- Email: [email protected]
Percutaneous removal using Perclose ProGlide closure devices versus surgical removal for weaning after percutaneous cannulation for venoarterial extracorporeal membrane oxygenation
Percutaneous removal using Perclose ProGlide closure devices versus surgical removal for weaning after percutaneous cannulation for venoarterial extracorporeal membrane oxygenation Ji-won Hwang, MD,a Jeong Hoon Yang, MD, PhD,a,b Kiick Sung, MD, PhD,c Young Bin Song, MD, PhD,a Joo-Yong Hahn, MD, PhD,a Jin-Ho Choi, MD, PhD,a Hyeon-Cheol Gwon, MD, PhD,a and Seung-Hyuk Choi, MD, PhD,a Seoul, Korea Objective: The removal of arterial cannulas using a Perclose device (Abbott Vascular, Santa Clara, Calif) has not been reported in patients undergoing venoarterial extracorporeal membrane oxygenation (ECMO). We investigated the procedural outcomes and complications of percutaneous device closure vs surgical repair for hemostatic control of the arterial access site in weaning from venoarterial ECMO. Methods: Between September 2012 and December 2014, 115 patients with ECMO weaned by percutaneous or surgical access were enrolled. The percutaneous technique used two ProGlide devices (Abbott Vascular) by direct puncture of an arterial cannula at the time of weaning off ECMO. The primary outcomes were composite complications of open repair at the insertion site, limb ischemia after removal of the arterial cannula, removal site infection, pseudoaneurysm, distal part embolization, and 10 minutes or more manual compression at the weaning site. Results: The percutaneous technique was performed on 56 patients, and the surgical exposure technique was performed on 59. Technical success was not significantly different between the percutaneous and surgical groups (85.7% vs 86.4%; P [ 1.0) although the procedure duration (17.15 6 9.38 minutes vs 64.33 6 31.67 minutes; P < .001) was shorter in the percutaneous access group. A composite of procedure-related complications and length of stay in the intensive care unit after weaning was not significantly different between groups (17.9% vs 28.8%; P [ .19 and 16.82 6 38.53 days vs 19.69 6 21.40 days; P [ .62). Conclusions: Percutaneous access using two Perclose ProGlide devices was a feasible and safe strategy for weaning from ECMO. (J Vasc Surg 2016;-:1-6.)
The percutaneous approach to endovascular repair of aortic pathology is becoming increasingly common. Suture-mediated closure devices (SMCDs) are often used for percutaneous management of small-diameter (6-10F) sheaths because they do not require prolonged manual compression or bed rest.1,2 The Perclose ProGlide (Abbott Vascular, Santa Clara, Calif) is an SMCD originally licensed for 5- to 8F access sites and was recently licensed to close large artery access sites using the “preclose technique.”3-5
From the Division of Cardiology, Department of Medicine,a Department of Critical Care Medicine,b and Department of Thoracic and Cardiovascular Surgery,c Samsung Medical Center, Sungkyunkwan University School of Medicine. Author conflict of interest: none. Additional material for this article may be found online at www.jvascsurg.org. Correspondence: Seung-Hyuk Choi, MD, PhD, Division of Cardiology, Department of Medicine, Cardiac and Vascular Center, Samsung Medical Center, Sungkyunkwan University School of Medicine, 81 Irwon-ro, Gangnam-gu, Seoul 135-710, Korea (e-mail: sh1214.choi@samsung. com). The editors and reviewers of this article have no relevant financial relationships to disclose per the JVS policy that requires reviewers to decline review of any manuscript for which they may have a conflict of interest. 0741-5214 Copyright Ó 2016 by the Society for Vascular Surgery. Published by Elsevier Inc. http://dx.doi.org/10.1016/j.jvs.2015.10.067
Venoarterial extracorporeal membrane oxygenation (ECMO) is mainly used for access of the femoral artery and vein because the equipment can be placed quickly and easily in emergencies.6,7 Generally, because a largesize sheath of 14 to 20F is used for ECMO insertion, surgical vascular repair is used as a standard weaning strategy. Maintaining stable vital signs at weaning is difficult, however, because of a decrease in systemic vascular resistance from use of sedatives and a relatively long operating time for critically ill patients with peripheral ECMO. As an alternative method to gain access at weaning, we performed weaning ECMO as a postclosure technique with direct puncture to the arterial cannula using the Perclose ProGlide device. The hypothesis of the present study is that percutaneous removal may not be inferior to standard surgical removal for procedural outcomes and complications in critically ill patients who underwent ECMO. We compared procedural outcomes and complications between patients receiving percutaneous device closure and patients undergoing surgical repair for hemostatic control of the arterial access site when weaning from peripheral ECMO. METHODS Study population. We retrospectively reviewed our registry of ECMO patients between September 2012 and December 2014. Patients were enrolled if they presented 1
2 Hwang et al
JOURNAL OF VASCULAR SURGERY --- 2016
Fig. The proximal portion of the arterial cannula by Seldinger needle (yellow arrow) was directly punctured (A). A 0.035-inch coated guidewire (Curved Terumo Wire; Terumo, Tokyo, Japan) (yellow arrows) was placed through the arterial cannula (B). The arterial cannula (yellow arrow) was pulled back and removed as guidewire was left in place (C). The first ProGlide device was inserted through the guidewire, and suturing was performed (yellow arrow) (D). The guidewire was reinserted through the hole of the first ProGlide to insert a second ProGlide device for suturing again.
