Oxygen Saturation

Pulse Oximetry

  • Gives you the oxyhemoglobin saturation or SpO2
  • Based on difference in light absorption between oxyhemoglobin and deoxyhemoglobin. 
  • HbSat= [OxyHb]/([OxyHb] + [DeoxyHb])
  • Typically utilize 660 nm and 940 nm
  • Do not differentiate between OxyHb and CarboxyHb, hence, in carbon monoxide poisoning, will give falsely elevated OxyHb Saturation reading (typically ~ 100%)
  • Cannot differentiate MetHb, hence in methemoglobinemia, typically SpO2 reading is ~ 85%
  • Less accurate at oxygen saturations <75%
  • Averaging time can be changed on the monitor (ie 2-12 seconds). Lower averaging time allows you to pick up a desaturation faster but introduces more noise (increased sensitivity for oxyHb desaturation but less specificity)
  • Sources of error include motion artifact, ambient light, and poor tissue perfusion
  • Not affected by fetal Hb or bilirubin


  • Gives you the oxygen tension (ie PaO2) as well as oxyhemoglobin saturation (ie SaO2)
  • Accurately measures other hemoglobins (ie methemoglobin and carboxyhemoglobin), thus provides the most accurate measure of oxygen saturations

Regional Tissue Oxygenation

  • Gives you the tissue oxygen saturation (StO2)
  • Near infra-red spectroscopy (NIRS) used to measure cerebral and somatic (ie renal) tissue oxygen saturation
  • More widely used in cardiac intensive care unit to monitor cerebral tissue oxygenation
  • No clear evidence to suggest improved outcomes



  • Reflects CO2 at end of exhalation
  • In the normal physiologic state, very closely related to PaCO2. The small difference that exists can be thought of as primarily representing anatomic dead space.
  • The Ae Difference, or PaCO2-ETCO2, reflects dead space in the lung (areas of lung that are ventilated but not perfused). Increases in dead space lead to increases in the Ae difference
  • Vd/VT= (PaCO2-PECO2)/PaCO2 or 1- (PeCO2/PaCO2)
  • Factors that reduce the PeCO2 include conditions that lead to increased dead space ventilation. This could include pulmonary embolism, cardiac arrest, and pulmonary hypoperfusion. 
  • A useful quick primer:

  • The waveform also provides information, with "ramping" suggesting the presence of airway obstruction 

Transcutaneous CO2 Monitoring

  • Utilize a Stowe-Severinghaus electrode, heats the skin to allow for rapid diffusion of carbon dioxide to the electrode
  • Parallel but generally overestimate PaCO2 values
  • Poor perfusion increases the amount of overestimation (CO2 accumulates in areas that are poorly perfused)
  • Generally utilized more often in neonates as ETCO2 devices can cause issues with weight on the small endotracheal tube

Blood Pressure


  • Most common method utilizes oscillometric method (ie Dinamap). 
  • As the blood pressure cuff is deflated, blood flow causes oscillations to occur in the artery, with maximal oscillations occurring at mean arterial blood pressure. An algorithm is used to estimate the SBP and DBP.
  • MAP most accurate, DBP least accurate
  • Blood pressure cuff size matters. The cuff should occupy at least 2/3 of the upper arm. The bladder should be at least 80% of the circumference of the arm.
  • With an undersized blood pressure cuff, you overestimate blood pressure. With an oversized blood pressure cuff, you underestimate blood pressure.
  • In general, these devices overestimate blood pressure in the context of hypotension and underestimate blood pressure in the context of hypertension (when compared with direct arterial measurements). Hence, they tend to squeeze everything towards the middle or make things look slightly better than they may be. (Holt, PCCM 2011)

Direct ABP Measurement

  • Involves direct measurement via a catheter typically placed in the radial, ulnar, dorsalis pedis or posterior tibial artery. Less common sites for placement/measurement include the femoral and axillary arteries. 
  • Considered gold standard for blood pressure measurement
  • SBP increases the further you go out into the periphery. ie Femoral SBP>Radial SBP> Aortic BP due to pulse amplification/reflection from arterial branch points
  • Can be subject to overdamping or underdamping. Overdamped values underestimate the amplitude of the signal whereas underdamped signals overestimate the amplitude of the signal
  • Underdamping and overdamping can be evaluated using the "fast flush" or "square wave" test

  • Although SBP and DBP can be over/underestimated due to damping, in general, the MAP should be accurate. Hence, some clinicians advocate primarily utilizing MAP as a target for blood pressure
  • Resonance can also overestimate the amplitude

Interpreting the CVP waveform

Central Venous Pressure Monitoring
Figure 6

-The normal CVP waveform (Figure 6) consists of various waves: 
    1) a wave: due to atrial contraction at the end of diastole
    2) c wave: due to the tricuspid valve bulging into the RA during isovolumic contraction
    3) x descent: due to atrial relaxation as the ventricle contracts
    4) v wave: due to filling of the atrium during systole
    5) y descent: due to emptying of the atria into the ventricle during diastole 

Various pathological processes can alter the CVP waveform:

    1) canon a waves:  large a waves seen with AV dissociation (ie JET) as the atrium contracts against a closed tricuspid valve
    2) loss of a waves: occurs in atrial fibrillation as there is no longer a significant atrial contraction
    3) prominent CV wave: occurs with tricuspid regurgitation; as the ventricle contracts, the regurgitation causes an increase in the C and V waves


  • Core temperature is the critical value as it primarily controls thermoregulation
  • Can be measured at tympanic membrane, distal esophagus, and nasopharynx. Rectal and bladder temperatures also reflect core temperature although they change less rapidly
  • Peripheral skin, while the easiest to measure, is the least reliable as it is altered by skin perfusion