DIAGNOSTIC METHODS
DOPPLER ECHOCARDIOGRAPHY
Pulsed Doppler echocardiographic determination of
stroke volume and cardiac output: clinical
validation of two new methods using the apical
window
JANNET F. LEWIS, M.D., LAWRENCE C. Kuo, M.D., JEAN G. NELSON, R.D.M.S.,
MARIAN C. LIMACHER, M.D., AND MIGUEL A. QUINONES, M.D.
Downloaded from http://circ.ahajournals.org/ by guest on June 14, 2017
ABSTRACT Two methods of measuring stroke volume and cardiac output with pulsed Doppler twodimensional echocardiography were developed and validated against the thermodilution technique in
39 patients, 33 of which were in an intensive care unit. With the use of the apical four-chamber view, a
mitral inflow method combined the velocity of left ventricular inflow at the mitral anulus with the crosssectional area of the anulus calculated from its diameter at middiastole (area = z r2). From the apical
five-chamber view a left ventricular outflow method combined the velocity of left ventricular outflow
with the cross-sectional area of the aortic anulus calculated from its diameter during early systole
(parastemal long-axis view). Measurements with the mitral inflow and left ventricular outflow methods
were obtained in 35 of 39 (90%) and 39 of 39 (100%) patients, respectively. Validation of the mitral
method excluded patients with mitral regurgitation (n = 11) and validation of the left ventricular
outflow method excluded those with aortic regurgitation (n = 4). Good correlations were observed
between thermodilution and Doppler measurements of stroke volume and cardiac output for both the
mitral anulus method (R = .96 and .87, respectively) and the left ventricular outflow method (R = .95
and .91, respectively). The results of the two methods correlated well with each other in patients
without regurgitant valve lesions. A greater interobserver variability was observed with the mitral
anulus method, which was related solely to greater variability in measuring the annular diameter. In
patients with mitral regurgitation, left ventricular inflow volume was always greater than left ventricular outflow stroke volume while the inverse was true in those with aortic regurgitation. Thus, stroke
volume and cardiac output can be accurately measured from the cardiac apex with mitral inflow or left
ventricular outflow methods when applicable. Comparison of volumes obtained with these two methods may prove valuable in quantitating the severity of mitral or aortic regurgitation.
Circulation 70, No. 3, 425-431, 1984.
RECENT technological developments have made possible the application of Doppler echocardiography to
the measurement of stroke volume and cardiac output.
Methods previously validated consist of measuring ascending aortic flow from the suprasternal window or
pulmonary arterial flow from the parasternal window.1-' These methods work on the premise that the
velocity of blood flow determined from the Doppler
From the Section of Cardiology, Baylor College of Medicine, The
Methodist Hospital, Houston.
Computational assistance was provided by the CLINFO Project,
funded by grant RR-00350, Division of Research Resources, National
Institute of Health, Bethesda.
Address for correspondence. Miguel A. Quinones, M.D., Section of
Cardiology, The Methodist Hospital, 6535 Fannin -MS, F-1001,
Houston, TX 77030.
Received Jan. 9, 1984; revision accepted April 26, 1984.
Presented at the 56th Annual Scientific Sessions of The American
Heart Association, November 1983, Anaheim, CA.
Vol. 70, No. 3, September 1984
shifts of the reflected sound waves are uniformly distributed throughout the cross section of the vessel so
that the product of the area under the velocity curve
times the cross-sectional area of the vessel is equal to
the volume of blood passing through the vessel.
Among other factors, the velocity profile within a
vessel and the accuracy of the measurements of the
cross-sectional area of the vessel affect the accuracy of
Doppler flow measurements. For instance, in patients
with aortic sclerosis or stenosis the velocity profile in
the ascending aorta becomes nonlaminar and a greater
dispersion of velocities is observed within the vessel,
invalidating the use of ascending aortic flow velocity
for measuring cardiac output. When pulsed Doppler
echocardiography is used, the vessel cross-sectional
area should ideally be measured at the site of sample
volume position for greater accuracy. Images adequate
425
LEWIS et al.
