CLIN. CHEM. 28/6,
1287-1292
(1982)
OxygenDissociationCurvesfor Whole Blood,Recordedwith an Instrument
That ContinuouslyMeasuresPo2 and S02Independentlyat Constantt, p,
and pH
A. Zwart, G. Kwant, B. Oeseburg, and W. G. Zijlstra’
We describe a method for recording oxygen dissociation
curves for whole-blood specimens. The blood sample is
placed in a thermostated measuring chamber, and Po2 and
S02 are measured continuously by polarography and by
reflectometry,
respectively.
During the recording of an
oxygen dissociation curve, the po2 and S02 signals are
stored in a data-acquisition
system, while pH, Pco2 and
temperature are kept constant. Determination of precision
and error discussion indicated that the coefficient
of
variation (CV) of the determination of the oxygen dissociation curve is mainly determined by the error in the
measurement of S02. The overall CV of Po2 values belonging to the lower, mid-, and upper parts of the S02 range
is estimated to be about 2.6, 3.1, and 2.1 %, respectively.
In practice the measurements are about 30% more precise than estimated. With our method, the fixed-acid-induced Bohr effect (H+ factor) can be determined over
the entire S02 range with much greater precision than
hitherto.
AdditIonalKeyphrases: blood gases
polarography
.
reflectometry
-
analytical error
Bohr effect
.
Although
the oxygen affinity of hemoglobin
is frequently
expressed
as P50, the oxygen tension (po2) at 50% oxygen
saturation
(S02), the best description
is still the complete
oxygen dissociation
curve (ODC). During the past decades
numerous
methods
have been developed
for ODC determination in whole blood. The methods
can be roughly divided
into three groups.
First, there are the discontinuous
point-by-point
methods,
which split the ODC into several distinct points. For these
points Po2 So2, Pco2 and pH are determined.
Examples
are
the mixing method of Edwards and Martin (1), later modified
by Blunt (2); the tonometry method of Duc and Engel (3); and
the method of Kernohan
and Roughton
(4), the last being
especially
suited for very low So2 values. These methods
are
rather time consuming,
because many data points are needed
if the fitting for the whole curve is to be reliable.
The second group of methods
is characterized
by continuous recording
ofpo2 or So2 only, together with measurement
or control of pH and Pco2. Measuring
just one of the two signals is allowed only when the other one changes in a predictable way or can be calculated
reliably. Continuous
measurement of So2 only, by spectrophotometry
beginning
with a
deoxygenated
sample and linearly increasing po2 by diffusion,
is described
by Niesel and Thews (5). Duvelleroy
et al. (6)
developed
performed,
a method in which two po2 measurements
one measuring blood po2 and one measuring
are
po2
in a connected
gas reservoir of known volume. The reading of
the latter electrode
is a function
of the oxygen content
and
Department of Physiology, University of Groningen,
10, 9712 KZ Groningen, The Netherlands.
Address correspondence
to this author.
Received Dec. 4, 1981; accepted Mar. 9, 1982.
Bloemsingel
thus also of the oxygen saturation
of the blood if one presumes
the blood is totally deoxygenated
before it contacts the gas
reservoir. Before and after recording the ODC, one measures
blood pH and p Co2. Probably because commercial
apparatus
(DCA-1; Radiometer, Copenhagen, Denmark) based on this
approach is available, this method has been widely used.
Later, Teisseire
et al. (7) adapted
this method
for microsamples
by diluting
the blood sample in a buffer solution,
which gives better control of pH and PCO2. In a recently developed method described
by Thibault
et al. (8), a chamber
filled with blood is connected
to a gas source through a silicon
rubber membrane.
Blood po2 is measured continuously,
while
So2 is calculated
at discrete intervals by means of a diffusion
equation.
PCo2 but not pH, is controlled
during the ODC
determination.
Neville (9) described
a similar method
involving
bioto-
nometry:
a fully oxygenated
blood sample is rapidly deoxy-
genated
by mixing with a yeast-cell
suspension.
Another
biotonometry
method was described
by Longmuir
and Chow
(10), based on the work of Colman and Longmuir
(11); the
blood sample
is deoxygenated
by adding
a heart-muscle
preparation.
With either method the decrease in P02 is monitored and, because the oxygen consumption
is assumed to be
linear with time, So2 can be determined
at any P02. The pH
is recorded
during determination
of the ODC, whereas PC02
has to be estimated
from the pH readings.
Rossi-Bernardi
et
al. (12) presented
a continuous
method in which a completely
deoxygenated
blood sample, with catalase (EC 1.11.1.6) added
to it, is gently titrated
with diluted H202. Catalase
promotes
the reaction 2 H202
2 H20 + 02, gradually oxygenating
the
blood. Blood P02 is measured
continuously,
and S02 is calculated from the amount of H202 added. During the measurement of an ODC, the pH or PCO2 of the sample is controlled
by titration.
