Definitions and Results

Timing: Identified during ECMO support (A-K) or the first intracranial imaging post-ECMO or autopsy (A-F)*

Reporting: Events are graded according to Designation of event (A-K) and Type of patient consequence.

A. Ischemic stroke: evidence of new focal neurological deficits corresponding to a vascular territory or territories within the brain meeting 1 major or 2 minor criteria.

  MAJOR: Radiographic or pathologic confirmation of ischemic stroke†

MINOR: Focal neurological signs without radiographic or pathologic confirmation AND with 2 or more of the following:

·         Persistent focal neurological dysfunction localizing to vascular territory/territories (can include new focal electroencephalography changes)

·         Persistent focal neurological dysfunction temporally correlated with known clot dislodgement

·         Expert vascular neurologist or pediatric neurocritical care specialist diagnosis of stroke

B. Intra-axial hemorrhage without mass effect†: includes petechial bleeding into the brain parenchyma, microhemorrhage, and hemorrhagic conversion of ischemic stroke without space-occupying effect.‡ Includes low grade intraventricular hemorrhage (IVH) Grade I-II.

C. Intra-axial hemorrhage with mass effect†: includes parenchymal hemorrhage and intraventricular hemorrhage (IVH) Grade III-IV resulting in mass effect or obstructive hydrocephalus.‡

D. Extra-axial hemorrhage†: Bleeding into the subarachnoid, subdural, or epidural space that is confined within defined compartment.‡

E. Mixed intra-axial and extra-axial hemorrhage†: includes components of D and either B or C.‡

F. Diffuse ischemic injury†: acute diffuse brain injury. Patterns of injury may include global, cortical, subcortical and watershed distribution.

G. Seizure: Evidence of seizure not attributable to preexisting epilepsy, meeting 1 major or 2 minor criteria

MAJOR: Electroencephalographic (EEG) seizure

MINOR: A clinical event of transient signs or symptoms with a clear start and finish, presumed to be due to abnormal excessive or synchronous neuronal activity in the brain. This event occurred at a time when EEG was not in place and requires confirmation with 2 of the following:

·         EEG with evidence of focal neuronal dysfunction and cortical hyperexcitability (e.g., recurrent regional epileptiform abnormalities)

·         Diagnosed by expert neurologist based on semiology and/or response to treatment.

I. Status epilepticus

·         ≥5 minutes of continuous convulsive activity (clinical or EEG)

·         ≥10 minutes continuous non-convulsive seizure (clinical or EEG) or ≥12 minutes total in a 60-minute epoch (EEG)

J. Neuropathy: new neuropathy or plexopathy attributable to ECMO. Includes femoral, sciatic or phrenic neuropathy; brachial or lumbosacral plexopathy; 2nd or 3rd order Horner syndrome. 

K. Death by neurologic criteria: meeting criteria per institutional policy or applicable legal statute.

* This excludes pre-existing strokes, intracranial hemorrhage and diffuse ischemic injury. This is particularly relevant to neonatal patients in whom extra-axial hemorrhage may be a complication of a difficult or traumatic delivery.

† Radiographically by cranial ultrasound, computed tomography, or magnetic resonance imaging or pathologically on autopsy

‡ Not caused by trauma

Patient Consequence for ECMO-CENTRAL ARC Neurologic Adverse Event

Type 1: No change in PCPC

Type 2: 1-point increase in PCPC

Type 3: 2-point increase in PCPC, or increase to PCPC >4

Type 4: Fatal

PCPC, Pediatric Cerebral Performance Category.  A PCPC of 4 is a patient with severe disability who is conscious but dependent on others for daily support because of impaired brain function.

* PCPC Baseline defined as pre-morbid PCPC (may be obtained retrospectively) prior to hospital encounter for ECMO. Comparison PCPC defined as PCPC at hospital discharge.

† If a patient would qualify for multiple types, select the most severe type. For example, if a patient has a 1-point increase from a baseline PCPC of 4 and at hospital discharge has a PCPC of 5, classify as a Type 3 consequence rather than Type 2. If a patient’s pre-ECMO PCPC was 5 and their outcome is death, classify as Type 4 rather than Type 2 (increase of 1 point) or Type 3 (PCPC>4).

Timing: Adjudicated daily (Q24) or more frequently during ECMO support and for up to 6 hours post-decannulation.

Reporting: Each event will be graded according to Type of event.

