Measurement of blood gases

Poliomyelitis was endemic in the first half of the 20th century; it arrived every summer, causing panic among parents and closing playgrounds, swimming pools and cinemas. ¹ There were two particularly severe global outbreaks. The first, in 1916, infected approximately 27,000 people in the USA alone, with a death rate of 35%; the second, in 1952, was the final peak before vaccines were developed, with 57,879 cases and 3145 deaths in the USA. The much lower mortality rate in 1952 was due to better supportive care, notably the introduction of the negative pressure ventilator known as the ‘iron lung’.

In Denmark however, there were very few negative pressure ventilators and, since the 1952 epidemic saw an unusually high incidence of bulbar palsy, they were only partially effective. ² By August 1952, 31 patients had been admitted to Blegdam Hospital in Copenhagen with bulbar poliomyelitis; 27 of these had died, but the cause of death was not clear as all the patients appeared to be well oxygenated. An urgent meeting of the senior hospital clinicians was called, including the recently appointed medical superintendent of the hospital laboratory, Poul Astrup, and the anaesthetist, Bjorn Ibsen.^3,4 While Astrup had been on holiday, the clinicians had used his laboratory equipment to determine that the patients had high bicarbonate levels, and concluded that patients were dying of an alkalosis of unknown origin. Ibsen suggested it was more likely this was due to inadequate ventilation and raised carbon dioxide (CO₂) levels so Astrup quickly performed pH readings at 38°C, establishing that the patients were actually acidotic with elevated CO₂ levels. ⁵ Until this moment, the clinical relationship between high partial pressure carbon dioxide (pCO₂), acidosis and compensatory metabolic alkalosis had been unknown.

The following day, they performed a tracheostomy on a severely cyanosed 12-year-old girl ‘gasping for air and drowning in her own secretions’. ⁴ Two experimental pieces of equipment were attached to the patient: a Brinkman Carbovisor, which sampled CO₂ directly from one of the bronchi, and a Millikan oximeter. These confirmed the laboratory findings that oxygenation was adequately maintained by the negative pressure ventilator, but hypercarbia remained a significant problem. Ibsen began squeezing the bag attached to the tracheostomy tube and subsequent readings showed that CO₂ and bicarbonate levels dropped with adequate positive pressure ventilation.

This was a turning point in their management of severely ill polio patients. As the epidemic spread in Denmark, a routine was quickly established. Very sick patients were anaesthetised and intubated prior to tracheostomy. As they had no positive pressure ventilators, patients were manually ventilated by teams of medical and dental students working six- to eight-hour shifts. ⁵ The logistics of this operation were enormous, with as many as 90 patients at a time requiring ventilation, and 102 patients being ventilated for more than a month. Each patient had their own team of students and often became extremely distressed when new recruits appeared. ⁶ They learned to communicate with eye movements and other gestures, and the students frequently read to the patients as they rhythmically squeezed the bag.

Poul Astrup also volunteered to ventilate patients, pondering how to provide real-time information to the students to optimise ventilation. ⁷ He began doing arterial punctures on the patients, wrapping the samples in a blanket to rush them up to a room permanently heated to 38°C. Initially he measured the pH and CO₂ content, using these to calculate pCO₂ using the Henderson–Hasselbalch equation.^8,9 This was a cumbersome and inefficient method, so he devised an equilibration technique which relied on the fact that the titration line between pH and the log of the pCO₂ was a straight line within physiological parameters.^3,10 The patient’s blood sample was divided between three tubes, two of which were equilibrated with known high and low concentrations of CO₂, respectively. The pH of all three was then measured and the unknown pCO₂ of the third tube was calculated from the titration line. Remarkably, the students were able to adjust their ventilation rates accordingly and achieve consistent results, despite the incredibly challenging circumstances. Poliomyelitis is a complex disease and although the mortality rate dropped dramatically, many patients still died, leaving the students emotionally and physically exhausted.^5,11

Astrup subsequently worked with Svend Schrøder, brother of one of the founders of the medical equipment company Radiometer, to design and construct a complex thermostatic anaerobic glass equilibration chamber containing a pH and reference electrode. ¹⁰ Their understanding of blood biochemistry expanded rapidly. Kjeld Jørgensen and Astrup coined the term ‘standard bicarbonate’ to indicate the bicarbonate of the plasma in fully oxygenated blood at 38°C and pCO₂ of 40 mmHg, ¹² and nomograms created by Ole Siggaard-Andersen allowed readings of real base excess, buffer base and standard bicarbonate.^13,14

Further developments came from the USA with the development of the oxygen electrode by Leland Clark in 1953 ¹⁵ and the CO₂ electrode by Richard Stow and John Severinghaus a few years later.^16,17 In December 1958, researchers came together at a Ciba-sponsored symposium on pH and blood gases in London where Astrup presented his equilibration equipment, incorporating new work from Siggaard-Andersen, Jørgensen and others. Severinghaus, also at the meeting, had just published his work combining both electrodes with a pH electrode to produce the first three function blood gas analyser. ¹⁷

As the science advanced, commercial interest also grew. The Yellow Spring Instrument Company displayed a combined thermostatically controlled water bath with both oxygen and CO₂ electrodes at the Ciba meeting. Subsequently Radiometer produced a manual system with pH, partial pressure oxygen and pCO₂ electrodes, and then in 1973, introduced the Acid-Base Laboratory (ABL1), the first commercially available, fully automated microprocessor-controlled blood gas machine.^7,18

Cover photo.

An Astrup blood gas analyser. Radiometer. Copenhagen. Image courtesy of Geoffrey Kaye Museum of Anaesthetic History, VGKM3994.