This chapter is very, very loosely related to Section F8(vii) from the 2017 CICM Primary Syllabus, which expects the exam candidates to be able to "describe physiology and consequences of abnormal haemoglobin". Instead of addressing these examinable syllabus items directly, the focus here is on the detection of normal and abnormal haemoglobin species. In brief, haemoglobin concentration (normal and abnormal) is measured directly, using visible absorption spectroscopy. The local unit features a 128-wavelength spectrophotometer with a measuring range of 478-672 nm. Indeed, it is pleasing to point out that the reference manual for this machine indulges in a digression on the theoretical foundations of absorption spectroscopy.
Absorption spectroscopy is based on Lambert-Beer's law, which relates the properties of transmitted light to the properties of the substance through which it is transmitted. Specifically, the law can be defined as a statement...
The measured absorbance for a single compound is directly proportional to the concentration of the compound and the length of the light path through the sample
...or as an equation...
The absorbance, in this scenario, is defined as the logarithm of the ratio of the light intensity before and after transmission through the compound. Each haemoglobin species has a different extinction coefficient (ε) and these are known from empiric measurements.
Obviously if you have a whole host of weird compounds all absorbing light, the total absorbance is the sum of the absorbances of each compound.
The blood gas analyser liberates haemoglobin from red cells by shredding them with a 30Hz ultrasound beam, and then passes the resulting puree through a beam of near-infrared light.
The light is then separated into 128 beams using a concave grate/mirror, and sent into 128 individual light-sensitive diodes. Each diode is responsible for measuring a specific wavelength, within a range of 478-672 nm. Each diode subsequently reports on the light intensity at that wavelength. The higher the light intensity, the lower the absorbance (i.e. less of that wavelength was absorbed by the sample).
Thus, for each blood gas sample, 128 separate wavelength-absorbance measurements are taken, and a graph can be generated, where absorbance can be plotted as a function of wavelength:
The total absorbance is the sum of absorbance from all the haemoglobin species, all together.
Thus, total absorbance correlates with the total haemoglobin concentration.
The ABG machine reports this variable as ctHb, the concentration of total haemoglobin.
However, when the extinction coefficient (ε) of a haemoglobin species is known at every given wavelength, one can separate the species.
The chapter by van Kampen and Zijlstra ("Determination of haemoglobin and its derivatives") from the 8th volume of "Advances in Clinical Chemistry" presents several graphs of various haemoglobin species, of which I have selected a particularly effective representative:
Hb in the legend above is obviously haemoglobin, HbO2 is obviously oxyhaemoglobin and HbCO is carboxyhaemoglobin. Hi is "hemiglobin", which is what the men of learning used to call methaemoglobin. HiCN is hemiglobin cyanide, and coproporphyrin III is a breakdown product of bilirubin which is excreted in urine and faeces.
With some modifications, a messy graph of ε/λ relationships can be created:
In a slightly neater interpretation, the graphs still don't gain any additional educational value. See?
That's right, useless. But the message (if there was one) is that the different haemoglobin species will each have some sort of influence on the total absorbance. According to Lambert-Beer's law, the contribution of each species at any given wavelength will be proportional to that species' concentration in the sample.
If one knows the expected contribution from known empirical measurements, one can determine the contribution from each haemoglobin species individually:
Rather than discussing maths which I don't fully understand, I will instead quote the ABL800 FLEX reference manual which says that the specific constants are "determined using Multivariate Data Analysis where the spectra of the calibration compounds were considered together with the reference values of the calibration compounds."