A STUDYOF THE VOLUMEOF THE BLOODIN CONGESTIVEHEARTFAILURE. RELATION TO OTHERMEASURE-

MENTSIN FIFTEEN PATIENTS

BY GEORGER. MENEELY1 AND NOLANL. KALTREIDER(From the Department of Medicine of the School of Medicine and Dentistry of the University

of Rochester and the Medical Clinics of Strong Memorial and Rochester MunicipalHospitals, Rochester, New York)

(Received for publication November 10, 1942)

While the volume of the blood in health is quiteconstant and closely related to the size, age, andsex of the individual, it has long been recognizedthat in congestive heart failure, this relationship isdisturbed. Therefore, investigations of the vol-ume of the blood in this clinical condition shouldthrow light on the mechanism of normal controlof the volume of the blood as well as on the na-ture of the physiological disturbances involved inheart failure itself. Studies of this nature, sum-marized by Gibson and Evans (1), indicate thatthere is an increase in the volume of the blood inthis condition. While this generality is well sup-ported, there is a striking variability in individualcases. It appeared likely that this variabilitymight be accounted for if quantitative measure-ment of factors which might influence blood vol-ume were made at the time of the determinationof the blood volume. Since the association be-tween anoxemia and compensatory polycythemiais well established, measurements of arterial andvenous oxygen were made and it proved possibleto show a rough quantitative relationship betweenanoxemia and cellular increase. On the otherhand, while the statement has been made (1, 2)that there is a positive relationship between thevenous pressure and the change in blood volumein congestive heart failure, it was not possible tofind such a relationship within this group ofseverely congested cardiac patients.

Although a number of the lines of investigationpursued did not lead to the establishment of anysimple correlation with plasma or cell volumechanges, these data presented are of interest, notonly in themselves, but also because certain in-teresting relations obtain between them. The re-sults of these studies conducted on cardiac pa-tients are therefore presented in full.

1 Louisiana State University School of Medicine, NewOrleans, La.

MATERIAL

Fifteen patients were selected who had clinical evidenceof severe congestive heart failure (class IV). All weredyspneic or orthopneic, all but one had pitting edema, allbut one had hepatomegaly, most had chronic passive con-gestion of the lungs, a few had hydrothorax, several hadascites. The clinical findings in these patients are sum-marized in Table I. Every effort was made to select un-complicated cases, who were untreated, or who were in asteady state despite digitalis. Etiology of the heart dis-ease in these cases included rheumatic heart disease, ar-teriosclerosis, hypertension, thyrotoxicosis, syphilis, andheart disease secondary to pulmonary disease. The dura-tion of failure ranged from a few weeks to several years.Patient number seven had evidence of a moderate degreeof renal insufficiency and he also had hemoptysis. Noneof the others had evidence of chronic blood loss. Fivepatients were female and ten were male. The average agewas 542 years, with one subject in the third decade, fivein the fourth decade, four in the fifth, and five in thesixth decades.

METHODS

The patient was brought to the laboratory in bed inthe post-absorptive state. The bed was adjusted to asflat a position as the subject could tolerate, and he wasallowed to rest in this position for more than thirty min-utes. Then the blood volume was determined by themethod of Gibson and Evans (3). Predictions of blood,plasma, and cell volumes, normal for these patients werebased on Gibson and Evans data (4), using the height ofthe subject, because edema rendered the weight an un-satisfactory reference value. Predicted values were sub-tracted from observed values, and the difference expressedas a percentage of the normal.

The same venipuncture made for the injection of thedye was used to determine the arm to tongue circulationtime (5) and the venous pressure (6). The venous pres-sures were referred to a point six centimeters below theAngle of Louis. A period of rest followed, then samplesof blood were taken from the radial artery and the ante-cubital vein simultaneously in oiled syringes and storedover mercury with refrigeration. Dry potassium oxalatewas used as the anti-coagulant. The vital capacity wasdetermined with a spirometer, recording the best of threetrials. Several hematocrit determinations were made bythe wet technique of Gibson (3), using 1.6 per cent potas-

