ASAIO Journal 2021 Guidelines

Guidelines for Pediatric Cardiac Failure Extracorporeal Life

Support Organization ECLS Guidelines 2020
Reviewers: GEORGIA BROWN, MBBS (Hons), FCICM,* KATIE M. MOYNIHAN, MBBS, FRACP, FCICM,§¶
KRISTOPHER B. DEATRICK, MD,† APARNA HOSKOTE, MBBS, DCH, MD, MRCP,‡ HITESH S. SANDHU , MBBS, MRCPCH,∥
DEVON AGANGA , MD,#** SHRIPRASAD R. DESHPANDE, MBBS, MS,†† ANURADHA P. MENON, MBBS, MRCPCH,‡‡
THOMAS ROZEN, MBBS, FRACP, FCICM,* LAKSHMI RAMAN, MD,§§
PETA M.A. ALEXANDER, MBBS, FRACP, FCICM,§¶Reviewers: RAVI THIAGARAJAN,§¶ PETER ROELEVELD,¶¶
ROBERTO CHILETTI,* GRAEME MACLAREN,∥∥ GILES PEEK##

These guidelines are applicable to neonates and children with Extracorporeal membrane oxygenation (ECMO) is the
cardiac failure as indication for extracorporeal life support most commonly utilized mechanical circulatory support
(ELSO). These guidelines address patient selection, manage- in neonates and children with refractory cardiac failure
ment during extracorporeal membrane oxygenation, and (Figure  1).1,2 Veno-arterial ECMO (VA ECMO) augments
pathways for weaning support or bridging to other therapies. systemic cardiac output and respiratory gas exchange to
Equally important issues, such as personnel, training, creden- facilitate adequate tissue oxygen delivery (DO2). Survival
tialing, resources, follow-up, reporting, and quality assurance, for children with heart disease supported with VA ECMO
are addressed in other ELSO documents or are center-specific. has improved over the past decade, despite expanding
indications and increasing patient complexity.2 Increased
Key Words: pediatrics, heart failure, extracorporeal membrane utilization and experience of pediatric cardiac ECMO is
oxygenation reflected in a number of publications, but large evidence
gaps remain. High-quality VA ECMO support for pediatric
This guideline is informed by available evidence and based on cardiac indications necessitates systems, protocols, interdis-
ciplinary teams, and training.
expert opinion, with targeted clinical recommendations for emerg-
ing centers and small volume programs as institutional standards
are developed. The guideline may also be of benefit to experienced
providers and centers in the process of reviewing local protocols. Patient Selection, Modes of Support,
Circuit configuration, equipment specifications, anticoagulation and Technical Considerations
recommendations, extracorporeal cardiopulmonary resuscitation Patient Selection
(ECPR), and some specific patient populations are presented in
other Extracorporeal Life Support (ELSO) guidelines. The indication for the use of VA ECMO for cardiac indica-
tions in children is cardiogenic shock unresponsive to standard
medical therapies. Persistent systemic systolic pressure less
than 50 mm Hg, urine output <1 ml/kg/h, lactic acidosis, cen-
From the *Cardiac Intensive Care Unit, The Royal Children’s
Hospital, Melbourne, Australia; †Division of Cardiac Surgery, tral venous oxygen saturation (SVO2) <60% or arteriovenous
University of Maryland School of Medicine, Baltimore, Maryland; oxygen saturation difference (AVO2) >30% in cyanotic congeni-
‡Cardiorespiratory and Critical Care Division, Great Ormond tal heart disease, an altered mental status due to low cardiac
Street Hospital for Children NHS Foundation Trust, London, United output may all be indicators for such a cardiogenic shock in
Kingdom; §Department of Cardiology, Boston Children’s Hospital,
Boston, Massachusetts; ¶Department of Pediatrics, Harvard Medical
children. Examples of pathologies causing shock are listed in
School, Boston, Massachusetts; ∥Department of Pediatrics, Critical Table  1. Consideration for early initiation of ECMO is impor-
Care Division, Le Bonheur Children's Hospital, University of tant as delayed initiation (beyond 6 hours of cardiogenic shock
Tennessee, Memphis, Tennessee; #Department of Anesthesiology state) is associated with worse outcomes.1,3–25 Local resources
and Perioperative Medicine, Mayo Clinic, Rochester, Minnesota; should be taken into account when determining ECMO candi-
**Department of Pediatric and Adolescent Medicine, Mayo Clinic,
Rochester, Minnesota; ††Pediatric Cardiology Division, Heart dacy (Figure 2).
Transplant and Advanced Cardiac Therapies Program, Children’s Veno-arterial ECMO should be considered with four primary
National Heart Institute, Washington, D.C.; ‡‡Children’s Intensive strategies for ECMO support:
Care Unit, Department of Paediatric Subspecialties, KK Women’s
and Children’s Hospital, Singapore; §§Department of Critical Care, 1. Bridge to recovery: In patients with reversible underlying dis-
University of Texas Southwestern Medical Center, Texas; ¶¶Leiden ease processes where cardiac function recovery can occur
University Medical Centre, Leiden, Netherlands; ∥∥National University with time, medical interventions, or surgical correction;
Heart Centre, Singapore; and ##UF Shands Children’s Hospital,
Gainesville, Florida. 2. Bridge to bridge: In patients with acute single organ dis-
Submitted for consideration January 2021; accepted for publication ease who can be supported to a ventricular assist device
in revised form January 2021. (VAD);
Disclosure: The authors have no conflicts of interest to report. 3. Bridge to organ transplantation: In patients who may require
Correspondence: Peta Alexander, Department of Cardiology, Boston
Children’s Hospital, Boston, MA 02115. Email: peta.alexander@car-
cardiopulmonary support until heart transplantation;
dio.chboston.org. 4. Bridge to decision: In patients who may recover end-
Copyright © ASAIO 2021 organ function, facilitate diagnosis, or determine candi-
DOI: 10.1097/MAT.0000000000001431 dacy for alternative support/transplantation.

1
Copyright © Extracorporeal Life Support Organization. Unauthorized reproduction of this article is prohibited.
 2 BROWN ET AL.

Figure 1. Decision-making flowsheet for VA ECMO for pediatric cardiac indications. VA ECMO, veno-arterial extracorporeal membrane
oxygenation.

Patient Selection Practice Points Mode of Support

1. Institute ECMO before evidence of severe oxygen defi- In children with a cardiac disease not responding to maxi-
ciency and end-organ damage. Early initiation is impor- mal medical therapy, VA ECMO support facilitates respiratory
tant for developing and low-volume ECMO centers gas exchange and augments cardiac output for DO2 while
without capacity for rapid initiation of ECPR. allowing time for myocardial recovery or diagnosis and repair
2. Individual institutions should consider their center of anatomical lesions. Effective VA ECMO support should
experiences and resources when evaluating indications be assessed by surrogates of tissue oxygen delivery includ-
and contraindications for ECMO. ing blood lactate, SVO2 or AVO2, near-infrared spectroscopy
3. Extracorporeal membrane oxygenation following pedi- (NIRS), measures of end-organ function, for example, urine
atric cardiac surgery should prompt early investigation output, creatinine, liver function tests, and adequacy of car-
and management of possible residual lesions. diac decompression assessed by echocardiography, chest radi-
ography, and ultimately by cardiac catheterization.

Copyright © Extracorporeal Life Support Organization. Unauthorized reproduction of this article is prohibited.
 ELSO GUIDELINES FOR PEDIATRIC CARDIAC FAILURE 3

Table 1.  Indications, Contraindications, and Special Considerations for Cardiac Indications for Extracorporeal Membrane
Oxygenation

INDICATIONS

Periprocedural Cardiac Surgery and Catheterization

  Preprocedural stabilization for Preoperative neonates with obstructed total anomalous pulmonary venous return or transposition of
inadequate systemic cardiac the great arteries with inadequate mixing or persistent pulmonary hypertension
output—in cases where Circular shunt with Ebstein’s anomaly or pulmonary regurgitation following balloon dilation
physiologic stability is likely to Anomalous left coronary artery from the pulmonary artery
be achieved over time or early Restrictive pulmonary blood flow in tetralogy of Fallot or obstructed Blalock-Taussig shunt
operative repair is likely to have Tetralogy absent pulmonary valve with airway or lung parenchymal compromise and inadequate
a successful outcome pulmonary blood flow
Undifferentiated congenital heart disease with cardiopulmonary compromise
  Failure to wean from Ischemic reperfusion injury following cardiopulmonary bypass or inadequate cardioplegia
cardiopulmonary bypass or low Myocardial edema related to the inflammatory bypass process or ventriculotomy
cardiac output in the postoperative
period
  Postoperative arrhythmia Junctional ectopic tachycardia with Tetralogy of Fallot with hemodynamic compromise refractory to
antiarrhythmic measures
Circulatory Failure Due To Other Etiologies
 Cardiogenic Myocarditis
Cardiomyopathy
Postcardiac arrest ventricular dysfunction
Intractable tachyarrhythmia or bradycardia
 Obstructive Pulmonary hypertension
Pulmonary embolus
 Distributive Sepsis
Note: May be challenging to achieve Anaphylaxis
adequate ECMO flow for delivery
of oxygen to tissues in the setting
of vasoplegia
Cardiopulmonary Arrest—see Extracorporeal Cardiopulmonary Resuscitation (ECPR) Guideline.