with acute cardiopulmonary failure that indicated ECMO. ECMO was performed when patients in shock were unresponsive to the administration of a vasopressor after correction for hypovolemia or hypoxemia or when arrest was prolonged or recurrent. Percutaneous cannulation was performed using the Seldinger technique. Cannula sizes were 14 to 20F for the femoral artery and 21 to 28F for the femoral vein. Conditions for exclusion were age <18 years of age, implantation of central ECMO, and death before weaning. The study was approved by the Institutional Review Board of Samsung Medical Center and waived the requirement for informed consent. Postclosure technique. Percutaneous access was performed using two Perclose ProGlide devices for removal of a cannula via the femoral artery. After direct puncture in the proximal portion of the arterial cannula and placing a 0.035-inch coated guidewire (Curved Terumo Wire; Terumo, Tokyo, Japan) through the arterial cannula, the ECMO arterial cannula was pulled back and removed by manual compression at the insertion site. The first ProGlide device was inserted through the guidewire, and suturing was performed. The guidewire was reinserted through the hole of the first ProGlide to insert a second ProGlide device for suturing again. A guidewire was then inserted through the hole of the second ProGlide, which was pulled back and removed. An assistant maintained pressure while knots were tightened to keep the guidewire within the artery. Once a second knot was tightened with a knot pusher
and after confirming adequate hemostasis, the wire was removed, and pressure was applied (Fig).2,4,8 Surgical access was through the common femoral artery, exposed using an incision at the cannulation site, and sutured. Definitions and outcomes. The technical success of the double ProGlide-assisted closure technique and surgical exposures was defined as hemostatic control; no sign of immediate adverse events such as additive manual compression, dissection, occlusion, or stenosis; and unimpaired limb perfusion at the arterial cannulation site without need for access site-related adjunctive surgical or endovascular procedures from hemorrhagic, infectious, or ischemic complications.2,9-14 The primary outcomes were composite complications of open repair at the insertion site, limb ischemia after removal of the arterial cannula, removal site infection, pseudoaneurysm, distal part embolization, or 10 minutes or more manual compression at the weaning site. Secondary outcomes were in-hospital mortality and length of intensive care unit (ICU) stay after weaning. Detailed clinical data during ECMO-inserted and laboratory and procedural parameters were obtained from medical record review. Statistical analysis. All values were presented as numbers with percentages for categorical variables and means with standard deviation for continuous variables. Comparisons between continuous variables were made using a t-test or Mann-Whitney U test, as appropriate. Categorical data were analyzed using the c2 test or Fisher exact
JOURNAL OF VASCULAR SURGERY Volume -, Number -
Hwang et al 3
Table I. Baseline clinical and laboratory characteristics
Age, years Gender (male) Body mass index, kg/m2 Body mass index > 25 Procedure duration, minutes Success of weaning Department to perform insertion Cardiologist Thoracic surgeon Same groin of A & V cannula Size of arterial cannula, F 14 15 16 17 18 20 Comorbidities Hypertension Diabetes mellitus Current smoker Chronic kidney disease Previous stroke Atrial fibrillation Type of shock Cardiogenic Respiratory failure Septic shock Hypovolemic shock Cardiopulmonary resuscitation at insertion Complications at insertion Continuous renal replacement therapy use Medication Heparin Aspirin P2Y12 inhibitors Warfarin Laboratory findings Prehemoglobin Posthemoglobin Platelet Prothrombin time, INR Activated clotting time Creatinine
Percutaneous removal (n ¼ 56)
Surgical removal (n ¼ 59)
P value
53.34 6 15.45 19 (33.9) 23.41 6 3.98 12 (21.4) 17.15 6 9.38 48 (85.7)
49.76 6 17.91 24 (40.7) 22.93 6 4.83 18 (30.5) 64.33 6 31.67 52 (88.1)
.26 .56 .56 .30 <.001 1.00 .46
31 (55.4) 25 (44.6) 40 (71.4)
28 (47.5) 31 (52.5) 27 (45.8)
2 4 35 13 2
(3.6) (7.1) (62.5) (23.2) (3.6) 0
3 8 26 14 7 1
(5.1) (13.6) (44.1) (23.7) (11.9) (1.7)
18 19 9 4 4 5
(32.1) (33.9) (16.1) (7.1) (7.1) (8.9)
19 12 9 4 3 10
(32.2) (20.3) (15.3) (6.8) (5.1) (16.9)
44 4 7 1 26 2 19
(78.6) (7.1) (12.5) (1.8) (46.4) (3.6) (33.9)
44 4 8 3 25 10 21
(74.6) (6.8) (13.6) (5.1) (42.4) (16.9) (35.6)
.71 .03 1.00
48 24 23 1
(85.7) (42.9) (41.1) (1.8)
49 14 11 1
(83.1) (23.7) (18.6) (1.7)
.80 .047 .014 1.00
10.77 11.85 112.09 1.21 150.50 1.32
6 6 6 6 6 6
1.20 2.86 60.27 0.26 34.38 0.90
10.87 10.70 91.71 1.41 147.07 1.26
6 6 6 6 6 6
1.63 2.18 52.97 0.51 3617 0.78
.008 .24
1.00 .14 .90 1.00 .71 .27 .88
.70 .017 .056 .008 .70 .72
INR, International normalized ratio. For laboratory findings, pre means prestate of ECMO weaning, and post means poststate of ECMO weaning. Values are mean 6 standard deviation or number (%).
test, as appropriate. All tests were two-tailed, and P < .05 was considered statistically significant. SPSS version 20 (IBM, Armonk, NY) was used for statistical analysis. RESULTS Patient characteristics. Between September 2012 and December 2014, 115 patients who underwent ECMO weaning by percutaneous or surgical access were analyzed, with 56 undergoing percutaneous and 59 receiving the surgical exposure technique. Baseline patient clinical characteristics and laboratory and procedural parameters for percutaneous and surgical access for weaning ECMO are in Table I and Supplementary Table I (online only). The incidence of the arterial and venous sheaths in the same
groins in the percutaneous access group were more than the surgical-access group (71.4% vs 45.8%; P ¼ .008). There was no significant difference in the rate of successful weaning between patients with the arterial and venous sheaths in the same groin (85.1%) and different groins (89.6%) (P ¼ .58) (Table I). We also evaluated the size of the arterial cannula used in two groups of percutaneous access and surgical access: 14F (n ¼ 2 [3.6%] vs 3 [5.1%]), 15F (n ¼ 4 [7.1%] vs 8 [13.6%]), 16F (n ¼ 35 [62.5%] vs 26 [44.1%]), 17F (n ¼ 13 [23.2%] vs 14 [23.7%]), 18F (n ¼ 2 [3.6%] vs 7 [11.9%]), and 20F (n ¼ 0 [0%] vs 1 [1.7%]), respectively (Table I). The 16F and 17F sizes were mostly used in both groups. The frequency of success of weaning was as
JOURNAL OF VASCULAR SURGERY --- 2016
4 Hwang et al
Table II. Overall clinical outcomes after venoarterial extracorporeal membrane oxygenation (ECMO)
In-hospital mortality Length of stay in ICU Length of stay in ICU after weaning Length of stay in hospital after weaning Overall complications Open repair in insertion site Limb ischemia Infection of removal site Pseudoaneurysm Distal part embolization Manual compression in weaning site
Percutaneous removal (n ¼ 56)
Surgical removal (n ¼ 59)
P value
14 (25.0) 23.82 6 41.12 16.82 6 38.53 42.51 6 74.33 8 (14.3) 1 (1.8) 2 (3.6) 0 1 (1.8) 1 (1.8) 5 (8.9)
15 (25.4) 28.10 6 25.25 19.69 6 21.40 50.50 6 89.89 7 (11.9) 4 (6.8) 3 (5.1) 2 (3.4) 0 0 2 (3.4)
1.00 .50 .62 .61 .79 .37 1.00 .50 .49 .49 .26
ICU, Intensive care unit. Values are mean 6 standard deviation or number (%).