Downloaded from http://circ.ahajournals.org/ by guest on June 14, 2017
for determination of cross-sectional area of the pulmonary artery from the parasternal window or of the ascending aorta from the suprasternal notch cannot always be obtained in adult patients, thus limiting the
applicability of these techniques in clinical practice.
The cardiac apex provides access to the evaluation
of flow velocities through both mitral and aortic valves
with minimal angulation between flow and the ultrasound beam and, theoretically, may be ideal for flow
calculations. Recently, Fisher et al.8 accurately measured mitral inflow volume from the cardiac apex in
experimental animals. They recorded mitral inflow velocity in the left ventricle just distal to the mitral valve
leaflets and calculated a mean mitral valve area by
combining the maximal area of the valve orifice visualized from a parasternal short-axis view with an M
mode tracing of the valve motion.
In this investigation we have developed and validated two other approaches to measuring intracardiac
flows from the apical window: (1) a modification of
Fisher's method for measuring mitral inflow consisting of placing the sample volume at the level of the
mitral anulus and (2) a left ventricular outflow method
with the sample volume positioned at the aortic anulus
immediately proximal to the aortic valve leaflets. We
have found that results of both methods correlate well
with measurements by thermodilution and that both are
applicable to a large number of critically ill adult
patients.
Methods
Clinical population. The clinical population consisted of 39
patients, 10 women and 29 men, whose ages ranged from 35 to
86 years and who had undergone determination of cardiac output by thermodilution either while in an intensive care unit or as
part of diagnostic cardiac catheterization. In none of the patients
was clinical, echocardiographic, or Doppler evidence of aortic
or mitral stenosis found. Thirty-three patients were studied
while in the intensive care unit being treated for either a myocardial infarction or severe congestive heart failure. In the majority
of patients the clinical diagnosis (table 1) was coronary artery
disease. Mitral regurgitation, diagnosed by Doppler echocardiography as a systolic wide frequency dispersion in the left
atrium behind the mitral valve, was present in 11 patients, and
aortic regurgitation, detected as a diastolic wide frequency
dispersion in the left ventricular outflow, was present in four
patients.
All of the echocardiographic studies were performed within
minutes before the determination of cardiac output by thermodilution. Both the two-dimensional imaging and the Doppler studies were done with an Electronics for Medicine/Honeywell sector scanner equipped with 2.25 and 3.5 MHz mechanical
transducers that oscillate through an angle of 60 to 75 degrees.
The system has a movable cursor that allows Doppler sampling
anywhere along the plane of the image. When the Doppler mode
is activated, the transducer stops oscillating, the last visualized
two-dimensional image is frozen, and the crystal is converted
fully into a pulsed Doppler system. The depth of the sample
volume is adjustable to a maximum of 16 cm from the trans-
426
ducer and the length of the volume is adjustable from 2 to 20
(set at 5 mm for this investigation). The two-dimensional
image is updated automatically every fifth cardiac cycle to allow
maintenance of the sample volume in the desired position. In
addition to the audio output, the frequency shifts (AF) are processed through a fast-Fourier transform spectral analyzer and
expressed graphically as flow velocity (V) by solving the Doppler equation:
mm
AF X Vm
2Fo x CosO
(1)
where Vm = the speed of sound in the medium (1540 M/sec);
Fo = emitting frequency of the transducer; 0
the angle of
incidence between sound waves and flow. When solving the
Doppler equation the instrument assumes that 0 = 0 (cos 0
1.00). However, within the plane of the two-dimensional image, this angle may be estimated with a movable cursor and
corrections to the above equation can be made if desired (see
=
below).
