ECMO

Carrying Capacity of Oxygen

CaO2 = 1.34 * SaO2 * Hg + PaO2 * 0.003

** 13.4 must be used in Fick's rather than 1.34 to convert Hg units from g/dl to g/mL

** 13.4 is actually somewhat debatable. If there was no carboxy or methem then this constant is actually closer to 13.9

CaO2 (normal) = 1.34 * 15g/dL + 90 * 0.003 = 20 g/dL

Cardiac Output

Two ways to estimate "basal" cardiac output

1) CO = CI * BSA

** basal CI thought to be 2.5; below 2.2 is considered cardiogenic shock though these thresholds vary by individual

** BSA = sqrt (height in cm x weight in kg) / 60

2) CO = 70ml/kg/min

Delivery of Oxygen

DO2 = CO * CaO2

DO2 (normal) = 70ml/kg/min * 20 g/dL = 0.7dL/kg/min * 20g/dl = 14g/kg/min = 14ml/kg/min

** in a 70kg male this approximates to 1000ml/min

** normal varies widely: 10ml/kg/min to 20ml/kg/min - or - 650ml/min to 1400ml/min

Consumption of Oxygen

VO2 = 3-5 ml/kg/min in adults

4-6 ml/kg/min in children

5-8 ml/kg/min in infants

DO2 : VO2 is 14:(3-5) = 3-5 : 1

** anaerobic metabolism starts at ratio of 2:1

** SvO2 coincidentally = VO2: DO2

Choosing ECMO Blood Flows

1) VA ECMO for cardiogenic shock. The goal is to have the ability to provide 100% of the cardiac output.

CO = 2.5 * sqrt (height in cm x weight in kg) / 60

2) VA ECMO for distributive shock. As high as technically feasible.

It is difficult to predict how much supplemental CO is necessary to match distributive physiology because SVR changes based on CO, BP, and arterial diameter (which changes with changes in arterial blood volume). There is also the consideration of reactance and impedance because pulsatility will be intact. These variables makes the math infinitely more complicated.

3) VV ECMO with hypoxemia with normal cardiac output

Method 1: [(Target SpO2 - Estimated SvO2) x Estimated cardiac output] / [100 * (1 - Estimated SvO2)]

if the target arterial sat is 90%, estimated SvO2 is 70% and estimated CO is 5lpm then targeted VV ECMO flow should be (0.9-0.7) x 5 /(1-0.7) = (2/3) x 5 = 3.33lpm

Notice how the required blood flow rates for VV ECMO is less than VA ECMO (2/3). This makes obesity less of a challenge with VV ECMO and much more of a challenge with VA ECMO. Arteries are also smaller which makes it harder to accommodate a large enough return cannula.

However you will frequently fall below this target because of recirculation

Method 2 : 55-80ml/kg/min

notice how this closely matches with 2/3 x "normal" cardiac output of 70ml/kg/min

but of course this falls apart if the patient is in a high cardiac output state (distributive shock)

4) VV ECMO for hypercarbic respiratory failure and normal oxygenation

10 ml/kg/min

Choosing Venous Cannula Size

1) Drainage Cannula. Pick the cannula that will have a delta pressure < 100 for the desired blood flow.

Typical sizes are 21 - 28 | 50-70cm

Below is a study of pressure flow characteristics of different drainage cannulas (pay attention to length as well as gauge) on Hct 27% in vitro

2) Choosing a return cannula. Same concept as above though cannulas tolerate higher positive pressure (less likely to explode) than negative pressure (more likely to collapse). So you can stretch the limits to +150, +200 pretty safely. Unfortunately, these graphs don't even go that high but one can imagine ...

VV ECMO: Typical sizes are 19-21FR | 16-18cm

VA ECMO: Typical sizes are 15-17Fr | 16-18cm; might need to make a chimney graft if artery is too small or use two smaller cannulas in different groins returning blood in parallel.

ECMO Specific Calculations

ECLS DO2 = CpostoxygenatorO2 x circuit blood flow

** but delivery of oxygen in the patient follows the above formula of CaO2 x CO

Recirc: SpostoxO2 - ScvO2/SpreoxO2-ScvO2

SaO2 (if lungs neither consuming nor reabsorbing O2) = (EF/CO)SpostoxO2 + (1-EF/CO)SvO2 + 0.003*(PecmoO2 + PvO2)

Hypoxemia on ECMO

SpostoxO2 failure: gas failure, oxygenator failure, flow higher than rated blood flow on oxygenator

** oxygenator rated blood flow is defined as maximal blood flow for SpreO2 75 to SpostO2 95% at a Hg 12

Gas failure

switch to wall oxygen and dial it to match previous gas flow

check connections

Oxygenator failure

check color change

check postox gases

calculate VoxygenatorO2 = (CpostoxO2 - CpreoxO2) * BF

V'O2 of 200-300 is normal; < 150 is concerning for failure

EF failure: inadequate ECMO BF, drainage insufficiency, recirculation

EF = BF * (1-recirc)*(1-duration of pauses/total duration)

Recirc failure

Drainage insufficiency failure

BF failure

SVO2 failure

Hg, temperature, agitation

When ruled out all of the above, the most common cause of hypoxemia on ECMO is that the native CO outpaces the ECMO BF. Assuming lungs are not contributing to gas exchange at all, the formula for SaO2 would be:

SaO2 = EF x SmO2 /Co + SvO2 x (1 - EF/CO) + 0.03 x PmO2

Or more simply:

Eq1: SaO2 = (EF + SvO2 x (CO - BF))/CO

So if CO goes up, the SaO2 will go down. But DO2 is dependent on both SaO2 and CO so if one goes down while other goes up, which wins?

DO2 = 13.4 x Hg x CO x SaO2

Sub Fick's equation in for SaO2:

Eq2: DO2= 13.4 x Hg x Co x (SvO2 + [VO2/(13.4xHgxCO)]) = 13.4 x Hg x CO x SvO2 + VO2

Solve SvO2 in Fick's equation with Eq1 and you get

Eq3: SvO2 = 1 - [VO2 / (13.4 x Hg x EF)]

Merge Eq2 and Eq3 and you get:

UltEq: DO2 = 13.4 x Hg x CO x [1 - [VO2 / (13.4 x Hg x EF)]] + VO2

This equation shows DO2 independent on SaO2 or SvO2. Decreasing CO will increase SaO2 but will decrease DO2.

Anticoag notes on ECMO

UFH binds to AT, inhibits fXa and thrombin; UFH also releases TFPI which inhibits TF-FVIIa; half life 30-60min, liver metabolized renally secreted

Following the principle of treating the patient, not the machine, you run into more trouble with over-anticoagulation than under anticoagulation

A cautious approach to heparin is to eliminate boluses, start gtt at 6U/kg/hr and increasing by 1U/kg/hr to goal

Bivalirudin directly inhibits thrombin; 25-35min half life; 80% secretion by proteolytic enzymes 20% by kidneys

Argatroban directly inhibits thrombin; 40-50min half life; liver metabolized

aPTT is falsely shortened by acute phase reactants

antiXa can be falsely low with elevated plasma hemoglobin and hyperbilirubinemia; assay must not add AT or dextra sulfate which can falsely overestimate; neext AT > 50-80%

1U RBC has 130-240ml of packed RBC in 225-350ml volume (Hct is 55-65%); 50-80g Hg (23g/dl) and 150-250mg of Fe

Cryo has 150mg of finbrinogen and 80IU of fVIII in 5-20ml of plasma; 5-6U per dose; 1 dose will raise fibrinogen in a 70kg person by 35mg/dl

respscore.com

Google Sites

Report abuse