follows for each size of arterial cannula in the percutaneous access group (n ¼ success cases, success rate): 14F (n ¼ 2; 100%), 15F (n ¼ 2; 50%), 16F (n ¼ 31; 88.6%), 17F (n ¼ 11; 84.6%), and 18F (n ¼ 2; 100%), respectively. The frequency of complications was as follows for each size of arterial cannula: 14F (n ¼ 0), 15 (n ¼ 3), 16F (n ¼ 6), 17F (n ¼ 4), 18F (n ¼ 2), and 20F (n ¼ 0), respectively (Supplementary Table II, online only). Patients in the percutaneous access group were more likely to receive aspirin (42.9% vs 23.7%; P ¼ .047) and P2Y12 inhibitors (41.1% vs 18.6%; P ¼ .014) than the surgical-access group. For laboratory findings, international normalized ratio was higher in the surgical group (1.41 6 0.51) than in the percutaneous group (1.21 6 0.26; P ¼ .008). Hemoglobin after catheter removal was higher in the percutaneous group (11.85 6 2.86) than in the surgical group (10.70 6 2.18; P ¼ .017). Procedure duration was shorter in the percutaneous access group than in the surgical group (17.15 6 9.38 minutes vs 64.33 6 31.67 minutes; P < .001). Success of ECMO weaning was similar between the two groups (85.7% for percutaneous vs 86.4% for surgery; P ¼ 1.0). Clinical outcome and complications. The prognosis and overall incidence of access site-related complications are shown in Table II. In-hospital mortality (23.2% vs 22%; P ¼ .86) and total length of ICU stay (23.82 6 41.12 days vs 28.10 6 25.25 days; P ¼ .50) were not significantly different between the groups. The length of ICU stay after ECMO weaning (16.82 6 38.53 days vs 19.69 6 21.40 days; P ¼ .62) and length of hospital stay after weaning (42.51 6 74.33 days vs 50.50 6 89.89 days; P ¼ .61) were also not significantly different. Overall complications for percutaneous and surgical access for ECMO weaning were not significantly different between the groups (14.3% vs 11.9%; P ¼ .79). Among complications, incidence of open repair at the insertion site, limb ischemia, and infection of the insertion site were numerically high in the surgical-access group, but the difference was not significant. The incidence of 10 minutes or more manual compression at the weaning site was higher in the percutaneous group, but the difference with the surgery group was not significant.
DISCUSSION We investigated the procedural outcomes of percutaneous device closure vs surgical repair for hemostatic control of arterial access site for weaning from peripheral ECMO. No significant differences between a percutaneous group and a surgical group were observed for rates of technical success, procedure-related complications, or length of ICU stay. However, procedure duration was shorter in the percutaneous access group. The Perclose ProGlide Suture-Mediated Closure System became available in 2004. It was developed for femoral artery closure after diagnostic cardiac catheterization or coronary artery intervention using 5F to 8F sheaths.9 For transfemoral coronary angiography, vascular closure devices are not inferior to manual compression measured by access-site complications and reduction in time to hemostatic control.15 In the percutaneous endovascular aortic aneurysm repair (PEVAR) trial, a preclose technique using ProGlide closure devices was safe and effective for closure of a 21F profile delivery system containing an integrated 19F introducer sheath. This method was not inferior to open femoral exposure.11 We applied our endovascular aortic aneurysm repair (EVAR) experience to remove ECMO arterial cannulas using a double ProGlide-assisted closure technique at an arterial access site and felt safe to proceed with larger sheath closures. When comparing ECMO insertion with EVAR, the large cannula was not immediately removed after ECMO insertion and the period with large cannula of ECMO placed was longer than EVAR. Thus, the preclose technique may be not useful for the removal of an arterial cannula at the time of weaning off ECMO. To decrease the invasiveness of SMCD procedures and groin-related wound complications, total percutaneous EVAR using SMCD or minimal incisions without femoral artery exposure have been developed.4 Operators can use SMCD to comfortably repair vessels. The use of closure devices correlates with reduced procedure time, less blood loss, increased patient comfort with less pain at the entry site, shorter time to ambulation after intervention, and decreased bed rest and hospital stay.12,16,17 Therefore, as a
JOURNAL OF VASCULAR SURGERY Volume -, Number -
simple alternative to ECMO weaning, we used a postclosure technique with direct puncture to the arterial cannula using the closure device as an application of the preclose technique in EVAR. Our study showed that percutaneous removal was not inferior to standard surgical removal for procedural outcomes and complications in critically ill patients who underwent ECMO. Percutaneous access reduced times to hemostasis and procedure completion but not the length of ICU stay. Surgical access including preparing surgical instruments and creating an incision at the access site might take more time than percutaneous access based on puncture only. We are uncertain why no significant differences were observed between groups for length of ICU stay and total hospitalization. Underlying comorbidities and disease severity in patients undergoing ECMO might explain the similar length of stay regardless of the ECMO weaning strategy. Hemoglobin monitored after weaning from ECMO was lower in the surgical group than in the percutaneous group. For the procedure duration and weaning method, this might mean more blood was lost during the surgical procedure, although most patients received red blood cell transfusions for blood loss from the oxygenator. In addition, complications during the initial insertion were higher in the surgical-access group. For access sites that initially had an injury at insertion, the simple method of two SMCDs at the physician’s discretion was not sufficient for hemostatic control and prevention of additional complications. These findings suggest that bleeding tendency and the presence of complications at insertion should be considered for choosing a weaning method. Technically, being thoroughly and appropriately skilled with the Perclose ProGlide device including using careful movements to optimize the stability of the knot was vital to the success of the percutaneous technique. In addition, the wire should always be kept in the common femoral artery until complete hemostasis was confirmed (in case of device malfunction or if using the third device for complete hemostasis was necessary). In our institution, all percutaneous access was performed by two operators (J.H.Y., S.-H.C.) with extensive experience in the preclose technique in EVAR procedures. To best of our knowledge, a double ProGlide-assisted closure strategy for peripheral ECMO has not been reported in a cohort thus far. Compared with surgical access, the double ProGlideassisted closure technique can be simple, safe, and effective in patients undergoing ECMO. Limitations. This study has several limitations. First, its design was nonrandomized, retrospective, and observational, so confounding factors might have affected the results. The sample size was relatively small, and the study did not have sufficient power to reveal a difference in events or attribute them to surgery or percutaneous access. In our study, 12 patients had procedure-related complications such as open repair for hemostasis, limb ischemia, and hematoma at insertion time. Among these 12 patients, 10 patients underwent surgical removal and the other two