Each patient was examined while in a lateral recumbent position, with the transducer at the point of the apical impulse or
slightly to the left of this area. The transducer was manually
rotated to obtain an apical four-chamber view of the heart that
provided good visualization of the left ventricular cavity with
maximal excursion of the mitral valve leaflets. The sample
volume was placed at the level of the mitral valve anulus with
the cursor line oriented as parallel as possible to an imaginary
line transversing the left ventricle from apex to mitral valve
(figure 1). Slight adjustments in transducer angulation or sample volume position were at times required to maximize the
audio and graphic quality of the Doppler signal. The velocity of
mitral inflow was recorded over several cardiac cycles at a paper
speed of 100 mm/sec. The position of the sample volume was
moved slightly from one corner of the anulus to the other in the
majority of studies so that any significant change in morphology
or amplitude of the velocity tracings that would suggest a nonlaminar velocity profile could be detected; this was not observed
in any patient. The sample volume was then gradually moved
through the leaflets and into the inflow region of the left ventricle to ensure the absence of inflow stenosis appearing as a highvelocity flow disturbance with frequency dispersion. Although
inflow stenosis was not present in any of the patients studied, it
was not uncommon to see a modest increase in peak early
diastolic velocity (for example from 50 to 70 cm/sec) when
sampling in the left ventricular inflow region. The magnitude of
this increase was greater in patients with low cardiac output and
poor mitral leaflet separation.
For the recording of left ventricular outflow, the transducer
was rotated slightly with a superior tilt until the aortic valve and
the ascending aorta were visualized and the sample volume was
placed in the middle of the left ventricular outflow immediately
proximal to the leaflet of the aortic valve (figure 2). As with the
mitral inflow method, slight adjustments in either transducer or
cursor angulation were at times required to optimize the orientation between the sound waves and flow. This was assessed by
the quality of the Doppler tracing. However, once this position
was obtained, minimal displacement of the sample volume laterally did not appear to influence the morphology or amplitude
of the flow-velocity curve. Several cycles of left ventricular
outflow velocity were recorded at a paper speed of 100 mm/sec.
The possibility of aortic stenosis was excluded by placing the
sample volume through the aortic valve and in the ascending
aorta using the parasternal and suprasternal windows and
searching for wide systolic frequency dispersion of the flowvelocity curve as an indicator of stenosis. Finally, the transducer
was placed in the parasternal position to obtain a long-axis view
CIRCULATION
DIAGNOSTIC METHODS-DoPPLER
ECHOCARDIOGRAPHY
(time-velocity integral) by the cross-sectional area of the mitral
anulus. Curves from five to 10 cardiac cycles were digitized
following the contour of the darkest portion of the velocity
tracing and an average of the time-velocity integrals was obtained.
The middiastolic transverse diameter of the mitral anulus was
measured from the second or third video frame after the initial
maximal opening motion of the anterior leaflet. Measurements
were taken from the inner edge of the lateral bright corner of the
anulus to the inner edge of the medial corner just below the
insertion of the mitral leaflets (figure 1). Measurements from a
minimum of five cardiac cycles were averaged and the crosssectional area of the anulus was derived as FT x r2, where r
of the left ventricle and the aortic valve, as shown in figure 2.
This view was selected over the apical view because the structures needed for adequate measurements of the aortic anulus
were not always properly visualized from the apical window.
Measurements and calculations. All measurements were
made with the aid of an off-line computerized-analysis station
equipped with internal calipers and a programmable graphic
analyzer (Digisonics EC-200). The recorded two-dimensional
images ('/2 inch VHS) were played back through a videocassette
system equipped with a frame-by-frame bidirectional search
module (JVC BR6400 U).
Mitral inflow method. The mitral inflow volume was determined by multiplying the area under the diastolic inflow curve
TABLE 1
Doppler and thermodilution data
No.
