assessment ofcardiac output by the doppler ... - heart

Br HeartJf 1985; 53: 123-9

Assessment of cardiac output by the Dopplerultrasound technique aloneNEVA E HAITES,* FIONA M McLENNAN,t DONALD H R MOWAT,t JOHN M RAWLES*From the *Department ofMedicine, University ofAberdeen, and the tAberdeen Maternity Hospital, Aberdeen

suMMARY The normal range of aortic blood velocity was established in 140 healthy adults, using anon-invasive Doppler ultrasound technique. Integration of the area under the velocity-time curvefor each heart beat gave stroke distance, which, when multiplied by heart rate, gave minute dis-tance. Stroke distance and minute distance are an indication of stroke volume and cardiac outputrespectively and both show a progressive decline with age of about 1% per annum of adult life.Stroke distance gave the greatest discrimination between patient groups and, compared with agecorrected normal values, was increased by 43% in 12 pregnant women, normal in 16 patientsconvalescing after myocardial infarction, and decreased by 14% in 15 patients with hypertension, by31% in 12 patients with atrial fibrillation, and by 43% in 13 patients with cardiac failure.Measurement of aortic blood velocity and its derivatives alone, without determination of aortic

size, is a safe, simple, and physiologically valid way of assessing cardiac output.

Cardiac output is the ultimate expression ofcardiovascular function, yet we learn little abouteither from clinical examination. Even the physicalsigns attributed to cardiac failure are not pathog-nomonic of impaired cardiac function but may also befound when the heart is functioning normally withvolume overload of the circulation. In a whole varietyof clinical situations there is therefore a need to beable to assess cardiac output, but until now there hasnot been a satisfactory non-invasive method.The Doppler ultrasound technique offers the pos-

sibility of a simple to use, non-invasive bedsidemethod of measuring cardiac output, giving absolutevalues of blood flow for a modest capital outlay andlow running cost. Blood velocity is determined ultra-sonically at a point in the circulation where the crosssectional area may be measured. Three sites have beendescribed: the mitral valve orifice,' the root of theaorta,2 and the aortic arch.34 Integration of the areaunder the velocity-time curve for each heart beat givesa distance that, if multiplied by the cross sectionalarea at the point of measurement of velocity, givesstroke volume. When the measurements are made inthe aorta there are several assumptions implicit in the

Requests for reprints to Dr J M Rawles, Department of Medicine,Aberdeen Royal Infirmary, Foresterhill, Aberdeen AB9 2ZB.

Accepted for publication 2 August 1984

calculation: non-turbulent flow with a flat velocityprofile, constant aortic diameter throughout systole,and total stroke volume passing the point of velocitymeasurement. If velocity is measured in the aorticroot no allowance is made for flow to the coronaryarteries or to the head and neck if the measurement ismade in the aortic arch. The most difficult part tech-nically is measurement of aortic cross sectional area,but even with satisfactory measurements volumetriccardiac output may be considerably overestimated.5The technique we describe uses blood velocity

measurements alone without estimation of cross sec-tional area. The arch of the aorta is used, where a flatvelocity profile is a reasonable assumption,6 and bloodflow is less likely to be turbulent than in the aorticroot,7 although a substantial and unknown proportionof stroke volume has left the aorta by the time itreaches the point where velocity is measured. Clinicalexperience suggests, however, that measuring thecross sectional area is unnecessary for converting thevelocity-time integral into a volume, the velocitymeasurements by themselves closely reflecting cardiacoutput. This is fortunate as velocity measurementsmay be made with ease even in very sick patients,whereas measurement of the cross sectional areas ofeither the aorta or mitral valve may be technicallydifficult in a substantial proportion of subjects. If vel-ocity measurements are used by themselves to assess

123

on January 8, 2022 by guest. Protected by copyright.

