Everything ECMO – Part Two

Natalie Kruit on Troubleshooting ECMO Retrieval

This second post from the Sydney HEMS ECMO education days summarises a great session by Retrieval Doctor and all-round legend Natalie Kruit on Troubleshooting in ECMO Retrieval.

What’s the role of the Retrieval Doctor?

As mentioned in Part One, the retrieval doctor is essentially there to look after the patient as a whole, while the separate ECMO team will oversee the function of the ECMO pump.

The first thing to know as a retrieval doctor is that the pump is primed with crystalloid – there is approx. 1L in circuit in total – so we should expect patients to be a bit unstable after transition onto pump!


There is an access line in or near vena cava – this draws blood into the pump. This is the negative pressure part of the circuit.

Blood drains into centrifugal pump, is oxygenated and is returned to the right atrium.

VV ECMO inclusion criteria

The Murray score is used for this: for a score >3 – consider referral to ECMO centre

Murray score: VV ECMO inclusion criteria

Pathologies that will do well: asthma, ARDS. The longer they’ve been on a ventilator, the less likely they are to do well on ECMO.

What can you as the retrieval doctor do?

  • Optimise
  • Sedate and paralyse
  • Find best available vent
  • Recruit and PEEP
  • Diurese
  • HRCT
  • Right radial arterial line for sampling etc

Troubleshooting 1: Access Insufficiency

This occurs when the vein the cannula is sitting in is collapsing around the cannula.

  • Sucking down onto cannula
  • “Kicking” of the line
  • Fluctuating circuit flows

Causes: hypovolaemia, anything altering venous return (eg coughing, sedation bolus, increased abdominal pressures)

Treatments: drop the revs, give volume, remove obstruction, consider and exclude tamponade

Ventilation strategies on VV ECMO

Once on VV ECMO – target lung protective ventilation: rest the lungs, minimise alveolar strain

Reduce FiO2 (target around 0.6 if possible: O2 toxicity, diffusion atelectasis from reabsorption), reduce the RR, increase PEEP (prevent alveolar closing and opening – maintain alveolar opening), smaller TVs (4mls/kg or even lower) (letting the ECMO circuit do the oxygenation)

Troubleshooting 2: Bleeding on ECMO

Caused by

  • Fibrinolysis
  • Circuit driven fibrinolysis
  • Mucosal bleeding initially
  • Strong correlation with mortality
  • Heparin resistance

Targets: Platelets >50, Fibrinogen >1.0, INR <2.0, APTT 1.5x normal

If bleeding: TXA, FFP, platelets (prepare to do an ECMO circuit change as may clot) – don’t give protamine (you will get clots and that is maximum badness!

Troubleshooting 3: Hypoxia

When SpO2 increasing on ECMO machine, the patient becoming hypoxic – this suggests recirculation (blood not getting to the patient)

Shunting on VV ECMO

The circuit may not be able to capture full cardiac output, so some deoxygenated blood is circulating. Preferentially this seems to go to the brain and heart. This can happen in someone whose circulation is hyperdynamic (eg sepsis, morbid obesity). The drainage cannula is a good determinant of flow (a bigger drainage cannula gives better flows).

Cardiac output >ECMO flow (if you set flow at 5L/min and the patient is hyperdynamic at 10L/min, only half blood vol will be going through circuit and being oxygenated)


  • Increase ECMO flow (but you risk suckdown)
  • High flow configuration – consider a second access cannula (can add third access line in neck if in fem/fem configuration)
  • Increase O2 carrying capacity: transfuse if Hb is low
  • Check oxygenator (may be cause due to clots) – measure pre and post oxygenator O2. Look for delta P <50
  • Check for recirculation (see below)
  • Check ventilator settings

Troubleshooting 4: Recirculation

This occurs when the access line and drainage line are sitting too close together so the flow is bypassing patient altogether! Or if there is increased intrathoracic/intracardiac pressure (e.g. tamponade)

You may see a low arterial SaO2 and high access cannula SaO2.


  • CXR to check pipe position
  • Could withdraw appropriately to increase distance between drainage and return lines
  • Check for tamponade


There are two different cannulation options for VA ECMO.

  • Central – directly cannulated via sternotomy
  • Peripheral – femoral vessels (blood from the femoral vein flows out to pump then back into femoral artery and pumped UP – this is non-physiological! Blood will be flowing against the normal flow direction – towards heart. This configuration needs a backflow cannula to ensure distal limb perfusion)

You’ll get some mixing of the deoxygenated blood from the ventricle with the oxygenated blood from the ECMO circuit – the more contractile the ventricle, the more distal the mixing point. There’s a nice interactive critical care case here that covers some of this.

If the lungs are not working and the heart is starting to recover and therefore has some recovering ejection fraction, the heart will preferentially eject to coronary and cerebral circulations (so check the ECG for signs of cardiac ischaemia) – it’s also helpful to have a  SpO2 probe on right hand to see differential hypoxia early.

Troubleshooting 1: Differential hypoxia (AKA harlequin syndrome)

In this scenario the heart is pumping (so there is a native cardiac output) and that is picking up and distributing deoxygenated blood that we can’t capture with ECMO flow. You may see demarcation of skin tone change across the trunk and/or the ECG may show ischaemic changes. There’s a nice diagram here which comes from this open access paper  in Circulation (clicking the link will download the PDF).