Downloaded from http://circ.ahajournals.org/ by guest on June 14, 2017
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
Age/sex
48
62
58
56
64
53
68
77
65
72
45
65
76
58
63
54
62
65
52
76
66
43
77
83
72
71
52
64
86
63
47
50
67
57
60
61
83
M
M
M
M
M
M
M
M
M
F
M
F
M
M
F
M
M
M
M
M
F
M
M
F
M
M
F
M
F
M
F
M
F
M
M
M
F
35 M
69 M
Dx
Doppler
Doppler
mitral
CSA
TVI
SV
CO
3.8
4.2
3.6
5.3
21.9
14.9
12.5
8.8
83
61
44
47
4.6
6.0
4.7
5.6
5.7
1.5
10.5
12.1
60
19
5.7
1.9
2.9
6.1
17
1.9
4.5
3.1
6.2
3.1
8.5
4.2
7.8
8.0
3.7
3.7
4.7
5.1
2.8
3.5
4.9
4.9
4.5
4.0
6.2
7.1
11.3
10.2
7.3
5.7
5.7
3.1
7.7
3.7
6.2
8.9
12.7
14.8
6.9
12
10.5
9.2
7.8
18.7
6.9
8.8
8.7
11.2
8.7
9.7
10.6
11.5
6.6
23.2
16.8
15.3
22.0
23.6
15.9
12.4
11.3
9.5
28
27
79
46
59
50
82
3.6
3.8
4.5
Patiente
CRF
CAD
CAD
AMI
AMI
CAD
AMI
CM
CAD
CAD
CAD
CAD
CAD
CAD
HTN
AMI
AMI
AMI
AMI
AMI
HTN
CM
AMI
CAD
CAD/MR
AMI/MR
AMI/MR
CAD/MR
CAD/MR
CAD/MR
MR
MR
MR
MR
MR
AR
AR
AR
AR
74
4.7
5.1
4.1
6.7
5.6
3.7
5.0
3.7
3.6
29
69
33
45
25
39
43
48
48
46
41
164
190
156
160
135
91
39
87
4.6
5.4
5.8
4.9
4.7
3.8
10.3
17.5
21.2
12
11.7
5.9
4.7
6.3
35
3.9
3.0
Thermodilution
LVO
CSA
TVI
SV
CO
sv
CO
5.0
4.2
3.1
4.4
3.8
4.9
4.9
2.5
3.6
2.8
3.8
3.1
3.0
3.8
2.8
3.5
3.8
5.7
4.5
2.9
3.2
2.7
3.4
2.7
3.5
3.8
2.8
2.8
17
19.3
19.7
10.7
12.5
18.3
12.0
7.3
16.3
6.8
13.9
12.1
10.4
20.5
12.1
16.8
13
13.5
12.7
11.4
18.2
7.5
12.6
9.4
8.1
8.0
11.5
85
80
62
47
48
90
58
19
59
19
53
38
31
78
34
59
49
77
58
33
58
20
43
25
28
30
33
33
32
13
53
25
26
78
38
127
44
171
52
9.6
4.4
6.0
4.9
5.7
6.7
5.6
1.9
5.6
2.1
3.1
4.9
4.4
4.4
3.5
5.1
4.0
6.3
4.4
4.2
4.2
2.3
3.4
3.0
3.4
3.8
4.0
3.5
3.3
1.2
3.4
2.2
3.6
5.9
3.3
8.3
5.3
12.3
5.7
86
82
68
46
49
81
57
24
65
24
56
9.5
4.9
6.6
4.8
5.4
6.1
5.4
2.4
6.1
2.8
3.9
4.0
2.8
3.1
3.5
2.8
3.1
4.5
3.3
4.9
2.5
7.7
3.7
11.9
11.5
4.1
15.3
8.9
8.3
17.4
11.5
25.9
17.7
22.3
14.3
31
36
84
45
48
56
87
75
37
55
28
41
26
38
28
37
50
27
25
57
38
33
79
40
89
31
79
38
5.1
4.8
4.6
4.2
4.4
7.1
5.5
4.7
4.0
3.1
3.9
3.1
4.5
3.2
4.5
5.3
2.7
2.1
3.6
3.4
4.5
7.0
4.8
5.8
3.4
5.7
3.9
Dx = diagnosis; LVO = left ventricular outflow; CSA = cross-sectional area; TVI = time-velocity integral; SV stroke volume; CO cardiac
output; CRF = chronic renal failure; CAD = coronary artery disease; AMI = acute myocardial infarction; CM = cardiomyopathy; HTN
hypertension; MR = mitral regurgitation; AR = aortic regurgitation.