http://heart.bmj.com

/B

r Heart J: first published as 10.1136/hrt.53.2.123 on 1 F

ebruary 1985. Dow

nloaded from

http://heart.bmj.com/

124cardiac output none of the assumptions mentionedabove need be made. Velocity alone may be measuredwith equipment used by relatively unskilled staffwhereas if combined with measurement of cross sec-tional area it is best done with a cross sectionalechocardiograph with a Doppler mode. Such equip-ment is commercially available and may give accuratevalues for volumetric cardiac output8 9 but is muchmore expensive and less portable than instrumentsmeasuring velocity alone, which have been used inconjunction with echocardiography.' O11 Considerableskill is required to get good quality cross sectionalechocardiograms for the measurement of cross sec-tional areas, and, because the equipment is multipur-pose, it is less likely to be available for repeated meas-urements at the bedside of a sick patient.

Reproducibility of the technique of transcutaneousaortovelography described here is good, with a 6*6%deviation from exact agreement with one operator andone of 7'3% between operators.12 While a satisfactorystudy showing correlation between aortovelographyand cardiac output measurement has yet to bereported, several comparisons have been madebetween aortovelography and cardiac output meas-ured by thermodilution or dye dilution, which showclose proportionality between changes in aortovelo-graphic variables and changes in cardiac output. 13- 15The safety and ease of use of the technique of trans-

cutaneous aortovelography makes the assessment ofcardiac output feasible in a wide range of situations,but validation is needed by direct comparisonbetween aortovelography measurements and cardiacoutput and by performing aortovelography in patientsknown to have normal and abnormal cardiac outputs.We have previously reported the range of aortic bloodvelocity in normal subjects and shown that it is inde-pendent of sex, body surface area, and blood pressurebut declines progressively with age.16 We reportresults from normal subjects, pregnant women, andpatients with hypertension, myocardial infarction,atrial fibrillation, or cardiac failure.

Subjects and methods

NORMAL SUBJECTSOne hundred and seventy five apparently healthy staffand patients attending a general practice surgery wereinitially selected, but 35 were excluded after a medicalhistory, physical examination, and electrocardiogra-phy because of known or suspected heart disease,hypertension, thyrori disease, concurrent drugtreatment, or recent intercurrent disease. The remain-ing 140 subjects ranged in age from 16 to 74 yearswith a balanced representation of all ages and bothsexes.

Haites, McLennan, Mowat, RawlesPREGNANT WOMENThis group comprised 12 healthy women with a meanduration of pregnancy of 18*6 weeks and a mean ageof 29 years.

PATIENT GROUPSHypertension-In the search for normal subjects 15

patients with known or newly diagnosed essentialhypertension were identified (mean age 53 years).Their mean blood pressure was 165/103 mm Hgcompared with the age corrected value of 136/83mm Hg in the normal subjects. Six were receiving notreatment, five were taking beta blockers, two withthe addition of a diuretic, and four were taking adiuretic alone.

Cardiac failure-Thirteen patients with cardiac dis-ease were studied (mean age 57 years) who wereadmitted to hospital with clinical signs of right, left,or biventricular failure for which they were stillrequiring treatment.Myocardial infarction-Sixteen patients recuperat-

ing from acute myocardial infarction within the previ-ous 10 days were studied shortly before dischargefrom hospital. None had clinical evidence of cardiacfailure at the time of aortic blood velocity measure-ments. The mean left ventricular ejection fraction ofeight patients from this group was normal at 0-62when measured at rest by a radionuclide method.

Atrial fibrillation-Twelve patients (mean age 64years) were studied: three had mitral valve disease,five ischaemic heart disease or hypertension, onetreated thyrotoxicosis, and three unexplained atrialfibrillation. None was taking diuretics or had a historyof cardiac failure, eight were taking digoxin, and oneamiodarone for the control of ventricular rate.

