Secondary Logo

Journal Logo


Retrospective Observational Review of Percutaneous Cannulation for Extracorporeal Membrane Oxygenation

Burns, Janis; Cooper, Eve; Salt, Gavin; Gillon, Stuart; Camporota, Luigi; Daly, Kathleen; Barrett, Nicholas A.

Author Information
doi: 10.1097/MAT.0000000000000339
  • Free


The use of venovenous extracorporeal membrane oxygenation (VV-ECMO) for patients with severe respiratory failure has increased significantly1 since the publication of the Conventional ventilatory support versus Extracorporeal membrane oxygenation for Severe Adult Respiratory Failure (CESAR) study2 and the 2010 influenza A H1N1 pandemic.3,4 Extracorporeal membrane oxygenation is now commonly provided using miniaturized centrifugal pumps with low-resistance membrane gas exchangers and heparin-bonded circuit components, which have contributed significantly to improving outcomes.5 Miniaturization has also facilitated the development of centralized ECMO services by allowing the development of safe mobile ECMO.6 Cannulation can be achieved through either the open or the percutaneous approach. The open approach typically requires surgical exposure with subsequent venotomy and cannula insertion, followed by formal wound closure.7 Problems with the open approach are the increased risk of bleeding and infection, as well as the need for surgical decannulation.7 Percutaneous approaches to vessel cannulation were first described more than 60 years ago8 and were described for ECMO in the 1990s.9,10 Percutaneous techniques have previously been reported to be safe when carried out by both surgeons and intensivists using a variety of imaging modalities.6,11–14

The aim of the current study was to describe one approach to cannulation and its complications in a large cohort of adults managed with VV-ECMO for severe respiratory failure.


We conducted a retrospective, observational, cohort, electronic, case note review study in our tertiary referral intensive care unit (ICU) of all patients receiving VV-ECMO from December 2011 to March 2015. Extracorporeal membrane oxygenation referrals came predominantly from the South East England. Patients were identified from a prospectively maintained registry of all patients cannulated for VV-ECMO. Data were retrieved retrospectively from our electronic charting system (ICIP Philips, Amsterdam, Netherlands) and included demographics, mortality, cannula location, size, duration of cannulation, and complications of cannulation (bleeding requiring transfusion of 2 units of packed red blood cells or more, arterial injury, venous insufficiency, death, rupture of the great vessels, cardiac tamponade, and infection). Institutional approval was given for this study, and the requirement for consent was waived (Institutional governance reference: 4272).

Patients received VV-ECMO if they met the English national respiratory ECMO criteria: a lung injury (Murray et al.15) score of 3 or more; less than 7 days conventional mechanical ventilation; and a cause of respiratory failure deemed reversible by the treating physician. Patients receiving VV-ECMO all had either Rotaflow (Maquet, Rastatt, Germany) or Cardiohelp (Maquet, Rastatt, Germany) consoles with a respective permanent life support (PLS) or heart-lung support (HLS) circuit both with centrifugal pumps: Quadrox membrane gas exchangers and Bioline-coated circuits (Maquet, Rastatt, Germany). All cannulations were performed by intensive care consultants with specific ECMO cannulation training or by senior intensive care trainees directly observed and supervised by an intensive care consultant.

Our standard cannula configuration is a multistage drainage cannula and a single-stage infusion cannula (Bio-Medicus; Medtronic, Minneapolis, MN) via the femoral veins. Alternative configurations include a femoral multistage drainage cannula with a right internal jugular vein single-stage infusion or a right internal jugular vein dual-lumen cannula (Maquet). Before cannulation, our usual practice is to give 5 mg/kg gentamicin and 400 mg teicoplanin as antibiotic prophylaxis. Any hair is removed from the cannula insertion site, and the site is cleaned using 4% chlorhexidine with 70% alcohol; the patient is subsequently covered with a full sterile drape (Angio Drape; Kimal, Dormagen, Germany). The vessels are assessed with ultrasound, and the diameter of the chosen cannula is no more than two-third of the vessel diameter (external cannula diameter is French gauge [Fr] divided by three). The vessel is then cannulated using the Seldinger technique with either out-of-plane or in-plane ultrasound guidance to reduce the risk of injury to other vessels. Once the needle is giving a free flow of venous blood, a J-tipped guidewire (Amplatz Extra Stiff; Cook Vascular, Vandergrift, PA) is inserted. Using fluoroscopy, the wire is followed up into the inferior vena cava (IVC) to the level of the right atrium. Once both wires are in place, 50 units/kg of unfractionated heparin is administered intravenously, unless there is a contraindication to anticoagulation. The skin and soft tissues are then dilated using serial tapered dilators (6, 8, 10, 12, 14, 16, 18, 20, 22, and 24 Fr) (Coon’s dilators; Cook Vascular, Vandergrift, PA). A small skin incision is performed to facilitate cannula insertion if required; although this is infrequently required. The cannulae are positioned using fluoroscopy to identify tip position. The tip of a femoral drainage cannula is positioned in the mid-hepatic IVC and the tip of a femoral infusion cannula at the IVC/right atrial junction or in the mid-right atrium. When using a single-stage jugular infusion cannula, it is positioned at the superior vena cava (SVC)/right atrial junction or in the mid-right atrium. Each cannula is then flushed with 50 ml of 10 unit/ml heparinized 0.9% saline. Final adjustments to the position of each cannula are made after injection of 50 ml contrast media under continuous fluoroscopy. The technique for jugular cannulation using a dual-lumen single cannula is modified by ensuring that the guidewire is directed through the SVC and the right atrium and into the IVC under continuous fluoroscopy. The dual-lumen cannula is then placed under fluoroscopy with contrast used to demonstrate infusion port position. After insertion of the dual-lumen cannula, fine adjustments are performed using transthoracic echocardiography. All cannulae are secured using multiple 1-0 silk ties (minimum 3 points) and dressed with chlorhexidine sponge discs (Biopatch; Ethicon, Somerville, New Jersey) and occlusive dressings.