Vol. 70, No. 3,
September
1984
427
LEWIS et al.
C¢.CS
JLL CORR
4cf*/3
}1
1 1 1 1
1
A
1
t
i
Downloaded from http://circ.ahajournals.org/ by guest on June 14, 2017
FIGURE 1. Apical four-chamber view of the heart with the sample volume (SV) positioned at the plane of the mitral anulus,
indicated by the large arrows. The Doppler cursor is aligned parallel to mitral inflow. The Doppler recording of flow velocity is
shown at a paper speed of 100 mm/see. The time-velocity integral (TVI) is derived by digitizing the velocity curve as outlined on
left atrium; RA
the second cardiac cycle. LV
left ventricle; LA
right atriuLm.
=
represents half of the annular diameter. This method assumes a
circular shape for the mitral anulus and a constant cross-sectional area throughout diastole.
In normal subjects. Ormiston et al.9 performed a two-dimensional echocardiographic reconstruction of the mitral anulus
=
from multiple calibrated apical views and found the shape of the
anulus to be nearly circular during diastole. A 12% gradual
increase in cross-sectional area was observed from early diastole to end-diastole.
In the first 15 patients the cross-sectional area of the mitral
FIGURE 2. Still-frame images ot the heart during systole in the parasternal long -axis view (toJp lcft), showing where the aortic
anulus diameter is measured (white arrow), and in the apical five-chamber view (tolp right), illustrating the position of the sample
volume immediately proximal to the aortic valve. Doppler recording of left ventricular outflow velocity is shown in the I/oer
paniel at a paper speed of 100 mm/sec. AV - aortic valve; Ao = ascending aorta; other abbreviations as in figure 1.
428
CIRCULATION
DIAGNOSTIC METHODS-DOPPLER ECHOCARDIOGRAPHY
anulus was also derived by combining annular diameters (D)
from the four-chamber and two-chamber apical views as (w x
D, x D2)/4. The results were nearly identical to those derived
from the four-chamber view alone and therefore the single measurement method was selected for the investigation in order to
increase the clinical applicability of the method. Stroke volume
(SV) was calculated as
SV
=
TVI
x
M-CSA
(2)
where M-CSA is the cross-sectional area of the mitral anulus.
Cardiac output was calculated as SV x heart rate derived from
the five to 10 cardiac cycles digitized.
Left ventricular outflow method. Left ventricular outflow
volume was determined as the product of the time-velocity
integral (average of five to 10 cardiac cycles) and the crosssectional area of the aortic anulus. The outflow velocity curves
were digitized following the contour of the darkest portion of
the curve in a manner similar to the mitral inflow method.
The cross-sectional area of the aortic anulus was calculated as
x r2, where r represents half of the annular diameter (average
of five to 10 cardiac cycles) measured immediately proximal to
the points of insertion of the aortic leaflets one or two video
frames after maximal systolic leaflet separation (figure 2).
Whenever possible, the aortic anulus was also imaged in the
short-axis plane, confirming its circular shape and the lack of
significant change in size during systole. However, the diameter
measured from the parasternal long-axis plane was used to
calculate the annular cross-sectional area in order to increase the
success rate in patients in whom there were technical
difficulties.
Correction for angle. As evident by equation 1, the angle of
incidence between blood flow and sound is an important determinant of the accuracy of Doppler measurements of blood velocity, and ideally it should either be zero or well known. In
practice, this angle cannot be measured precisely from a twodimensional image for two reasons. First, total flow within a
vessel or through an orifice is not seen, and therefore, to measure an angle it has to be assumed that flow is directed parallel to
a visualized anatomic landmark, such as the walls of the aortic
root or the long-axis plane of the ventricle. Second, in a given
two-dimensional plane the angle may be significantly underestimated due to the orientation of the sound waves with flow on the
orthogonal (nonvisualized) plane. Fortunately, 0 can vary by as
much as 20 degrees with an error in underestimating flow velocity of no more than 6%. Use of the apical window has the
advantage of providing a shallow ('20 degrees) angle between
the sound beam and both mitral inflow and left ventricular
outflow (figures 1 and 2). Thus, during this investigation we
elected not to correct for the angle but rather to optimize transducer angulation so that the transducer would be as parallel to
flow as possible.