This system has been improved
and interfaced
with computer
control by Winslow et al. (13).
In the third group of methods, both Po2 and So2 are measured continuously.
The blood sample
is oxygenated
and
deoxygenated
mostly with 02 and N2 gases, respectively.
Simply by adding an equal volume fraction of CO2 to each gas
supply, the pco2 is kept constant during an ODC registration.
Clerbaux et al. (14), using undiluted
whole blood and a titration setup, measured
and controlled
blood pH as well during
the ODC determination.
Recently,
Reeves (15) described
a
method in which a thin film of blood is used. Given the known
PCO2
in the tonometry
gases, the pH is read from a nomogram.
This method is similar in many ways to the method used in
the commercially
available Hemoscan
(American
Instrument
Co., Silver Spring, MD 20910). Imai et al. (16) described
a
-
method that resulted
in a commercially
the Hemox TM analyzer
Southampton,
PA 18966).
in a buffer, keeping the pH
possibility
of temperature
On reviewing
available
apparatus,
(Technical
Consulting
Services,
A small amount of blood is diluted
constant.
In all these methods the
control is available.
the literature
it becomes clear that an ideal
method for recording ODC in whole blood should measure po2
S02, and pH independently
and continuously,
while pH, PCo2,
and temperature
can be kept at any desired value. Besides
CLINICAL CHEMISTRY,
Vol. 28, No. 6, 1982
1287
that, blood must be kept in a state resembling the in vivo situation as closely as possible. An in vivo method for recording
ODC in dogs (17), in use for 10 years, is the model for the development of the new in vitro method presented
here. In the
in vivo method,
p2
is measured
with a Clark electrode
mounted on a catheter tip (18), and So.2 is measured
by reflection oximetry,
with use of optical fibers for light conduction (19,20); blood Pco2 and pH are controlled by means of
the inspired gas mixture, while body temperature
is kept
constant
with the help of a heating blanket. The in vitro
method is based on the same techniques
for measuring
P02
and So2. Until recently we used and tested the method only
for P50 measurements.
The results compared
so favorably
with those in the literature (21-23) that we have extended the
measuring system with a data-acquisition
system for studying
the complete ODC.
In this paper we describe the new method and its performance. To demonstrate
the possibilities of the method, we
have recorded the ODC from nine blood donors and calculated
the H factor, defined as 8 logpo2/8pH,
as a function
So2 for six blood donors, as measured
under standard
tions.
of the
condi-
sample
Materials and Methods
Measuring
system.
An outline of the measuring system is
given in Figure 1, a and b. The temperature
in the doublewalled stainless-steel
cylindrical measuring chamber, with a
magnetic stirrer at the bottom, is maintained
with a constant-temperature
waterbath (0.02 #{176}C),
which is checked
with a calibrated
certified
mercury
thermometer.
A P02
electrode, an So2 catheter, a pH electrode, and a buret outlet
are connected through ports in the side wall. The Perspex lid
of the measuring chamber contains holes through which
preheated and water-saturated
gases enter. During the determination and recording of an ODC, the signals of the P02
and S02 measuring units are sampled by a data-acquisiton
2
system.
Measurement
of p02. The P02 of the blood is measured
continuously
with a fast-responding
Clark-type
electrode
(Eakis;
Eschweiler,
Kiel, F.R.G.) prepared
as described
pre-
viously (24). The electrode is calibrated by using two exactly
known pc values, one of about 20 kPa (150 mmHg), the other
being zero. The former calibration value is obtained by equilibrating water with air at 37 #{176}C;
the zero calibration is with
pure nitrogen. To check the linearity of the P02 electrode, we
frequently used a third po2 value of about 4 kPa (30 mmHg)
with the aid of gas-mixing pumps
(W#{246}sthoff,
Bochum,
F.R.G.). We detected no deviation from linearity. The maximum error in blood P02 so measured
is no more than 0.2%.
Measurement
of So2. The S02 of the blood is measured
continuously
and instantaneously
by means of a fiber optic
reflection oximeter (20). By means of a catheter containing
two strands of optical fibers, light is guided from an incandescent lamp to the blood and the reflected light is guided
back to a detection and processing unit. With the help of appropriate filter-photocell
combinations,
two signals are obtained. The first (ei) represents
the amount of reflected light
at 630 nm and the second (e2) the amount of reflected light
at 920 nm. At 630 nm, light absorption by oxyhemoglobin
is
much less than by deoxyhemoglobin,
whereas at 920 nm there
is little difference, the absorption by oxyhemoglobin slightly
exceeding that by deoxyhemoglobin.