521

GEORGER. MENEELYAND NOLAN L. KALTREIDER

uopsjnJqn J* .-0pS8 0100 qi

-lip ieaq a u6cpU8 010Jp04lOgJBmJy

uoisjntUqgnlno u _A; -nfu 6~0uS!0flln pu s

_ < SpOU"sl 1184!m agelp .8Oq opsmneqI

uopsjuiqg

I.m oaIoiloIalJv

uops~qv

e

xognod7

_~Imeqopoj.jovl% V

210o0R qousNq

*WMp luopnuqg

^ ast jw'ewosPo uso°"2_n Lnuoholosoyo

uovI fqlgCo. aspnolpns'I A'ujp_~~~~-M Paug0 'Pop

pu lo us

t oDPos pus 18410M '113

qpvPq ooos.iol-spoloqdgun po-ao

-Ypz s.qopoxspiopsn puse

uopsw°Pq° aDnoJJnDSlAguapWf

-u pusjpol l401m. 'essasup !8sq "nu"

-adA_q pus onpuP~~~~OR IUDpWWp8I

. foow atnnI

-pVq' soq "u- tdgiC pU8e o#su=qU

-* I Izoupusgpous mu

-o pu piwq °.

R ImP

-si Pso

Vusmpll@"O511ip q

i I1

to ?;- o4 +

14 +

C+

0l+0e +0

+

:1:0

m ++ ++

op4eqj2q 0

Ieq I

cot1 +oo+ ++o

CI + 0++ ++0

00X'00 8,t@ot | ++o

| 0+ + + N

I 400+++ ++

-+ +

co;g;1 +++++++++oo

+ ++++

10+

~~+0~0,

@0+

+ + +

~ ++++ ++

IMI .- a A d I..

oiv et- t-e c'

t o o ovo noq o O

°0° 00 -i

| | | . .e 8 | g |! |i sCi cm

odetoooo'4s-zlaf°C4 4te9o 1

oI0Q0II "lW C410lo-t e V v

I 131+++ 1XO" °" U60%

coI4IIi 6&I I eqe e I

Cl . ..+81_ . .C . . e. . e U

coeq4z~~ ~ ~ ~ ~ ~ ~

0~~~~~~~~~~Wao49

*11 ro'o eOaO toG

0a_ ;

1

g odj; i32?2d ¢gdd {

++"

r0(-10 to co*eq

VDC~ ~ ~ ~

t C ; 0; co

0o'0040'4

0~c-1c s a

3zt-J ~ : g

.4 u

i --4

522

a

VOLUMEOF THE BLOOD IN CONGESTIVEHEART FAILURE

TABLE II

Determinations repeated after clinical change

Patient number 12 13 14 15

Sex F M M M

Clinical change Improved Improved Worse Improved

Interval between determinations 18 days 13 days 6 days 4 days

Venous pressure, mm. of H20 144 107 64 109 270 244 114Arm to tongue time, seconds 13 13 26 35 37 33Vital capacity, liters 2.0 2.6 2.9 2.5 1.4 1.7Arterial 02 saturation, per cent 92.5 92.5 68.7 90.3 97.1 94.0 85.9 91.4Venous 02 saturation, per cent 75.0 60.0 53.6 60.6 62.3 16.2 65.1 70.7Capillary unsaturation, volumes per cent 2.5 3.9 7.2 5.2 3.6 7.1 .4.5 4.2Hemoglobin, grams 12.3 12.5 13.7 16.3 13.8 12.1 13.7 17.4Volume packed red cells, per cent 35.4 38.9 43.2 47.0 46.5 42.2 44.8 51.8Plasma proteins, grams per cent 5.7 6.2 6.1 6.8 6.8 5.9 6.0 4.7Blood volume, per cent change +34 +21 +41 +21 +8 +25 +14 -8Cell volume, per cent change +19 +18 +36 +27 +13 +19 +15 +7Plasma volume, per cent change +44 +23 +45 +16 +4 +31 +14 -20Weight, kgm. 62 57 56 62 91 79

sium oxalate. The arterial blood pressure and the heightand weight were determined.

Analyses of the blood samples for oxygen and carbondioxide content and oxygen combining power were made(7) and the hemoglobin content calculated. Capillary un-saturation was calculated from Lundsgaard's (8) formulaA + V where A and V are the arterial and venous un-2saturation respectively. The carbon dioxide combiningpower (9) and the plasma total protein (micro-Kjeldahl)were determined. In nine patients, satisfactory teleo-roentgenograms of the heart were taken. These weremeasured to obtain the internal diameter of the chest atthe level of the fourth costochondral junction and fortransverse diameter of the heart, taking the horizontaldistance from the outermost right to the outermost leftborder. The ratio of the transverse diameter of the heartto the internal diameter of the chest was calculated.