Interobserver variability. All measurements and calculations were done by an observer with no knowledge of the thermodilution data. To test interobserver variability all the measurements in 15 patients were repeated by a second observer.
This variability was expressed as a percent error for each measurement and was determined as the difference between the two
observers divided by the mean value of the two observations.
Thermodilution-derived cardiac output. The thermodilution cardiac outputs were obtained with an Edward cardiac
output instrument model 9520-A by injecting 10 ml of 5%
dextrose in water at 0° C through the proximal port of a thermodilution Swan-Ganz catheter placed in the main pulmonary artery of each patient. Cardiac output was computed as the average of several determinations. If the difference between the
lowest and highest value of the first three determinations was
7w
Downloaded from http://circ.ahajournals.org/ by guest on June 14, 2017
Vol. 70, No. 3, September 1984
less 10%, the average of the three was taken. If the difference
was more than 10%, two more outputs were obtained and the
values at the two extremes were discarded before averaging.
Stroke volume was calculated as cardiac output divided by heart
rate.
Data were analyzed by linear regression analysis.
Results
Table 1 lists the data obtained for each of the patients studied and the statistical correlations are shown
in table 2. The mitral inflow method could not be used
in four patients because the quality of the two-dimensional images was not adequate for measurements of
mitral annular diameter. The left ventricular outflow
method was successfully applied in all patients.
Mitral inflow vs thermodilution. In 24 patients with
clinically adequate mitral inflow measurements and
without Doppler evidence of mitral regurgitation, a
high correlation was observed between thermodilution-derived stroke volume and Doppler-determined
mitral inflow volume over a wide range of stroke volumes, with an R value of .96 and an SEE of 5.9 ml
(figure 3, A). There was no consistent over- or underestimation of thermodilution stroke volume by the
Doppler method, as shown by the regression equation
y = 0.91x + 5.1. Comparison of cardiac outputs
obtained with the two methods revealed an R value of
.87 and an SEE of 0.59 liters/min.
Left ventricular outflow vs thermodilution. In the 35
patients without aortic insufficiency a high correlation
was observed between thermodilution stroke volume
and Doppler-determined left ventricular outflow volume over a wide range of stroke volumes, with an R
value of .95 and an SEE of 6.4 ml (figure 3, B). As
with the mitral method, the regression equation (y =
0.91x + 7.8) indicated no significant over- or underestimation of stroke volume by the Doppler method.
Cardiac output determined by the Doppler left ventricular outflow method correlated with that by thermodilution cardiac output, with an R value of .91 and an
SEE of 0.63 liters/min.
TABLE 2
Linear regression analysis
MVI-SV
MVI-CO
LVO-SV
LVO-CO
LVO-SV
LVO-CO
TD-SV
TD-CO
TD-SV
TD-CO
MVI-SV
MVI-CO
TD
x
vs
y
=
n
R
24
24
35
35
20
20
0.96
0.87
0.95
0.91
0.95
0.87
Regression
equation
SEE
5.9 ml
0.59 1/min
6.4 ml
0.63 1/min
6.6 ml
0.64 1/min
y
y
y
y
y
y
=
=
=
=
=
=
0.91x
+ 5.1
0.80x + 0.94
0.91x + 7.8
0.85x + 1.1
1.05x -0.192
0.89x + 0.624
stroke volume; MVI = mitral valve
cardiac output; LVO = left ventricular outflow.
thermodilution; SV
inflow; CO
=
429
LEWIS et al.