The two signals are
processed according to equation 1, which differs from the
procedure
described
earlier
The relationship
1288
I.-
4M
-1
b
Fig. 1. a. Schematic representation of the measuring system
and pH of the stirred blood are measured continuously. The SrJ and
p signals are stored ina data-acquisition system, while the pH sIgnal controls
an automatic titration ttilt. Temperature is controlled by means of a thermostated
water bath. By changing the ratio between the two gas supplies, the Po2 of blood
can be varied while p Is kept constant. The sample Is used for measurement
of total hemoglobin, 2,3-dlphosphoglycerate, hernatocrft, carboxyhemoglobln.
methemoglobln, and sulfhemoglobin
b. Stainless steel measuring chamber with fiber optic catheter
for measuring S02 (A), membrane-covered Pt electrode for p02
(B), outlet of titration system (C), combined glass-reference
electrode for measuring pH (D), and two ports that are not in
use (E)
The chamber Is double-walled and perfused with water from a thermostated
water bath. The magnetic stirrer at the bottom Is not shown. The Perspex lid with
the gas Inlets and outlets has been removed
(20).
e2
=
C
k
e2
+ e1
between e and So2 can be made linear by
CLINICALCHEMISTRY,Vol. 28, No. 6, 1982
substituting
the appropriate
value of k. This value, which
must be determined exactly for each So2 measuring system,
was 0.35 for our setup. A two-point calibration of the So2
measuring system is carried out after an ODC has been recorded. The stored e0 signals at P02 = 0, are averaged and set
to correspond to 502 = 0, while the stored e0 signals at P02>
55 kPa are averaged in the same way and set to correspond
to
S02 = 100%. In addition, the 02 value is displayed on a digital
panel, with a least significant
digit reading of 0.1%.
The performance
of the So2 measuring
system is checked
at the end of each ODC determination.
To this end, the gas
supply is adjusted
so that the S02 reading becomes stable at
50%. A sample is taken and analyzed with an OSM2-Hemoximeter (Radiometer),
calibrated
according to the manual, for
each donor individually. If a deviation between the reflection
oximeter and the OSM2-Hemoximeter
exceeds 1% So2, the
whole ODC measurement
is discarded and repeated. This
occurs in about 10% of the cases. Results of the OMS2-Hemoximeter were previously compared with those of a spectrophotometric
two-wavelengh
method (25) over the entire
S02 range. Because the OSM2-Hemoximeter
proved to be just
as reliable as the spectrophotometric
method, we chose the
former for convenience.
Because So2 = 0 and So2 = 100% readings are absolute by
definition, and a deviation of >1%
at So2 = 50% in the
control samples
is not accepted, the maximum
error in 02
measurement
is ±1% So2 at So2 = 50%, and decreases linearly
502
to zero at both
Measurement
extremes
of the So2 range.
and setting
of pH. Blood pH is measured
sition
system
(MINC11;
Digital
Equipment
Corp.,
Maynard,
MA 01754). Every 0.2 s, P02 and So2 signals are sampled.
Measurement of an ODC takes about 8 mm, which means that
the total ODC is based on about 2500 data points.
Measuring
procedure.
About 8 mL of heparinized whole
blood is placed in the measuring chamber. To prevent foaming
of the blood during the subsequent rapid stirring (ca. 2500
r.p.m.), a few drops of a 10% antifoam silicon emulsion (S.L.E.;
Wacker Chemie, Munich, F.R.G.) are added. This very small
amount
of silicon
emulsion
has no influence
on hemoglobin
oxygen affinity or on any of the measurements.
Stirring with
a speed of about 2500 r.p.m. causes no detectable
hemolysis
of the blood.
The gas phase above the stirred blood is kept at P02 =0 and
after about 20 mm the blood also has a p02 of zero. At this
moment the blood pH is set at the desired value by adding
either NaOH or HC1. The value is checked by means of the
capillary pH electrode. Before the p02 is slowly increased for
the ODC run, 100 data points are sampled at p02 = 0. After
the
blood
has reached
p02
a value
>55
kPa,
another
set
of 100 data points is sampled and then data storage is stopped.
Blood pH is checked again and the gas supply is adjusted so
that the reading of So2 = 50% can be checked by taking a
sample. The whole measuring procedure, including the control
continuously
with a combined
glass-reference
electrode
(7GR241; Electrofact,
Amersfoort, The Netherlands)
connected to a pH meter (PHM27; Radiometer)
that is frequently
measurements,
calibrated
Table 1 presents results of measurements
we carried out
under standard conditions (pH 7.4, pco2 5.33 kPa, and 37
#{176}C)
for three different blood samples, to determine precision.
Each P02 value and its standard
deviation
(SD) bear upon
seven independent
measurements.
Figure 2 shows a typical
direct result (without any fitting procedure) of seven ODC
runs at blood pH ranging from 7.0 to 7.6. Table 2 shows a set
ofp02 values based on the average ODC at standard conditions
of nine different blood samples for 10 So2 values. The mean
H+ factor with its SD bears upon measurements on six donor
with
two
U.S. National
Bureau
of Standards
pH 6.841 and 7.383 at 37 #{176}C.
Comparing
phosphate buffers:
measurements
of blood pH with those of a device for measuring blood pH with a capilllary electrode (G297/G2; Radiometer), we can trace and correct an occasional shift in
liquid-liquid
junction potential between buffers and blood.