The data obtained are presented in Table I. In severalof the patients, the whole procedure was repeated after aclinical change in the patient, and the relevant data show-ing these changes are presented in Table II.

RESULTS

The blood volume was increased in all casesand in the great majority, the increase was sub-stantial. The mean increase over predicted was46 + 27 2 per cent. The plasma volume was in-creased with a mean of 44 + 272 per cent for thegroup as was the cell volume with a mean of 48+ 43 2 per cent. While the mean increase in the

cell volume is greater than the increase in theplasma volume, in the light of the variability

2 One standard deviation not corrected for small size ofsample.

within the group, there is no statistical signifi-cance in the difference between these means.Even when the more sensitive comparison in eachindividual case is made, the mean difference be-tween the cell volume and the plasma volume isonly plus 5.5 per cent. This mean is not sig-nificantly different from zero by Fisher's methodof t (11).

The mean of the venous pressures of the groupis 203 + 45 2 mm. of water in the fourteen caseswhere measurements were made. In nearly allthe cases, right upper quadrant pressure for aboutten seconds produced a rise in venous pressure.This test has often been discussed in connectionwith congestive heart failure, but in this study, itwas performed primarily for the purpose of as-suring a free connection between the manometerand the venous circuit.

The venous pressure was compared with theblood volume, the plasma volume, and the cellvolume, by graphic and statistical methods. (Fig-ures 1 and 2 and Table III.) There is no sig-nificant correlation between the blood volume orthe plasma volume and the venous pressure. Theinterpretation of this statement must be strict if itis to have meaning. These patients all suffer fromsevere congestive heart failure. If a group ofcases ranging from normal through mild insuffi-ciency to severe heart failure were studied, therewould in all probability be a high correlation be-tween venous pressure and the blood volume orplasma volume. The finding means simply that

523

GEORGER. MENEELYAND NOLAN L. KALTREIDER

within a group of patients with severe failurethere is no correlation between venous pressureand blood volume or plasma volume. There is anegative correlation of moderate significance be-tween the cell volume and the venous pressure.

The circulation time for the group is increasedto a mean of 31 + 18 2 seconds. Two patients(case number 3 and case number 12) had shortcirculation time from arm to tongue. Both ofthese had hyperthyroid heart disease. The cir-

0

0

(4DG%+q)10Q)

A@ Ie I I I

0 0 20 40 60 60 100 120% CHANGE CELL VOLUME

14a

%V S 19-0z2 lo4a

<

4

z

12

6I-114

-S4

0s a i £

140 160 t0 0 20 40 60 60% CHANGE PLASMA VOLUME

43

+

0

260

z

.240a

.200:

0160

01200

z

usI> so0

20 40 60 6O 100 120

% CHANGE CELL VOLUME140 160

0z00hi40

60

50

hiX 40Pz0

49-J

30

01 C) 20®-12/

a

a 00Os'0

t @~Oo o

0 0 20 40 60 60% CHANGE PLASMA VOLUM9

100

0) 3

@9)0'60 0 t0 40 60 60 100 1t0 140 160 -20 0 40 60 s0 100

% CHANGECELL VOLUME % CHANGEPLASMA VOLUME

FIG. 1. CORRELATIONS BETWEEN CAPILLARY UNSATURATION, VENOUSPRFESSURE, AND CIRCULATIONTIME, AND THE PER CENT CHANGEIN THE BLOOD VOLUMEAND ITS COMPONENTSIN 15 PATIENTS

WITH SEVERE CONGESTIVE HEART FAILURE

'41

I.0I22 t

0 10

49a

I 8490z

,, 6

4

IL40)

.0

*0

100

0 0

280

0 240

200hia00 160wai> 120m0z

I0 60

o 0

a 60z00

e

a 50

hi*- 40

z2) 30I-49

O20£ .a

I-

a a a t a I 2 a a 2 a a a t a a

524

(DI02daa.a

6)() 10

. (D

VOLUMEOF THE BLOOD IN CONGESTIVE HEART FAILURE

.0

0 10 20 30 40 .50 60 70 0

% CHANGE BLOOD VOLUME

1.7

hIda 1.60.-z

;: 3.54

-i0g 1.4aaI.0 1.3

_ _ 1.2

. 1 0 II .90 100 110 * 54

0

@0P)