Hwang et al 5
patients underwent percutaneous removal. These findings suggest that physicians may prefer surgical removal in patients with procedure-related complications at the time of insertion. Lee et al reported anterior or near-circumferential calcific disease, small iliofemoral arteries, proximal iliac occlusive disease, and a high femoral bifurcation as contraindications for the preclose method.2 Detailed evaluation of preprocedural (EVAR) cross-sectional imaging is necessary for selecting proper candidates for percutaneous femoral access.14 We did not fully evaluate these parameters using computed tomography or conventional angiography before removal of the arterial cannula. However, we routinely performed plain X-ray (both hips, anteroposterior) to evaluate the location of the femur head and the arterial cannula. Although a plain X-ray would show the cannula position in relation to the femoral head, we could not predict the punctured site of arterial cannula because we did not evaluate the femoral arteries with ultrasound before cannula removal. CONCLUSIONS Weaning from venoarterial ECMO using ProGlide closure devices was a feasible and safe strategy with minimal access-related complications in particular patients with smaller cannula ranged from 14F to 18F. The method was not inferior to standard open femoral exposure. Large-scale, prospective, randomized, controlled trials are needed to clarify the efficacy and safety of the double ProGlide-assisted closure strategy in patients with ECMO support. AUTHOR CONTRIBUTIONS Conception and design: JY, SC Analysis and interpretation: JWH, JY Data collection: JWH, JY, KS, YS, JYH, JC, HG Writing the article: JWH, JY Critical revision of the article: JY, KS, YS, JYH, JC, HG, SC Final approval of the article: JY, SC Statistical analysis: JWH, JY Obtained funding: JY, SC Overall responsibility: SC JWH and JY contributed equally to this article and share co-first authorship. REFERENCES 1. Lee WA, Brown MP, Nelson PR, Huber TS, Seeger JM. Midterm outcomes of femoral arteries after percutaneous endovascular aortic repair using the Preclose technique. J Vasc Surg 2008;47:919-23. 2. Lee WA, Brown MP, Nelson PR, Huber TS. Total percutaneous access for endovascular aortic aneurysm repair (“Preclose” technique). J Vasc Surg 2007;45:1095-101. 3. Patel R, Juszczak MT, Bratby MJ, Sideso E, Anthony S, Tapping CR, et al. Efficacy and safety of augmenting the preclose technique with a collagen-based closure device for percutaneous endovascular aneurysm repair. Cardiovasc Intervent Radiol 2015;38:821-6. 4. Dosluoglu HH, Cherr GS, Harris LM, Dryjski ML. Total percutaneous endovascular repair of abdominal aortic aneurysms using Perclose ProGlide closure devices. J Endovasc Ther 2007;14:184-8.
6 Hwang et al
5. Lonn L, Larzon T, Van Den Berg JC. From puncture to closure of the common femoral artery in endovascular aortic repair. J Cardiovasc Surg (Torino) 2010;51:791-8. 6. Cove ME, MacLaren G. Clinical review: mechanical circulatory support for cardiogenic shock complicating acute myocardial infarction. Crit Care 2010;14:235. 7. Ganslmeier P, Philipp A, Rupprecht L, Diez C, Arlt M, Mueller T, et al. Percutaneous cannulation for extracorporeal life support. Thorac Cardiovasc Surg 2011;59:103-7. 8. Ramponi F, Yan TD, Vallely MP, Wilson MK. Total percutaneous cardiopulmonary bypass with Perclose ProGlide. Interact Cardiovasc Thorac Surg 2011;13:86-8. 9. Griese DP, Reents W, Diegeler A, Kerber S, Babin-Ebell J. Simple, effective and safe vascular access site closure with the double-ProGlide preclose technique in 162 patients receiving transfemoral transcatheter aortic valve implantation. Catheter Cardiovasc Interv 2013;82: E734-41. 10. Leon MB, Piazza N, Nikolsky E, Blackstone EH, Cutlip DE, Kappetein AP, et al. Standardized endpoint definitions for Transcatheter Aortic Valve Implantation clinical trials: a consensus report from the Valve Academic Research Consortium. J Am Coll Cardiol 2011;57:253-69. 11. Nelson PR, Kracjer Z, Kansal N, Rao V, Bianchi C, Hashemi H, et al. A multicenter, randomized, controlled trial of totally percutaneous access versus open femoral exposure for endovascular aortic aneurysm repair (the PEVAR trial). J Vasc Surg 2014;59:1181-93.
JOURNAL OF VASCULAR SURGERY --- 2016
12. Torsello GB, Kasprzak B, Klenk E, Tessarek J, Osada N, Torsello GF. Endovascular suture versus cutdown for endovascular aneurysm repair: a prospective randomized pilot study. J Vasc Surg 2003;38:78-82. 13. Kim WH, Shin S, Ko YG, Hong MK, Jang Y, Choi D. Efficacy and safety of the preclose technique following percutaneous aortic stentgraft implantation. J Endovasc Ther 2013;20:350-5. 14. Al-Khatib WK, Zayed MA, Harris EJ, Dalman RL, Lee JT. Selective use of percutaneous endovascular aneurysm repair in women leads to fewer groin complications. Ann Vasc Surg 2012;26:476-82. 15. Schulz-Schupke S, Helde S, Gewalt S, Ibrahim T, Linhardt M, Haas K, et al. Comparison of vascular closure devices vs manual compression after femoral artery puncture: the ISAR-CLOSURE randomized clinical trial. JAMA 2014;312:1981-7. 16. Georgiadis GS, Antoniou GA, Papaioakim M, Georgakarakos E, Trellopoulos G, Papanas N, et al. A meta-analysis of outcome after percutaneous endovascular aortic aneurysm repair using different size sheaths or endograft delivery systems. J Endovasc Ther 2011;18: 445-59. 17. Schwartz BG, Burstein S, Economides C, Kloner RA, Shavelle DM, Mayeda GS. Review of vascular closure devices. J Invasive Cardiol 2010;22:599-607. Submitted Aug 14, 2015; accepted Oct 15, 2015.
Additional material for this article may be found online at www.jvascsurg.org.