Type 1. Overt bleeding associated with any one of the following:

• Measurable blood loss of <2 mL/kg/h (up to 20 mL/kg/24-hour period)* or bloody dressings changed no more frequently than every 6 hours

• Clinically insignificant bleeding (e.g., from the endotracheal tube, nasogastric-tube, in-dwelling urinary catheter, mucosal, cutaneous including petechiae or purpura)

• Actions included observation, except for the use of local hemostatic agents only (no administration of systemic pharmacologic agents or blood products)

Type 2. Overt bleeding associated with any one of the following:

• Measurable blood loss of 2-5 mL/kg/h for 2 hours or more (up to 40 mL/kg/24-hour period)†

• Hb-drop: Hb-drop <3 g/dL (<30 g/L) over < 12 hours but with no impaired O2-delivery capacity: lactate < 3 mmol/L (<27 mg/dL) or falling over the last 12 hours

• GI-bleeding, pulmonary hemorrhage, mediastinal cannulation site bleeding, or surgical site bleeding more than Type 1 but does not meet the criteria of Type 3 bleeding

Type 3. Overt bleeding associated with any two of the following:

• >5 mL/kg/h for 2 hours or more (or >40 mL/kg/24-hour period)‡

• Hemodynamic instability associated with the bleeding in absence of other causes

• Increase in vasoactive agents in response to hemodynamic instability

• Volume administration to maintain pump preload

• Severe Hb-drop: Hb-drop (absolute value) >3 g/dL (>30 g/L) within 12 hours OR Impaired O2-delivery capacity evidenced by lactate > 3 mmol/L (> 27 mg/dL) and increasing in two consecutive measurements at least 2 hours apart

• Causing cardiac tamponade or compartment syndrome (abdomen or muscular compartment)

Type 4. Fatal

* For patients >50 kg, total measurable blood loss of <100 mL/hr for 2 hours or more (up to 1000 mL/24-hour period)

† For patients >50 kg, total measurable blood loss of 100-250 mL/hr for 2 hours or more (up to 2000 mL/24-hour period)

‡ For patients >50 kg, total measurable blood loss of >250 mL/hr for 2 hours or more (or >2000 mL/24-hour period) 

Timing: During ECMO support only. The timing of device adverse event should include documentation of the number of hours of support at which the adverse event occurred.

ECMO device malfunction is organized into four hierarchical reasons for device exchange according to degree of the failure, relatedness, and abruptness. As such, Type 1 (structural failure) is the loss of separation of blood and air from compromised structural integrity of the device and is likely attributable to a given circuit component. Type 2 (functional failure) is a complete and abrupt loss of mechanical operation, but it may be more difficult to attribute to a given circuit component. Type 3 (device dysfunction) is the reduced performance of the component device’s dedicated task at optimal parameters; the attribution to a given circuit component is variable. Type 4 (device replacement) is the replacement of the device without identified failure, but this does not rule out dysfunction and could still result in patient consequence. Patient consequence as result of the device exchange is measured separately.

a.   Blood pump or whole system replaced due to blood pump

b.  Membrane lung or whole system replaced due to membrane lung

c.       Cannula or whole system replaced due to cannula*

d.      Tubing or whole system replaced due to tubing†

e.   Tubing connectors or whole system replaced due to tubing connector

f.        Heater unit or whole system replaced due to heater unit

g.   ECMO console or whole system replaced due to ECMO console

h.      Whole system replaced due to an identified dysfunction that is not able to be localized to a specific ECMO component device.

Reason for Device Replacement

Type 1: Structural failure (integrity) is anything that impacts the integrity of the device separation of blood and air leading to blood leak, air entrainment or embolus and managed with replacement or separation from ECMO, or resulting in patient death.

Type 2: Functional failure

• (mechanical operation) is failure or presumed failure of normal mechanical operation and managed with replacement or separation from ECMO, or resulting in patient death.

• (electrical operation) is failure or presumed failure of normal operation of the blood pump, some membrane lungs, heat exchanger, ECMO console or the whole system (when integrated).

Type 3: Device dysfunction determined by failure of the device to perform its dedicated task at optimal parameters, managed with replacement or separation from ECMO, or resulting in patient death.

Type 4: Device replacement (without evidence of dysfunction) is replacement of a device without documented failure of the device. Often may be due to a center-specific protocol regarding longevity of use.