The least significant
digit reads 0.001 pH unit. The pH signal
from the combined electrode in the measuring chamber is sent
to a titrator (TTT11),
which in turn controls an autoburet
(ABU12;
solution.
both from Radiometer)
filled with 2 mol/L NaOH
By titrating
the oxygen-linked
protons
liberated
during
oxygenation, blood pH can be kept constant at any
desired preset value during the complete recording of an ODC.
Just before the beginning and just after the end of an ODC
determination,
blood
trode. The maximum
pH is checked
with the capillary
elec-
error in blood pH measured and checked
this way is ±0.002 units.
Gas mixture supply. Three gas cylinders are used: pure 02,
N2, and CO2. Using the set of gas-mixing
pumps, we mix CO2
both with the 02 and the N2 gas supply to obtain the following
mixtures, C02/02 (5.6/94.4 by vol) and C02/N2 (5.6/94.4 by
vol). The mixtures are saturated with water at 37#{176}C.
The flow
of each gas mixture
to the measuring
by needle valves. By changing
chamber
is controlled
the ratio of the two supplies,
p02 can be varied from zero to about 80 kPa without
changing
Pco2. The Pco2, depending on the actual barometer reading,
was 5.33 kPa ± 0.07 kPa (range) during the experiments described below.
Hematological
measurements.
All blood samples were
taken by venipuncture
from apparently
healthy human
subjects. The samples were anticoagulated
with heparin and
stored on ice directly after collection. Total hemoglobin concentration (Cub) was measured by the HiCN method (26) and
2,3-diphosphoglycerate
concentration
according to a kit
modification
(No.35 U.V.; Sigma Chemical Co., St. Louis, MO
63178) of the enzymic endpoint method of Keitt (27). Carboxyhemoglobin,
methemoglobin,
and sulfhemoglobin were
measured by a multiple-wavelength
method (28). The hematocrit of the samples was determined with a microhematocrit centrifuge.
Data storage. During the run of an ODC, the signals from
the units measuring
po2 and So2 are stored in a data-acqui-
takes
about
30 to 35 mm for each ODC run.
Results
samples
is also presented,
with some hematological
data.
Figure 3 shows a plot of mean H factor vs S02. In the 02
range 1%-90%, the H factor is calculated at 1% intervals,
while in the S02 range 90%-99%; the H factor is calculated
at 0.1% So2 intervals.
Discussion
Error evaluation.
The total standard
deviation
of the
measured p02 at fixed S02 values is the accumulated result of
uncertainties
in all experimentally determined variables: Po2
pH, Pco2, and temperature
(t). The estimated
contribution to the standard deviation in P02 of each of the uncertainties in the determination
of these variables is taken to be
independent of each other. po2 itself can be measured with a
coefficient of variation of 0.2% (CV1).
The error in the determination
of S02 is assumed to be a
linear function of S02 between 0 and 50% S02 as well as between 50 and 100% So2. The maximum error is 1% 02 at S02
=
50% and zero for So2 = 0 and S02 = 100%. With the aid of
a digital printout of the data on which also Table 1 is based,
the effect of the maximum error in So2 upon the corresponding
P02 value has been estimated
over the entire So2
range in steps of 1% 02. This results in a contribution of about
2.5, 3.0, and 2.0% to the coefficient of variation of po for the
lower, mid, and upper part of the So2 range, respectively
(CV2).
If for the entire 502 range an H factor of -0.45 (cf. Figure
3) is assumed, a maximum uncertainty
of ±0.002 pH unit will
02,
contribute
about
P02.
Since within
0.4% (CV3) to the coefficient
a working
day usually
of variation
no appreciable
of
changes
CLINICAL CHEMISTRY, Vol. 28, No. 6, 1982
1289
Table 1. PrecisIon of p Measurements (Mean ± SD, kPa) at Fixed Values of $02 under Standard
Conditions (pH 7.4, Pco2 5.33 kPa, and 37 #{176}C)
for Three Different Human Blood Samples (n = 7 each)
so2
AZ.,
10
D.A.,
1.37± 0.03
20
1.98 ± 0.03
30
2.49± 0.03
40
80
2.99 ± 0.04
3.52 ± 0.06
4.12 ± 0.06
4.85 ± 0.07
5.86 ± 0.08
90
7.77 ± 0.11
95
Erythrocyte 2,3-DPG, mmol/L
2,3-DPGIHb4, mol/mol
Carboxyhemoglobin
Methemoglobin
Sulthemoglobin
10.13± 0.25
50
60
70
2,3-DPG, 2,3-diphosphoglycerate;
2.98 ± 0.06
3.26
3.81
4.42
5.16
3.50± 0.07
4.10 ± 0.06
4.84 ± 0.08
5.88 ± 0.09
7.82± 0.11
10.32± 0.17
n.d.
n.d.
n.d.
nd., not detectable.