AORTIC BLOOD VELOCITY MEASUREMENTIn subjects in sinus rhythm aortic blood velocity wasrecorded over 10 cardiac cycles using the Trans-cutaneous Aortic Velograph (TAV) Type 1006(Muirhead Medical Ltd), a restyled version of theequipment described by Sequeira and colleagues.13The recording was made with the subject supine'afterfive minutes' rest. Longer recordings of up to 200complexes were recorded from patients with atrialfibrillation as part of another study.A transducer, emitting continuous wave ultrasound

at 2 MHz, is held in the suprasternal notch anddirected at the arch of the aorta. The same transducerdetects back scattered ultrasound which generatesDoppler shift frequencies proportional to the vel-ocities of blood in the ultrasound pathway. Theresults of spectral analysis of the Doppler frequenciesare displayed on a paper trace, which may be directlycalibrated as aortic blood velocity against time. Angu-lation of the transducer is not critical, and help in



/B


ebruary 1985. Dow

nloaded from


Assessment of cardiac output by the Doppler ultrasound technique alone

'E 200 ArealStroke distance or* 200 Area -- --------::Pasystolicvelocity integra-----------o --: --:----:.bloodvelocity---> 100 _- -

*- o._ __L

.0

° -I -- -

4 is Flow time Flow timecurve-following triangulatiorl

Fig. 1 A typical aortovelography trace showing measurement of stroke distance by curve-following and triangulation.

positioning it is obtained by maxmsmg the pitch ofthe Doppler frequencies played through a loud-speaker, at which point the ultrasound beam is in linewith the direction of blood flow.

INTERPRETATION OF VELOCITY MEASUREMENTFig. 1 shows a typical trace. The outline of thevelocity-time complex represents the highest velocityrecorded in the ultrasound path at any instant. Thecomplexes are approximately triangular, and the apexrepresents peak aortic blood velocity occurring in aor-tic midstream in midsystole, and the length of thebase of the triangle represents the duration of leftventricular ejection, or flow-time. The area within thevelocity-time complex is dimensionally a distance andis recorded in centimetres as "stroke distance,"analogous to stroke volume. Stroke distance multi-plied by heart rate in beats per minute gives "minutedistance," analogous to cardiac output. The velocity-time complex is not exactly triangular, with somerounding of the apex and non-linearity of leading andtrailing edges. In our normal subjects we compare twomethods of calculating stroke distance, by measuringthe base and height of a superimposed best fittriangle, and by a curve-following programme with adigitiser and computer. Triangulated measurementsof stroke distance were used in patients with atrialfibrillation and their controls; in other groups theresults were calculated by curve-following. Measure-ments from 10 consecutive complexes were averagedin sinus rhythm, but up to 200 complexes were aver-aged in atrial fibrillation.

STATISTICAL METHODSBlood velocity measurements in five patient groupswere compared with predicted means and theirresidual standard deviations, obtained from the nor-mal series by using the mean ages of the patientgroups in the linear regression equations for strokedistance and minute distance against age. Unpaired

t tests were used for comparison of means which aregiven with 1 standard deviation.

Results

STROKE DISTANCE BY TRIANGULATION ANDCURVE-FOLLOWINGIn 140 healthy subjects the mean peak aortic bloodvelocity was 102 cm s-I as measured by a curve-following programme with a digitiser. This was over-estimated by 14% by triangulation, but triangulationslightlly underestimates stroke distance, with meanstroke distances of 19-4 and 19-7 cm by triangulationand curve-following respectively. For practical pur-poses either method may be used to measure strokedistance, depending on the availability of digitisingequipment, the close agreement between these twomethods of measurement being shown in Fig. 2.

E 40-u

0PC.30=230-0

4,

v 20-

.0

o -

0 10-U'

y= 1.55094xr = 0.98P<0.001

6 lb io 30Stroke distance by triangulation (cm)

40

Fig. 2 Stroke distance measured by curve-following and bytriangulation. The regression line and 95% confidence limits aregiven for data from 140 normal subjects.

125



/B


ebruary 1985. Dow

nloaded from


12640*

- 30-

6 20

do

U,, 10.y=30.5 -0.24xr = -0.66P<0.0o1

O00 20 40 60 80

Age (yr)

Fig. 3 Regression ofstroke distance on age. The regression lineand 95% confidence limits are given for data from 140 normalsubjects.