JOURNAL OF VASCULAR SURGERY Volume -, Number -
Hwang et al 6.e1
Supplementary Table I (online only). Overall clinical outcomes after venoarterial extracorporeal membrane oxygenation (ECMO) between two groups as body mass index (BMI) (kg/m2) # 25 and BMI > 25
In-hospital mortality Length of stay in ICU Length of stay in ICU after weaning Length of stay in hospital after weaning Success of weaning Overall complications
BMI # 25 (n ¼ 85)
BMI > 25 (n ¼ 30)
P value
22 (25.9) 25.09 6 25.61 16.79 6 21.67 42.47 6 62.84 76 (89.4) 8 (9.4)
7 (23.3) 28.63 6 50.89 22.57 6 48.51 59.18 6 125.18 23 (76.7) 7 (23.3)
1.00 .50 .62 .38 .35 .06
ICU, Intensive care unit. Values are mean 6 standard deviation or number (%).
Supplementary Table II (online only). The frequency of overall complications by the size of arterial cannula Size of arterial cannula, F 14 15 16 17 18 20
Percutaneous removal complication cases/total cases (the complication rate), % 0/2 (0) 2/4 (50.0) 4/35 (11.4) 2/13 (15.4) 0/2 (0) 0/0
Surgical removal complication cases/total cases (the complication rate), % 0/3 1/8 2/26 2/14 2/7 0/1
(0) (12.5) (7.7) (14.3) (28.6) (0)
The percutaneous approach to endovascular repair of aortic pathology is becoming increasingly common. Suture-mediated closure devices (SMCDs) are often used for percutaneous management of small-diameter (6-10F) sheaths because they do not require prolonged manual compression or bed rest.1,2 The Perclose ProGlide (Abbott Vascular, Santa Clara, Calif) is an SMCD originally licensed for 5- to 8F access sites and was recently licensed to close large artery access sites using the “preclose technique.”3-5
From the Division of Cardiology, Department of Medicine,a Department of Critical Care Medicine,b and Department of Thoracic and Cardiovascular Surgery,c Samsung Medical Center, Sungkyunkwan University School of Medicine. Author conflict of interest: none. Additional material for this article may be found online at www.jvascsurg.org. Correspondence: Seung-Hyuk Choi, MD, PhD, Division of Cardiology, Department of Medicine, Cardiac and Vascular Center, Samsung Medical Center, Sungkyunkwan University School of Medicine, 81 Irwon-ro, Gangnam-gu, Seoul 135-710, Korea (e-mail: sh1214.choi@samsung. com). The editors and reviewers of this article have no relevant financial relationships to disclose per the JVS policy that requires reviewers to decline review of any manuscript for which they may have a conflict of interest. 0741-5214 Copyright Ó 2016 by the Society for Vascular Surgery. Published by Elsevier Inc. http://dx.doi.org/10.1016/j.jvs.2015.10.067
Venoarterial extracorporeal membrane oxygenation (ECMO) is mainly used for access of the femoral artery and vein because the equipment can be placed quickly and easily in emergencies.6,7 Generally, because a largesize sheath of 14 to 20F is used for ECMO insertion, surgical vascular repair is used as a standard weaning strategy. Maintaining stable vital signs at weaning is difficult, however, because of a decrease in systemic vascular resistance from use of sedatives and a relatively long operating time for critically ill patients with peripheral ECMO. As an alternative method to gain access at weaning, we performed weaning ECMO as a postclosure technique with direct puncture to the arterial cannula using the Perclose ProGlide device. The hypothesis of the present study is that percutaneous removal may not be inferior to standard surgical removal for procedural outcomes and complications in critically ill patients who underwent ECMO. We compared procedural outcomes and complications between patients receiving percutaneous device closure and patients undergoing surgical repair for hemostatic control of the arterial access site when weaning from peripheral ECMO. METHODS Study population. We retrospectively reviewed our registry of ECMO patients between September 2012 and December 2014. Patients were enrolled if they presented 1
2 Hwang et al
JOURNAL OF VASCULAR SURGERY --- 2016
Fig. The proximal portion of the arterial cannula by Seldinger needle (yellow arrow) was directly punctured (A). A 0.035-inch coated guidewire (Curved Terumo Wire; Terumo, Tokyo, Japan) (yellow arrows) was placed through the arterial cannula (B). The arterial cannula (yellow arrow) was pulled back and removed as guidewire was left in place (C). The first ProGlide device was inserted through the guidewire, and suturing was performed (yellow arrow) (D). The guidewire was reinserted through the hole of the first ProGlide to insert a second ProGlide device for suturing again.
with acute cardiopulmonary failure that indicated ECMO. ECMO was performed when patients in shock were unresponsive to the administration of a vasopressor after correction for hypovolemia or hypoxemia or when arrest was prolonged or recurrent. Percutaneous cannulation was performed using the Seldinger technique. Cannula sizes were 14 to 20F for the femoral artery and 21 to 28F for the femoral vein. Conditions for exclusion were age <18 years of age, implantation of central ECMO, and death before weaning. The study was approved by the Institutional Review Board of Samsung Medical Center and waived the requirement for informed consent. Postclosure technique. Percutaneous access was performed using two Perclose ProGlide devices for removal of a cannula via the femoral artery. After direct puncture in the proximal portion of the arterial cannula and placing a 0.035-inch coated guidewire (Curved Terumo Wire; Terumo, Tokyo, Japan) through the arterial cannula, the ECMO arterial cannula was pulled back and removed by manual compression at the insertion site. The first ProGlide device was inserted through the guidewire, and suturing was performed. The guidewire was reinserted through the hole of the first ProGlide to insert a second ProGlide device for suturing again. A guidewire was then inserted through the hole of the second ProGlide, which was pulled back and removed. An assistant maintained pressure while knots were tightened to keep the guidewire within the artery. Once a second knot was tightened with a knot pusher
