Clinical Assessment of Oxygen Transport-Related Quantities
By 1950, oxygen transport by the blood was well under- sion that still hinders the evaluation of oxygen transport
stood. The oxygen-carrying properties of the blood had disturbances in patients. Contrary to the prevailing con-
been shown to be determined by oxygen capacity, oxygen vention and disregarding the physiologic basis of the
saturation, and oxygen affinity, the latter expressed by a definition of So2, the instrument was programmed to
graph relating oxygen saturation (So2) to oxygen tension display O2Hb as a percentage of the total hemoglobin
(Po2), the oxygen saturation curve. Although these quan- concentration. Thus, So2 cO2Hb/(cO2Hb cHHb) was
tities could be determined in the physiology laboratory replaced by FO2Hb cO2Hb/ctHb. This quantity was
and although various oximeters for measuring So2 in vivo called “oxygen saturation”, and it was plainly stated that
had been developed (1, 2 ), examination of the oxygen “defining saturation in terms of all Hb species present
transport status of patients was still based largely on gives a more exact and meaningful interpretation of the
clinical signs, such as cyanosis. The accurate evaluation of data” (8 ). This was followed by the historically incorrect
the oxygen capacity had to await the standardization of remark that formerly So2 was defined only in terms of
hemoglobinometry (3 ) and the complete determination of HHb and O2Hb because these hemoglobin species make
the composition of human hemoglobin (HbA) (4 ). up most of the hemoglobin and because COHb and
Standardization of hemoglobinometry on the basis of MetHb could not be easily measured.
spectrophotometric determination of methemoglobincya- In a report on the performance of the instrument, Zwart
nide, a stable hemoglobin derivative into which all hemo- et al. (9 ) explained why substituting FO2Hb for So2 was
globin derivatives usually present in the blood can be wrong. They advised reprogramming of the CO-oximeter
easily converted, involved an understanding that the total so that ctHb, So2, and dyshemoglobin fractions would be
hemoglobin concentration (ctHb) includes the inactive displayed, but in vain. Because the numeric difference
(non-oxygen-binding) derivatives. Consequently, even in between So2 and FO2Hb is usually small, FO2Hb is easily
healthy individuals, the oxygen capacity per gram of mistaken for So2, the more so because both are called
hemoglobin is slightly lower than the theoretical value of “saturation”. Initially, this went unnoticed. When an
1.39 mL/g, calculated by dividing the molar volume of appreciable amount of COHb or MetHb was present,
oxygen (22 394 mL at standard temperature and pressure however, So2 as calculated by a blood gas analyzer was
dry) by one fourth of the molar mass of human HbA different from FO2Hb as displayed by a CO-oximeter.
(16 114.5g). Oxygen capacity may be diminished in pa- This received little attention until pulse oximeters came
tients by an increased fraction of inactive hemoglobin, into general use, and the arterial So2 measured in vivo
later called dyshemoglobin (5 ). The Beckman DU spec- was compared with the analysis of arterial samples by a
trophotometer soon enabled the measurement of the CO-oximeter (10 ).
common dyshemoglobins, methemoglobin (MetHb) and Through the introduction of the pulse principle (1, 2 ),
carboxyhemoglobin (COHb), in the clinical chemical lab- oximetry in vivo became suitable for routine clinical
oratory (6 ). Determination of ctHb by the standard application. A pulse oximeter is a two-wavelength pho-
method and correction for the fractions of MetHb and tometer that determines arterial So2 by measuring light
COHb thus made simple determination of the oxygen absorption in a piece of well-perfused tissue. Through
capacity possible. The validity of this procedure has been proper wavelength selection, photometric interference by
confirmed experimentally (2, 5 ). other hemoglobin derivatives can be minimized. When
Meanwhile, the methodology for measuring oxygen wavelengths of 660 and 940 nm are used, COHb causes a
saturation in the clinical laboratory had progressed slight underestimation of So2 and MetHb causes moder-
through the development of numerous (spectro)photo- ate underestimation in the higher So2 range and overes-
metric methods, using Van Slyke’s manometric procedure timation in the lower So2 range; when So2 70%, the
as the reference method. Using this method, So2 is de- error caused by MetHb is negligible (2, 11 ).