* Cannula dysfunction does not require replacement; but requires demonstration of visual evidence of thrombus within the cannula leading to an intervention for thrombi removal.

† Tubing dysfunction can include replacement, segment exchange or cutout.

Patient Consequence for ECMO-Central ARC Device Malfunction Adverse Event

Grade 1: Transient, asymptomatic or mild symptoms; Clinical or diagnostic observations only.

Grade 2: Maintenance of patient vital signs that require an escalation in non-extracorporeal life support.

Grade 3: Hypotension or hypoxemia despite medical interventions.

Grade 4: Cardio-pulmonary arrest requiring cardiopulmonary resuscitation.

Grade 5: Fatal 

Timing: Adjudicated daily (Q24) or more frequently during ECMO support only.

Reporting: Each event will be graded according to Stage of severity (1-3) and chronicity of event (acute injury to chronic disease).

AKI is defined as any of the following occurring during ECMO Support:

• Increase in serum creatinine by ≥ 0.3 mg/dL (≥ 26.5 micromol/L) within a 48-hour period
  OR

• Increase in serum creatinine to ≥ 1.5 times baseline* (for measured serum creatinine ≥ 0.5 mg/dL or ≥ 44.2 micromol/L) over 7 days
  OR

• Urine output < 0.5 ml/kg/hour for 6 hours in children, or ≤ 1 ml/kg/hour over 24 hours in neonates
  OR

• Initiation of renal replacement therapy (RRT) only for the treatment or prevention of complications of kidney injury†

AKI is staged for severity according to the following criteria:

Stage 1

• Increase in serum creatinine to 150% to 199% (1.5-1.99x increase compared with baseline*) or increase of > 0.3 mg/dL (> 26.4 micromol/L)
  OR

• Urine output < 0.5 ml/kg/hour for > 6 hours but <12 hours in children, and 0.5 - 1 ml/kg/hour over 24 hours in neonates

Stage 2

• Increase in serum creatinine to 200% to 299% (2.0-2.99x increase compared with baseline*)

OR

• Urine output < 0.5 ml/kg/hour for ≥ 12 hours but < 24 hours in children, or 0.3 - ≤ 0.5 ml/kg/hour over 24 hours in neonates

Stage 3

• Increase in serum creatinine to > 300% (> 3x increase compared with baseline*)

OR

• Serum creatinine of > 4.0 mg/dL (> 354 micromol/L) (or ≥  2.5 mg/dL in neonates) with an acute increase ≥  0.5 mg/dL (44 micromol/L)

OR

• Decrease in eGFR to <35 mL/min over 1.73 m2 

OR

• Urine output < 0.3 ml/kg/hour for ≥ 24 hours or anuria for ≥ 12 hours

OR

• Initiation of renal replacement therapy (RRT) only for the treatment or prevention of complications of kidney injury†

Acute Kidney Disease

The condition in which AKI stage 1 or greater is present < 90 days after ECMO start time, including continued RRT

Chronic Kidney Disease

The condition in which AKI stage 1 or greater is present ≥ 90 days after ECMO start time, including continued RRT

* Baseline defined as the lowest measured creatinine prior to admission in the previous 90 days for children. Baseline defined as the lowest measured creatinine prior to admission in the previous 7 days for neonates. If measured creatinine is unavailable, utilized standardized age- and sex-based measurements in PODIUM.

† Treatment or prevention of complications of kidney injury to not include toxic ingestion, hyperammonemia, or standard initiation as a component of ECMO configuration.

Timing: Adjudicated daily (Q24) during ECMO support.

Reporting: Each event will have an associated Type, reflecting severity as well as proximity to cardiopulmonary bypass.

Type 1a. Moderate hemolysis: Peak plasma free hemoglobin 50-100 mg/dL (500-1000 mg/L) occurring at least once during the ECLS run.  Defined by one of the following:

• Plasma free hemoglobin measurements of 50-100 mg/dL (500-1000 mg/L) sustained for at least 24 hours
  OR

• One measurement 50-100 mg/dL (500-1000 mg/L) and a major circuit component change, namely the membrane lung, blood pump or entire circuit.

Type 1b. Moderate hemolysis (as Type 1a) for children within the first 72 hours after exposure to cardiopulmonary bypass (CPB)*.