Normal subjectsFigs. 3 and 4 show the 95% confidence limits for thedecline of stroke distance and minute distance withage. Both measurements fall by about 45% betweenthe ages of 20 and 70 years. The systolic-velocityintegral or stroke distance is highly correlated withpeak aortic velocity (r=0*91, p<0-001) but not withflow-time (r=0.08, 0.02<p<0.4), indicating that innormal subjects changes in stroke distance arebrought about by variation in blood velocity ratherthan in ejection time.

Patient groupsFig. 5 shows the mean values of stroke distance andminute distance in pregnancy, hypertension, cardiacfailure, myocardial infarction, and atrial fibrillationcompared with age corrected normal values. Peak aor-tic blood velocity and flow time, the apex and base of

26007

2200*

Q 1800-

; 1400-

S 1000-

600-y=2060-15.5xr=-0.55p<0 001

Haites, McLennan, Mowat, Rawlesthe triangular velocity-time complex, though not cor-related with each other in normal subjects, arenevertheless altered in the same direction in differentpatient groups. Stroke distance, the area of the com-plex, being closely related to the product of peak aor-tic blood velocity and flow-time, shows a greater sep-aration between groups than either of these measure-ments. Since a reduced stroke volume or stroke dis-tance will lead to an increased heart rate and a propor-tionately smaller reduction in cardiac output minutedistance therefore shows less discrimination betweengroups than stroke distance.Pregnancy-In pregnancy there is a 43% increase in

stroke distance (p<0O001) resulting from a 300/oincrease in peak aortic blood velocity (p<0-001) and a12%o increase in flow-time (p<0.01). Minute distanceshows an even greater increase over normal at 56%(p<0001) since mean heart rate is five beats perminute faster in pregnancy (0 1<p<02).Hypertension-Although neither peak aortic blood

velocity or flow time are significantly reduced, strokedistance and minute distance are 13% and 15% lower

30001 p<<0001 p<005 p<0 Ol p>0 3 p>0 05

2000 T

(n=15)I1

15iooo ln=13 (n=12i

I

A B C D E

0 20 40 60 80

Age (yr)

Fig. 4 Regression ofminute distance on age. The regression lineand 95% conidence limits are given for data from 140 normalsubjects.

Fig. 5 (a) Mean (SD) stroke distance and (b) mean (SD)minute distance compared with age corrected normal (0) values(n=140) infive patien groups (-): (A) pregnancy (mean age29 yr); (B) hyperension (mean age 53 yr); (C) cardiac failure(mean age 57 yr); (D) myocardial infarction (mean age 60 yr);(E) atrialfibriUation (mean age 64 yr).



/B


ebruary 1985. Dow

nloaded from



than normal respectively (p<0.05). The number ofsubjects in this group is too small to enable us to saywhether this reduction in cardiac output is due tohypertension itself or its treatment.

Cardiacfailure-Peak aortic blood velocity and flowtime are reduced by 23% and 25% leading to 43%reduction in stroke distance. Heart rate is significantlyincreased by 20 beats per minute to 90 beats/min(p<0001), compensating to some extent for thereduced stroke distance, minute distance beingreduced by 28%.Myocardial infarction-Aortic blood velocity and its

derivatives were normal in this group of convalescentpatients.

Atrial fibrillation-Peak aortic blood velocity andflow time are both reduced by 17% leading to 31%reduction in stroke distance. Ventricular rate issignificantly increased to 87 beats per minute(p<0-00l), compensating partially for the reducedstroke distance, minute distance being reduced butnot significantly.