.56 .62 .66 .70 .74TRANSVERSE DIAMETERHEART

INTERNAL DIAMETER CHEST

0

00

0D

stoo

60

240

200

IOO620

0 10 20 30 40 50 60 70 60 90 100 1 10

% CHANGE BLOOD VOLUME

7or

0D at0) 60laa-i0

50

0-354 40re

30

a

+2@0

@)

10 20 30 40 50 60

CIRCULATION TIME, SECONDS

0

II-

00 (~~~~~SID|

IF

I v0 10 20 30 40 50 60 70 0 90 100 110 40 50 60 70 90

S CHANGE BLOOD VOLUME % ARTERIAL 0v SATURATION

FIG. 2. CORRELATIONS BETWEENVARIOUS CARDIOVASCULARMEASUREMENTSIN 15 PATIENTS WITHSEVERE CONGESTIVE HEART FAILTURE

culation time was compared with the blood vol-ume, plasma volume, and cell volume (Figures 1and 2 and Table III). There is no simple rela-tion, the coefficients of correlation all being non-

significant. As in the case of the venous pressure,

the interpretation must be strict: within a group

of patients with severe heart failure, there is no

simple relation between the blood volume or itsfractions and the arm to tongue circulation time.

Because both the venous pressure and the cir-

14

I 9

I1I0

6

4

525

as0

:0

:i0-490

49

3-I

49

0

2600.1

. 240

S10

N

, 200

0

.0

0 120

60

0az00aia

hia

F40

<-i0if

50

30

t0 A

a a 1-60.1 %.VI a a a 2 --I-

GEORGER. MENEELYAND NOLAN L. KALTREIDER

culation time usually rise as congestive heart fail-ure develops, it might be anticipated that therewould be a simple relationship between these two.Within this group, such is not the case. There isno significant correlation between the venous pres-

TABLE III

Correlation between various measurementsCoefficients italicized are significant

Correlation Cas coeff Sccient (-

Venous presure-blood volume 14 -0.418 noVenous p ure-plasma volume 14 -0.400 noVanous pressure-cell volume 14 -0.54 lessthan0.06Circulation time-blood volume 12 -0.105 noCirculation timeplasma volume 12 -0.199 noCirculation time-cell volume 12 +0.231 noCapillary unsaturation-blood volume 15 +0.264 noCapillary unsaturation-plasma volume 15 -0.257 noCapillary unsaturation-call volume 15 +0.6S0 lessthan0.02Venous pressure-circulation time 12 +0.099 noArterial C02 content-arterial 02 saturation 15 -0.856 lessthan 0.01

T. D. HeartLog circulation time- 8 +0.959 leuthan0.01

I. D. Chest

sure and the circulation time, another statementwhich must be interpreted strictly as applying tothis group of patients.

The saturation of arterial blood with oxygenranged from normal levels to the low level of 41.8per cent. The majority of the patients, eleven innumber, show values between 80 and 94 per centsaturation, which is the common range in heartfailure. Only four patients show values above 90per cent, indicating that most of these patientshave anoxic anoxia in varying degree.

While venous blood from the antecubital veindoes not represent true mixed venous blood, it is,when taken under standard conditions, a usefulindex. The venous oxygen saturations observedrange from 75 per cent to the very low value of16.2 per cent. This lowest venous oxygen satura-tion occurred in the individual (number 14) whohad the highest arterial saturation, representingpurely stagnant anoxia. However, for the mostpart, the subjects studied showed mixtures of thestagnant and the anoxic forms of anoxia.

The best formulation of the degree of anoxiaavailable from the data obtained is the capillaryunsaturation expressed by Lundsgaard (8), as

mentioned above. The mean capillary unsatura-tion in this group was 6.6 + 3.42 volumes percent. Normal capillary unsaturation is 3.5 vol-umes per cent, and cyanosis appears at the levelof 6.0.

To determine the relation of anoxia to theblood volume, the capillary unsaturation was com-pared with the blood volume and its fractions.There was no significant correlation with totalblood volume nor with plasma volume, but therewas a highly significant correlation with cell vol-ume. The coefficient of correlation between thecell volume and the capillary unsaturation was+ 0.630.8

The carbon dioxide content of the arterial bloodcovers a wide range, from marked degrees of re-tention to quite low values. Taking 42 to 54 vol-umes per cent as the normal range (10), there arefive cases below this range and two above it.The patients with marked carbon dioxide reten-tion (numbers 2 and 11) had associated, markedanoxia. Both these patients had primarily pul-monary disease.