(intra-individual
variation)
may
variation)
have
been
60
shown
in Table
probably
1, where
5.33 kPa,
cerate 5.0 mmol/L
0
15
20
curves (donor A.Z.) at seven dif-
The pH range from left to right Is from 7.6 to 7.0 in steps of 0.1. The curves are
not &awn by any fitting procedure, but are directly measured
CLINICAL CHEMISTRY,
Vol. 28, No. 6, 1982
DA has a right-shifted
ODC,
2,3-diet al. (31) eliminated
two
and HbCO
state pH 7.4, Pco2
2,3-diphosphogly-
1%. They reduced
their
P02
that was somewhat shifted to the right as compared with the
one presented here.
The H+ factor is determined over almost the entire So2
range,
from 1% to 99% (cf. Figure
3). At So2
=
50% the mean
value of the H factor is -0.44 (SD 0.02, n = 6) which correlates very well with values determined earlier, when only P50
had been measured (22, 23). Our values for the H+ factor
Table 2. Po2 Values (Mean ± SD, kPa) from Nine
Different Human Blood Samples, and H+ Factor
Values from Six Different Human Blood Samples
at Fixed Values of $02 and Standard Conditions
(pH 7.4, Pco2 5.33 kPa, and 37 #{176}C)
95
12
in (e.g.)
values to these standard conditions with the help of correction
factors determined earlier (32). Ultimately,
they found a curve
80
90
(kPo)
P02
donor
content. Arturson
by accepting
as a standard
37 #{176}C
as well as erythrocyte
40
50
60
70
20
2 (interindividual
by differences
as a result of the higher intra-erythrocytic
phosphoglycerate
of these factors
30
40
and Table
caused
HbCO, 2,3-diphosphoglycerate,
fetal hemoglobin, and glycosylated hemoglobin in the blood samples. An example is
10
20
Fig. 2. Oxygen dissociation
ferent blood pH values
± 0.14
8.17
10.59± 0.17
5.11
1.02
so2
8
6.20 ±
0.07
0.08
0.09
0.10
0.12
0.92
80
4
±
±
±
±
5.02
(z-)
1290
1.52 ± 0.03
2.21 ± 0.05
2.75 ± 0.06
4.60
0.86
in atmospheric
pressure will occur, PC02 changes during the
experiments can be neglected.
If for the entire So2 range a temperature
factor (8 log
po.J8T) of 0.024(22,29) is assumed, a maximum uncertainty
of ±0.02 #{176}C
will contribute 0.2% (CV4) to the coefficient of
variation of Po2 The direct influence of these temperature
changes on the measurement
of P02 itself is negligible.
Taking the aforementioned
influences together, the estimated total coefficient
of variation
of po2 determination
(CVi = v’(CV1)2 + (CV2)2 + (CV3)2 + (CV4)2) is about 2.6,
3.1, and 2.1% for the lower, mid, and upper part of the S02
range, respectively. This shows that it is mainly the uncertainty in So2 measurement
that is responsible for the total
standard
deviation. As can easily be calculated from the data
in Table 1, the measured CVs for each of the three blood
samples are equal to or smaller than the estimated ones.
Especially in the mid-saturation
range, the calculated CV is
about 30% greater than that found experimentally.
Discussion
of results. The mean P02 standard values presented in Table 2 compare favorably with the standard ODC
as tabulated by Severinghaus
(30). The differences in standard deviation of the measured po2 values between Table 1
0
2
1.38 ± 0.03
1.99 ± 0.04
2.48 ± 0.05
P02
H+ facf
1.39
± 0.08
-0.32 ± 0.06
2.01
2.52
3.01
3.53
4.12
4.84
5.86
±
±
±
±
±
±
±
0.11
-0.39
± 0.02
0.13
0.15
0.17
0.19
0.21
0.23
-0.42
± 0.01
-0.43
± 0.01
-0.44
-0.45
-0.44
-0.44
-0.43
±
±
±
±
±
7.76 ± 0.29
10.14
± 0.33
0.02
0.02
0.02
0.02
0.02
-0.40 ± 0.03
Mean sample erythrocyte 2.3-DPG was 4.64 (SD 0.37) mmol/L (n = 9); mean
2,3-DPG/lb4 was 0.90 (SD 0.08) mol/mol (n 9). Each sample contained <1%
each of carboxyhemoglobin
and methemoglobin; sulthemoglobin was not detectable.
P50 read from the whole ODC
equilibration
method (22,23).
ence between these two values,
that the P02 electrode
indeed
-.6
with the P50 measured
by the
There proved to be no differand it was therefore concluded
records
the real po2.
For oxygen affinity measurements
in whole blood from
newborns, the amount of blood needed may be too large in
some cases, so that dilution with plasma or saline becomes
necessary.
Dilution to CHb = 30 g/L is possible without interfering with the reflection So2 measuring system.