CONTRAINDICATIONS
Patient-level factors—ECMO Prolonged state of cardiogenic shock (over 6 hours) unlikely to benefit from initiation of ECMO
support would be unlikely to Relative prematurity or low birth weight in neonates (<34 weeks of gestational age or <2.0 kg) with
facilitate survival without likelihood significant morbidity and mortality
of major morbidity Extremes of prematurity or low birth weight (<32 weeks of gestational age or <1.5 kg)
Severe chromosomal abnormalities (e.g., Trisomy 13 or 18)
Irreversible brain damage or Intracranial hemorrhage (grade III or IV IVH)
Uncontrollable hemorrhage should be considered a contraindication for ECMO unless cannulation
to ECMO would assist in source control
Procedural factors Inability to achieve vascular or central access for cannulation

SPECIAL CONSIDERATIONS
Aortic regurgitation VA ECMO flow results in increased afterload on the left ventricle, and even mild aortic regurgitation
progress to become clinically significant
Attention to left heart decompression or early transition to cardiopulmonary bypass may be required
Interrupted aortic arch Careful attention to the anatomy of head and neck vessels (i.e., location of interruption of arch) is
required before ECMO cannulation to ensure brain perfusion with oxygenated ECMO flow
Stage 1 palliative surgery (S1P) ECMO support after S1P for hypoplastic left heart syndrome is the most frequent postoperative
ECMO indication in neonates
In S1P with systemic to pulmonary shunt, higher ECMO flow may be required (150–200 ml/kg/min)
Temporary shunt restriction to limit pulmonary blood flow and promote systemic blood flow may be
necessary
In patients with RV-PA conduit, maintaining cardiac ejection (and thus flow across the conduit) may
prevent shunt thrombosis
Stage 2 and 3 palliative surgery Infants and children after surgical palliation with cavopulmonary anastomoses (Glenn and Fontan
circulations) represent a complex physiologic group in whom stable support with ECMO can be
difficult to establish given the separation of systemic venous return (Table 2)

ECMO, extracorporeal life support; RV-PA, right ventricle to pulmonary artery; VA ECMO, veno-arterial extracorporeal membrane oxygenation.

Cannulation a recent sternotomy (i.e., less than 10–14 days), with right atrial
access for venous drainage and cannulation of the aorta for arte-
In children supported with ECMO for cardiac indications, rial return.19,20,26 Neck access via the internal jugular vein and
cannulation site, and strategy are determined by the patient’s common carotid artery is the favored peripheral cannulation
size, underlying cardiac anatomy, the anatomy and surgical sites in many smaller children (<5–6 years or <30 kg), balancing
palliation of congenital heart disease, and any recent surgical optimized ECMO flow through the larger upper body vessels
intervention (Table  2). Central cannulation is commonly used against possible increased risk of neurologic adverse events.27,28
in the postcardiopulmonary bypass period or in the presence of For older children, the femoral vein and femoral artery approach

Copyright © Extracorporeal Life Support Organization. Unauthorized reproduction of this article is prohibited.
 4 BROWN ET AL.

Figure 2. ICU ECMO cannulation preparation. ECMO, extracorporeal membrane oxygenation; ICU, intensive care unit.

may be utilized.26 Neck vessel cannulation in younger children Children with congenital heart disease are at risk of periph-
typically necessitates open surgical access, while femoral ves- eral vessel occlusion due to cardiac catheterization access,
sels may be cannulated using either open or Seldinger tech- and knowledge of vessel patency is important before attempted
nique. When femoral vessels are used, limb ischemia is avoided cannulation.10 This is particularly relevant for children with
by using the smallest arterial cannula for desired flow rate, distal single-ventricle physiology palliated to cavopulmonary circu-
reperfusion cannula insertion, and opposing femoral arterial and lations who often require multisite cannulation to maximize
venous vessel site cannulation strategies.29,30 venous drainage (Table 2).

Table 2.  Cannulation Strategy in Children With Cardiac Disease

Central Cannulation Peripheral Cannulation

Anatomy or Surgical Venous Arterial Venous Arterial

Palliation Access Access Access Access Additional Strategies

TWO VENTRICLES
Biventricular circulation or Right atrium Aorta Internal jugular or Common Left atrial decompression may need to be
structurally normal heart femoral carotid or considered
femoral
SINGLE VENTRICLE
Shunted or RV-PA Common Aorta Internal jugular Common *Peripheral = neck access due to
conduit physiology atrium carotid patient size. Care re: cannula position
(stage 1) with respect to shunt—may result
in overcirculation to lungs or shunt
occlusion
Superior SVC or Aorta Internal jugular or Common *If femoral approach only used, passive
cavopulmonary common femoral carotid venous return must flow through
anastomosis atrium lungs—ventilation must be optimized.
(stage 2) Additional venous cannula may be
required
Fontan (stage 3) Fontan baffle Aorta Internal jugular or Common Additional venous cannula often required
or common femoral carotid or
atrium femoral

RV-PA, right ventricle to pulmonary artery; SVC, superior vena cava; VA ECMO, veno-arterial extracorporeal membrane oxygenation.

Copyright © Extracorporeal Life Support Organization. Unauthorized reproduction of this article is prohibited.
 ELSO GUIDELINES FOR PEDIATRIC CARDIAC FAILURE 5

Figure 3. Determinants of tissue oxygen delivery on veno-arterial ECMO. ECMO, extracorporeal membrane oxygenation.

The goal during VA ECMO support is to maintain DO2 as close

If ventricular function is inadequate to open the aortic valve to normal as possible—at least three times VO2 (ratio of >3:1).
during systole, left ventricular diastolic pressure and left atrial During VA ECMO, the systemic oxygen extraction is continu-
pressure increase resulting in inadequate myocardial decom- ously monitored via the drainage cannula (SVO2), and as the
pression, subendocardial ischemia, and pulmonary venous arterial oxygen concentration and hemoglobin are known, a
congestion. Decompressing the hypertensive left atrium can be DO2:VO2 ratio can be monitored. If the arterial saturation is
achieved by atrial septostomy, left atrial cannulation (directly 100% and the SVO2 is 80% the ratio is 5:1. So adjusting flow
or via catheter crossing the atrial septum), left ventricular vent- and hemoglobin to maintain SVO2 >66% assures that the goal
ing via an open approach,31–34 or an axial transaortic valve of DO2:VO2 >3 is met. Typically, this is achieved with a car-
pump (Impella Abiomed, Danvers, MA) device.35 Left atrial diac index of 2.5–3 L/min/m2 or 100–150 ml/kg/min in infants,
decompression performed early (<18 hours postcannulation) 70–100 mL/kg/min in larger child. Further details available in
minimizes the duration of ECMO and mechanical ventilation.34 the ELSO Red Book.38
In children with congenital heart disease, predominant To commence VA ECMO support after cannulation, pump
right ventricular failure can be observed in patients during flow is gradually increased until adequate flow (as above) is
separation from perioperatively (e.g., patients with Tetralogy achieved. Blood flow is subsequently decreased to the lowest
of Fallot or Ebstein’s anomaly). Here, a special case can level that will provide adequate support to meet cellular met-
be made for VA ECMO for right-sided support. This can be abolic demands (typically a cardiac index of 2.5–3 L/min/m2
achieved via standard cannulation or with targeted support or 100–150 ml/kg/min in infant, 70–100 mL/kg/min in larger
by cannulation of the right atrium (venous cannula) and child). Ideally, arterial pulse pressure will be at least 10 mm
the pulmonary artery (arterial cannula), facilitating oxygen- Hg, indicating systemic ventricular ejection, which reduces
ation and supplementing subpulmonary ventricle function. the risk of systemic ventricular thrombosis. This may not be
Isolated right ventricular support, therefore, warrants central achieved if ventricular function is poor, despite inotropic sup-
cannulation.36,37 port, and should prompt consideration of left atrial decompres-
sion or alternate support strategies. Inability to achieve desired
Mode of Support Practice Points blood flow or ongoing evidence of inadequate cardiac output
necessitates urgent consideration of adjustment or additional
1. Central cannulation is used following recent sternotomy. drainage cannulae. Because the pulse pressure is low, the
2. Peripheral cannulation in infants is via the neck and, for mean systemic arterial pressure will be somewhat lower than
older children, the femoral vessels. normal. In addition, patients placed on ECMO for cardiac sup-
3. Consider preplanning a cannulation strategy in patients port are often managed with inotropes before ECMO initiation.
with congenital heart disease and difficult vascular As these drugs are titrated down, systemic vascular resistance
access. (SVR) decreases, and systemic pressure falls proportionately.
4. Early evaluation for the need for cardiac decompression If the perfusion pressure is inadequate (low urine output and
in the setting of severe myocardial dysfunction without poor perfusion), systemic arterial pressure can be increased by
native ejection. increasing pump flow, transfusing blood products, or titrating
vasopressor infusions.
Management During Extracorporeal Life Support
Oxygenation
Tissue oxygen delivery is determined by the sum of patient
and ECMO output which independently contribute to oxygen The rated flow is the blood flow rate at which venous blood
delivery (Figure 3). with a saturation of 75% and hemoglobin of 12 g/dl exits the
gas exchange device with a saturation of 95%. The rated flow
is a standard to compare the maximum oxygenation capacity
Systemic Blood Flow
of gas exchange devices and is determined by surface area and
Normal homeostatic mechanisms maintain DO2 to oxygen blood path mixing. As long as the blood flow is below recom-
consumption (VO2) at a ratio of 5:1 (20% extraction). During mended rated flow for that gas exchange device (and the inlet
shock states, a DO2:VO2 ratio of less than 2:1 (50% extrac- saturation is 70% or higher), the oxyhemoglobin saturation
tion) leads to anaerobic metabolism and metabolic acidosis. at the outlet should be greater than 95%. Usually, the outlet