Type 2a. Severe hemolysis: Peak plasma free hemoglobin > 100 mg/dL (>1000 mg/L) occurring at least once during ECLS run.  Defined by one of the following:

• Plasma free hemoglobin measurements of > 100 mg/dL (>1000 mg/L) sustained for at least 24 hours
  OR

• One measurement > 100 mg/dL (>1000 mg/L) and a major circuit component change, namely the membrane lung, blood pump or entire circuit.

Type 2b. Severe hemolysis (as Type 2a) for children within the first 72 hours after exposure to cardiopulmonary bypass (CPB)*.

* If the patient was most recently exposed to CPB >72 hours before hemolysis definition met, then hemolysis is assessed based on Type 1a or Type 2a.

Timing: Must occur after ECMO cannulation – specified below.

Reporting: Each event will be graded according to Designation of event (A-C) and Type of patient consequence.

A. Peripheral cannulation site infection: A positive culture from the skin surrounding the ECMO cannula when there is clinical evidence of infection such as pain, drainage, erythema or leukocytosis.

• • • Must occur after ECMO cannulation and within 30 days of ECMO decannulation.

    • The peripheral cannulation site may have drainage and/or the surrounding skin may have erythema or inflammation. 

B. Central cannulation site infection or mediastinitis: A positive culture from the mediastinum when centrally cannulated for ECMO when there is clinical evidence of infection such as drainage, erythema, or leukocytosis.

• Must occur after ECMO cannulation and within 30 days of ECMO decannulation.

• Deep sternal wound infection (isolated), not a direct extension from ECMO components.

• Deep sternal wound infection involving ECMO components (continuous with mediastinum or already situated in the mediastinum). Maybe be contiguous with ECMO cannulas or graft material.

C. Bloodstream Infection (BSI): Positive blood cultures with new pathogens and no other source identified before ECMO.

• Must occur >48 hours after ECMO initiation and while on ECMO or within 24 hours of decannulation.

Patient Consequence for ECMO-CENTRAL ARC Infection Adverse Event

Type 1: Minimal, local or noninvasive intervention indicated. 

Type 2: Non-urgent intervention indicated.

Type 3: Life-threatening consequences; urgent intervention indicated.

Type 4: Fatal

Timing: Injury caused in the process of ECMO cannulation, related to cannula manipulation/movement, or decannulation. Do not use this definition to describe injuries that occurred during vascular access unrelated to ECMO cannulation.

Reporting: Each event will include a Designation of event (A-H) and Type of patient consequence.

A. Cardiac or caval injury. Penetrating injury related to vascular access for ECMO that damages the cardiac structures or cava and is associated with a pericardial effusion. 

B. Pneumothorax. Penetrating injury related to vascular access for ECMO that damages the pleural cavity and is associated with pneumothorax.

C. Limb hypoperfusion. Reduced perfusion of a limb distal to a cannulation or vascular injury associated with vascular access for ECMO.

D. Other vascular. Vascular (arterial or venous) injury (dissection, stenosis, ischemia, rupture, arteriovenous fistula, pseudoaneurysm, retroperitoneal hematoma, hemothorax) related to ECMO cannulation.

E. Repositioning of cannula. Post-ECMO cannulation, an additional procedure is undertaken to reposition cannula. Clinical team response to inadequate flow, inadequate respiratory gas exchange or diagnostic imaging – not due to dislodgement of a previously adequately positioned cannula.

F. Cannula dislodgement. At any time during the planned ECMO support, displacement or dislodgement of an ECMO cannula.

G.  Arterial ligation. Permanent ligation of the artery in relation to ECMO

H.  Venous ligation.  Permanent ligation of the vein in relation to ECMO

Patient Consequence for ECMO-CENTRAL ARC Vascular Access Adverse Event

Type 1. Does NOT result in any of the following:

• Impeded ECMO flow, tamponade physiology, hypotension

• ≥ ECMO-CENTRAL type 2 bleeding, limb or visceral ischemia, or irreversible neurologic impairment

• Death

Type 2. Associated with one of the following:

• Pericardial effusion or mediastinal collection leading to impeded ECMO flow, tamponade physiology, hypotension

• Injury to cardiac structures or cava associated with placement of a pericardial drain

• Increased compartment pressure within an enclosed limb space (legs, feet, arms or hands) restricting blood flow to muscles and nerves from ECMO cannulation, without tissue necrosis

Type 3. Associated with one of the following:

• Surgical cardiovascular repair to address cardiac or caval injury leading to impeded ECMO flow, tamponade physiology, hypotension

• Increased compartment pressure within an enclosed limb space (legs, feet, arms or hands) restricting blood flow to muscles and nerves from ECMO cannulation associated with tissue necrosis OR managed with fasciotomy secondary to compartment syndrome from ECMO cannulation

Type 4. Associated with one of the following:

• Cardiopulmonary arrest of >5 minutes duration resulting from pericardial effusion or mediastinal collection

• Vascular injury or occlusion related that causes reduced perfusion leading to amputation

Type 5. Fatal.