The class of methods called the discontinuous
point-bypoint methods
(1-4) has the disadvantage
of being very laborious and it is difficult, if not impossible,
to keep the blood
pH identical in all samples. These methods involve the use of
correction
factors for pH differences.
Determination
of pH
modulation
of the oxygen affinity is therefore
virtually
impossible.
The class of continuous
measuring techniques,
in which only
P02 is measured,
has the advantage
of being able to also
20
0
40
60
90
measure
ODC’s
bloodneedsamples
containing
moglobmns
without in the
of knowing
their
100
Fig. 3. Mean H+ factor of samples from six different donors,
calculated and plotted for every 1% O2 in the S02 range from
1% to 90%, and for every 0.1% S02in the So2range from 90%
to 99%
The solid line was obtained by connecting the plotted mean H factor values;
no curve fittingwas applied. The dotted lines indicate ± 1 SD
abnormal prophespectral
erties. With the DCA1 apparatus,
however, it is
keep blood pH and (or) Pco2 constant
during
Hahn et al. (38) have equipped the machine with
correction
procedure
for pH changes,
using a
factor of -0.48.
However,
impossible
to
oxygenation.
an automatic
constant
H+
as can be read from Figure 3, this
certainly does not allow for accurate determinations.
A great
disadvantage
of this method is the requirement
that the two
compare
favorably
with those presented
in the papers of
Duhm (33), Hlastala and Woodson (34), and Severinghaus
(30), but less well with the result obtained by Garby et al. (35),
who found a value of -0.35.
The oxygen saturation
dependency
of the H+ factor
is
shown in Figure 3. At S02 = 5% the mean H factor is about
-0.2, decreasing
rapidly to -0.4 at So2 = 20%. From So2 = 20%
to So2 = 90% the value stays almost constant at -0.44. For 02
>95% the H+ factor increases
again, reaching about -0.2 for
So2 = 99%. Hlastala
and Woodson
(34) found values comparable with ours, though their scatter is rather high, especially
at the lower end of the saturation
range. In fact, their standard
error of the mean was much larger than our standard
devia-
tion. Garby et al. (35) found a much more pronounced
saturation
dependency
of the H
factor and higher
values
throughout
the 502 range.
Theoretically
(36) the H factor, integrated
over the entire
So2 range, should be equal to the Haldane factor, which is the
amount
of proton
equivalents
liberated
during full oxygen-
ation of a completely
deoxygenated
blood sample. Experimentally,
however, no complete
equivalence
of these factors
was observed
(37). This is due to the fact that the H factor
was determined
at So2 = 50% only and was then considered
to be constant
over the entire So2 range. Our results show,
however, that the H factor is not constant over the entire So2
range. Therefore,
figures comparable
with the Haldane factor
can only be obtained
by integrating
the H factor over the
entire 02
range. The value of the H+ factor integrated
over
the entire saturation
range (cf. Figure 3) is -0.40, which is 0.04
greater than the value at S02 = 50%. This increase is large
enough to explain the difference
in H+ factor and Haldane
factor as reported
by Siggaard-Andersen
et al. (37).
Discussion
of method.
Our method has the advantage
that
the measurements
are made on whole blood, under defined
conditions,
instead
of on hemolysates.
This yields directly
interpretable,
data.
physiological-and
thus clinically
relevant-
Because the registration
of a whole ODC takes only about
8 mm, the p02 measurement,
which is certainly
not instantaneous, should
be fast enough
for recording
the real p02
changes. The most sensitive
test for this was comparing
the
P02 electrodes
must have the same characteristics,
independent of the medium in which they are placed. Only by using
electrodes
with an extremely
long response time can this requirement
be fulfilled. In the method of Thibault
et al. (8),
So2 is calculated
by means of a diffusion
equation,
with the
assumption
that hemoglobin
is fully saturated
with oxygen
at P02 = 20 kPa. Furthermore,
it is also impossible
to
determine
the H factor by this method, because blood pH
cannot be kept constant
at any value.
Biotonometry
(9, 10) has about the same advantages
and
disadvantages
as the DCA1
apparatus,
but additional
uncertainties
are the assumed linear consumption
of oxygen
with respect to time and the pH disturbance
caused by the
metabolizing
oxygen consumer.
The technique
of RossiBernardi
et al. (12), with its controlled
addition of H2O2, has
the possibility
of keeping pH or Pco2 at any desired value. It
is also possible to measure the oxygen affinity of abnormalhemoglobins
without a new calibration
procedure.
The disadvantages
are the separate equilibration
of the blood sample
with high-po2 gas to achieve So2 = 100% and the likelihood
of methemoglobmn
formation.
Methods
using direct independent
measurement
of both
P02 and So2 are preferable
for the reasons already mentioned.
Reeves (15) uses a blood film, which is shown to behave like
whole blood in his setup. However, pH has to be read from a
nomogram
incorporating
the known PCO2, soit is not possible
to study the effect of fixed-acid
changes on oxygen affinity.