Descriptive statistics were performed using Excel (Microsoft Corporation, Redmond, Washington).


There were 348 cannulae placed in 179 patients admitted for VV-ECMO for a variety of indications from December 2011 to March 2015 (Table 1). Patients commencing ECMO had a mean (SD) age of 44.4 (14.4) years, a median (IQR) age of 45 (34–56) years, and 57% were male. The majority of patients (149/179) were cannulated at the referring hospital by our mobile ECMO team consisting of a consultant intensivist, ECMO nurse specialist, and perfusionist. The mean (SD) duration of ECMO was 15.1 (24) days, with a median (IQR) duration of ECMO of 9 (6–15) days. Overall survival to ECMO decannulation discharge was 79.9% (143/179). Six-month survival was 71.5% (128/179). Of the 348 cannulae inserted (Table 2), the most common drainage cannula was a 60 cm 25 Fr multistage, and the most common infusion cannula was a 50 cm 23 Fr single stage. Cannulation was successful in all cases. Cannula configuration was selected by the intensive care consultant performing the cannulation and was based on patient’s body habitus, venous anatomy, and the presence of local contraindications to cannulation. The median blood flow was 4.5 L/minute (IQR: 4.0–5.1). The blood flow varied depending on the size of drainage cannula (Table 3).

Table 1.
Table 1.:
Demographic Data of Entire Cohort
Table 2.
Table 2.:
Type, Size, and Location of Cannulae
Table 3.
Table 3.:
Blood Flow Through Drainage Cannulae

There were no deaths as a direct result of complications of cannulation. Complications recorded at insertion included one cardiac tamponade, secondary to a guidewire perforating the right ventricle (this occurred in case 15). During cannulation, the wire was noted to be curled in the right ventricle and was withdrawn. There was no evidence of contrast extravasation after cannulation. The tamponade was diagnosed 6 hours later with the evidence of cardiogenic shock and echocardiographic features of tamponade. The tamponade was managed using a percutaneous pericardial drain, cessation of anticoagulation, and return of the blood to the patient after cell salvage. The drain was removed 24 hours later, and there were no lasting adverse sequelae. In addition, there was one arterial injury to a branch of the femoral artery that crossed in front of the femoral vein that required replacement of the cannula and formal repair of the artery; there were no lasting adverse sequelae. No other complications were recorded related to cannula insertion, including the rupture of the great veins. There was no significant hemorrhage requiring transfusion related to cannulation. No cannula acutely thrombosed during cannulation. Four cannulae required subsequent replacement, one to manage the arterial injury, two for venous insufficiency within the first 24 hours of cannulation, and one for severe recirculation in a patient with a scoliosis. In this case, the drainage cannula was placed femorally (tip in the hepatic IVC) and the infusion cannula in the right internal jugular. Significant recirculation occurred with this configuration and was resolved by changing to a femorofemoral configuration. Cannula site infections occurred in 1.4% (5/348), all of which were managed successfully with intravenous antibiotics. The site infections were not associated with signs of systemic sepsis. There were no bloodstream infections secondary to cannula infection on cultures drawn from circuits where cannula infections were clinically apparent.


Percutaneous cannulation for VV-ECMO using ultrasound and fluoroscopic guidance undertaken by consultant intensivists is a successful technique with minimal complications. The high level of success and low complication rate reported in the current study is comparable with other reported series describing percutaneous cannulation.6,11,12,14 The current study reinforces the safety of percutaneous cannulation for VV-ECMO.