The vital capacity in all cases was low, rangingfrom 2.5 liters to 0.7 liters. The patient with thehighest vital capacity (number 14) had also thehighest oxygen saturation, and the individual withthe lowest (number 2) had the lowest arterialoxygen saturation, but among the others there isno relationship apparent.

In those patients where roentgenograms of thechest were taken, measurement of the heartshowed that all had large hearts. In order toallow for variation in stature of the individuals,the ratio of transverse heart diameter to internalthoracic diameter is used rather than the size ofthe heart itself. In normal patients, measured asin this study, this ratio is usually well below 0.5.It ranged in eight available cases between 0.564and 0.750 with a mean of 0.66 + 0.07.2 An in-teresting relation obtains between the T.D./I.D.ratio and the circulation time. A linear plot ofpairs of data suggested an exponential relationand a highly significant correlation was found be-tween the logarithm of the circulation time andthe T.D./I.D. ratio with a coefficient of correla-tion of 0.929.4

The hemoglobin concentration and the hemato-crit give very little information with regard to

8 A coefficient of correlation greater than 0.592 will oc-cur in only two per cent of cases by chance with fifteenpairs of data. Fisher's Table V.A (11).

4A coefficient of correlation greater than 0.834 will oc-cur by chance in only one per cent of cases with nine pairsof data. Fisher (11).

526

VOLUMEOF THE BLOOD IN CONGESTIVE HEART FAILURE

the actual volume of circulating blood or theabsolute increase of either fraction. While thelowest values observed for volume of packed redblood cells occurs in those patients with the small-est change of total cell volume from normal, andwhile those patients with the largest volume ofpacked red cells had substantial increases, yet thecorrelation is very loose. Between the values of40 and 45 volumes per cent of packed red cellsthere are five cases with red cell volumes rangingfrom normal to 106 per cent above predicted.

The plasma carbon dioxide combining powerobserved in these patients varies over a wide rangefrom 33 volumes per cent to 76 volumes per cent.It is closely related to the arterial carbon dioxidecontent. There is no relation between the carbondioxide content and the blood volume or its frac-tions within this group of patients.

Serum proteins averaged 5.91 + 0.47 2 gramsper cent, in the lower range, but not by any meansin the range of edema production. The volume ofthe blood and its fractions bore no simple rela-tion to the plasma proteins.

Three patients were studied after improvementof their condition had been observed clinically.Patient number 12 had little anoxia and verylittle increase in cell volume. As she improved,there was diminution of her blood volume, due en-tirely to reduction in plasma volume. In the othertwo, the blood volume also diminished but the cellvolume shared in the reduction, although not tothe same extent as the plasma volume.

Patient number 14 who became worse whentherapy was withdrawn and salt and fluids givenfreely, showed essentially the opposite change.Blood volume increased due to an increase in bothfractions but principally due to plasma increase.Hemodilution occurred. Capillary unsaturationdoubled due to stagnation, although changes inthis value had not been striking in the patientswho improved.

DISCUSSION

The blood volume method described by Greger-son, Gibson, and Stead (12), using the blue dyeT-1824 suggested by Dawson, Evans and Whip-ple (13), was adapted to clinical use by Gibsonand Evans (3). It is thought to be more accuratethan previous dye methods because of improvedcolorimetry, correction for hemolysis, allowance

of adequate time for mixing, and correction forthe disappearance of dye from the blood stream.Gibson and Evans found two-thirds of their nor-mal patients had blood volumes within plus andminus 10 per cent of the average for their sex,and that predictions of greater accuracy could bemade on the basis of sex and height (or surfacearea).

Freeman, Freedman, and Miller (14) foundthat under certain conditions, notably shock, theblue dye may escape from the blood stream, asevidenced by failure of the extrapolated dye lineto check with subsequent determinations, steepslope of subsequent dye lines, and the presence ofdye stained fluid in the tissue spaces, cisternachyli, pericardial sac, and the lumen of the in-testine. In the cases studied, there is no reasonto believe that permeability of the vascular tree tothe dye existed, because there was no unusualdye line slope and blue staining of transudateswas never observed although many were ex-amined. Weconcluded that determination of theplasma volume in these patients was as accurateas in Gibson and Evans' standard series, that is,better than 10 per cent in two-thirds of the casesstudied.