In the method of Clerbaux
et al. (14) So2 is measured
by absorption spectrophotometry
instead of by reflectometry.
The
latter method
has the advantage
of being less sensitive
to
variations in plasma turbidity,
in addition to its suitability
for
a simple and compact design of the measuring
chamber.
Because all relevant
variables
influencing
hemoglobin
oxygen
affinity are measured or controlled independently,
our method
is not only well suited for the study of the various factors which
modulate
hemoglobin
oxygen affinity, but also for the study
of the influence
of the interaction
between these factors.
This work was supported in part by grants from the Netherlands
Organization for the Advancement of Pure Research (Z.W.O.) received through the Foundation for Medical Research (Fungo). We
gratefully acknowledge the aid of Dr. P.H.W. van der Ploeg (Clinical
Chemical Laboratory, Diakonessenhuis,
Groningen, The NetherCLINICAL CHEMISTRY,
Vol. 28, No. 6, 1982
1291
lands) in measuring 2,3-diphosphoglycerate
and of Dr. V. J. Fidler
(Centre for Medical Informatics and Statistics, University of Groningen, The Netherlands)
in checking the error discussion.
References
1. Edwards, M. J., and Martin, R. J., Mixing technique for the oxygen
hemoglobin equilibrium
and Bohr effect. J. App!. Physiol. 21,
1898-1902 (1966).
2. Blunt, M. H., A micromixing method for the determination
of
oxygen equilibria in blood. J. App!. Physiol. 37, 123-125 (1974).
3. Duc, G., and Engel, K., A method for determination
of oxyhemoglobin dissociation curves at constant temperature,
pH and pCO2.
Respir. Physiol. 8, 118-126 (1970).
4. Kernohan, J. C., and Roughton, F. J. W., The precise determination
of the bottom of the oxyhaemoglobin
dissociation curve of human
blood. In Oxygen Affinity of Hemoglobin and Red Cell Acid-Base
Status, M. R#{246}rth
and P. Astrup, Eds., Munksgaard, Copenhagen,
1972, pp 54-64.
5. Niesel, W., and Thews, G., Em neues Verfahren zur schnellen und
genauen Aufnahme der Sauerstoffbindungskurve
des Blutes und
konzentrierter Hamoproteinlosungen.
Pfluegers Arch. Ges. Physiol.
273, 380-395 (1961).
6. Duvelleroy, M. A., Buckles, R. G., Rosenkaimer, S., et al., An oxyhaemoglobin dissociation analyzer. J. APP!. Physiol. 28, 227-233
(1970).
7. Teisseire,
B., Teisseire,
L., Lautier, A., et al., A method of continuous recording
on microsamples
of the Hb-O2 association curve. I.
Technique and direct registration of standard results. Bull. Physiol.
Pathol. Respir. 11,837-851(1973).
8. Thibault, L. E., Schuette, W. H., and Lees, D. E., Instrument to
determine the haemoglobin-oxygen
equilibrium curve based on the
calculated diffusion of oxygen across a semi-permeable
membrane.
Med. Biol. Eng. Comput. 18,401-405
(1980).
9. Neville, J. R., Hemoglobin oxygen affinity measurement
using
biotonometry. J. App!. Physiol. 37, 961-971 (1974).
10. Longmuir, I. S., and Chow, J., Rapid method for determining
effect of agents on oxyhemoglobin
dissociation curves. J. App!.
Physiol. 28, 343-345 (1970).
11. Colman, C. H., and Longmuir, 1. S., A new method for registering
oxyhemoglobin dissociation curves. J. App!. Physiol. 18, 420-423
(1963).
12.. Rossi-Bernardi,
L., Luzzana, M., Samaja, M., et al., Continuous
determination
of the oxygen dissociation curve for whole blood. Clin.
Chem. 21, 1747-1753 (1975).
13. Winslow, R. M., Morrissey, J. M., Berger, R. L., et al., Variability
of oxygen affinity of normal blood: An automated method of measurement. J. App!. Physiol. 45, 289-297 (1978).
14. Clerbaux, T., Fesler, R., and Bourgeois, J., A dynamic method for
continuous
recording
of the whole bood oxyhemoglobin dissociation
curve at constant temperature, pH and CO2. J. Med. Lab. Technol.
30, 1-9 (1973).
15. Reeves, R. B., A rapid micro method for obtaining oxygen equilibrium curves on whole blood. Respir. Physiol. 42, 299-315
(1980).
16. Imai, K., Morimoto, H., Kotani, M., et al., Studies on the function
of abnormal hemoglobins. I. An improved method for automatic
measurement
of the oxygen equilibrium curve of hemoglobin. Biochim. Biophys. Acta 200, 189-196 (1970).
17. Oeseburg, B., Landsman, M. L. J., Mook, G. A., and Zijlstra, W.
G., Direct recording of oxyhaemoglobin
dissociation curve in vivo.
Nature (London) 237, 149-150 (1972).