  • try to treat the lungs (eg bronchoscopy, increase the PEEP, add a venous catheter so some oxygenated blood is going to right heart)

Patient selection for VA ECMO

  • Patients unable to wean from bypass (post cardiotomy)
  • Awaiting heart transplant
  • RV failure
  • Overdose
  • AMI with multi-organ failure
  • Cardiomyopathy/myocarditis

When severe end organ damage  has occurred (e.g. AST/ALT very high, lactate >10 and not clearing) we have missed the boat!

Ventilation strategies on VA ECMO

We are aspiring to the least damaging lung ventilation

  • Maintain PEEP (aeration)
  • Measure SpO2 and PaO2 right arm
  • Minimise RV afterload and strain – avoid hypercapnia, hypoxia and complete lung collapse

Haemodynamic Management

  • Discuss haemodynamic goals with ECMO specialist
  • Inotropes to maintain ejection fraction
  • Maintain perfusion pressure for end organ perfusion
  • Promote forward flow: maintain MAP 65mmHg ideally (less than 90mmHg)

Troubleshooting 2: Loss of pulsatility


  • Tamponade
  • Myocardial depression
  • LV failure
  • Access insufficiency will be the first sign – blood is stuck in lungs not getting to circuit

You shouldn’t need to give a fluid bolus – so if doing this, consider LV failure is evolving (dilating LV with inability to pump out)

Offloading the LV – may need to decompress by creating a vent between L and circuit, R heart – there’s potential to use impella to suck blood out directly

Retrieval Final Checks Before Departure

  • Adequate oxygen and power supply
  • Confirm haemodynamic goals with ECMO specialist
  • All taps have stopcocks on
  • Access line dark, return line bright
  • Circuit flows stable
  • Lines secured, away from snag hazards
  • Patient sedated and paralysed

If you work for Sydney HEMS, take a look at the new ECMO class available through the virtual learning environment.

A/Prof Paul Forrest: VA ECMO in Sepsis

A/Prof Forrest talked a little bit about VA ECMO in sepsis.

The key to the septic shock definition is MAP 60mmHg (or inotropes required to keep it here).

Most adults dying from sepsis do so in a high output state (increased cardiac output with decreased systemic vascular resistance): this paper from 1984 looked at 20 patients with severe septic shock – 50% had moderate to severe decrease in left ventricular function. Notably they found the left ventricular ejection fraction had returned to normal within ten days in those who survived.

This leads us to consider whether low ejection fraction might be a good thing in patients with septic shock – this review by Vieillard-Baron et al from 2001 seems to support the hypothesis.

It’s possible, then, that the relationship between survival and non-survival depends on systemic vascular resistance in the context of vasoplegia – this allows for left ventricular ejection fraction to be maintained against lower resistance

There’s some evidence we might consider in paediatric patients:

MacLaren et al described their experiences with VA ECMO for septic shock in children.

They looked at 45 patients, of whom 21 survival (47%). In terms of central vs peripheral VA ECMO, survival proportions were 73% vs 44% and no survivors severely disabled. The authors felt that their data suggested that central configuration might provide a survival advantage.

The same team published a further paper in 2011: in this paper they studied 23 patients, 17  of whom survived to discharge (74%): they noted that a high lactate predicted worse survival.

As a result of this work, the American College of Critical Care Medicine’s paediatric sepsis guidelines published in 2017 include ECMO if the child has persistent catecholamine refractory shock.

For adults the data is lacking; literature only describes around 200 adults on VA ECMO when case series are combined and there are considerable different baseline characteristics between patient groups.

Cheng et al studied adult patients requiring ECMO for the management of sepsis in their paper on predictors of survival: they found that gram +ve sepsis does better than gram –ve or fungal.

There is also the possibility (in adult patients) of using V-AV ECMO: this is when a second return line is sited in the internal jugular (or similar) with the aim of providing oxygenated blood to the right heart. The femoral artery is part of a higher pressure circuit than the internal jugular line and it can be tricky to regulate flow between the two lines. As yet there is little data on this: this case series by Yeo published in Critical Care in 2016 reports a 50% survival with this technique for adults with severe ARDS and septic shock.

Indications for VA ECMO in refractory septic shock

For paediatric patients – central VA ECMO seems the best option

For adult patients:

  • In the presence of severe ARDS: V-AV ECMO
  • In the absence of severe ARDS: VA ECMO

VA ECMO in septic shock

In summary, VA ECMO in septic shock seems to be good for:

  • Kids with or without CA
  • Adults with low cardiac output/high systemic vascular resistance state

but more evidence is needed.

Where can I find out more?



Loads of resources over at IntensiveCareNetwork.com


On the literature

ECMO literature summaries – Life in the Fast Lane

About Natalie May

EM and PEM-trained UK doctor working in Prehospital & Retrieval Medicine in Australia. Evolving medical education interest, running enthusiast, karaoke queen. Here for the #FOAMed.
This entry was posted in ECMO, General PH&RM and tagged , , , , , , , , , , . Bookmark the permalink.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s