and after confirming adequate hemostasis, the wire was removed, and pressure was applied (Fig).2,4,8 Surgical access was through the common femoral artery, exposed using an incision at the cannulation site, and sutured. Definitions and outcomes. The technical success of the double ProGlide-assisted closure technique and surgical exposures was defined as hemostatic control; no sign of immediate adverse events such as additive manual compression, dissection, occlusion, or stenosis; and unimpaired limb perfusion at the arterial cannulation site without need for access site-related adjunctive surgical or endovascular procedures from hemorrhagic, infectious, or ischemic complications.2,9-14 The primary outcomes were composite complications of open repair at the insertion site, limb ischemia after removal of the arterial cannula, removal site infection, pseudoaneurysm, distal part embolization, or 10 minutes or more manual compression at the weaning site. Secondary outcomes were in-hospital mortality and length of intensive care unit (ICU) stay after weaning. Detailed clinical data during ECMO-inserted and laboratory and procedural parameters were obtained from medical record review. Statistical analysis. All values were presented as numbers with percentages for categorical variables and means with standard deviation for continuous variables. Comparisons between continuous variables were made using a t-test or Mann-Whitney U test, as appropriate. Categorical data were analyzed using the c2 test or Fisher exact
JOURNAL OF VASCULAR SURGERY Volume -, Number -
Hwang et al 3
Table I. Baseline clinical and laboratory characteristics
Age, years Gender (male) Body mass index, kg/m2 Body mass index > 25 Procedure duration, minutes Success of weaning Department to perform insertion Cardiologist Thoracic surgeon Same groin of A & V cannula Size of arterial cannula, F 14 15 16 17 18 20 Comorbidities Hypertension Diabetes mellitus Current smoker Chronic kidney disease Previous stroke Atrial fibrillation Type of shock Cardiogenic Respiratory failure Septic shock Hypovolemic shock Cardiopulmonary resuscitation at insertion Complications at insertion Continuous renal replacement therapy use Medication Heparin Aspirin P2Y12 inhibitors Warfarin Laboratory findings Prehemoglobin Posthemoglobin Platelet Prothrombin time, INR Activated clotting time Creatinine
Percutaneous removal (n ¼ 56)
Surgical removal (n ¼ 59)
P value
53.34 6 15.45 19 (33.9) 23.41 6 3.98 12 (21.4) 17.15 6 9.38 48 (85.7)
49.76 6 17.91 24 (40.7) 22.93 6 4.83 18 (30.5) 64.33 6 31.67 52 (88.1)
.26 .56 .56 .30 <.001 1.00 .46
31 (55.4) 25 (44.6) 40 (71.4)
28 (47.5) 31 (52.5) 27 (45.8)
2 4 35 13 2
(3.6) (7.1) (62.5) (23.2) (3.6) 0
3 8 26 14 7 1
(5.1) (13.6) (44.1) (23.7) (11.9) (1.7)
18 19 9 4 4 5
(32.1) (33.9) (16.1) (7.1) (7.1) (8.9)
19 12 9 4 3 10
(32.2) (20.3) (15.3) (6.8) (5.1) (16.9)
44 4 7 1 26 2 19
(78.6) (7.1) (12.5) (1.8) (46.4) (3.6) (33.9)
44 4 8 3 25 10 21
(74.6) (6.8) (13.6) (5.1) (42.4) (16.9) (35.6)
.71 .03 1.00
48 24 23 1
(85.7) (42.9) (41.1) (1.8)
49 14 11 1
(83.1) (23.7) (18.6) (1.7)
.80 .047 .014 1.00
10.77 11.85 112.09 1.21 150.50 1.32
6 6 6 6 6 6
1.20 2.86 60.27 0.26 34.38 0.90
10.87 10.70 91.71 1.41 147.07 1.26
6 6 6 6 6 6
1.63 2.18 52.97 0.51 3617 0.78
.008 .24
1.00 .14 .90 1.00 .71 .27 .88
.70 .017 .056 .008 .70 .72
INR, International normalized ratio. For laboratory findings, pre means prestate of ECMO weaning, and post means poststate of ECMO weaning. Values are mean 6 standard deviation or number (%).
test, as appropriate. All tests were two-tailed, and P < .05 was considered statistically significant. SPSS version 20 (IBM, Armonk, NY) was used for statistical analysis. RESULTS Patient characteristics. Between September 2012 and December 2014, 115 patients who underwent ECMO weaning by percutaneous or surgical access were analyzed, with 56 undergoing percutaneous and 59 receiving the surgical exposure technique. Baseline patient clinical characteristics and laboratory and procedural parameters for percutaneous and surgical access for weaning ECMO are in Table I and Supplementary Table I (online only). The incidence of the arterial and venous sheaths in the same
groins in the percutaneous access group were more than the surgical-access group (71.4% vs 45.8%; P ¼ .008). There was no significant difference in the rate of successful weaning between patients with the arterial and venous sheaths in the same groin (85.1%) and different groins (89.6%) (P ¼ .58) (Table I). We also evaluated the size of the arterial cannula used in two groups of percutaneous access and surgical access: 14F (n ¼ 2 [3.6%] vs 3 [5.1%]), 15F (n ¼ 4 [7.1%] vs 8 [13.6%]), 16F (n ¼ 35 [62.5%] vs 26 [44.1%]), 17F (n ¼ 13 [23.2%] vs 14 [23.7%]), 18F (n ¼ 2 [3.6%] vs 7 [11.9%]), and 20F (n ¼ 0 [0%] vs 1 [1.7%]), respectively (Table I). The 16F and 17F sizes were mostly used in both groups. The frequency of success of weaning was as
JOURNAL OF VASCULAR SURGERY --- 2016
4 Hwang et al
Table II. Overall clinical outcomes after venoarterial extracorporeal membrane oxygenation (ECMO)
In-hospital mortality Length of stay in ICU Length of stay in ICU after weaning Length of stay in hospital after weaning Overall complications Open repair in insertion site Limb ischemia Infection of removal site Pseudoaneurysm Distal part embolization Manual compression in weaning site
Percutaneous removal (n ¼ 56)
Surgical removal (n ¼ 59)
P value
14 (25.0) 23.82 6 41.12 16.82 6 38.53 42.51 6 74.33 8 (14.3) 1 (1.8) 2 (3.6) 0 1 (1.8) 1 (1.8) 5 (8.9)
15 (25.4) 28.10 6 25.25 19.69 6 21.40 50.50 6 89.89 7 (11.9) 4 (6.8) 3 (5.1) 2 (3.4) 0 0 2 (3.4)
1.00 .50 .62 .61 .79 .37 1.00 .50 .49 .49 .26
ICU, Intensive care unit. Values are mean 6 standard deviation or number (%).