Only one case of arterial injury and two cases of venous insufficiency subsequent to cannulation were reported in the current cohort. One of the key components of the cannulation strategy was to use either in-plane or out-of-plane ultrasound (Figure 1) for all cases. Ultrasound has been previously shown to reduce the complications of central venous catheter insertion in the general ICU population.16 Although the prevalence of arterial and venous complications is low in the literature, previous studies where ultrasound was used11 do appear to have a comparable vascular complication rate with the current series and a lower rate than studies where ultrasound was not used.14 Another element that may have assisted in producing the low complication rate was the use of fluoroscopy. Fluoroscopy allowed manipulation of the wire to ensure that it remained in the intended vessel. In our experience, there is a possibility of the wire passing into a small side branch, particularly in the iliac vessels, rather than passing into the IVC. Fluoroscopy immediately identified this complication and allowed repositioning of the wire into the correct location and thereby prevention of vessel rupture.17 Finally, local damage to the vein at the site of cannulation was avoided by the combination of a nitrile-stiffened guidewires and tapered dilators. When used together, in our experience, they reduce the likelihood of guidewire kinking and subsequent perforation of the vessel wall. This modification of the cannulation technique requires equipment that is not in the prepacked cannula insertion kits supplied by manufacturers.

Figure 1.
Figure 1.:
Ultrasound images of femoral vessels: (A) out-of-plane ultrasound of femoral vessels; (B) in-plane ultrasound of femoral vein; (C) in-plane ultrasound demonstrating guidewire within femoral vein.

In our opinion, tapered dilators also add to the safety of this approach, as less force is required to dilate the soft tissues overlying the vein and the nitrile-stiffened wire is resistant to kinking, resulting in a lower likelihood of forming false tracts. A flat approach to cannulation also reduces the risk of vascular damage by ensuring that the dilator is being used closer to the line of the vein rather than a more perpendicular, traditional approach.

There was no major cannula insertion associated with bleeding requiring transfusion either during or after cannulation, despite the use of heparin after guidewire insertion. This is in part because of the use of the approach described previously reducing the risk of major vessel injury.

Our approach using sequential tapered dilators means that the skin and soft tissues can be gradually dilated without the need to routinely incise the skin, and consequently, the skin sits snugly around the cannula and bleeding is minimized. Cannulations were also undertaken or directly observed by a small cohort of six intensivists. There is some evidence that outcomes for complex therapies relate to volume.18–20 It is possible that the outcomes seen in the current study are the result of the combination of high patient volume (179 patients in 40 months) and a small group of intensive care consultants performing cannulation. Although the only tamponade occurred early in this series, during case 15, it is not possible to say from the current data whether there is a learning curve for cannulation because complications were so few.

Although femoral cannulae have been historically associated with higher rates of infection than jugular or subclavian cannulae,21,22 only 1.4% of cannulae demonstrated local infection in the current study, which is similar to previous reports for percutaneously inserted ECMO cannulae.11 Furthermore, there was no evidence of cannula-related bloodstream infection despite the fact that they remained in situ for a mean of 15 days, with the longest cannula in situ for 231 days. We used strict aseptic technique, routinely gave prophylactic antibiotics, used 4% chlorhexidine with 70% alcohol for skin preparation, full bed drapes, and chlorhexidine-impregnated dressings, all of which, when used together as a bundle have been shown to reduce the catheter-related bloodstream infection rate.23 We have previously reported that in our institution, ECMO cannulation is associated with an incidence of cannula-associated deep vein thrombosis of 8/1,000 cannula days.24 Given the paucity of literature on ECMO cannula-associated deep vein thrombosis, it is not possible to know whether the cannulation approach has any impact on the incidence of this complication.

The current study has limitations common to retrospective observational studies, including selection bias, incomplete data sets, and the lack of strict protocolization of many aspects of care. Complications were only evident if they were recorded in the contemporaneous medical notes, and it is possible that additional complications occurred but were not recorded.


Percutaneous cannulation for VV-ECMO undertaken by intensivists is safe, with very few complications and a very high success rate, particularly when undertaken using ultrasound and fluoroscopic guidance and with strict aseptic precautions.