Copyright © Extracorporeal Life Support Organization. Unauthorized reproduction of this article is prohibited.
 6 BROWN ET AL.

saturation will be 100%, and the pO2 will be over 300 mm goal is to maintain circuit flow without thrombosis despite
Hg. If the fraction of oxygen on the sweep gas (FdO2) is 100%, artificial circuit exposure. Anticoagulation therapy and mon-
at or below the device rated flow, and the outlet saturation itoring strategies vary between institutions and according to
is less than 95%, the gas exchange device is not working at patient diagnosis and thrombotic versus bleeding consider-
full efficiency (due to irregular flow, clotting). It may be neces- ations.43 Neonates and infants pose particular challenges due
sary to replace it. Oxygen delivery from the circuit should be to physiologic differences in hemostasis.44,45 Extracorporeal
adequate for full support (systemic saturation greater than 95% membrane oxygenation programs must develop an institu-
at low ventilator settings and FiO2). AVO2 difference should tional approach to monitoring and managing anticoagula-
be less than 20–30%. In states of high oxygen demand or tion for patients.46
poor oxygen delivery (low cardiac output, impaired lung gas In recent years, some programs have evolved protocols
exchange), maintaining the hematocrit over 40% (hemoglobin incorporating direct thrombin inhibitors (DTIs) such as bivali-
~12 g/dl) can optimize oxygen delivery. Patients with already rudin and argatroban, citing some practical and theoretical
impaired oxygen delivery, such as those with palliated single- benefits.47,48 While initially limited to specific instances in
ventricle physiology, should be maintained with a hematocrit which UFH was specifically contraindicated (e.g., heparin-
above 40%, particularly during partial VA ECMO support and induced thrombocytopenia, heparin-induced thrombocytope-
in preparation for weaning. Target paO2 should be maintained nia [HIT], and heparin resistance) the relative predictability of
within normal limits, as hyperoxia during VA ECMO support response and stability in patients on long-term support have
has been associated with mortality.39 led to increased adoption and preferential use in some pro-
grams. Advantages of the use of DTIs include the ability to treat
Carbon Dioxide Clearance or prevent HIT, the lack of dependence on antithrombin, and
reduced time to achieving steady-state concentrations due to
CO2 transfer across the membrane lung is more efficient their short half-life (25 minutes for bivalirudin and 45 minutes
than that of oxygen, and CO2 removal will exceed oxygen for argatroban).47,49 Relative disadvantages of DTIs include
uptake. CO2 clearance is controlled by the sweep gas flow higher cost (10–70× that of UFH in some institutional com-
rate. Increasing sweep gas flow rate increases CO2 clear- parisons) and the lack of reversal agents (mitigated by short
ance but does not affect oxygenation. Initial gas flow to half-life). Due to the lack of reversal agents, DTIs are less com-
blood flow ratio varies across institutions. As a guide, gas monly utilized for cardiopulmonary bypass, and transition
flow rate less than blood flow rate can be used to begin from heparin as used in the operating room must be managed
support, with early arterial blood gas monitoring to avoid carefully, especially for patients with recent cardiotomy and
hypocarbia, aiming for a gradual reduction (over 4–8 hours) central cannulation.
in PaCO2 to minimize rapid shifts in cerebral perfusion in Viscoelastic tests, including thromboelastography or rota-
patients who are hypercapnic at the time of cannulation. If tional thromboelastometry are whole-blood assays measuring
the initial PaCO2 is greater than 70 mm Hg, the PaCO2 should the rapidity and strength of thrombus formation. These tests
be normalized over several hours to avoid rapid changes of provide additional data to standard coagulation panels on clot-
cerebral perfusion related to CO2 and pH, which are associ- ting time, clot formation time, the firmness of clot formation,
ated with neurologic injury and mortality.40,41 If CO2 clear- and clot lysis. They are particularly beneficial in diagnosing
ance is decreased, but oxygenation is adequate, the cause specific aspects of coagulopathy in bleeding patients, such
is usually water accumulation within the gas compartment as those supported for failure to wean from cardiopulmonary
of the membrane lung. This may be cleared by intermittently bypass after complex reconstruction. These tests can identify
increasing sweep gas flow to a higher rate. states of fibrinolysis and help guide targeted treatment of spe-
cific factor or component deficiencies.
Goals Following Cannulation Practice Points Indications for blood product administration relate to patient
and circuit factors, specifically bleeding, evidence of thrombus
1. Following ECMO initiation, inability to achieve desired formation, and flow rates. Centers should refer to the ELSO
blood flow or ongoing evidence of inadequate tissue anticoagulation guideline for detail.
oxygen delivery requires an urgent reassessment of
ECMO strategy/cannulation.
2. With myocardial recovery, the patient's contribution to Blood Product Administration
systemic oxygen content and oxygen delivery increases Patients often receive blood products to maintain hemoglo-
and may require adjustments to ventilation/FiO2. bin and product targets on ECMO.50–52 In the recent Pediatric
3. Increasing the sweep flow will increase CO2 clearance Critical Care Transfusion and Anemia Expertise Initiative
but will not improve oxygenation. consensus statements regarding red blood cell transfusion in
critical illness, there was insufficient evidence for a specific
Anticoagulation; See Extracorporeal Life red blood cell transfusion threshold for children on ECMO,53
Support Anticoagulation Guideline and instead recommending transfusion decision-making to take
Extracorporeal Life Support Red Book, Ch 7 into account evidence of inadequate systemic or regional
oxygen delivery, adoption of blood conservation procedures
Unfractionated heparin (UFH) remains the most commonly and minimization of donor exposure.54 Many center protocols
utilized anticoagulant for pediatric ECMO.38,42 A bolus dose maintain hemoglobin between 10–12 g/dl. There is some evi-
of UFH (50–100 units/kg) is given at cannulation and subse- dence that children on ECMO are more likely to have bleeding
quently administered as a continuous infusion (Figure 2). The complications with a platelet count below 80,000; however,

Copyright © Extracorporeal Life Support Organization. Unauthorized reproduction of this article is prohibited.
 ELSO GUIDELINES FOR PEDIATRIC CARDIAC FAILURE 7

Table 3.  Indications for vasoactive infusion during VA ECMO support for cardiac indications

Vasoactive
support Indication and Benefits and Specific Risks Medication Starting Dose Range

Inotrope Enhancement of contractility in a patient with severe cardiac dysfunction Epinephrine 0.02–0.05 µg/kg/min
to facilitate aortic valve opening and prevent stasis of blood in the Dobutamine 5 µg/kg/min
systemic ventricle and aortic root
To optimize blood pressure and end-organ perfusion
Does not facilitate myocardial rest
Vasopressor Peripheral vasoconstriction is indicated in a patient on VA ECMO for Norepinephrine 0.02–0.05 µg/kg/min
distributive shock, on maximal circuit blood flow with inadequate Vasopressin 0.01–0.06 IU/kg/hr
cardiac output to optimize blood pressure and end-organ perfusion.
Vasodilator Peripheral vasodilation will reduce systemic afterload improving circuit Sodium nitroprusside 0.5–3 µg/kg/min
blood flow and systemic perfusion as well as decreasing left ventricle Milrinone 0.25–1.0 µg/kg/min
afterload, promoting ejection

VA ECMO, veno-arterial extracorporeal membrane oxygenation.