Timing: New thromboses identified during ECMO support, or the first diagnostic testing within 30 days post-ECMO decannulation, or autopsy.

Reporting: Each event will include a Designation of event (A-B) and Type of patient consequence.

A.   Venous thrombosis or thromboembolic event (e.g., deep vein thrombosis, pulmonary embolism) confirmed by diagnostic testing* or autopsy findings.

Type 1. Systemic venous thrombosis or thromboembolism detected by diagnostic testing* without compromising organ or limb perfusion (i.e., no clinical signs or symptoms)

Type 2. Systemic venous thrombosis or thromboembolism detected by diagnostic testing* causing patient harm as distal venous thrombosis or thromboembolism compromising organ or limb perfusion.

Type 3. Systemic venous thrombosis or thromboembolism detected by diagnostic testing* causing patient harm as distal venous thrombosis or thromboembolism resulting in fatality.

B.   Arterial thromboembolism resulting in an acute systemic arterial perfusion deficit in any non-cerebrovascular organ system or thrombosis of left ventricle or central aorta confirmed by diagnostic testing* or autopsy findings.

Type 1. Systemic arterial thrombosis or thromboembolism detected by diagnostic testing* without compromising organ or limb perfusion (i.e., no clinical signs or symptoms)

Type 2. Left ventricular† OR aortic root thrombus OR systemic arterial thrombosis or thromboembolism causing patient harm as distal thromboembolism compromising organ or limb perfusion

Type 3. Left ventricular† OR aortic root thrombus OR systemic arterial thrombosis or thromboembolism resulting in fatality.

*Diagnosis of thrombosis or thromboembolism using any of the following modalities: ultrasound (including echocardiography), or CT angiography (artery or vein), or intravascular angiography (artery or vein).  Pending timing, MRI angiography (artery or vein) may also apply.

† Systemic ventricle in patients with congenital heart disease

Timing: New dysfunction occurring during ECMO, assessed at least 72 hours post-cannulation

Reporting: This event will be identified as Grade 3 if the definition is met during ECMO support*

No evidence of chronic liver disease AND evidence of new or worsening acute hepatic dysfunction 72 hours post-cannulation by presence of

No evidence of chronic liver disease AND evidence of new or worsening acute hepatic dysfunction 72 hours post-cannulation by presence of AT LEAST 2 of the following:

Aspartate Transferase (AST) or Alanine Amino Transferase (ALT):

• • • ≥ 3-fold increase from level at least 24 hours after initiation of ECMO AND rising at ≥ 72 hours on ECMO
      AND

    • Absence of hemolysis, myositis, or rhabdomyolysis

Total bilirubin or direct (conjugated) bilirubin

• • • ≥ 1.5-fold increase from level at least 24 hours after initiation of ECMO AND rising at ≥ 72 hours on ECMO*
      AND

    • Absence of suspected Gilbert’s Disease, biliary obstruction or hemolysis

Coagulopathy

• • • International normalized ratio (INR) ≥ 3 despite Vitamin K administration within 24 hours on-ECMO accounting for direct thrombin inhibitors (DTI) associated changes in INR.

*Not applicable for neonates <7 days of age

Timing: NEW DIAGNOSIS of Right Heart* Failure during VV-ECMO

Reporting: This event will be identified as Grade 3 if the definition is met during ECMO support

Definition is detailed below. It requires either:

• Combination of echocardiographic evidence AND clinical/hemodynamic findings;
  OR

• Decision to switch to VA or VP ECMO OR death due to right ventricular (RV) dysfunction

Echocardiographic evidence of (subpulmonic) RV dysfunction

Qualitative ≥ mild-moderate RV systolic dysfunction PLUS RV enlargement or ³moderate tricuspid valve regurgitation†

AND

At least 2 of the following clinical/hemodynamic findings (not due to other identified causes e.g. volume overload)

• New evidence of atrial tachyarrhythmia

• Increase in right heart filling pressure‡ by ≥ 5mmHg over 24 hours to ≥ 14 mmHg

