Respiration‎ > ‎


Oxygenation is objectively measured by both SaO2 and PaO2. However, SaO2 carries greater significance as exemplified by the oxygen carrying capacity formula:

CaO2 = (SaO2 x Hb x 1.34) + .003(PaO2 mmHg)

SaO2 can be approximated by SpO2 with reasonable accuracy with some exceptions which will be detailed below: leukocyte larsony and saturation gap.

At room air with Hb of 14, 180mL of O2 is bound to Hb while 3mL is dissolved in blood (total 183mL). Place the pt on 100% FiO2 and dissolved O2 increases from 3mL to 14mL, which isn’t physiologically beneficial as it reacts rapidly with surrounding tissues before reaching capillary beds.



The gold standard for measuring % shunt is:

\dfrac {Q_s} {Q_t} = \dfrac {Cc_{O_2} - Ca_{O_2}} {Cc_{O_2} - Cv_{O_2}}

But this requires the patient to have a central line. Surrogates for estimating shunt include:

  1. a/FiO2 (abnl <200)
  2. a/(FiO2 x mean airway pressure)
  3. a/A (abnl: <0.77)
  4. A-a (nl: 2.5 + 0.21 x age)
  5. (A-a)/a.

Calculating A-a gradient was drilled into every medical student. It is useful to understand the variables that enter the equation. But in truth, A-a gradient is rarely used in clinical practice because PAO2 is annoying to calculate and is inferior to a/FiO2.

Unfortunately, a/FiO2 does not linearly correlate with degree of shunt (see below). Additionally, it does not take positive pressure into account: a patient requiring 50% FiO2 and PEEP of 10 may have the same extent of pulmonary pathology as a person requiring 100% FiO2 and PEEP of 5 though their a/FiO2 will be drastically different. New studies suggests that a/(FiO2 x mean airway pressure) may be more accurate because it accounts for this variable.




In the presence of shunt, mismatch between utilization and delivery of oxygen can also decrease SaO2. PmvO2 <40mmHg is highly suggestive of this and optimization of oxygen delivery can improve SaO2.

Leukocyte Larsony

Spurious hypoxemia is low SaO2 with normal SpO2. This is associated with leukocytosis or thrombocytosis though the mechanism is unknown as platelets do not consume O2.

Saturation Gap

In contrast to leukocyte larsony, saturation gap occurs with low SpO2 and normal SaO2. Hypoperfusion, hypothermia, anemia, venous congestion causing pulsatile venous flow, and nail polish can confound measurements. More interestingly, abnormal hemoglobins: methemoglobin, and glycohemoglobins (A1c > 7) can produce saturation gap.



PEEP as hypoxemia therapy makes rational sense because it addresses the cause of the hypoxemia. PEEP recruits collapsed alveoli, which is the source of shunt. Using a mode that sounds similar to APRV with Phigh at 40, Rothen demonstrates dramatic reduction in the degree of atelectasis seen on CT.

Increased alveolar pressure produces a gradient for moving alveolar interstitial fluid to to peribronchial interstitium (Zone 4 turns to Zone 3), which improves diffusion impairment.

Increasing PEEP also increases mean airway pressure, generating a higher gradient for oxygen diffusion by increasing Patm.


An added bonus is that PEEP promotes left heart forward flow by increasing transmural pressure. It simultaneously ebbs right heart forward flow by reducing RV venous return and increasing RV afterload. All together, PEEP is ideal for left sided heart failure at the expense of RV strain.

Detriments include decreasing LV output from reduced preload, which is seen in states of hypovolemic or distributive shock. PEEP may also be detrimental in patients with existing RV strain, such as PE or pHTN, by increasing RV afterload. PEEP increases Zone 2 conversion to Zone 1, which raises PaCO2 through dead space. Patients with cardiac shunt may experience R>L shunting from elevated pulmonary vascular resistance, worsening hypoxemia.



Oxygen is an obvious and mainstay therapy in hypoxemia. But it is important to recognize that oxygen, per se, does not address the pathology. In severe shunt, in fact, supplemental oxygen does nothing.

In the absence of hypoxemia, there is little benefit to increasing FiO2. It is important to titrate FiO2 down if it is to be used for prolonged periods of time and preferentially rely on PEEP.

ARDSnet Titration Ladder

FiO2 30 40 40 50 50 60 70 70 70 80 90 90 90 100
loPEEP 5 5 8 8 10 10 10 12 14 14 14 16 18 18-24
FiO2 30 30 30 30 30 40 40 50 50 50-80 80 90 100 100
hiPEEP 5 8 10 12 14 14 16 16 18 20 22 22 22 24