this may depend on the age of the patient and other factors Ventilator Management
such as recent cardiac surgery and cannulation strategy.50,55
Ventilation management on ECMO should minimize lung
Centers should refer to the ELSO anticoagulation guideline for
injury and optimize lung function to facilitate ECMO separa-
suggested transfusion targets and thresholds.
tion with cardiac recovery. Of note, for patients cannulated
via femoral vessels with poor lung function, as cardiac func-
Patient Management During Extracorporeal Life Support tion recovers, it is native lung function that determines oxygen
Cardiovascular content of antegrade flow from the heart into coronary arteries
and head and neck vessels. These patients may consequently
Heart rate and rhythm. Veno-arterial ECMO is capa- require higher FiO2 ± positive end-expiratory pressure (PEEP)
ble of providing full cardiac output during cardiac arryth- than the “lung rest” settings used when patients are supported
mias. However, arrhythmia can increase myocardial oxygen on ECMO for pulmonary indications.
demands, delay ventricular recovery, and if associated with a Suggested protective lung strategy: Pressure-limited ventila-
lack of cardiac ejection, can lead to ventricular distension and tion, PEEP (8–10 cm H2O), tidal volume <6–8 ml/kg ideal body
inadequate myocardial decompression. Restoration of sinus weight, peak inspiratory pressures <18–20 cm H2O, with a low
rhythm and atrioventricular synchrony with antiarrhythmic rate (10 bpm).57–60
therapy, overdrive pacing, direct current cardioversion, and No single ventilation strategy is universally practiced, and
electrophysiology ablation should be considered. the suggested strategy may need modification in patients with
Blood pressure and vasoactive medications. Patient an open sternum or pulmonary/intrathoracic pathology.61,62
blood pressure is determined by two modifiable factors: Adjuncts to mechanical ventilation may include suctioning,
blood flow (pump flow plus native cardiac output, Figure 3), use of ventilator, bronchoscopy, and prone positioning.63,64
and SVR. The pulse pressure is narrow when the native car- It may be appropriate in carefully selected patients to allow
diac output is minimal on full VA ECMO support. Ideally, spontaneous breathing or extubate in the absence of lung
the arterial pulse pressure is at least 10 mm Hg, with left pathology.
ventricle ejection reducing intracardiac thrombosis risk.
Severely reduced left ventricular systolic function may pre-
Fluid Management, Blood Volume, and Fluid Balance
vent ejection against the VA ECMO flow, even with inotropic
support. This should prompt early consideration of left atrial After stability has been achieved following cannulation,
decompression or alternate support strategies. Adequacy of resuscitation, and blood product replacement, diuresis should
systemic cardiac output is determined by an assessment of be instituted to target euvolemia. If pharmacologic diuresis is
clinical parameters of tissue oxygen delivery: warmth and ineffective, continuous ultrafiltration or renal replacement ther-
color of extremities, urine output, lactate, premembrane apies can be used, often incorporated into the ECMO circuit.
oxygen saturation or AVO2 difference, and NIRS.56 In the
setting of inadequate cardiac output with hypotension, after Neurology and Sedation (Also See Extracorporeal
addressing hypovolaemia, VA ECMO flow can be increased Life Support Sedation Guideline)
or inotropes or vasopressors commenced (Table 3). Signs of
low cardiac output with hypertension on ECMO may neces- Neurologic complications (hemorrhage, thrombosis, and
sitate optimization of sedation and use of vasodilator infu- seizures) are relatively common and adversely affect morbidity
sions. To maintain some intracardiac and pulmonary blood and survival outcomes of children on ECMO for cardiac indi-
flow, target VA ECMO flow can be around 80% of venous cations.21,65–67 Risk factors include weight <3 kg, gestational age
return, reflected by pulse pressure of approximately 10 mm less than 34 weeks, CPR, degree of precannulation vasoactive
Hg monitored with an invasive arterial line. Changes in support, and acidosis.21,68 Surveillance for neurologic events
arterial waveform may reflect myocardial recovery or an includes daily clinical examination if feasible, screening cra-
acute complication (circuit dysfunction, reduced preload, nial ultrasounds, and EEG with more definitive neuroimaging
increased afterload). (CT brain) recommended to detect intracranial hemorrhage

Copyright © Extracorporeal Life Support Organization. Unauthorized reproduction of this article is prohibited.
 8 BROWN ET AL.

Figure 4. ECMO weaning flowsheet. ECMO, extracorporeal membrane oxygenation.

or embolic events if heightened concern. Intermittent or con- Infection

tinuous EEG monitoring is recommended as seizures may
Infectious complications are frequent and are associated
only be detected electrographically, but this is not universally
with mortality.3,19,24 The most important risk factor is sup-
available.67
port duration.74 Vigilance is essential, as standard markers
Analgesia and sedation strategies for patients supported
(temperature, white cell count, inflammatory markers) may
on ECMO reflect strategies for critically ill children, trending
not adequately reflect the presence (or absence) of infection.
towards minimizing sedation, permitting spontaneous move-
There is little evidence to support routine surveillance cul-
ment, and facilitating neurologic assessment. While permit-
tures or the use of prophylactic antibiotics outside of that
ting awake ECMO is not practically possible in neonates on
for administration within 30 minutes of emergent surgical
VA ECMO, it may be carefully considered in a select group
procedures.75
of older children who have secure peripherally placed ECMO
cannulae, good circuit function, closed-chest, and a level of
understanding that will permit wakefulness without discom- Endocrine
fort and anxiety, along with full engagement of wider multi- Normoglycemic targets are effective for pediatric cardiac
disciplinary team and caregivers including parents/guardians. patients on ECMO.76,77
A multidisciplinary approach with protocolized management
of sedation as per institutional preferences is recommended.
Opioids have been the mainstay of sedation management, but Patient Management Practice Points
increasingly dexmedetomidine is being used over benzodi- 1. Arrhythmias should be addressed to promote myocar-
azepines. Dosing requirements may be elevated, as drugs are dial recovery and cardiac ejection.
adsorbed into the circuit, tolerance can develop, and hemofil- 2. Vasoactive medications can be useful to manipulate
tration may remove administered drugs.69,70 For example, mid- SVR, complement VA ECMO flow and optimize tissue
azolam, lorazepam, and fentanyl can show reductions greater oxygen delivery (Table 3).
than 50% in levels due to adsorption to the ECMO circuit.71 3. Inotropes may promote cardiac ejection and decom-
A practice of daily interruption of sedation medications on pression but do not facilitate myocardial rest.
ECMO in neonates and infants, particularly if impaired renal 4. Attention should be given to oxygenating blood travers-
and liver function, may be considered to prevent the accu- ing the pulmonary vasculature in the setting of femoral
mulation of sedative agents and reduce tolerance and later VA ECMO to avoid differential hypoxemia.
effects of sedation withdrawal. Assessment and management
of delirium and overlap with signs of sedation withdrawal is an
important consideration and should not be missed in children Weaning Off Extracorporeal Life Support and Decannulation
supported on ECMO.
Successful VA ECMO wean is defined as discontinuation of
mechanical circulatory support because the patient improved
Nutrition
and is expected to recover.78 Assessment of readiness to wean
Enteral nutrition in patients receiving VA ECMO is well tol- should occur at regular intervals of 24–48 hours from the time
erated, provides adequate nutrition, is cost-effective, and has of ECMO cannulation (Figure  1). Factors for consideration
minimal risk.72,73 before weaning VA ECMO (shown in Figure  4) include the

Copyright © Extracorporeal Life Support Organization. Unauthorized reproduction of this article is prohibited.
 ELSO GUIDELINES FOR PEDIATRIC CARDIAC FAILURE 9