18. Janssen, T. C., Lafeber, H. N., Visser, H. K. A., et al., Construction
and performance of a new catheter-tip oxygen electrode. Med. Biol.
Eng. Comput. 16, 274-277 (1978).
19. Mook, G. A., Osypka, P., Sturm, R. E., and Wood, E. H., Fiber
1292
CLINICAL CHEMISTRY, Vol.28,No. 6, 1982
optic reflection photometry on blood. Cardiovasc.
Res. 2, 199-209
(1968).
20. Landsman, M. L. J., Knop, N., Kwant, G., et al., A fiberoptic reflection oximeter. Pt luegers Arch. 373, 273-282 (1978).
21. Oeseburg, B., p(50) and Bohr-factor determination
by direct simultaneous measurement of SO2, PO2 and pH in blood samples. Proc.
Workshop on Blood pH and Gases, University Press, Utrecht, The
Netherlands,
1979, pp 63-68.
22. Oeseburg, B., Determination
of 02 affinity of hemoglobin by independent continuous measurement of SO2, p02, and pH with control
of pH, pCO2, and temperature.
Crit. Care Med. 7, 396-398 (1979).
23. Oeseburg, B., Kwant, G., and Zwart, A., Determination of factors
modulating haemoglobin oxygen affinity and the significance of these
factors for the calculation of oxygen saturation by automated blood
gas machines. Proc. 5th Meeting of the IFCC Expert Panel on pHand
Blood Gases, Copenhagen, 1981, pp 139-150.
24. Oeseburg, B., Kwant, G., Schut, J. K., and Zijlstra, W. G., Measuring oxygen tension in biological systems by means of Clark-type
polarographic
electrodes. Proc. Kon. Ned. Akad. Wet., Ser. C 82,
83-90 (1979).
25. Mock, G. A., Buursma, A., Gerding, A., et al., Spectrophotometric
determination
of oxygen saturation of blood independent
of the
presence of indocyanine
green. Cardiouasc.
Res. 13, 233-237
(1979).
26. International
Committee for Standardization
in Haematology.
Recommendations
for reference method for haemoglobinometry
in
human blood (ICSH Standard EP 6/2: 1977) and specifications
for
international
haemiglobincyanide
reference preparation
(ICSH
Standard EP 6/3: 1977). J. Clin. Pathol. 31, 139-143 (1978).
27. Keitt, A. S., Reduced nicotinamide adenine dinucleotide linked
analysis of 2,3-DPG: Spectrophotometric
and fluorometric procedures. J. Lab. Clin. Med. 77, 470-475 (1972).
28. Zwart, A., Buursma, A., Van Kampen, E. J., et al., A multiwavelength spectrophotometric
method for the simultaneous determination of five haemoglobin derivatives. J. Clin. Chem. Clin. Biochem. 19,457-463
(1981).
29. Reeves, R. B., The effect of temperature on the oxygen equilibrium curve of human blood. Respir. Physiol. 42, 317-328 (1980).
30. Severinghaus, J. W., Blood gas calculator. J. App!. Physiol. 21,
1108-1116(1966).
31. Arturson, G., Garby, L., Robert, M., and Zaar, B., The oxygen
dissociation curve of normal human blood with special reference to
influence of physiological effector ligands. Scand. J. Clin. Lab. Invest.
34, 9-13 (1974).
32. Arturson, G., Garby, L., Wranne, B., and Zaar, B., Effect of 2,3DPG on the oxygen affinity and on the proton- and carbamino-linked
oxygen affinity of hemoglobin in human whole blood. Acta Physiol.
Scand. 92, 332-340 (1974).
33. Duhm, J., Dual effect of 2,3-diphosphoglycerate
on the Bohr effects of human blood. Pfluegers Arch. 363, 55-60 (1976).
34. Hlastala, M. P., and Woodson, R. D., Saturation dependency of
the Bohr effect: Interactions among H4, C02, and DPG. J. App!.
Physiol. 38, 1126-1131 (1975).
35. Garby, L., Robert, M., and Zaar, B., Proton- and carbaminolinked oxygen affinity of normal human blood. Acta Physiol. Scand.
84, 482-493 (1972).
36. Wyman, J., Linked functions and reciprocal effects in hemoglobin:
A second look. Ado. Protein Chem. 19, 223-286 (1964).
37. Siggaard-Andersen,
0., SaIling, N., Norgaard-Pedersen,
B., and
R#{246}rth,
M., Oxygen-linked hydrogen ion binding of human hemoglobin. Effects of carbon dioxide and 2,3-diphosphoglycerate.
III.
Comparison of the Bohr effect and the Haldane effect. Scand. J. Clin.
Lab. Invest. 29, 185-193 (1972).
38. Hahn, C. E. W., Fo#{235}x,
P., and Raynor, M., A development of the
oxyhemoglobin
dissociation
curve analyzer.
J. App!. Physiol. 41,
259-267 (1976).