follows for each size of arterial cannula in the percutaneous access group (n ¼ success cases, success rate): 14F (n ¼ 2; 100%), 15F (n ¼ 2; 50%), 16F (n ¼ 31; 88.6%), 17F (n ¼ 11; 84.6%), and 18F (n ¼ 2; 100%), respectively. The frequency of complications was as follows for each size of arterial cannula: 14F (n ¼ 0), 15 (n ¼ 3), 16F (n ¼ 6), 17F (n ¼ 4), 18F (n ¼ 2), and 20F (n ¼ 0), respectively (Supplementary Table II, online only). Patients in the percutaneous access group were more likely to receive aspirin (42.9% vs 23.7%; P ¼ .047) and P2Y12 inhibitors (41.1% vs 18.6%; P ¼ .014) than the surgical-access group. For laboratory findings, international normalized ratio was higher in the surgical group (1.41 6 0.51) than in the percutaneous group (1.21 6 0.26; P ¼ .008). Hemoglobin after catheter removal was higher in the percutaneous group (11.85 6 2.86) than in the surgical group (10.70 6 2.18; P ¼ .017). Procedure duration was shorter in the percutaneous access group than in the surgical group (17.15 6 9.38 minutes vs 64.33 6 31.67 minutes; P < .001). Success of ECMO weaning was similar between the two groups (85.7% for percutaneous vs 86.4% for surgery; P ¼ 1.0). Clinical outcome and complications. The prognosis and overall incidence of access site-related complications are shown in Table II. In-hospital mortality (23.2% vs 22%; P ¼ .86) and total length of ICU stay (23.82 6 41.12 days vs 28.10 6 25.25 days; P ¼ .50) were not significantly different between the groups. The length of ICU stay after ECMO weaning (16.82 6 38.53 days vs 19.69 6 21.40 days; P ¼ .62) and length of hospital stay after weaning (42.51 6 74.33 days vs 50.50 6 89.89 days; P ¼ .61) were also not significantly different. Overall complications for percutaneous and surgical access for ECMO weaning were not significantly different between the groups (14.3% vs 11.9%; P ¼ .79). Among complications, incidence of open repair at the insertion site, limb ischemia, and infection of the insertion site were numerically high in the surgical-access group, but the difference was not significant. The incidence of 10 minutes or more manual compression at the weaning site was higher in the percutaneous group, but the difference with the surgery group was not significant.
DISCUSSION We investigated the procedural outcomes of percutaneous device closure vs surgical repair for hemostatic control of arterial access site for weaning from peripheral ECMO. No significant differences between a percutaneous group and a surgical group were observed for rates of technical success, procedure-related complications, or length of ICU stay. However, procedure duration was shorter in the percutaneous access group. The Perclose ProGlide Suture-Mediated Closure System became available in 2004. It was developed for femoral artery closure after diagnostic cardiac catheterization or coronary artery intervention using 5F to 8F sheaths.9 For transfemoral coronary angiography, vascular closure devices are not inferior to manual compression measured by access-site complications and reduction in time to hemostatic control.15 In the percutaneous endovascular aortic aneurysm repair (PEVAR) trial, a preclose technique using ProGlide closure devices was safe and effective for closure of a 21F profile delivery system containing an integrated 19F introducer sheath. This method was not inferior to open femoral exposure.11 We applied our endovascular aortic aneurysm repair (EVAR) experience to remove ECMO arterial cannulas using a double ProGlide-assisted closure technique at an arterial access site and felt safe to proceed with larger sheath closures. When comparing ECMO insertion with EVAR, the large cannula was not immediately removed after ECMO insertion and the period with large cannula of ECMO placed was longer than EVAR. Thus, the preclose technique may be not useful for the removal of an arterial cannula at the time of weaning off ECMO. To decrease the invasiveness of SMCD procedures and groin-related wound complications, total percutaneous EVAR using SMCD or minimal incisions without femoral artery exposure have been developed.4 Operators can use SMCD to comfortably repair vessels. The use of closure devices correlates with reduced procedure time, less blood loss, increased patient comfort with less pain at the entry site, shorter time to ambulation after intervention, and decreased bed rest and hospital stay.12,16,17 Therefore, as a
JOURNAL OF VASCULAR SURGERY Volume -, Number -
simple alternative to ECMO weaning, we used a postclosure technique with direct puncture to the arterial cannula using the closure device as an application of the preclose technique in EVAR. Our study showed that percutaneous removal was not inferior to standard surgical removal for procedural outcomes and complications in critically ill patients who underwent ECMO. Percutaneous access reduced times to hemostasis and procedure completion but not the length of ICU stay. Surgical access including preparing surgical instruments and creating an incision at the access site might take more time than percutaneous access based on puncture only. We are uncertain why no significant differences were observed between groups for length of ICU stay and total hospitalization. Underlying comorbidities and disease severity in patients undergoing ECMO might explain the similar length of stay regardless of the ECMO weaning strategy. Hemoglobin monitored after weaning from ECMO was lower in the surgical group than in the percutaneous group. For the procedure duration and weaning method, this might mean more blood was lost during the surgical procedure, although most patients received red blood cell transfusions for blood loss from the oxygenator. In addition, complications during the initial insertion were higher in the surgical-access group. For access sites that initially had an injury at insertion, the simple method of two SMCDs at the physician’s discretion was not sufficient for hemostatic control and prevention of additional complications. These findings suggest that bleeding tendency and the presence of complications at insertion should be considered for choosing a weaning method. Technically, being thoroughly and appropriately skilled with the Perclose ProGlide device including using careful movements to optimize the stability of the knot was vital to the success of the percutaneous technique. In addition, the wire should always be kept in the common femoral artery until complete hemostasis was confirmed (in case of device malfunction or if using the third device for complete hemostasis was necessary). In our institution, all percutaneous access was performed by two operators (J.H.Y., S.-H.C.) with extensive experience in the preclose technique in EVAR procedures. To best of our knowledge, a double ProGlide-assisted closure strategy for peripheral ECMO has not been reported in a cohort thus far. Compared with surgical access, the double ProGlideassisted closure technique can be simple, safe, and effective in patients undergoing ECMO. Limitations. This study has several limitations. First, its design was nonrandomized, retrospective, and observational, so confounding factors might have affected the results. The sample size was relatively small, and the study did not have sufficient power to reveal a difference in events or attribute them to surgery or percutaneous access. In our study, 12 patients had procedure-related complications such as open repair for hemostasis, limb ischemia, and hematoma at insertion time. Among these 12 patients, 10 patients underwent surgical removal and the other two
Hwang et al 5
patients underwent percutaneous removal. These findings suggest that physicians may prefer surgical removal in patients with procedure-related complications at the time of insertion. Lee et al reported anterior or near-circumferential calcific disease, small iliofemoral arteries, proximal iliac occlusive disease, and a high femoral bifurcation as contraindications for the preclose method.2 Detailed evaluation of preprocedural (EVAR) cross-sectional imaging is necessary for selecting proper candidates for percutaneous femoral access.14 We did not fully evaluate these parameters using computed tomography or conventional angiography before removal of the arterial cannula. However, we routinely performed plain X-ray (both hips, anteroposterior) to evaluate the location of the femur head and the arterial cannula. Although a plain X-ray would show the cannula position in relation to the femoral head, we could not predict the punctured site of arterial cannula because we did not evaluate the femoral arteries with ultrasound before cannula removal. CONCLUSIONS Weaning from venoarterial ECMO using ProGlide closure devices was a feasible and safe strategy with minimal access-related complications in particular patients with smaller cannula ranged from 14F to 18F. The method was not inferior to standard open femoral exposure. Large-scale, prospective, randomized, controlled trials are needed to clarify the efficacy and safety of the double ProGlide-assisted closure strategy in patients with ECMO support. AUTHOR CONTRIBUTIONS Conception and design: JY, SC Analysis and interpretation: JWH, JY Data collection: JWH, JY, KS, YS, JYH, JC, HG Writing the article: JWH, JY Critical revision of the article: JY, KS, YS, JYH, JC, HG, SC Final approval of the article: JY, SC Statistical analysis: JWH, JY Obtained funding: JY, SC Overall responsibility: SC JWH and JY contributed equally to this article and share co-first authorship. REFERENCES 1. Lee WA, Brown MP, Nelson PR, Huber TS, Seeger JM. Midterm outcomes of femoral arteries after percutaneous endovascular aortic repair using the Preclose technique. J Vasc Surg 2008;47:919-23. 2. Lee WA, Brown MP, Nelson PR, Huber TS. Total percutaneous access for endovascular aortic aneurysm repair (“Preclose” technique). J Vasc Surg 2007;45:1095-101. 3. Patel R, Juszczak MT, Bratby MJ, Sideso E, Anthony S, Tapping CR, et al. Efficacy and safety of augmenting the preclose technique with a collagen-based closure device for percutaneous endovascular aneurysm repair. Cardiovasc Intervent Radiol 2015;38:821-6. 4. Dosluoglu HH, Cherr GS, Harris LM, Dryjski ML. Total percutaneous endovascular repair of abdominal aortic aneurysms using Perclose ProGlide closure devices. J Endovasc Ther 2007;14:184-8.