patient phenotype being compatible with recovery, preserved conducted over several days. The purpose of weaning ECMO
lung function, end-organ function recovering, and improved is to determine if the patient is ready to trial-off ECMO. The
myocardial function with vasopressors and inotropes at low weaning process varies according to institutional and patient
levels (e.g. epinephrine/norepinephrine ≤0.02–0.05 μg/kg/min factors but would commonly occur over 4–8 hours with a
or dopamine/dobutamine <5 μg/kg/min).79,80 gradual decrease in ECMO flows while optimizing lung ven-
tilation strategy, intravascular volume status, and vasoactive
Timing of Myocardial Recovery medications. The purpose of the trial-off (15 minutes to 2
hours) is to determine if the patient is ready for decannula-
The clinical course of myocardial improvement may be rapid,
tion. Clinical examination, targeted echocardiography, and
for example, 24 hours to 5 days, postcardiotomy or cardiopul-
serial laboratory parameters should be used to assess the
monary bypass, or more prolonged in the case of primary myo-
adequacy of cardiac output and ventricular function during
cardial dysfunction or prolonged ischemia before correction of
ECMO weaning. Once the patient has demonstrated satisfac-
the cardiac lesion.81–85 While the former may be well supported
tory hemodynamics on minimum ECMO support, typically
with ECMO, those with prolonged myocardial dysfunction (>7–
around 50 ml/kg/min, it is reasonable to trial the patient off
10 days) should be considered for other forms of mechanical
ECMO. The main risk during ECMO Trial-Off is circuit throm-
circulatory support as end-organs recover (Figure 4).81,86
bus. Extracorporeal membrane oxygenation should not be
reduced below the lowest compatible with the oxygenator in
Signs of Myocardial Recovery the circuit. The risk of clot formation depends on anticoagula-
While supported on VA ECMO, evidence of myocardial tion, circuit size, existing clot burden, and circuit complex-
recovery includes increasing pulse pressure, increasing sys- ity. Adequate anticoagulation during ECMO trial-off must be
tolic pressure, rising end-tidal CO2, and improving ventricu- maintained (using either infusions or intermittent boluses). It
lar systolic function on echocardiography.80,87–89 Loading is usual to carry out a trial-off ECMO for 1–2 hours before
conditions impact assessment of ventricular systolic function decannulation.
on ECMO, but other parameters (aortic velocity time integral There are two broad trial-off strategies:
≥10 cm, left ventricular ejection fraction >20–25%, and lat- 1. Clamp trial. This is the classic approach. Clamping the
eral mitral annulus peak systolic velocity ≥6 cm/s) under low cannula proximal to the patient (clamp trial) allows com-
flow conditions have been predictive of successful ECMO plete separation of the patient from the circuit; circuit
decannulation in adult patients with cardiogenic shock.80,89 flow is maintained around the bridge. Adequate antico-
agulation should be established in the ECMO circuit to
Special Considerations slow thrombosis of the cannulae and lines during the no-
Left atrial decompression. In patients whose support flow state. Every 10–15 minutes, the cannulae are inter-
includes a left atrial vent or atrial septostomy, lower vent flows, mittently flashed by releasing the clamps and clamping
and a drop in circuit, mixed venous oxygenation coincide the bridge for 15–30 seconds.
with left ventricular recovery. Tolerating left atrial or left ven- 2. Pump-controlled retrograde trial-off is a more recently
tricular vent clamping and subsequent removal are initial steps described technique that has been shown to be a safe,
towards weaning. simple, and reproducible approach. This relies on the
Systemic to pulmonary artery shunts.  Some patients require retrograde flow generated by the patient’s native cardiac
partial occlusion of their systemic to pulmonary artery shunt or output to maintain circuit integrity, provides a ‘stress
right ventricle to pulmonary artery (RV-PA) conduit to achieve test’ to evaluate cardiorespiratory reserve during the trial
sufficient systemic perfusion on ECMO. In these patients, the sur- period off ECMO by creating an obligatory left to right
geon may need to adjust or remove the partial occlusion to bal- shunt.90,91 This method avoids manipulation of the ECMO
ance pulmonary and systemic blood flow during weaning and circuit without insertion of an arteriovenous bridge and
trialing off ECMO. RV-PA conduits, valved or otherwise, usually circuit clamping.
do not require any restriction to balance systemic and pulmonary 3. Some centers omit the trial-off completely and just dis-
blood flows on ECMO. continue ECMO after weaning but leave the cannula
in place, fill them with heparinized saline and run a
heparin flush through the cannula to maintain patency.
Weaning With LA Decompression Practice Points Decannulation is then carried out after several hours if
1. If myocardial recovery has occurred, clamping of the patient remains stable.
the LA/LV line should improve native cardiac output
(increase pulse pressure). Preparation for Extracorporeal Membrane
2. Clamping and removal of the LA/LV line and confir- Oxygenation Wean, Trial-Off, and Decannulation
mation of sustained LV decompression can be the first
component of a staged ECMO wean. Physiologic conditions during the wean should closely
approximate those after decannulation, including optimized
volume status and consideration of inotropy commencing sev-
Weaning and Trailing Off Extracorporeal eral hours before weaning. Preparing for ECMO trial-off and
Membrane Oxygenation decannulation should include:
Weaning is the term applied to the reduction of ECMO 1. Confirm endotracheal tube position. Optimize ventila-
flow, which accompanies myocardial recovery. It may be tion and lung recruitment for adequate oxygenation and

Copyright © Extracorporeal Life Support Organization. Unauthorized reproduction of this article is prohibited.
 10 BROWN ET AL.

ventilation. Consider transition to VV-ECMO if ventilation Risk Factors for Death

is contributing to the failure to wean (Figure 4);
The mortality associated with pediatric cardiac ECMO is
2. Correct hematological and metabolic abnormalities;
between 45% and 50% depending on the indication.1,19,20 Risk
3. If temporary pacing wires are available, connect and test
factors for mortality in children on ECMO for cardiac indica-
pacemaker;
tions include acute renal failure, bleeding, and pre-ECMO
4. Consider the use of inhaled nitric oxide or pulmonary
lactatemia and acidosis with delayed resolution post-ECMO
vasodilator therapy in lesions that may predispose to pul-
initiation and ECMO duration.19,96–101 Primary cardiac diagno-
monary hypertension;
sis has a significant bearing on mortality, in particular patients
5. Optimize preload (central venous pressure [CVP] > 5mm
with single-ventricle physiology.6 If myocardial function remains
Hg) and determine inotropic medication plan (often epi-
poor despite optimization of ECMO support and identification
nephrine 0.02–0.05 µg/kg/min or dopamine/dobutamine
and management of residual lesions, consideration may be
3–5 µg/kg/min);
given to transition to longer-term device support (Figure 4). Such
6. ECMO flows are reduced (by 20 ml/kg/min increments).
transition to longer-term VAD should occur along with assess-
Clinical hemodynamic assessment of HR, BP, and CVP
ment for cardiac transplantation. Transparent discussions with
should determine the timing of further decrements inflow.
patient families and decision-makers must occur regarding the
If hemodynamic parameters remain within acceptable
risk of mortality increasing over the course of longer duration
limits with minimal change post wean, flows can con-
VA ECMO. Setting expectations and goals of care in the event of
tinue to be further reduced;
failure of organ recovery should be built over the ECMO course,
7. Assessment of clinical status, lactate, AVO2 difference,
with shared values informing end-of-life care.
and respiratory gas exchange (arterial blood gas) should
be made at low ECMO flow (50 ml/kg/min);
8. Echocardiography may be used for assessment of sys-
tolic ventricular function, estimation of RV pressures, Acknowledgment
adequacy of volume status, and additional evaluation for The authors thank Elaine Cooley, MSN RN, and Peter Rycus, MPH
valvular regurgitation/stenosis; for the help in the overall process.
9. During weaning, the adequacy of cardiac output and
respiratory gas exchange should be assessed, and ongoing
requirement for support should be determined (Figure 4). References
1. Barbaro RP, Paden ML, Guner YS, et al; ELSO member centers:
Veno-arterial ECMO decannulation can be performed in the
Pediatric Extracorporeal Life Support Organization Registry
operating room or the intensive care unit according to local International Report 2016. ASAIO J 63: 456–463, 2017.
protocols. Potential complications should be considered, 2. Organization ELS: Registry Report International Summary.
including low cardiac output, pulmonary hypertension, bleed- Extracorporeal Life Support Organization. 2020. Available at:
ing, arrhythmias, and the need to rapidly reinitiate VA ECMO https://www.elso.org/Registry/Statistics/InternationalSummary.
aspx. Accessed September 6, 2020.
support. Candidacy for recannulation should be discussed 3. Mascio CE, Austin EH III, Jacobs JP, et al: Perioperative mechanical
before attempting weaning and decannulation since survival circulatory support in children: An analysis of the Society of
declines with subsequent ECMO runs.82,92,93 Thoracic Surgeons Congenital Heart Surgery Database. J Thorac
Decannulation is a surgical operation that should be super- Cardiovasc Surg 147: 658–64: discussion 664, 2014.
4. Misfeldt AM, Kirsch RE, Goldberg DJ, et al: Outcomes of single-
vised by a suitably qualified surgeon and undertaken with for-
ventricle patients supported with extracorporeal membrane
mal general anesthesia and aseptic technique. The circuit tubing oxygenation. Pediatr Crit Care Med 17: 194–202, 2016.
proximal to the cannula should be prepared and placed in the 5. Sherwin ED, Gauvreau K, Scheurer MA, et al: Extracorporeal mem-
operative field. When removing a venous cannula, venous air brane oxygenation after stage 1 palliation for hypoplastic left heart
embolism may occur through the cannula side holes if the syndrome. J Thorac Cardiovasc Surg 144: 1337–1343, 2012.
6. Kulik TJ, Moler FW, Palmisano JM, et al: Outcome-associated fac-
patient is breathing spontaneously. This is prevented by either a tors in pediatric patients treated with extracorporeal membrane
Valsalva maneuver on the ventilator or by the use of short-term oxygenator after cardiac surgery. Circulation 94(Suppl 9): II63–
pharmacological paralysis. For peripheral cannulation, vascular II68, 1996.
reconstruction may be required depending on the cannulation 7. Thiagarajan RR: Extracorporeal membrane oxygenation for car-
technique. For central cannulation, chest closure can be consid- diac indications in children. Pediatr Crit Care Med 17(8 Suppl
1): S155–S159, 2016.
ered if the ventricular function has improved significantly and 8. Jolley M, Thiagarajan RR, Barrett CS, et al: Extracorporeal mem-
indices of end-organ oxygen delivery are reassuring. brane oxygenation in patients undergoing superior cavopulmo-
nary anastomosis. J Thorac Cardiovasc Surg 148: 1512–1518,
2014.
Failure to Wean
9. Rood KL, Teele SA, Barrett CS, et al: Extracorporeal membrane
Failure to progress according to expected timing for etiology oxygenation support after the Fontan operation. J Thorac
Cardiovasc Surg 142: 504–510, 2011.
of cardiovascular failure should be investigated early to assess 10. Marino BS, Tabbutt S, MacLaren G, et al; American Heart

