How the blood gas analyser measures ABG variables

Question 9.1 from the second paper of 2008 asks of the bewildered candidates:

"Outline  how pH, PCO2 and PO2  are measured in a blood gas analyser and briefly state the underlying principle behind  each of those measurements."

The pass rate of 47% was pleasantly surprising, suggesting that many (if not all) of the candidates arose from an intellectually robust anaesthetic background. Judging by the college model answer, the bare minimum of expected knowledge resembles the contents of the following three paragraphs:

pH measurement

  • pH is measured with a glass electrode suspended in the blood sample.
  • The blood sample acts a a conducting electrolyte.
  • The potential difference across the electrode is proportional to the pH difference, and this can be measured.

PaCO2 measurement

  • PaCO2 is measured with a modified glass electrode.
  • The electrode is bathed in a solution which contains some sodium bicarbonate, and generates a known potential difference.
  • The CO2 from the blood sample diffuses across a semipermeable membrane into the bicarbonate solution,
  • The reaction changes the pH in the electrode, which corresponds to a change in potential difference, and this is measured.
  • The CO2 is then inferred from the change in pH.

PaO2 measurement

  • PaO2 is measured with a Clark electrode
  • The Clark electrode measures the change in current flowing through a reaction chamber where O2 is reduced to OH- ions by a change in voltage.
  • O2 from the blood sample diffuses through a semipermeable membrane into an aqueous buffer.
  • In the aqueous buffer it is reduced to OH- ions with the application of a potential difference (600-800mV); this causes a current to flow between two submerged electrodes. Increasing the voltage across this system also increases the current - up to a plateau. The plateau level depends upon, and is proportional to, the concentration of oxygen.
  • The rate of increase of current in proportion to increase in voltage becomes non-linear at a PaO2 above 150mmHg, and the ABG machine is usually clever enough to compensate for this known fact.

Derived measurements

  • HCO3- is calculated from the Henderson-Hasselbalch equation.


LITFL give a good overview, in sufficient detail for the time-poor exam candidate:

The best resource for this sort of stuff is the handbook and user guide for the blood gas machine. The brief brochure for our hometown machine can be found here, at the Radiometer website. There is also a comprehensive operations manual.

If one wishes for more information than LITFL, but less than the 160-page operations manual, one may find the desired level of detail at The Deep Picture; theirs is an excellent professional resource on this topic.

The Radiometer Blood Gas Handbook was once a valuable resource. However, it tried to explain what the parameters mean for clinical decisionmaking, rather than how they are measured. In any case, it is now gone; and the handbook-shaped hole it left in the Radiometer website now generates 404 errors.

If one wants to go to town on this topic, Severinghaus (yes, THE Severinghaus) has published a series of articles on this theme:

Severinghaus, John W., and Paul B. Astrup. "History of blood gas analysis. I. The development of electrochemistry." Journal of clinical monitoring 1.3 (1985): 180-192.

Severinghaus, John W., and Poul B. Astrup. "History of blood gas analysis. II. pH and acid-base balance measurements." Journal of clinical monitoring 1.4 (1985): 259-277.

Severinghaus, John W., and Poul B. Astrup. "History of blood gas analysis. III. Carbon dioxide tension." Journal of clinical monitoring 2.1 (1986): 60-73.

Severinghaus, John W., and Poul B. Astrup. "History of blood gas analysis. IV. Leland Clark's oxygen electrode." Journal of clinical monitoring 2.2 (1986): 125-139.

Severinghaus, John W., and Poul B. Astrup. "History of blood gas analysis. V. Oxygen measurement." Journal of clinical monitoring 2.3 (1986): 174-189.

Severinghaus, John W., and Poul B. Astrup. "History of blood gas analysis. VI. Oximetry." Journal of clinical monitoring 2.4 (1986): 270-288.

Severinghaus, John W., and Yoshiyuki Honda. "History of blood gas analysis. VII. Pulse oximetry." Journal of clinical monitoring 3.2 (1987): 135-138.

Lastly, a digression of truly ridiculous proportions exists in this very website. It is a diverticular sack-like outpouching in the form of a chapter on arterial blood gas interpretation.