 3015 were randomly assigned to a VCD group (1509 received intravascular VCD and 1506 received extravascular VCD) and 1509 patients were randomly assigned to the manual compression group. Before hospital discharge, duplex sonography of the access site was performed in 4231 (94%) patients. The primary end point was observed in 208 patients (6.9%) assigned to receive a VCD and 119 patients (7.9%) assigned to manual compression (difference, -1.0% [1-sided 97.5% CI, 0.7%]; P for noninferiority<.001). Time to hemostasis was significantly shorter in patients with VCD (1 minute [interquartile range {IQR}, 0.5-2.0]), vs manual compression (10 minutes [IQR, 10-15]; P < .001). Time to hemostasis was significantly shorter among patients with intravascular VCD (0.5 minute [IQR, 0.2-1.0]), vs extravascular VCD (2.0 minutes [IQR, 1.0-2.0]; P <.001) and closure device failure was also significantly lower among those with intravascular vs extravascular VCD (80 patients [5.3%], vs 184 patients [12.2%]; P < .001).<bold>Conclusions and Relevance: </bold>In patients undergoing transfemoral coronary angiography, VCDs were noninferior to manual compression in terms of vascular access-site complications and reduced time to hemostasis.<bold>Trial Registration: </bold>clinicaltrials.gov Identifier: NCT01389375. [ABSTRACT FROM AUTHOR]