for residual lesions, inadequacy of hemodynamic or pulmonary Association Congenital Cardiac Defects Committee of the
support, or alternate diagnoses.20,31 Cardiac catheterization may Council on Cardiovascular Disease in the Young; Council
reveal residual lesions amenable to surgical or interventional cor- on Clinical Cardiology; Council on Cardiovascular and
rection not apparent on echocardiography.94,95 With each attempt Stroke Nursing; Council on Cardiovascular Surgery and
Anesthesia; and Emergency Cardiovascular Care Committee:
at weaning, the underlying cause of failure to wean (pulmonary, Cardiopulmonary resuscitation in infants and children with car-
cardiovascular, or cardiopulmonary) should be determined to diac disease: A scientific statement from the American Heart
target optimal mechanical support strategies (Figure 4). Association. Circulation 137: e691–e782, 2018.

Copyright © Extracorporeal Life Support Organization. Unauthorized reproduction of this article is prohibited.
 ELSO GUIDELINES FOR PEDIATRIC CARDIAC FAILURE 11

11. Girotra S, Spertus JA, Li Y, Berg RA, Nadkarni VM, Chan PS;
supported with venoarterial extracorporeal membrane oxygen-
American Heart Association Get With the Guidelines– ation*. Pediatr Crit Care Med 15: 355–361, 2014.
Resuscitation Investigators: Survival trends in pediatric 28. Werho DK, Pasquali SK, Yu S, et al; ELSO Member Centers:

in-hospital cardiac arrests: An analysis from Get With the Epidemiology of stroke in pediatric cardiac surgical patients
Guidelines-Resuscitation. Circ Cardiovasc Qual Outcomes 6: supported with extracorporeal membrane oxygenation. Ann
42–49, 2013. Thorac Surg 100: 1751–1757, 2015.
12. Gupta P, Jacobs JP, Pasquali SK, et al: Epidemiology and outcomes 29. Schad CA, Fallon BP, Monteagudo J, et al: Routine use of distal
after in-hospital cardiac arrest after pediatric cardiac surgery. arterial perfusion in pediatric femoral venoarterial extracorpo-
Ann Thorac Surg 98: 2138–43; discussion 2144, 2014. real membrane oxygenation. Artif Organs 41: 11–16, 2017.
13. Lasa JJ, Rogers RS, Localio R, et al: Extracorporeal cardiopulmonary 30. Fraser CD III, Kovler ML, Guzman W Jr, et al: Pediatric femoral
resuscitation (E-CPR) during pediatric in-hospital cardiopulmo- arterial cannulations in extracorporeal membrane oxygenation:
nary arrest is associated with improved survival to discharge: A review and strategies for optimization. ASAIO J 65: 636–641,
A report from the American Heart Association’s Get With The 2018.
Guidelines-Resuscitation (GWTG-R) Registry. Circulation 133: 31. Callahan R, Trucco SM, Wearden PD, Beerman LB, Arora G,

165–176, 2016. Kreutzer J: Outcomes of pediatric patients undergoing cardiac
14. Meert K, Slomine BS, Silverstein FS, et al; Therapeutic Hypothermia catheterization while on extracorporeal membrane oxygen-
after Paediatric Cardiac Arrest (THAPCA) Trial Investigators: ation. Pediatr Cardiol 36: 625–632, 2015.
One-year cognitive and neurologic outcomes in survivors 32. Hacking DF, Best D, d’Udekem Y, et al: Elective decompression of
of paediatric extracorporeal cardiopulmonary resuscitation. the left ventricle in pediatric patients may reduce the duration
Resuscitation 139: 299–307, 2019. of venoarterial extracorporeal membrane oxygenation. Artif
15. Church JT, Kim AC, Erickson KM, et al: Pushing the boundaries of Organs 39: 319–326, 2015.
ECLS: Outcomes in <34 week EGA neonates. J Pediatr Surg 52: 33. Eastaugh LJ, Thiagarajan RR, Darst JR, McElhinney DB, Lock JE,
1810–1815, 2017. Marshall AC: Percutaneous left atrial decompression in patients
16. Gupta P, Gossett JM, Rycus PT, Prodhan P: Extracorporeal mem- supported with extracorporeal membrane oxygenation for car-
brane oxygenation in children with heart disease and down diac disease. Pediatr Crit Care Med 16: 59–65, 2015.
syndrome: A multicenter analysis. Pediatr Cardiol 35: 1421– 34. Zampi JD, Alghanem F, Yu S, et al: Relationship between time to
1428, 2014. left atrial decompression and outcomes in patients receiving
17. Furlong-Dillard JM, Amula V, Bailly DK, Bleyl SB, Wilkes J, Bratton venoarterial extracorporeal membrane oxygenation support:
SL: Use of extracorporeal membrane oxygenation and mortality A multicenter pediatric interventional cardiology early-career
in pediatric cardiac surgery patients with genetic conditions: A society study. Pediatr Crit Care Med 20: 728–736, 2019.
multicenter analysis. Pediatr Crit Care Med 18: 850–858, 2017. 35. Parekh D, Jeewa A, Tume SC, et al: Percutaneous mechanical cir-
18. Bembea MM, Ng DK, Rizkalla N, et al; American Heart
culatory support using Impella devices for decompensated car-
Association’s Get With The Guidelines – Resuscitation diogenic shock: A pediatric heart center experience. ASAIO J
Investigators: Outcomes after extracorporeal cardiopulmonary 64: 98–104, 2018.
resuscitation of pediatric In-hospital cardiac arrest: A report 36. Di Nardo M, MacLaren G, Marano M, Cecchetti C, Bernaschi P,
from the get with the guidelines-resuscitation and the extracor- Amodeo A: ECLS in pediatric cardiac patients. Front Pediatr 4:
poreal life support organization registries. Crit Care Med 47: 109, 2016.
e278–e285, 2019. 37. Bauser-Heaton H, Nguyen C, Tacy T, Axelrod D: A novel approach
19. Ford MA, Gauvreau K, McMullan DM, et al: Factors associated to the management of critically ill neonatal Ebstein’s anomaly:
with mortality in neonates requiring extracorporeal membrane Veno-venous extracorporeal membrane oxygenation to pro-
oxygenation for cardiac indications: Analysis of the extracor- mote right ventricular recovery. Ann Pediatr Cardiol 8: 67–70,
poreal life support organization registry Data. Pediatr Crit Care 2015.
Med 17: 860–870, 2016. 38. Brogan TV LL, Lorusso R, MacLaren G, Peek G: Extracorporeal
20. Howard TS, Kalish BT, Wigmore D, et al: Association of extracor- Life Support: The ELSO Red Book. 5th ed. Ann Arbor, MI,
poreal membrane oxygenation support adequacy and residual Extracorporeal Life Support Organization, 2017.
lesions with outcomes in neonates supported after cardiac sur- 39. Cashen K, Reeder R, Dalton HJ, et al; Eunice Kennedy Shriver
gery. Pediatr Crit Care Med 17: 1045–1054, 2016. National Institute of Child Health and Human Development
21. Polito A, Barrett CS, Rycus PT, Favia I, Cogo PE, Thiagarajan RR: Collaborative Pediatric Critical Care Research Network
Neurologic injury in neonates with congenital heart disease (CPCCRN): Hyperoxia and hypocapnia during pediatric
during extracorporeal membrane oxygenation: An analysis of extracorporeal membrane Oxygenation: Associations With
extracorporeal life support organization registry data. ASAIO J Complications, Mortality, and Functional Status Among
61: 43–48, 2015. Survivors. Pediatr Crit Care Med 19: 245–253, 2018.
22. MacLaren G, Butt W, Best D, Donath S: Central extracorporeal 40. Diehl A, Burrell AJC, Udy AA, et al: Association between arterial
membrane oxygenation for refractory pediatric septic shock. carbon dioxide tension and clinical outcomes in venoarterial
Pediatr Crit Care Med 12: 133–136, 2011. extracorporeal membrane oxygenation. Crit Care Med 48: 977–
23. Zaleski KL, DiNardo JA, Nasr VG: Bivalirudin for pediatric proce- 984, 2020.
dural anticoagulation: A narrative review. Anesth Analg 128: 41. Cavayas YA, Munshi L, Del Sorbo L, Fan E: The early change in
43–55, 2019. PaCO2 after extracorporeal membrane oxygenation initiation
24. Brunetti MA, Gaynor JW, Retzloff LB, et al: Characteristics, risk fac- is associated with neurological complications. Am J Respir Crit
tors, and outcomes of extracorporeal membrane oxygenation use Care Med 201: 1525–1535, 2020.
in pediatric cardiac ICUs: A report from the pediatric cardiac crit- 42. Bembea MM, Annich G, Rycus P, Oldenburg G, Berkowitz I,