More serious objection exists to the determina-tion of, the cell volume. Smith, Arnold, andWhipple (15) ascribed the lack of agreement be-tween blood volume methods which depend onplasma volume measurements and those which de-pend on cell volume measurements to variation inthe ratio of cells to plasma in different parts ofthe vascular tree. Ebert and Stead (16), by theirbleeding experiments, have recently produced evi-dence that this is indeed the case as has Hahn(17), using radioactive red cells. It is evident,then, that in the absolute values reported for redcell volume, there is an error of unknown magni-tude which arises from the assumption that thehematocrit ratios obtained with venous blood rep-resent the average body hematocrit. However,the relative changes in the total cell volume prob-ably are significant.

The blood volumes found in the nineteen casesstudied were all above normal, and in 13 cases,the increase was over 20 per cent. This is in ac-cord with the finding of Gibson and Evans (1, 2)that in congestive heart failure, patients withdiminished blood volumes do not occur. It was

527

GEORGER. MENEELYAND NOLAN L. KALTREIDER

not found, however, that there was any significantdifference in the change in the two fractions.Great variability exists in the elevation of theblood volume itself, and in the plasma and cellvolumes in the individual cases, although, in themajority, elevation of both components occurred.

Consideration of changes observed in the vol-ume of the blood may well involve separate con-sideration of the two fractions, for there is evi-dence that control of the watery and of the cellularfractions of the blood are mediated in differentways.

Anoxemia has been considered as a factor inchanges in hemoglobin and red cells since the workof Bert in 1882 (18), and Wintrobe and Harrop(19) have reviewed the hematologic effects ofanoxemia. In brief, there is an increase in the cir-culating red cell volume in chronic anoxemia dueto altitude, with an increased total blood volume,due largely to increase in the cells. Upon returnto normal levels, the cell volume falls and there isa slight compensatory rise in the plasma volume.Compensatory polycythemia is thus a cellular af-fair, with essentially normal plasma volume. Forreasons not understood, anoxemia does not alwaysproduce a polycythemia as noted by Kaltreider,Hurtado, and Brooks (20) and Lemon (21).

In heart disease, however, the change in theblood volume involves both fractions. From thework of Bolton (22), it would appear that theearlier change is in the plasma volume. It mightbe supposed that anoxemia does not become a fac-tor until stagnation and pulmonary congestionproduce it, and in the train of anoxemia, increasedhematopoietic activity ensues.

There is increased hematopoietic activity in con-gestive heart failure: Weil in 1901 (23) andFromherz in 1903 (24) noted virtual replacementof the yellow marrow by red in polycythemic con-genital cardiac patients. Waller and Blumgart(25) showed a definite reticulocytosis, diminishedexcretion of urobilin, and increased red cell massbefore congestive heart failure becomes advanced.With onset of clinical improvement, when reduc-tion of the total blood volume ensues, the red cellsundergo hemolysis (due in part to the observedincreased fragility), the level of bilirubin in theblood rises, the excretion of urobilinogen increases,and the reticulocyte percentage becomes smaller.

From these findings and the data reported

above, it appears that the cellular portion of theblood responds to the observed anoxemia in heartfailure according to the degree in which it is pres-ent, due to increase or decrease in hematopoieticactivity and changes in blood destruction as thepatient becomes more or less anoxemic. More sat-isfactory formulation of the degree of anoxemiathan the "capillary unsaturation" should show acloser relation, for it is well recognized that it isthe oxygen tension, not the concentration, which isof primary importance.

Nothing in the data accumulated bears a simplerelation to the plasma volume. It does not cor-relate with circulation time, anoxemia, plasma car-bon dioxide combining power, nor plasma pro-teins. It is more labile than the cell volume asevidenced by the four repeated studies, a phenome-non observed by Gibson and Evans (1) who notedthe more rapid decline of the plasma volume withclinical improvement, but also a subsequent cellvolume decline.

The negative correlation observed between thevenous pressure and the cell volume within thisgroup of patients is difficult to understand. Sur-vey of a group of patients such as Gibson's (1)shows general rise of the blood volume and itsfractions with rising venous pressures. Despitethe statistical indication of probable significance,it is most likely an artificial result of selection ofthese patients, who all had elevated venous pres-sures; for in other factors studied, a better distri-bution over the range from normal to marked ab-normality was obtained. Conceivable, however, isthe hypothesis that if the elevation of venous pres-sure is compensatory, then in those who respondless well, more anoxemia occurs and hence greaterhematopoietic stimulation. In the interpretationof these results, it is important to remember thatthese correlations were derived from a very limitedseries of patients, suffering from severe congestiveheart failure. It would appear that the control ofthe plasma volume in severe congested cardiacs ismediated by mechanisms not studied in this groupof patients.

