ECMO circuit configurations

The configuration of the ECMO circuit refers to the arrangement of the venous drainage and arterial return pathways, and is described by an increasingly confused nomenclature. This chapter is based mostly on Broman et al (2019), the paper announcing an agreement to name things consistently, which ELSO reached in 2019, and which was for some reason called "the Maastricht Treaty", like some kind of fearsome military pact between moustachioed imperialists in monocles and pickelhauben.  

In summary:

ECMO cannula configuration is ordered as follows:

  • Access cannulae in capitals
  • " - " to represent the membrane lung
  • Return cannulae in capitals
  • Cannulae capitalisation is:
    • V = venous
    • A = arterial 
    • P - pulmonary artery

When there are more cannulae:

  • An additional uppercase letter is placed accordingly to the outer side of the already existing cannulae, which constitute the core. For example:
    • VV-A is an additional drainage cannula
    • V-AV is where an additional venous return cannula is added to a V-A system to improve systemic oxygenation when the lungs go down but the cardiac output is improving
    • V-VA is where an arterial cannula is added to a V-V system to support the circulation when the heart goes down and the lungs are still terrible

When there are more, or fewer, oxygenators: 

  • " = " designates a second oxygenator in parallel,
    " + " designates a second oxygenator in series, 
    " x " designates a circuit without an oxygenator

Where two pumps are used simultaneously, 

  •  " / " separates the two device configurations,
    in chronological order of cannulation

There is also much more to it, but this level is probably already beyond what is expected from the CICM exam candidate. The reader is offered this brief glimpse of the level of complexity that can develop from the full classification system, which contains four levels of detail including cannula site , dimensions and tip position; but the image is intentionally left tiny and unreadable to discourage frivolously wasteful curiosity.

The table of ECMO nomenclature from the ELSO Maastricht Treaty for ECLS Nomenclature

For the vast majority of visitors here, the occasion to drop a fully configured four-level-deep ECMO classification on an unsuspecting colleague will never arrive; and so no further time will be spent on this here. Instead, some justifications and uses cases for specific ECMO configurations will be described in terms most suitable to a revising trainee.

ECMO cannula configurations and rationale for each

The objective here is not only to explain what the configuration does, but why, and the potential problems arising from that strategy. There is no one-stop resource for something like this, and the information included below was knitted together out of threads from multiple resources, of which the most notable were ECMO Cannulation and Configuration by Ihab Ahmed (2024), as well as The Alfred ECMO configuration page. The latter is especially valuable as it gives "advantages and disadvantages" of each mode, which seems almost as if it was created to answer an SAQ. There is also an excellent paper by Sorokin et al (2017), which presents not only a summary of configurations but also a series of gorgeous illustrations by Elizaveta Sorokina (then in high school) which describe the circuit configurations diagrammatically. Though the market for ECMO configuration diagrams is already vastly oversupplied with images ranging from slick polished digital art to everything else, the author felt an irresistable urge to add his own, and makes the original vector graphics file available for anyone who may need to reconfigure it for their own diagrams. In line with a general pattern of oversimplifying things, complicated dual lumen (eg. Avalon) and VADlike cannulae (Impella) will not be discussed here, out of respect for people who understand that subject better. 

Configuration Comments

V-V femoral-femoral configuration

A standard approach to veno-venous ECMO. Well suited to smaller patients and those with a normal cardiac output.

Advantages

  • Easy to insert (only one anatomical region needs to be prepared and cleaned)
  • Easily compressible sites

Caveats

  • This patient will not be able to mobilise, and even sitting up to 45 degrees is questionable
  • Awkward turns or patient misbehaviour could cause the atrium to become impaled by the return cannula 
  • The total flow is limited by the blood flow of the IVC, which may not be much over 3L/min. With increasing cardiac output, eg. in the hyperdynamic phase of sepsis, this configuration is likely to become inadequate because of access limitations (i.e. the drainage cannula will not get enough blood flow to compensate for the increased cardiac output)
  • May become problematic in patients with IVC flow restriction, eg. due to increased intra-abdominal pressure or during pregnancy

V-V, femoral-jugular configuration

Increased access capacity, allowing higher flow rates. The femoral cannula is the access cannula, and the right IJ cannula is the return. 

Advantages

  • Capable of high flow with only two veins 
  • The right IJ return cannula naturally points the jet of blood towards the tricuspid valve, which means most of the oxygenated blood ends up entrained into the right heart, and recirculation is minimised. Rich et al (1998) attributed the greatly improved performance of a fem-jug circuit to this idea, and even though they did not report on recirculation directly, their paper is usually quoted to support this assertion.
  • The patient could potentially mobilise, with some attention to securing the femoral cannula
  • The right IJ can also be used for access if the femoral return cannula is pushed into the RA 

Caveats

  • Cannulation is technically more challenging and the whole process requires two separate sterile fields
  • The right IJ is a dominant drainage site for cerebral venous blood, and one would not wish for it to be severely mangled thrombosed or compressed if the ICP is at all important

VV-V, fem-fem-jug configuration

Greatly increased access capacity, allowing higher flow rates. The femoral cannulae are for access and the right IJ cannula is the return. 