Discussion

Cardiac output is the ultimate expression of car-diovascular function, so it is surprising that its meas-urement is made so seldom in clinical practice. Thisrelates to difficulties of measurement but also to itsinterpretation. Volumetric cardiac output has to becorrected for variation in body size and by conventionthis is done by division with an estimate of body sur-face area, even though the statistical basis for thismanoeuvre is suspect, and the correction breaks downin neonates.'7 Cardiac output diminishes with age,'8but there are no generally accepted normal valueswhich take age into consideration.Another problem is that homeostatic mechanisms

seek to stabilise cardiac output, the most important ofthese being alteration of heart rate. Cardiac output perbeat, or stroke volume may therefore be a more fun-damental expression of cardiovascular function, andthat too is usually corrected for body size by divisionby body surface area. By dimensional analysis,volume/area=distance and stroke volume/body sur-face area=stroke volume index, the latter thereforebeing a linear measurement of distance.

Non-invasive measurement of cardiac output bycombined Doppler and echocardiography depends onmeasurement of the systolic velocity integral, orstroke distance, and the cross sectional area at thepoint where the velocity measurement is made. Thesystolic velocity integral is then multiplied by thecross sectional area to give stroke volume. The rela-tion between stroke volume and stroke distancemeasured in the aorta is thus approximately as fol-lows: stroke volume/aortic cross sectional area=stroke

distance. Put in this way we see the close analogybetween stroke volume index and stroke distance.Both are unidimensional and corrected for variation inbody size, the former by body surface area and thelatter by aortic cross sectional area, which may reflectbody size and metabolic requirements rather betterthan body surface area. Having measured stroke dis-tance directly we consider it unnecessary to convert itto a volumetric measurement by multiplication withaortic cross sectional area only to correct for body sizeby dividing arbitrarily by another area, that of thebody surface.

If stroke distance is biologically corrected for bodysize there should be no correlation with body surfacearea, which is indeed the case,'6 and aortic blood vel-ocity might be expected to have a similar valuethroughout infancy and childhood, which is alsotrue,'9 which cannot be said of the cardiac index.20 21These results indicate that once a normal range of

values is established, aortic blood velocity measure-ments may be used by themselves to assess cardiacoutput without conversion to a volumetric measure-ment or correction for body size. The pronounceddecline of aortic blood velocity and its derivatives withage means that age correction is necessary, however,but prediction of mean values at any adult age under75 years is readily done from our data using theregression equations relating velocity measurementsto age.There are two possible reasons for the reduction of

aortic blood velocity and its derivatives with age:reduction of cardiac output and enlargement of theaorta. Volumetric cardiac output decreases by about1% per annum between the ages of 20 and 80,18 acomparable figure to the decline of minute distance of0-88% a year in our series. Furthermore, during adultlife there may be some passive dilatation of the aorta,but evidence on this point is scanty. One sourcereports an increase in external aortic diameter of0*75% a year.22

In pregnancy, an increase in both stroke volumeand heart rate contribute to an increased cardiac out-put of about 40% in mid-trimester.23 Our finding of a56% increase in minute distance is a higher percentageincrease than has been recorded for cardiac output,but we have confirmed this finding in a separate groupof patients at an earlier stage of pregnancy.

In the early stages of hypertension cardiac outputmay be raised, but later with increasing peripheralresistance it falls below normal.24 Treatment eitherwith beta blockers or diuretics may also cause a reduc-tion of cardiac output,25 26 so our finding of asignificant reduction of minute distance in the hyper-tensive group is not unexpected.The definition of cardiac failure is elusive,27 but it

has long been known that cardiac output is reduced in

127



/B


ebruary 1985. Dow

nloaded from


128

the syndrome of congestive cardiac failure, althoughthere is some overlap with the range of values found innormal subjects.28 In our patients with cardiac fail-ure, the mean values of aortic blood velocity and itsderivatives were all significantly reduced. Stroke dis-tance showed the greatest percentage reduction, anincreased heart rate providing partial compensationleading to a smaller proportional fall in minute dis-tance. As might be expected, the range of values forstroke distance found in cardiac failure overlappedwith normal age corrected values, but its measure-ment gives an indication of the severity of cardiacdysfunction that is not available from the clinicalexamination.At the time of discharge all aortic blood velocity

measurements were normal in the patients withmyocardial infarction that we studied, who had hadan uncomplicated course with no clinical evidence ofcardiac failure and, where measured, normal left ven-tricular ejection fractions. Stroke volume and cardiacoutput have been measured in patients with myocar-dial infarction,29 and while stroke volume and oftencardiac output were diminished on the first days afterinfarction, both were normal by the time of dischargefrom hospital.