ical care consortium registry. Pediatr Crit Care Med 19: 544–552, Pronovost P: Variability in anticoagulation management of
2018. patients on extracorporeal membrane oxygenation: An interna-
25. Alore EA, Fallon SC, Thomas JA, Vogel AM: Outcomes after extra- tional survey. Pediatr Crit Care Med 14: e77–e84, 2013.
corporeal life support cannulation in pediatric patients with tri- 43. Bembea MM, Schwartz JM, Shah N, et al: Anticoagulation moni-
somy 13 and trisomy 18. J Surg Res 257: 260–266, 2021. toring during pediatric extracorporeal membrane oxygenation.
26. Johnson K, Jarboe MD, Mychaliska GB, et al; ELSO/Euro-ELSO ASAIO J 59: 63–68, 2013.
Neurologic Outcomes Working Group: Is there a best approach 44. Monagle P, Barnes C, Ignjatovic V, et al: Developmental haemo-
for extracorporeal life support cannulation: A review of the stasis. Impact for clinical haemostasis laboratories. Thromb
extracorporeal life support organization. J Pediatr Surg 53: Haemost 95: 362–372, 2006.
1301–1304, 2018. 45. Andrew M, Vegh P, Johnston M, Bowker J, Ofosu F, Mitchell L:
27. Teele SA, Salvin JW, Barrett CS, et al: The association of carotid Maturation of the hemostatic system during childhood. Blood
artery cannulation and neurologic injury in pediatric patients 80: 1998–2005, 1992.

Copyright © Extracorporeal Life Support Organization. Unauthorized reproduction of this article is prohibited.
 12 BROWN ET AL.

46. Northrop MS, Sidonio RF, Phillips SE, et al: The use of an extra- Centers: Lung rest during extracorporeal membrane oxygen-
corporeal membrane oxygenation anticoagulation labora- ation for neonatal respiratory failure-practice variations and
tory protocol is associated with decreased blood product use, outcomes. Pediatr Crit Care Med 18: 667–674, 2017.
decreased hemorrhagic complications, and increased circuit 62. Marhong JD, Telesnicki T, Munshi L, Del Sorbo L, Detsky M, Fan
life. Pediatr Crit Care Med 16: 66–74, 2015. E: Mechanical ventilation during extracorporeal membrane
47. Hamzah M, Jarden AM, Ezetendu C, Stewart R: Evaluation of
oxygenation. An international survey. Ann Am Thorac Soc 11:
bivalirudin as an alternative to heparin for systemic antico- 956–961, 2014.
agulation in pediatric extracorporeal membrane oxygenation. 63. Kimmoun A, Roche S, Bridey C, et al: Prolonged prone positioning
Pediatr Crit Care Med 21: 827–834, 2020. under VV-ECMO is safe and improves oxygenation and respira-
48. Snyder CW, Goldenberg NA, Nguyen ATH, Smithers CJ, Kays DW: tory compliance. Ann Intensive Care 5: 35, 2015.
A perioperative bivalirudin anticoagulation protocol for neo- 64. Haefner SM, Bratton SL, Annich GM, Bartlett RH, Custer JR:

nates with congenital diaphragmatic hernia on extracorporeal Complications of intermittent prone positioning in pediatric
membrane oxygenation. Thromb Res 193: 198–203, 2020. patients receiving extracorporeal membrane oxygenation for
49. Coughlin MA, Bartlett RH: Anticoagulation for extracorporeal life respiratory failure. Chest 123: 1589–1594, 2003.
support: Direct thrombin inhibitors and heparin. ASAIO J 61: 65. Dalton HJ, Garcia-Filion P, Holubkov R, et al; Eunice Kennedy
652–655, 2015. Shriver National Institute of Child Health and Human
50. Dalton HJ, Reeder R, Garcia-Filion P, et al; Eunice Kennedy Shriver Development Collaborative Pediatric Critical Care Research
National Institute of Child Health and Human Development Network: Association of bleeding and thrombosis with out-
Collaborative Pediatric Critical Care Research Network: Factors come in extracorporeal life support. Pediatr Crit Care Med 16:
associated with bleeding and thrombosis in children receiving 167–174, 2015.
extracorporeal membrane oxygenation. Am J Respir Crit Care 66. Polito A, Barrett CS, Wypij D, et al: Neurologic complications in
Med 196: 762–771, 2017. neonates supported with extracorporeal membrane oxygen-
51. Guimbretiere G, Anselmi A, Roisne A, et al: Prognostic impact of ation. An analysis of ELSO registry data. Intensive Care Med 39:
blood product transfusion in VA and VV ECMO. Perfusion 34: 1594–1601, 2013.
246–253, 2018. 67. Lin JJ, Banwell BL, Berg RA, et al: Electrographic seizures in chil-
52. O’Halloran CP, Alexander PMA, Andren KG, et al: RBC expo- dren and neonates undergoing extracorporeal membrane oxy-
sure in pediatric extracorporeal membrane oxygenation: genation. Pediatr Crit Care Med 18: 249–257, 2017.
Epidemiology and factors associated with large blood transfu- 68. Cengiz P, Seidel K, Rycus PT, Brogan TV, Roberts JS: Central ner-
sion volume. Pediatr Crit Care Med 19: 767–774, 2018. vous system complications during pediatric extracorporeal life
53. Bembea MM, Cheifetz IM, Fortenberry JD, et al; Pediatric Critical support: Incidence and risk factors. Crit Care Med 33: 2817–
Care Transfusion and Anemia Expertise Initiative (TAXI); 2824, 2005.
Pediatric Critical Care Blood Research Network (BloodNet), 69. Himebauch AS, Kilbaugh TJ, Zuppa AF: Pharmacotherapy dur-
and the Pediatric Acute Lung Injury and Sepsis Investigators ing pediatric extracorporeal membrane oxygenation: A review.
(PALISI) Network: Recommendations on the indications for Expert Opin Drug Metab Toxicol 12: 1133–1142, 2016.
RBC transfusion for the critically ill child receiving support 70. Anton-Martin P, Modem V, Taylor D, Potter D, Darnell-Bowens C:
from extracorporeal membrane oxygenation, ventricular assist, A retrospective study of sedation and analgesic requirements
and renal replacement therapy devices from the pediatric criti- of pediatric patients on extracorporeal membrane oxygenation
cal care transfusion and anemia expertise initiative. Pediatr Crit (ECMO) from a single-center experience. Perfusion 32: 183–
Care Med 19(9S Suppl 1): S157–S162, 2018. 191, 2017.
54. Valentine SL, Bembea MM, Muszynski JA, et al; Pediatric Critical 71. Harthan AA, Buckley KW, Heger ML, Fortuna RS, Mays K:

Care Transfusion and Anemia Expertise Initiative (TAXI); Medication adsorption into contemporary extracorporeal mem-
Pediatric Critical Care Blood Research Network (BloodNet), brane oxygenator circuits. J Pediatr Pharmacol Ther 19: 288–
and the Pediatric Acute Lung Injury and Sepsis Investigators 295, 2014.
(PALISI) Network: Consensus recommendations for RBC trans- 72. Pettignano R, Heard M, Davis R, Labuz M, Hart M: Total enteral
fusion practice in critically ill children from the pediatric criti- nutrition versus total parenteral nutrition during pediatric extra-
cal care transfusion and anemia expertise initiative. Pediatr Crit corporeal membrane oxygenation. Crit Care Med 26: 358–363,
Care Med 19: 884–898, 2018. 1998.
55. O’Halloran CP, Andren KG, Mecklosky J, et al: Mortality and fac- 73. Hanekamp MN, Spoel M, Sharman-Koendjbiharie I, Peters JW,
tors associated with hemorrhage during pediatric extracorpo- Albers MJ, Tibboel D: Routine enteral nutrition in neonates on
real membrane oxygenation. Pediatr Crit Care Med 21: 75–81, extracorporeal membrane oxygenation. Pediatr Crit Care Med
2020. 6: 275–279, 2005.
56. Hoffman TM, Wernovsky G, Atz AM, et al: Efficacy and safety of 74. Hsu MS, Chiu KM, Huang YT, Kao KL, Chu SH, Liao CH: Risk fac-
milrinone in preventing low cardiac output syndrome in infants tors for nosocomial infection during extracorporeal membrane
and children after corrective surgery for congenital heart dis- oxygenation. J Hosp Infect 73: 210–216, 2009.
ease. Circulation 107: 996–1002, 2003. 75. Elerian LF, Sparks JW, Meyer TA, et al: Usefulness of surveillance
57. Brower RG, Fessler HE: Mechanical ventilation in acute lung
cultures in neonatal extracorporeal membrane oxygenation.
injury and acute respiratory distress syndrome. Clin Chest Med ASAIO J 47: 220–223, 2001.
21: 491–510, viii, 2000. 76. Agus MS, Steil GM, Wypij D, et al; SPECS Study Investigators: Tight
58. Brower RG, Matthay MA, Morris A, Schoenfeld D, Thompson BT, glycemic control versus standard care after pediatric cardiac
Wheeler A; Acute Respiratory Distress Syndrome Network: surgery. N Engl J Med 367: 1208–1219, 2012.
Ventilation with lower tidal volumes as compared with tradi- 77. Agus MS, Wypij D, Hirshberg EL, et al; HALF-PINT Study

tional tidal volumes for acute lung injury and the acute respira- Investigators and the PALISI Network: Tight glycemic control in
tory distress syndrome. N Engl J Med 342: 1301–1308, 2000. critically ill children. N Engl J Med 376: 729–741, 2017.
59. Bein T, Weber-Carstens S, Goldmann A, et al: Lower tidal volume 78. ELSO Registry Committee: ELSO Registry Data Definitions. Ann
strategy (≈3 ml/kg) combined with extracorporeal CO2 removal Arbor, MI, Extracorporeal Life Support Organization, 2020,
versus ‘conventional’ protective ventilation (6 ml/kg) in severe p. 113. Available at: https://www.elso.org/Portals/0/Files/PDF/
ARDS: The prospective randomized Xtravent-study. Intensive ELSO%20Registry%20Data%20Definitions%2001_13_2021.
Care Med 39: 847–856, 2013. pdf. Accessed March 17, 2021.
60. Schmidt M, Stewart C, Bailey M, et al: Mechanical ventilation 79. Howard TS, Kalish BT, Rajagopal SK, et al: Factors associated with
management during extracorporeal membrane oxygenation mortality in children who successfully wean from extracorporeal
for acute respiratory distress syndrome: A retrospective interna- membrane oxygenation. Pediatr Crit Care Med 19: 875–883,
tional multicenter study. Crit Care Med 43: 654–664, 2015. 2018.
61. Alapati D, Aghai ZH, Hossain MJ, Dirnberger DR, Ogino MT, 80. Fried JA, Masoumi A, Takeda K, Brodie D: How I approach wean-
Shaffer TH; Extracorporeal Life Support Organization Member ing from venoarterial ECMO. Crit Care 24: 307, 2020.

Copyright © Extracorporeal Life Support Organization. Unauthorized reproduction of this article is prohibited.
 ELSO GUIDELINES FOR PEDIATRIC CARDIAC FAILURE 13

81. Yarlagadda VV, Maeda K, Zhang Y, et al: Temporary circulatory weaning from venoarterial extracorporeal membrane oxygen-
support in U.S. children awaiting heart transplantation. J Am ation. Eur J Cardiothorac Surg 56: 277–284, 2019.
Coll Cardiol 70: 2250–2260, 2017. 92. Bohuta L, d’Udekem Y, Best D, Alexander P, Barozzi L, Butt W:

82. d’Udekem Y, Shime N, Lou S, MacLaren G: Recurrent or prolonged Outcomes of second-run extracorporeal life support in children: A
mechanical circulatory support: Bridge to recovery or road to single-institution experience. Ann Thorac Surg 92: 993–996, 2011.
nowhere? Pediatr Crit Care Med 14(5 Suppl 1): S69–S72, 2013. 93. Shuhaiber J, Thiagarajan RR, Laussen PC, Fynn-Thompson F, del
83. Rajagopal SK, Almond CS, Laussen PC, Rycus PT, Wypij D,
Nido P, Pigula F: Survival of children requiring repeat extracor-
Thiagarajan RR: Extracorporeal membrane oxygenation for poreal membrane oxygenation after congenital heart surgery.
the support of infants, children, and young adults with acute Ann Thorac Surg 91: 1949–1955, 2011.
myocarditis: A review of the Extracorporeal Life Support 94. Agarwal HS, Hardison DC, Saville BR, et al: Residual lesions in
Organization registry. Crit Care Med 38: 382–387, 2010. postoperative pediatric cardiac surgery patients receiving extra-
84. Wernovsky G, Giglia TM, Jonas RA, Mone SM, Colan SD, Wessel corporeal membrane oxygenation support. J Thorac Cardiovasc
DL: Course in the intensive care unit after ‘preparatory’ pul- Surg 147: 434–441, 2014.
monary artery banding and aortopulmonary shunt placement 95. Abraham BP, Gilliam E, Kim DW, Wolf MJ, Vincent RN, Petit CJ:
for transposition of the great arteries with low left ventricular Early catheterization after initiation of extracorporeal membrane
pressure. Circulation 86(Suppl 5): II133–II139, 1992. oxygenation support in children is associated with improved sur-
85. Wernovsky G, Wypij D, Jonas RA, et al: Postoperative course and vival. Catheter Cardiovasc Interv 88: 592–599, 2016.
hemodynamic profile after the arterial switch operation in neo- 96. Merrill ED, Schoeneberg L, Sandesara P, et al: Outcomes after pro-
nates and infants. A comparison of low-flow cardiopulmonary longed extracorporeal membrane oxygenation support in children
bypass and circulatory arrest. Circulation 92: 2226–2235, 1995. with cardiac disease–Extracorporeal Life Support Organization
86. Almond CS, Singh TP, Gauvreau K, et al: Extracorporeal membrane registry study. J Thorac Cardiovasc Surg 148: 582–588, 2014.
oxygenation for bridge to heart transplantation among children in   97. Duncan BW, Hraska V, Jonas RA, et al: Mechanical circulatory
the United States: Analysis of data from the Organ Procurement support in children with cardiac disease. J Thorac Cardiovasc
and Transplant Network and Extracorporeal Life Support Surg 117: 529–542, 1999.
Organization Registry. Circulation 123: 2975–2984, 2011.   98.  Kumar TK, Zurakowski D, Dalton H, et al: Extracorporeal mem-
87. Park BW, Seo DC, Moon IK, et al: Pulse pressure as a prognos- brane oxygenation in postcardiotomy patients: Factors influ-
tic marker in patients receiving extracorporeal life support. encing outcome. J Thorac Cardiovasc Surg 140: 330–336.e2,
Resuscitation 84: 1404–1408, 2013. 2010.
88. Naruke T, Inomata T, Imai H, et al: End-tidal carbon dioxide con-   99.  Fleming GM, Sahay R, Zappitelli M, et al: The incidence of acute
centration can estimate the appropriate timing for weaning off kidney injury and its effect on neonatal and pediatric extracor-
from extracorporeal membrane oxygenation for refractory cir- poreal membrane oxygenation outcomes: A multicenter report
culatory failure. Int Heart J 51: 116–120, 2010. from the kidney intervention during extracorporeal membrane
89. Aissaoui N, Luyt CE, Leprince P, et al: Predictors of successful oxygenation study group. Pediatr Crit Care Med 17: 1157–
extracorporeal membrane oxygenation (ECMO) weaning after 1169, 2016.
assistance for refractory cardiogenic shock. Intensive Care Med 100.  Lv L, Long C, Liu J, et al: Predictors of acute renal failure during
37: 1738–1745, 2011. extracorporeal membrane oxygenation in pediatric patients
90. Westrope C, Harvey C, Robinson S, Speggiorin S, Faulkner G, after cardiac surgery. Artif Organs 40: E79–E83, 2016.
Peek GJ: Pump controlled retrograde trial off from VA-ECMO. 101.  Alsoufi B, Wolf M, Botha P, et al: Late outcomes of infants sup-
ASAIO J 59: 517–519, 2013. ported by extracorporeal membrane oxygenation following
91. Pandya NR, Daley M, Mattke A, et al: A comparison of pump- the Norwood operation. World J Pediatr Congenit Heart Surg
controlled retrograde trial off to arterio-venous bridging for 6: 9–17, 2015.

Copyright © Extracorporeal Life Support Organization. Unauthorized reproduction of this article is prohibited.