• Renal failure indicated by ECMO-Central AKI definition and not explained by other etiologies

• Liver injury defined by the ECMO-Central liver injury definition and not explained by other etiologies

• Otherwise not explained increasing recirculation

OR

At least 1 of the following:

• Decision to switch to VA-ECMO or VP-ECMO due to RV dysfunction

• Death occurring in patients who have not switched to VA-ECMO or VP-ECMO but who remain on inotropes or vasopressors at the time of death AND have a diagnosis of RHF made by the clinical care team

* Subpulmonary ventricle

†Note that while some quantitative assessments of RV dysfunction are variably reported in the literature, none are well established, validated or standardly available in pediatric practice.

‡ Right heart filling pressure can be measured by superior vena cava (SVC) central venous pressure (CVP) or right atrial pressure (RAP). The elevation in right heart filling pressure is not due to another identified etiology (e.g., pericardial effusion, mediastinal collection).

Timing: New or progressive LA hypertension identified during VA ECMO support

Reporting: This event will be identified if the definition is met during ECMO support*

We define left atrial hypertension acquired after VA ECMO as including at least one characteristic from each of the following categories:

Support Strategy:

VA ECMO for cardiovascular failure or cardiopulmonary resuscitation (i.e., at-risk population)

AND

Signs or symptoms consistent with onset of LA hypertension:

             At least 1 of the following:

• Clinical evidence of pulmonary edema

• New radiographic evidence pulmonary edema following ECMO cannulation

• Clinical evidence of pulmonary hemorrhage

• Arterial line pulse pressure <10mmHg

AND

Echocardiographic OR Hemodynamic evidence:

Echocardiographic evidence of:

Qualitative assessment of ≥ moderately dilated LA or LV AND at least one of the following:

• Stasis (contrast “smoke” sign)

• Intracardiac thrombus

• LV 2D (or 3D) EF <-2 Z score for age (BCH Z score calculator)

• Intermittently or completely absent opening of the aortic valve

OR

Hemodynamic evidence of:

At least 1 of the following invasive pressure measurement of >=20mmHg

• Pulmonary capillary wedge*

• LA pressure

• LV end diastolic pressure

* in the absence of pulmonary vein stenosis  

Timing: Adjudicated daily (Q24) during ECMO support.

Reporting: This event will be identified if the definition is met during ECMO support

LACTIC ACIDEMIA secondary to inadequate tissue oxygen delivery acquired while on ECMO, not associated with metabolic disease, hepatic dysfunction or pre-existing tissue necrosis. This may be evidenced by increasing arterial-venous O2 (AVO2) difference or decreasing cerebral or renal Near Infrared Spectroscopy (NIRS).

AND 1 of the following:

• New lactate ≥ 5mmol/L (≥ 45 mg/dL) post-ECMO initiation and rising

•  If pre-ECMO lactate is between 5 - 10 mmol/L (45 – 90 mg/dL) and fails to fall by at least 2.5mmol/L (22.5mg/dL) over 24 hours on ECMO

•  If pre-ECMO lactate is ≥ 10 mmol/L (≥ 90 mg/dL) and fails to fall by at least 5mmol/L (45 mg/dL) over 24 hours on ECMO

Timing: Adjudicated daily (Q24) during ECMO support.

Reporting: This event will be identified if the definition is met during ECMO support

HYPOXEMIA Inability to achieve adequate oxygenation via ECMO circuit with persistent (pre-ductal) systemic saturation ≤ 85% for ≥ 24 hours on ECMO* AND at least 2 of the following:

• Use of continuous infusion of neuromuscular blockade to reduce oxygen consumption or improve oxygenation.

• Attempt to reduce oxygen consumption or improve oxygenation through one of the following: deepening sedation, cooling the patient by ≥ 3 degrees, or packed red blood cell transfusions to target increase in goal hemoglobin.

•  Increase in fraction of inspired oxygen (FiO2) on the ventilator above 0.5.

• Increase in ventilator settings greater than lung protective limits previously identified by PALICC-2. (These limits are identified as an inspiratory plateau or peak inspiratory pressure less than or equal to 28 cm of water.)

PALICC-2; Pediatric Acute Lung Injury Consensus Conference 2

* Persistent systemic saturations ≤ 65% for ≥ 24 hours on ECMO for those with cyanotic congenital heart disease