Advantages

  • Capable of the highest flow; adding the extra cannula tends to get 1-2/min of extra flow (i.e. it could theoretically approach the limits of the pump, which are usually stated to be 7-9 L/min)
  • Theoretically, could provide good oxygenation even with greatly increased cardiac output, eg. in sepsis

Caveats

  • The three cannulation sites increase the risks of vascular injury
  • All the disadvantages of IJ cannulation
  • Three of the greater vessels being occupied by large lines makes it challenging to resite central access later, as the vessels can be thrombosed or compressed by haematoma

V-P, fem-PA configuration

Indicated for severe hypoxemia and severe RV dysfunction, where VA ECMO is not a viable option.

Returning the blood into the pulmonary artery decreases RV workload.  

Advantages

  • Short of RVAD and VA ECMO, this is the only method of decompressing a dilated RV and resting it until recovery 
  • Minimises recirculation (cannulae are well-separated)
  • Bypasses intracardiac septal defects 

Caveats

  • Difficult to site (requires fluoroscopy)
  • Traverses two valves, which could lead to structural damage
  • Could cause extrinsic compression of the right coronary artery 
  • Not exactly a well-known strategy (case reports)
  • Hard to know where its role is, as V-A ECMO could achieve the same thing, and if the RV has quit so completely, likely the patient is also in cardiogenic shock and it would be hard to justify trying this trick first before cannulating for V-A.

V-A, fem-fem

The standard configuration for VA ECMO; access is from the SVC and the return cannula is in the lower aorta

Advantages

  • Quick to cannulate
  • Well away from the chest, can cannulate with CPR in progress

Caveats

  • Needs a femoral arterial backflow cannula
  • A-V recirculation: as native cardiac output becomes stronger, the arterial return to the lower body from the low aortic cannula improves the perfusion of the abdominal viscera and legs, increasing the venous saturation of the blood returning via the IVC (while the SVC remains 

V-A, jug-fem

A variant on the usual configuration for VA ECMO; access is from the SVC via the IJ, which leaves one of the groins free for things like angiography

Advantages

  • The IJ access drains the RA and SVC better than the fem-fem, reducing the A-V recirculation phenomenon

Caveats

  • Harder to cannulate
  • Also needs a femoral arterial backflow cannula

V-A, jug-subclavian or fem-subclavian

A variant that maximises the mobility of the lower limbs

Advantages

  • Reduced A-V recirculation phenomenon
  • The patient can mobilise

Caveats

  • Harder to cannulate (usually surgical)
  • Excessive perfusion of the subclavian artery; or deficient perfusion - in either case, requiring intervention
  • Incompressible site makes decannulation more interesting

V-AV, fem-fem-jug 

A venous return cannula is added to V-A ECMO, because the cardiac output is improved enough to create differential oxygenation, but not enough to be ready for weaning. 

Advantages

  • Supplies oxygenated blood into the RV, reducing the risk of upper body hypoxia
  • The new return cannula is being placed last, so it can be sized accordingly to increase its resistance and limit the flow diversion from the arterial circuit

Caveats

  • A third cannula is never welcome, from a vascular injury point of view
  • Increases RV preload, which may diminish the protective effects of decompressing it using V-A ECMO (one of the main advantages of V-A).
  • Hard to find a role for it, as often the patient may continue improving and be ready for weaning in the near future, making the extra cannula obsolete.

V-VA, jug-fem-jug 

An arterial return cannula is added to a fem-jug V-V ECMO, because the cardiac output collapses and the patient now also needs circulatory support. (Identical to V-AV, except in the timing of the interventions) 

Advantages

  • Complements the V-V oxygen delivery to the tissues by adding oxygenated blood flow into the lower aorta

Caveats

  • A third cannula is never welcome.
  • Occasionally the flow diversion into the new arterial return cannula can be disappointingly poor, as the arterial circulation is a higher resistance circuit and the blood will still preferentially flow into the venous return line. 
  • This might require a downsizing of a venous return line (never good; the hole stays the same size) or some kind of clamp (which can cause haemolysis). Belliato et al (2020) suggested that a good pairing was a 17Fr venous and 19Fr arterial cannula.

VV-A, jug-fem-fem

A common adjustment to the circuit which adds an additional access site, usually IJ.