In atrial fibrillation, loss of atrial function leads toreduction of stroke volume and cardiac output.3031Our results show a 31% reduction of stroke distancein atrial fibrillation with a 19% reduction of minutedistance, which does not, however, achieve statisticalsignificance. The increased ventricular rate in atrialfibrillation compensates for the fall in stroke volumeand may well be under automatic control, especiallyin patients taking digoxin.32 Measurement of cardiacoutput in atrial fibrillation could bring a new precisionto the control of the ventricular rate: the optimumrate is that which results in a normal cardiac output.The non-invasive assessment of cardiac output by

aortic blood velocity measurements by themselves hasgiven results that are wholly consistent with previ-ously published work where cardiac output has beenmeasured by conventional invasive techniques.Because of the safety and acceptability of the methodit is possible to collect a large series of results fromnormal subjects to compare with those from patientsof any age, in whom it is now feasible to make fre-quent sequential measurements at the bedside or inthe outpatient clinic. Cardiac output is such an impor-tant haemodynamic variable that its more frequentmeasurement-and a greater understanding of thefactors that influence it-must lead to improvedpatient care.

References

1 Fisher DC, Sahn DJ, Friedman MJ, et al. The mitral

Haites, McLennan, Mowat, Rawlesvalve orifice method for noninvasive two-dimensionalecho Doppler determinations of cardiac output. Circula-tion 1983; 67: 872-7.

2 Huntsman LL, Stewart DK, Barnes SR, Franklin SB,Colocousis JS, Hessel EA. Noninvasive Doppler deter-mination of cardiac output in man. Clinical validation.Circulation 1983; 67: 593-602.

3 Light H. Transcutaneous aortovelography. A new win-dow on the circulation? Br HeartJ 1976; 38: 433-42.

4 Rawles JM, Haites NE. Doppler ultrasound measure-ment of cardiac output: transcutaneous aortovelography.Br j Hosp Med 1984; 31: 292-7.

5 Ihlen H, Amlie JP, Dale J, et al. Determination of car-diac output by Doppler echocardiography. Br Heart J1984; 51: 54-60.

6 Schultz DL, Tunstall-Pedoe DS, Lee G de J, GunningAJ, Bellhouse BJ. Velocity distribution and transition inthe arterial system. In: Wolstenholme GEW, Knight J,eds. Circulatory and respiratory mass transport. London:J & A Churchill, 1969: 172-202.

7 Light LH, Cross G. Convenient monitoring of cardiacoutput and global left ventricular function by trans-cutaneous aortovelography-an effective alternative tocardiac output measurements. In: Spencer MP, ed. Car-diac Doppler diagnosis. The Hague: Martinus Nijhoff,1983: 69-80.

8 Fisher DC, Sahn DJ, Friedman MJ, et al. The effect ofvariations on pulsed Doppler sampling site on calculationof cardiac output: an expermental study in open-chestdogs. Circulation 1983; 67: 370-6.

9 Magnin PA, Stewart JA, Myers S, Von Ramm 0, KissloJA. Combined Doppler and phased-array echocardiog-raphic estimation of cardiac output. Circulation 1981; 63:388-92.

10 Loeppky, Greene ER, Hoekenga DE, Caprihan A, LuftUC. Beat-by-beat stroke volume assessment by pulsedDoppler in upright and supine exercise. 7 Appl Physiol1981; 50: 1173-82.

11 Alverson DC, Eldridge M, Dillon T, Yabek SM,Berman W Jr. Non-invasive pulsed Doppler determina-tion of cardiac output in neonates and children. J Pediatr1982; 101: 46-50.