1. Paden ML, Conrad SA, Rycus PT, Thiagarajan RR: Extracorporeal Life Support Organization Registry Report 2012. ASAIO J 2013.59: 202210.
2. Peek GJ, Mugford M, Tiruvoipati R, et al.: Efficacy and economic assessment of conventional ventilatory support versus extracorporeal membrane oxygenation for severe adult respiratory failure (CESAR): A multicentre randomised controlled trial. Lancet 2009.374: 13511363.
3. Davies A, Jones D, Bailey M, et al.: Extracorporeal membrane oxygenation for 2009 influenza A(H1N1) acute respiratory distress syndrome. JAMA 2009.302: 18881895.
4. Noah MA, Peek GJ, Finney SJ, et al.: Referral to an extracorporeal membrane oxygenation center and mortality among patients with severe 2009 influenza A(H1N1). JAMA 2011.306: 16591668.
5. Brodie D, Bacchetta M: Extracorporeal membrane oxygenation for ARDS in adults. N Engl J Med 2011.365: 19051914.
6. Sherren P, Shepherd S, Glover G, et al.: Capabilities of a mobile extracorporeal membrane oxygenation service for severe respiratory failure delivered by intensive care specialists. Anaesthesia 2015.70: 707714.
7. Bartlett R, Roloff D, Custer J, Younger J, Hirschl R: Extracorporeal life support: The University of Michigan experience. JAMA 2000.283: 904908.
8. Seldinger SI: Catheter replacement of the needle in percutaneous arteriography; a new technique. Acta Radiol 1953.39: 368376.
9. Mair P, Hoermann C, Moertl M, Bonatti J, Falbesoner C, Balogh D: Percutaneous venoarterial extracorporeal membrane oxygenation for emergency mechanical circulatory support. Resuscitation 1996.33: 2934.
10. Moreno-Cabral R, Dembitsky W, Adamson R, Daily P: Percutaneous extracorporeal membrane oxygenation. Adv Card Surg 1994.5: 163179.
11. Conrad S, Grier L, Scott L, Green R, Jordan M: Percutaneous cannulation for extracorporeal membrane oxygenation by intensivists: A retrospective single-institution case series. Crit Care Med 2015.43: 10101015.
12. Ganslmeier P, Philipp A, Rupprecht L, et al.: Percutaneous cannulation for extracorporeal life support. Thorac Cardiovasc Surg 2011.59: 103107.
13. Lamb K, Hirose H, Cavarocchi N: Preparation and technical considerations for percutaneous cannulation for veno-arterial extracorporeal membrane oxygenation. J Card Surg 2013.28: 190192.
14. Pranikoff T, Hirschl RB, Remenapp R, Swaniker F, Bartlett RH: Venovenous extracorporeal life support via percutaneous cannulation in 94 patients. Chest 1999.115: 818822.
15. Murray JF, Matthay MA, Luce JM, Flick MR: An expanded definition of the adult respiratory distress syndrome. Am Rev Respir Dis 1988.138: 720723.
16. Vezzani A, Manca T, Vercelli A, Braghieri A, Magnacavallo A: Ultrasonography as a guide during vascular access procedures and in the diagnosis of complications. J Ultrasound 2013.16: 161170.
17. Trimlett R, Cordingley J, Griffiths M, et al.: A modified technique for insertion of dual lumen bicaval cannulae for venovenous extracorporeal membrane oxygenation. Intensive Care Med 2011.37: 10361037.
18. Harrison A: Assessing the relationship between volume and outcome in hospital services: Implications for service centralisation. Health Serv Manag Res 2012.25: 16.
19. Kahn J, Goss C, Heagerty P, Kramer A, O’Brien C, Rubenfeld G: Hospital volume and the outcomes of mechanical ventilation. N Engl J Med 2006.355: 4150.
20. McCrum M, Lipsitz S, Berry W, Jha A, Gawande A: Beyond volume: Does hospital complexity matter? An analysis of inpatient surgical mortality in the United States. Med Care 2014.52: 235242.
21. Ge X, Cavallazzi R, Li C, Pan S, Wang Y, Wang F: Central venous access sites for the prevention of venous thrombosis, stenosis and infection. Cochrane Database Syst Rev 2012.3: CD004084.
22. Merrer J, De Jonghe B, Golliot F, et al.: Complications of femoral and subclavian venous catheterization in critically ill patients: A randomized controlled trial. JAMA 2001.286: 700707.
23. Entesari-Tatafi D, Orford N, Bailey M, Chonghaile M, Lamb-Jenkins J, Athan E: Effectiveness of a care bundle to reduce central line-associated bloodstream infections. Med J Aust 2015.202: 247249.
24. Cooper E, Burns J, Retter A, et al.: Prevalence of venous thrombosis following venovenous extracorporeal membrane oxygenation in patients with severe respiratory failure. Crit Care Med 2015.43: e581e584.

venovenous ECMO; percutaneous cannulation complications; severe respiratory failure; ARDS

Copyright © 2016 by the American Society for Artificial Internal Organs