fined as the ratio of the volume of hemoglobin-bound Comparing So2 obtained by pulse oximetry with
oxygen to the oxygen capacity. This shows that So2 FO2Hb obtained by “CO-oximetry” in patients with high
represents the fraction of the oxygen capacity that is fractions of COHb or MetHb led to numerous reports
occupied by oxygen, which is equivalent with cO2Hb/ erroneously stating that the oxygen saturation was greatly
(cO2Hb cHHb), the definition of So2 used in photomet- overestimated by the pulse oximeter in these patients.
ric methods where O2Hb and HHb are oxyhemoglobin This in turn led to many unnecessary experiments and
and de-oxyhemoglobin, respectively. to the “discovery” that there are two kinds of oxygen
Around 1980, spectrophotometric multicomponent saturation, subsequently called “functional saturation”
analysis (MCA) of hemoglobin derivatives became suit- (for So2) and “fractional saturation” (for FO2Hb). These
able for application in the clinical laboratory (7 ). The first new terms and the strange quantity “pulse oximeter gap”,
automated photometer for MCA of hemoglobin deriva- being the difference between FO2Hb and So2, only added
tives was the IL282 CO-Oximeter, a four-wavelength to the confusion.
instrument that measured HHb, O2Hb, COHb, and A case in point is the recent report presenting a patient
MetHb (8 ). This instrument was a real asset to the clinical with methemoglobinemia as “a woman with low oxygen
laboratory, but it also was the root of widespread confu- saturation” (12 ). Only after working through a long
Clinical Chemistry 51, No. 2, 2005 291
292 Zijlstra: Clinical Assessment of Oxygen Transport-Related Quantities
differential diagnosis was the tentative conclusion seem to have lost the ability to use them properly in the
reached that an abnormal hemoglobin might be present, examination of the oxygen transport status of patients.
influencing the light-absorbing properties of the blood. If Medical technology has dissociated from pathophysio-
the CO-oximeter had been programmed to display the logic knowledge. It was a mistake to change the definition
dyshemoglobin fractions instead of fractional saturation, of oxygen saturation and to substitute FO2Hb for So2 in
the diagnosis would have been obvious after analysis of the first automated photometer for MCA of hemoglobin
the first arterial blood sample. The fact that this report, (8 ). The crucial step out of this quagmire is the standard-
parading the whole array of incorrect ideas of CO-oxim- ization of the readout of such instruments so that unam-
etry and pulse oximetry, was considered fit for publica- biguous quantities are displayed, enabling clinicians to
tion in a peer-reviewed journal shows how widespread analyze oxygen transport disturbances in terms of oxygen
these misconceptions have become. capacity, oxygen saturation, and oxygen affinity. In most
Fortunately, the report by Hammond et al. in this issue cases, ctHb, arterial So2, and dyshemoglobin fractions will
of Clinical Chemistry (13 ) clarifies interpretation of oxygen suffice; in more complicated cases, arterial Po2 and half-
transport and clears up past errors. The authors compre- saturation tension (P50) may be required (15 ). There is no
hend that So2 is not method-dependent, that a pulse place for FO2Hb.
oximeter faithfully measures So2 in the presence of
COHb, and that the underestimation of So2 by MetHb is References
1. Severinghaus JW, Astrup PB. History of blood gas analysis. VI. Oximetry.
attributable to its color, which interferes with the mea- J Clin Monit 1986;2:270 – 88.
surement. They note that measuring So2 in dyshemoglo- 2. Zijlstra WG, Buursma A, van Assendelft OW. Visible and near infrared
binemia is of limited value because COHb and MetHb absorption spectra of human and animal haemoglobin: determination and
application. Utrecht: VSP, 2000:368pp.
affect oxygen capacity and oxygen affinity, not oxygen 3. International Committee for Standardisation in Haemotology (ICSH). Recom-
saturation. Figs. 3 and 4 [from the papers of Barker and mendations for haemoglobinometry in human blood. Br J Haematol 1967;
coworkers (Refs. 8 and 13 in the article by Haymond et ¨
4. Braunitzer G, Gehring-Muller R, Hilschmann N, Hilse K, Rudolf V, Wittmann-
al.)] are reinterpreted correctly: the upper graphs show ¨
Liebold B. Die Konstitution des normalen adulten Humanhamoglobins.
the photometric interference by COHb and MetHb, re- Hoppe-Seylers Z Physiol Chemie 1961;325:283– 8.