12 Fraser CB, Light LH, Shinebourne EA, Buchtal A,Healy MJR, Beardshaw JA. Transcutaneous aortovelog-raphy: reproducibility in adults and children. EurJ Car-diol 1976; 4: 181-9.

13 Sequeira RF, Light LH, Cross G, Raftery EB. Trans-cutaneous aortovelography. A quantitative evaluation.Br HeartJ 1976; 38: 443-50.

14 Light LH, Sequeira RF, Cross G, Bilton A, Hanson GC.Flow-orientated circulatory patients assessment andmanagement using transcutaneous aortovelography.J Nucl Med Allied Sci 1979; 23: 137-44.

15 Distante A, Moscarelli E, Rovai D, L'Abbate A.Monitoring changes in cardiac output by transcutaneousaortovelography a non-invasive Doppler technique:comparison with thermodilution. J Nucl Med Allied Sci1980; 24: 171-5.

16 Mowat DHR, Haites NE, Rawles JM. Aortic blood vel-ocity measurement in healthy adults using a simpleultrasound technique. Cardiovasc Res 1983; 17: 75-80.

17 Guyton AC, Jones CE, Coleman TG. Circulatory physiol-



/B


ebruary 1985. Dow

nloaded from



ogy: cardiac output and its regulation. 2nd ed. Philadel-phia: W B Saunders, 1973.

18 Brandfonbrener M, Landowne M, Shock NW. Changesin cardiac output with age. Circulaon 1955; 12: 557-66.

19 Sequeira RF. Transcutaneous aortovelography. In:Woodcock JP, Sequeira RF, eds. Doppler ultrasound inthe study of the cenral and peripheral circulation. Bristol:Bristol University Printing Unit, 1978: 30-8.

20 Brotmacher L, Deuchar DC. The systemic blood flow incongenital heart disease, with an mination of the val-idity of the cardiac index. Clin Sci 1956; 15: 441-56.

21 Jegier W, Sekelj P, Auld PAM, Simpson R, McGregorM. The relation between cardiac output and body size.Br HeartJ 1963; 25: 425-30.

22 Documenta Geigy. Sciennfic tables. Macclesfield: GeigyPharmaceuticals, 1970.

23 Hytten F, Leitch I. The physiology of human pregnancy.2nd ed. Oxford: Blackwell Scientific Publications, 1971:69-110.

24 Lund-Johansen P. Central haemodynamics in essentialhypertension.ActaMedScand [Suppl] 1977; 606: 35-42.

25 Conway J, Lauwers P. Hemodynamic and hypotensiveeffects of long term therapy with chlorothiazide. Circula-

tion 1960; 21: 21-7.26 Bevegird S, Castenfors J, Danielson M. Haemodynamic

effects of saluretic treatment and beta-receptor blockadein patients with essential hypertension. Acta Med Scand1977; 201: 99-104.

27 Harris P. Evolution and the cardiac patient. 6. Origins ofcongestive cardiac failure. Cardiovasc Res 1983; 17:440-5.

28 Stead EA Jr, Durham NC, Warren JV, Brannon ES.Cardiac output in congestive cardiac failure. Am HeartJ1948; 35: 529-41.

29 Malcrona R. Haemodynamics in myocardial infarction.Acta Med Scand [Suppl] 1964; 417: 1-54.

30 Ferrer MI, Harvey RM, Cathcart RT, Cournand A,Richards DW Jr. Hemodynamic studies in rheumaticheart disease. Circulation 1952; 6: 688-710.

31 Wade G, Werko L, Eliasch H, Gidlund A, Largerlof H.The haemodynamic basis of the symptoms and signs inmitral valvular disease. Q Jf Med 1952; 21: 361-83.

32 Ferrari A, Bonazzi 0, Gregorini L, Gardumi M,Perondi R, Mancia G. Modification of the baroreceptorcontrol of atrioventricular conduction induced bydigitalis in man. Cardiovasc Res 1983; 17: 633-41.

129



/B


ebruary 1985. Dow

nloaded from


assessment ofcardiac output by the doppler ... - heart

Documents