The role of the blood volume in the completepicture of congestive heart failure remains obscure.From the known response of the cell volume andthe hematopoietic system as a whole to anoxe-mia, and from the observed parallelism of increas-ing cell volume with increasing grades of anoxe-

528

VOLUMEOF THE BLOODIN CONGESTIVE HEART FAILURE

mia, the contribution of the cell volume to theincreased blood volume of congestive heart failureis comprehensible. It appears likely, however,that the changes in the plasma volume are pri-mary to the changes in cell volume, because theplasma volume is more labile, because heart failureis at least moderately advanced before anoxemiaof any significant degree obtains, and because riseof the plasma volume occurs first in experimentalcongestion. The data obtained and the analysesmade do not shed light upon the question ofwhether the increased blood volume is the resultof the increase in venous pressure or the causeof it.

Failure of the venous pressure to correlate withthe circulation time is at variance with the generalexperience that as heart failure advances, the circu-lation time and the venous pressure both increase.Two factors appear to contribute to the observa-tions made here: first, the selection only of se-verely congested cardiacs, and second, the impos-sibility of evaluation of right versus left heartfailure from the data obtained.

Most interesting is the relation between the cir-culation time and the transverse diameter of theheart. First mention of this relation appears inthe French literature in the work of David andBouvrain (26). Since there are only nine casesobserved in this study, it is not advisable to specu-late upon this relationship until further study addsa larger amount of data.

The plasma protein values found agree with thegenerally accepted view that in heart failure, theproteins are lower than normal, but not loweredto a degree sufficient to produce edema. Inter-pretation of the significance of plasma proteins inheart failure should only be made in the light ofthe state of fluid balance of the patient, accordingto Stewart (27), and as these patients were notfollowed in this regard, no comment should bemade upon the proteins.

SUMMARYANDCONCLUSIONS

1. In fifteen patients with severe congestiveheart failure, the blood volume, arterial and venousblood pressure, arm to tongue circulation time,vital capacity, arterial and venous oxygen andcarbon dioxide content and oxygen combiningpower, volume of packed red cells, plasma carbon

dioxide combining power, and plasma total pro-teins were measured and are reported.

2. The blood volume was increased in all cases,and in most, the increase was substantial.

3. There was great variability in the increaseand in the degree to which the plasma and thecells contributed to the increase, and no statisti-cally significant difference was observed betweenthe increase in cell and plasma volume.

4. There was no simple correlation between thetotal blood volume and the other measurementsmade, nor between the venous pressure and thecirculation time.

5. A negative correlation of moderate signifi-cance was found between the venous pressure andthe plasma volume.

6. The logarithm of the circulation time bore alinear relation to the ratio between the transversediameter of the heart and the internal diameterof the chest in eight cases where roentgenogramsof the chest were taken.

7. The venous hematocrit and hemoglobin gavelittle indication of the actual volume of circulatingcells.

8. There was a highly significant correlation be-tween the degree of anoxemia measured by Lunds-gaard's "capillary unsaturation" and the increasein cell volume.

9. In three patients who improved, all measure-ments tended towards normal, the plasma volumedecreasing more than the cell volume. In the pa-tient who grew worse, the reverse was the case.

BIBLIOGRAPHY

1. Gibson, J. G., Jr., and Evans, W. A., Jr., Clinical stud-ies of the blood volume. III. Changes in bloodvolume, venous pressure, and blood velocity rate inchronic congestive heart failure. J. Clin. Invest.,1937, 16, 851.

2. Gibson, J. G., Jr., The clinical significance of the bloodvolume. Ann. Int. Med., 1941, 14, 2014.

3. Gibson, J. G., Jr., and Evans, W. A., Jr., Clinical stud-ies of the blood volume. I. Clinical application ofa method employing the azo dye "Evans blue" andthe spectrophotometer. J. Clin. Invest., 1937, 16,301.

4. Gibson, J. G., Jr., and Evans, W. A., Jr., Clinical stud-ies of the blood volume. II. The relation of theplasma and total blood volume to venous pressure,blood velocity rate, physical measurements, age, andsex in ninety normal humans. J. Clin. Invest., 1937,16, 317.