lotr-
lOOr
A
0
80 F
80
/
B 1~~~~~~~~~~
..I
,/
.1
TD-SV
(cc)
60F
-
-
60 F
S
3
-1
.1,
40
40 F
0*
20 F
oL
R= 0.96, n= 24
SEE= 5.9 cc
y= 0.91X + 5.1
20
A, -'.S*
0 *
R= 0.95, n= 35
SEE= 6.4 cc
.
n al- + 71.0Q
v~~~~=
Y`
U.1x
L
IoILI
}O
I
I
0
40
80
20
40
60
100
60
80
100
MVI-SV (cc)
LVO-SV (cc)
FIGURE 3. Correlations between thermodilution (TD) and Doppler measurements of stroke volume (SV) using A, the mitral
inflow method (MVI) and B, the left ventricular outflow method (LVO). The broken lines indicate the 95% confidence limit of
the regression.
0
20
Downloaded from http://circ.ahajournals.org/ by guest on June 14, 2017
Mitral inflow vs left ventricular outflow. In the 20 patients with technically adequate mitral and left ventricular outflow measurements and without evidence of
mitral or aortic regurgitation by Doppler echocardiography, the stroke volumes determined by the two
Doppler methods correlated well with each other, with
an R value of .95 and an SEE of 6.6 ml. Likewise,
cardiac output determinations with the two methods
correlated, with an R value of .87 and an SEE of 0.64
liter/min. Neither method appeared to consistently
yield results higher or lower than the other method.
Interobserver variability. The percent error between
the two observers for the different measurements are
listed in table 3 as the means + SD. The variability in
measuring the time-velocity integral was low for both
the mitral and the left ventricular outflow methods. On
the other hand, a larger error was observed in the
determination of the cross-sectional area of the mitral
anulus (14.8%) when compared with the aortic anulus
(6.0%). This increased variability in measurement of
the mitral anulus resulted in a larger interobserver error
in the measurement of stroke volume by the mitral
method (16.4%) than by the left ventricular outflow
method (6.8%).
Discussion
This investigation clinically validates the use of
Doppler-determined flow velocity through the mitral
anulus or left ventricular outflow to calculate stroke
volume and cardiac output in patients without stenotic
or regurgitant lesions of the respective valves. The
primary advantage of this approach is the higher yield
of technically satisfactory studies in critically ill patients since the apical window allows visualization of
most of the structures involved in the measurements
and provides an optimal angle between the sound
waves and flow.
430
Both methods used in this study assumed uniform
velocities within the mitral or aortic anulus. At low
velocities, flow is usually laminar.11 In addition, the
velocity profile tends to flatten as blood converges into
the inlet of a conduit (such as from left atrium to mitral
anulus) and during rapid acceleration, e.g., during early ejection through the aortic anulus. In this study we
did not observe significant changes in velocity as we
moved the sample volume laterally within the area of
the mitral or aortic anulus and the dispersion of velocities was minimal. Thus, it appears that the velocity
profile within the normal mitral and aortic anulus were
appropriate for determination of blood flow by Doppler echocardiography.
Most Doppler methods assume that the cross-sectional area at which the sample volume is placed remains unchanged during the time period of flow. When
Doppler sampling is done in the left ventricular inflow
distal to the mitral valve leaflets, the velocity of flow
may be altered by instantaneous changes in the size of
the valve orifice. Fisher et al.8 therefore derived a
"mean mitral area" by combining two-dimensional
imaging of the orifice of the valve with a mean leaflet
diastolic separation derived from the M mode echocardiogram and were able to accurately compute cardiac
output in an experimental model. Unfortunately all the
measurements required to apply their method are freTABLE 3
Interobserver variability
Mean percent error ( SD)
CSA
TVI
SV/CO
Mitral method
LV outflow method
14.8 + 9.9
3.8+2.9
16.4 13.8
6.0 1.6
2.4+1.5
6.8+5.0
Abbreviations are as in table 1.