Advantages

  • Contributes a large amount of drainage and can therefore produce a large cardiac output, eg. where the VA patient has become septic and vasodilated, or where the patient is very large and has great VO2 demands.
  • When the time comes, and cardiac output improves, the old access jugular line can be repurposed as a return line for a V-AV configuration to prevent differential oxygenation

Caveats

  • Too many cannulae
  • The extra cannula might add 0.5-1 L/min of additional flow, which is not a huge return on the investment

VV-AV, jug-fem-fem-fem

Double cannulation of the two femoral veins, one up to the inferior vena cava and the other up to the iliac vein, with femoral artery as the arterial return and the IJ as venous return. This is the adaptation of V-AV for the differentially oxygenated patient with access insufficiency, or an adaptation of VV-A for the patient who is slowly recovering their cardiac output. 

Advantages

  • Able to respond to a higher cardiac output demand, but also accommodates the slow return of a native cardiac output
  • A solution for larger patients 

Caveats

  • Way too many cannulae
  • It is never good for two ECMO lines to be in the same groin 
  • There are other ways to reduce the cardiac output demand that would not require this much vessel damage

VVvent-A, jug-left atrium-fem
("venting")

Resorted to when the LV fails to reliably open the aortic valve, or there is aortic regurgitation,  resulting in ventricular dilatation and blood stasis. Possible options listed by Cevasco et al (2019) include septal puncture and cannulation from the RA side,  surgical placement of the cannula into the pulmonary vein, or by cannulatiung the LV directly through the LV apex via a mini-thoracotomy.

Advantages

  • Decompresses the LV and removes the threat of LV rupture and thrombosis 

Caveats

  • Very invasive, needs to be done surgically
  • Introduces the risk of LV rupture and thrombosis (apical approach)
  • Surely an IABP could be enough, one might argue

 The nature of ECMO being that of an improvised and ad-hoc solution adapted on the fly to catch the falling patient means that every possible configuration and cannula position technique cannot be captured in this resource, to interrupt the complaints from those readers who note that their favourite veno-pulmo-VSD-into-VA hybrid mode is not listed. That mode is probably well discussed in the excellent  paper by Camboni et al (2019), where seeming endless permutations of cannula arrangements are explored.

Reading the above, and particularly towards the end, the sane reader may reflect on these horrors and ask, is this really safe? Shin et al (2025) explored this question and concluded (albeit from a small series) that hybrid modes overall have more complications associated with their use, but that transitioning to a hybrid mode does not have any effect on complication rates or outcomes when compared to starting with an initially hybridised mode. 

References

Broman, Lars Mikael, et al. "The ELSO Maastricht Treaty for ECLS Nomenclature: abbreviations for cannulation configuration in extracorporeal life support-a position paper of the Extracorporeal Life Support Organization." Critical Care 23.1 (2019): 36.

Brasseur, Alexandre, et al. "Hybrid extracorporeal membrane oxygenation." Journal of Thoracic Disease 10.Suppl 5 (2018): S707.

Sorokin, Vitaly, et al. "Choosing the appropriate configuration and cannulation strategies for extracorporeal membrane oxygenation: the potential dynamic process of organ support and importance of hybrid modes." European journal of heart failure 19 (2017): 75-83.

Ahmed, Ihab. "ECMO Cannulation and Configuration." ECMO: A Practical Guide to Management. Cham: Springer International Publishing, 2024. 93-119.

Abrams, Darryl, Matthew Bacchetta, and Daniel Brodie. "Recirculation in venovenous extracorporeal membrane oxygenation." Asaio Journal 61.2 (2015): 115-121.

Rich, Preston B., et al. "A prospective comparison of atrio-femoral and femoro-atrial flow in adult venovenous extracorporeal life support." The Journal of thoracic and cardiovascular surgery 116.4 (1998): 628-632.

Rojas-Velasco, Gustavo, et al. "Pulmonary artery cannulation during venovenous extracorporeal membrane oxygenation: An alternative to manage refractory hypoxemia and right ventricular dysfunction." Respiratory Medicine Case Reports 38 (2022): 101704.

Unger, Erin D., Ranya N. Sweis, and Ankit Bharat. "Unusual complication of a right ventricular support–Extracorporeal membrane oxygenation cannula." JAMA cardiology 6.6 (2021): 723-724.

Shah, Aakash, et al. "Hybrid and parallel extracorporeal membrane oxygenation circuits." JTCVS techniques 8 (2021): 77-85.

Shin, Irin, et al. "Mortality and Complications in Hybrid Extracorporeal Membrane Oxygenation: A Meta-Analysis of Initial Use Versus Transition." ASAIO Journal 71.5 (2025): 363-369.

Cevasco, Marisa, et al. "Left ventricular distension and venting strategies for patients on venoarterial extracorporeal membrane oxygenation." Journal of thoracic disease 11.4 (2019): 1676.

Camboni, Daniele, et al. "Double, triple and quadruple cannulation for veno-arterial extracorporeal membrane oxygenation support: is there a limit?." Annals of Cardiothoracic Surgery 8.1 (2019): 151.

[Submit a comment or correction]