OR NE babel dem be Rad
R. WENNESLAND ET AL.
Nios and
‘pA
eee aT pe
ee
(Samet, Fritts, Fishman & Cournand 1957)
and 5.7 per cent of blood (Berlin, Hyde,
Parsons & Lawrence 1952),
The data presented earlier make it clear
that with our procedure used in healthy sub-
—
jects and animals, errors in Vrbc due to the
in vitro handling of blood and to the technique of administering the cells and of sampling are very small, totaling well under 1 per
cent (Tables I, III and IV). With an average counting error of 1.6 per cent (Nomof
et al. 1954), the error arising from estimating radioactivity of the average of two blood
specimens, compared to that of the tagged
cell suspension, is less than 2 per cent. At
least Z per cent of the average test-retest
variation remains to be explained.
The S.D. of the hematocrit of a single
blood specimen, measured in duplicate, was
0.27 scale divisions in the 103 experiments
where 3 samples were taken. The experience
is about the same as described by Wintrobe
(1934), from whose data it appears that the
maximum expected variation of the test as
performed in vitro is 0.86 per cent. Our
ptactice of averaging the duplicated hematocrits of 25- and 30-minute samples lowers
this source of error, An indeterminate error
in calculating Verbe from the radioactivity of
whole blood relates to the quantity of plasma trapped in the cell column. With normal
blood and uniform procedures of anticoagulation and centrifugation, this quantity might
be expected to represent a constant fraction
of the cell column. However, from the af
culties encountered in te
ts4
(Chaplin & Mollison 1952; |
1955; Furth 1956; Grae
1959), this expectation may, not i
Such error coul rey amount to
“
f
per cent or more. An error, probebly not tx.
ceeding one per cent, arises from assumjy,
constancy of the fraction of the cell columy
occupied by white cells and platelets, or from
errors in estimating the thickness gf the
buffy coat (Reeve 1952; Wintrobe 1961).
Vpl and Vwb are derived values and un.
certain to the extent that the observed hema-
tocrit of blood taken from a large vein o,
artery differs from “true” body hematocrit
(Chaplin, Mollison & Vetter 1953; Greger.
sen & Rawson 1959). Tourniquet stasis oy
dilution of samples with saline or anticga.
gulant solutions can seriously affect Vwh
and Vpl without affecting Vrbe. In repeated
determinations (Table V1) we found that
Vwb was less constant than Vrbe. This can
probably be accounted for by the lability of
Vpl and its dependence on body water and
cardiovascular phenomena. Because the ratio
of body hematocrit to large vessel hematocnt
is affected in several clinical conditions where
blood
volume
is
an
important
variable
(Brown, Hopper & Weunesland 1957), out
prediction standards for healthy men and
women were prepared without applying corrections for this discrepancy. If vcr isis.
thus considered to be Vwb, and Vpl C Vil
— Vrbe, observed values can be compared
to the values derived from the prediction
chasts and “ee (Wennesland et ol.
1959; Brown ¢
“True” Vwil vefapes
de
about 110 per cent
of VCr8!, although: the relationship is not
constant} (Chaplin ef al. 1953; Samet ¢t al.
1957). Failure to correct for the body hematocrit: venous hematocrit ratio, or the a%
sumption of a ratio not applicable under the
circumstances of the study, will lead 1
errors in estimating “true” Vwhbfrom vcr
god hematocrit,
AMOUNTtee
sy
tat ous
10 per cent in healthy subjects at rest, 20 pct
cent in cases Of congestive heart Ciluce
(Samet et al, 957; Brown et al 1957) and
so per cent with massive splenomepaly
(Fudenberg,
Baldini,
Mahoney
&
Dame.
gek 1961). Birkeland (1900) has pomted
et that whegp blood volume is calculated
from the hematocrit and a measurement of
ws
The niger sontees at error are ay the
Helcominiation ot cahoactivity of blood specimens aad kyged cell suspension, and 6) lelermination of the centrifuged hematocrit,
particularly wath respect te the percentage
Of Wapped plasma in the cell column. Failure
fo measure accuruely the volume of tagged
coll suspension delivered to the subject, a
serous potential source of error, proved to
only cell or plasma volume, errors due ty be relatively unimportant with the technique
mistaken assumptions about the bosly lemat- toed Errors rekuing to the collection aud
ant: venous hematocrit ratio wall be greater
handling ef blood for hematocrit determinawhen the hematocrit is lugh than when it 1 _ bony and ia the prediction of the “body
low. In tnany clinical situations ic is desire
hematocrit : venous
able to make separate measurements of cell
siscaterially affect the estimation of blood and
and plasma volumes.
SUMMARY
'
ratio
can
plasma volumes, but not cell volume.
|
REFERENCES
We have described tia detail a modifies
ton of Sterling and Gray’ > Gr*l method for
Horha,
blood volume determination with which we
have had considerable expericuce. The dose
of Crt needed for tagying 12 ml of the pea-
Bishelam
West blood need not exceed 50--75 uCif
a well-type’ sfintillation counter is used. The
tagged cell suspension is usually stored over-
mght so that the test can be done conven"daily before breakfast. The cells are celnered from a smal infusion apparatus
huugh an indwelling needle which is alsa
wed for sampling.
The over-all error of the measurement of
tell volume, as shown by repeating: the test
alter intervals of 3 to 31 weeks, averages 44
Ver cent. This compares favorably with rewalls obtain
With other madifications of
the Cys:
hetmatrocnt:
method and with 12, even though
the long time interval between tests iu this
uly allowed the possibility of wathinesub"I changes of cell volume.
N.
©,
Tyke,
SoM,
Parsom,
R.
J.
&
lawrence, J. HE: The blood volume in various
medical and sucgercul combtions. New Engl. J.
Med 247, O76, 142
oS:
Thast
volume
deferminagion
wn
surgcal paticws. Acta chir gcaml Suppl 253,
64, 1900.
Brown, E., Hopper, J. Jr, Sampson, J. J. &
Mudeck, C.: Vemuas ogngestion of the extrematies in relation to blood volume dotermihationy aal ly imixwee curves of carbon motoxide and T-1K24 tn nunnadl bumnan subjects.
J. clay bevesa 40 P44, 1951.
- & Wenwstimt Ro: Livod vohame and its
Fegulation Sam Kev Uhysiol. 19, 231, £957,
--~- Sampson JoJo & Mudeick, Cs The joss
Of laid ait protium [ram the blood durwnig a
systemic mee of cenuus pressure pruduced by
fepaated Vall inaneavers int man J. chm.
buves. 37, 105, 1958,
-Vbealges, Joo, Jr, Bradley, BH, WennesTum Ko Yamauchi, HL: Red cell and blood
volume mn telidion ta height and weight in 101
healthy women. a. ancasined with Cragged
cell aed heinateat — In preparatn.
aS edn,
364