529

GEORGER. MENEELYAND NOLAN L. KALTREIDER

5. Winternitz, M., Deutsch, J., and Brull, Z., Eine klin-isch brauchbare Bestimmungsmethode der Blutum-laufszeit mittels Dekholininjection. Med. Klin.,1931, 27, 986.

6. Griffith, G. C., Chamberlain, C. T., and Kitchell, J. R.,A simplified apparatus for direct venous pressuredetermination modified from Moritz and v. Tabora.Am. J. M. Sc., 1934, 187, 371.

7. Van Slyke, D. D., and Neill, J. M., The determinationof gases in blood and other solutions by vacuum ex-traction and manometric measurement. J. Biol.Chem., 1924, 61, 523.

8. Lundsgaard, C., Studies on cyanosis. I. Primarycause of cyanosis. J. Exper. Med., 1919, 30, 259.

9. Van Slyke, D. D., and Cullen, G. E., Studies of acido-sis. I. The bicarbonate concentration of the bloodplasma; its significance and its determination as ameasure of acidosis. J. Biol. Chem., 1917, 30, 289.

10. Peters, J. P., and Van Slyke, D. D., Quantitative Clin-ical Chemistry. Williams and Wilkins, Baltimore,1931.

11. Fisher, R. A., Statistical Methods for Research Work-ers. Oliver and Boyd, London, 1938

12. Gregerson, M. I., Gibson, J. G., Jr., and Stead, E. A.,Plasma volume determination with dyes; errors incolorimetry; use of the blue dye T-1824. Am. J.Physiol., 1935, 113, 54.

13. Dawson, A. B., Evans, H. M., and Whipple, G. H.,Blood volume studies. III. Behavior of a largeseries of dyes introduced into the circulating blood.Am. J. Physiol., 1920, 51, 232.

14. Freeman, N. E., Freedman, H., and Miller, C. C., Pro-duction of shock by prolonged continuous injectionof adrenalin in unanaesthetized dogs. Am. J. Phys-iol., 1941, 131, 545.

15. Smith, H. P., Arnold, H. R., and Whipple, G. H.,Blood volume studies, VII. Comparative values ofWelcker, CO, and dye methods for blood volumedeterminations. Accurate estimation of the abso-lute blood volume. Am. J. Physiol., 1921, 56, 336.

16. Ebert, R. V., and Stead, E. A., Jr., Demonstrationthat the cell plasma ratio of blood contained inminute vessels is lower than that of venous blood.J. Clin. Invest., 1941, 20, 317.

17. Hahn, P. F., et al., Red cell volume circulating andtotal as determined by radio iron. Science, 1941,93, 87.

18. Bert, P., Sur la richesse en hemaglobine du sang desanimaux vivant sur les haut lieux. Compt. rend.Acad. d. sc., 1882, 94, 805.

19. Harrop, G. A., and Wintrobe, M. M., Polycythemia.Handbook of Hematology, H. Downey, Editor.P. B. Hoeber, Inc., New York, 1938.

20. Kaltreider, N. L., Hurtado, A., and Brooks, W. D. W.,Study of the blood in chronic respiratory diseases,with special reference to the volume of the blood.J. Clin. Invest., 1934, 13, 999.

21. Lemon, W. S., A study of the effect of chronic pul-monary diseases on the volume and composition ofthe blood. Ann. Int. Med., 1929, 3, 430.

22. Bolton, C., The pathology of cardiac dropsy. Brit. M.J., 1917, 1, 642; and ibid., 678.

23. Weil, E., Note sur les organes hematopoietiques etl'hematopoiese dans la cyanose congenitale. Compt.rend. Soc. d. biol., 1901, 53, 713.

24. Fromherz, E., Zur Bedeutung der Hyperglobulie beikongenitalen Herzkrankheiten. Munchen. med.Wchnschr., 1903, 1, 1718.

25. Waller, J. V., Blumgart, H. L., and Volk, M. D.,Studies of the blood in congestive heart failure,with particular reference to reticulocytosis, erythro-cyte fragility, bilirubinemia, urobilinogen excretion,and changes in blood volume. Arch. Int. Med.,1940, 66, 1230.

26. David, C., and Bouvrain, Y., Contribution a l'etudephysiopathologique de la vitesse de circulation.Arch. d. mal. du coeur, 1940, 33, 147.

27. Stewart, H. J., Mechanism of diuresis: Alterations inthe specific gravity of the blood plasma with onsetof diuresis in heart failure. J. Clin. Invest., 1941,20, 1.

530