5. Dijkhuizen P, Buursma A, Fongers TME, Gerding AM, Oeseburg B, Zijlstra
spectively, which leads to some underestimation of So2 by ¨
WG. The oxygen binding capacity of human haemoglobin. Hufner’s factor
the pulse oximeter; the lower graphs are simply the ¨
redetermined. Pflugers Arch 1977;369:223–31.
expression of a mathematical relationship. Note that 6. van Kampen EJ, Zijlstra WG. Determination of hemoglobin and its deriva-
tives. Adv Clin Chem 1965;8:141– 87.
Barker considered these graphs to demonstrate that the 7. Zwart A, Buursma A, van Kampen EJ, Oeseburg B, van der Ploeg PHW,
pulse oximeter greatly overestimates the oxygen saturation. Zijlstra WG. A multiwavelength spectrophotometric method for the simulta-
An important observation is that in the presence of an neous determination of five haemoglobin derivatives. J Clin Chem Clin
Biochem 1981;19:457– 63.
abnormal hemoglobin (HbM Sasketoon), even multi- 8. Brown LJ. A new instrument for the simultaneous measurement of total
wavelength spectrophotometry may give erroneous re- hemoglobin, % oxyhemoglobin, % carboxyhemoglobin, % methemoglobin, and
oxygen content in whole blood. IEEE Trans Biomed Eng 1980;27:132– 8.
sults. However, in this case it cannot be ruled out that the 9. Zwart A, Buursma A, Oeseburg B, Zijlstra WG. Determination of hemoglobin
high COHb fraction that only slowly decreased on oxygen derivatives with the IL282 CO-oximeter as compared with a manual spectro-
therapy is at least partly real, although no external source photometric five-wavelength method. Clin Chem 1981;27:1903–7.
10. Zwart A. Spectrophotometry of hemoglobin: various perpectives [Editorial].
of CO was found. In normal hemoglobin, the bound O2 Clin Chem 1993;39:1570 –2.
forms a hydrogen bond with histidine at position E7. This 11. Zijlstra WG, Buursma A, Meeuwsen-van der Roest WP. Absorption spectra of
favors O2 binding with respect to CO (14 ). This relative human fetal and adult oxyhemoglobin, de-oxyhemoglobin, carboxyhemoglo-
bin and methemoglobin. Clin Chem 1991;37:1633– 8.
advantage of O2 is lost through the substitution of ty- 12. Hurford WE, Kratz A. Case 23-2004: a 50-year old woman with low oxygen
rosine for histidine, and the preference of the binding site saturation. N Engl J Med 2004;351:380 –7.
for CO becomes so strong that even part of the endog- 13. Haymond S, Cariappa R, Eby CS, Scott MG. Laboratory assessment of
oxygenation in methemoglobinemia. Clin Chem 2005;51:434 – 44.
enously produced CO is bound. 14. Mims MP, Porras AG, Olson JS, Noble RW, Peterson JA. Ligand binding to
Hammond et al. (13 ) would have provided an even heme proteins. An evaluation of distal effects. J Biol Chem 1983;258:
better service to the clinical community if they had 15. Wimberley PD, Burnett RW, Covington AK, Fogh-Andersen N, Maas AHJ,
advised against the use of FO2Hb. FO2Hb per se has no ¨
Muller-Plathe O, et al. Guidelines for routine measurement of blood hemo-
physiologic significance. It may be convenient that a globin affinity. International Federation of Clinical Chemistry, Scientific
Division, Committee on pH, Blood Gases and Electrolytes. Scand J Clin Lab
normal FO2Hb signals in a single figure that So2 is not too Invest Suppl 1990;203:227–34.
low and that no significant amount of dyshemoglobin is
present. However, when FO2Hb is subnormal, the cause Willem G. Zijlstra
may be a low So2, the presence of dyshemoglobin, or
both. FO2Hb cannot replace So2 and the dyshemoglobin Department of Paediatrics
fractions, and there is good reason to keep the latter University Hospital
quantities separate: So2 has a unique relationship with University of Groningen
Po2, whereas the dyshemoglobin fractions affect oxygen 9713 GZ Groningen, The Netherlands
capacity and oxygen affinity. E-mail firstname.lastname@example.org
Thus, we have run into a paradoxical situation—the
necessary quantities can now be determined in the clinical DOI: 10.1373/clinchem.2004.043638
laboratory, some of them even at the bedside, but we