6 Hwang et al
5. Lonn L, Larzon T, Van Den Berg JC. From puncture to closure of the common femoral artery in endovascular aortic repair. J Cardiovasc Surg (Torino) 2010;51:791-8. 6. Cove ME, MacLaren G. Clinical review: mechanical circulatory support for cardiogenic shock complicating acute myocardial infarction. Crit Care 2010;14:235. 7. Ganslmeier P, Philipp A, Rupprecht L, Diez C, Arlt M, Mueller T, et al. Percutaneous cannulation for extracorporeal life support. Thorac Cardiovasc Surg 2011;59:103-7. 8. Ramponi F, Yan TD, Vallely MP, Wilson MK. Total percutaneous cardiopulmonary bypass with Perclose ProGlide. Interact Cardiovasc Thorac Surg 2011;13:86-8. 9. Griese DP, Reents W, Diegeler A, Kerber S, Babin-Ebell J. Simple, effective and safe vascular access site closure with the double-ProGlide preclose technique in 162 patients receiving transfemoral transcatheter aortic valve implantation. Catheter Cardiovasc Interv 2013;82: E734-41. 10. Leon MB, Piazza N, Nikolsky E, Blackstone EH, Cutlip DE, Kappetein AP, et al. Standardized endpoint definitions for Transcatheter Aortic Valve Implantation clinical trials: a consensus report from the Valve Academic Research Consortium. J Am Coll Cardiol 2011;57:253-69. 11. Nelson PR, Kracjer Z, Kansal N, Rao V, Bianchi C, Hashemi H, et al. A multicenter, randomized, controlled trial of totally percutaneous access versus open femoral exposure for endovascular aortic aneurysm repair (the PEVAR trial). J Vasc Surg 2014;59:1181-93.
JOURNAL OF VASCULAR SURGERY --- 2016
12. Torsello GB, Kasprzak B, Klenk E, Tessarek J, Osada N, Torsello GF. Endovascular suture versus cutdown for endovascular aneurysm repair: a prospective randomized pilot study. J Vasc Surg 2003;38:78-82. 13. Kim WH, Shin S, Ko YG, Hong MK, Jang Y, Choi D. Efficacy and safety of the preclose technique following percutaneous aortic stentgraft implantation. J Endovasc Ther 2013;20:350-5. 14. Al-Khatib WK, Zayed MA, Harris EJ, Dalman RL, Lee JT. Selective use of percutaneous endovascular aneurysm repair in women leads to fewer groin complications. Ann Vasc Surg 2012;26:476-82. 15. Schulz-Schupke S, Helde S, Gewalt S, Ibrahim T, Linhardt M, Haas K, et al. Comparison of vascular closure devices vs manual compression after femoral artery puncture: the ISAR-CLOSURE randomized clinical trial. JAMA 2014;312:1981-7. 16. Georgiadis GS, Antoniou GA, Papaioakim M, Georgakarakos E, Trellopoulos G, Papanas N, et al. A meta-analysis of outcome after percutaneous endovascular aortic aneurysm repair using different size sheaths or endograft delivery systems. J Endovasc Ther 2011;18: 445-59. 17. Schwartz BG, Burstein S, Economides C, Kloner RA, Shavelle DM, Mayeda GS. Review of vascular closure devices. J Invasive Cardiol 2010;22:599-607. Submitted Aug 14, 2015; accepted Oct 15, 2015.
Additional material for this article may be found online at www.jvascsurg.org.
JOURNAL OF VASCULAR SURGERY Volume -, Number -
Hwang et al 6.e1
Supplementary Table I (online only). Overall clinical outcomes after venoarterial extracorporeal membrane oxygenation (ECMO) between two groups as body mass index (BMI) (kg/m2) # 25 and BMI > 25
In-hospital mortality Length of stay in ICU Length of stay in ICU after weaning Length of stay in hospital after weaning Success of weaning Overall complications
BMI # 25 (n ¼ 85)
BMI > 25 (n ¼ 30)
P value
22 (25.9) 25.09 6 25.61 16.79 6 21.67 42.47 6 62.84 76 (89.4) 8 (9.4)
7 (23.3) 28.63 6 50.89 22.57 6 48.51 59.18 6 125.18 23 (76.7) 7 (23.3)
1.00 .50 .62 .38 .35 .06
ICU, Intensive care unit. Values are mean 6 standard deviation or number (%).
Supplementary Table II (online only). The frequency of overall complications by the size of arterial cannula Size of arterial cannula, F 14 15 16 17 18 20
Percutaneous removal complication cases/total cases (the complication rate), % 0/2 (0) 2/4 (50.0) 4/35 (11.4) 2/13 (15.4) 0/2 (0) 0/0
Surgical removal complication cases/total cases (the complication rate), % 0/3 1/8 2/26 2/14 2/7 0/1
(0) (12.5) (7.7) (14.3) (28.6) (0)