CIRCULATION
DIAGNOSTIC METHODS-DOPPLER ECHOCARDIOGRAPHY
Downloaded from http://circ.ahajournals.org/ by guest on June 14, 2017
quently difficult to obtain in many adult patients, particularly in the critically ill.
In this investigation we assumed that flow through
the mitral anulus was dependent mostly on the crosssectional area of the anulus and less on the mobility of
the valve leaflets. In fact, we frequently observed an
increase in inflow velocity as the sample volume was
passed through the valve orifice into the body of the
left ventricle and the magnitude of this increase was
greater in patients with low cardiac outputs, suggesting
that this assumption was correct. Although an increase
in anular area averaging 12% has been documented
from early to end-diastole by Ormiston et al.,9 this
change represents a relatively small change in diameter
since it is related to the square root of area. The method
described by Ormiston for measuring mitral annular
area is complex, time consuming, and technically applicable to a small subgroup of patients studied. Thus,
at present, the only practical alternative is to calculate
the area from a diameter measurement. In our initial
efforts we did not find any significant difference between the use of two orthogonal diameters of the mitral
anulus vs the use of one derived from the apical fourchamber view, and therefore we selected this simplified approach in order to increase the clinical applicability of the method. The results correlated well with
those obtained with the thermodilution method.
There are potentially fewer theoretical problems
with the left ventricular outflow method since the aortic anulus is indeed circular and its size changes minimally during ejection. We again used one diameter
measurement from a long-axis view in order to simplify the technique and make it more clinically applicable. The results with the left ventricular outflow
method were equal in accuracy to those with the mitral
anulus method. However, this method was easier
to apply in all patients and, importantly, involved
less interobserver variability than the mitral anulus
method.
An additional advantage of these two new Doppler
methods of measuring stroke volume and cardiac output is the potential for calculating regurgitant volumes
by comparing one to the other in patients with isolated
Vol. 70, No. 3, September 1984
mitral or aortic regurgitation. As shown in table 1, the
mitral inflow volumes were greater than the left ventricular outflow volumes in all patients with mitral
regurgitation, while the inverse was true in the patients
with aortic regurgitation. A regurgitant fraction could
therefore be calculated as the difference between the
results of the two Doppler methods divided by the
volume derived from the respective regurgitant valve.
Future clinical studies are needed to validate this new
Doppler approach against hemodynamic-angiographic
calculations of regurgitant fractions. The increased
clinical yield provided by these two new methods
should expand the applicability of Doppler echocardiography for measuring cardiac output in routine
clinical practice.
We acknowledge the secretarial assistance of Almanubia Cespedes.
References
1. Huntsman LL. Stewart DK, Barnes SR, Franklin SB, Colocousis
JS, Hessel EA: Noninvasive Doppler determination of cardiac output in man. Circulation 67: 593, 1983
2. Goldberg SJ, Sahn DJ, Allen HD, Valdes-Cruz LM, Hoenecke H,
Carnahan Y: Evaluation of pulmonary and systemic blood flow by
2-dimensional Doppler echocardiography using fast Fourier transform spectral analysis. Am J Cardiol 50: 1394, 1982
3. Alverson DC, Eldridge M, Dillon T, Yabek SM, Berman W Jr:
Noninvasive pulsed Doppler determination of cardiac output in
neonates and children. J Pediatr 101: 46, 1982
4. Distante A, Moscarelli E, Rovai D, L'Abbate A: Monitoring of
changes in cardiac output by transcutaneous aortovelography, a
non-invasive Doppler technique: comparison with thermodilution.
J Nucl Med Allied Sci 24: 171, 1980
5. Loeppky JA, Greene ER, Hoekenga DE, Caprihan A, Luft UC:
Beat-by-beat stroke volume assessment by pulsed Doppler in upright and supine exercise. J Appl Physiol 50: 1173, 1981
6. Magnin PA, Stewart JA, Myers S, von Ramm 0, Kisslo JA:
Combined Doppler and phased-array echocardiographic estimation
of cardiac output. Circulation 63: 388, 1981
7. Schuster AH, Nanda NC: Doppler echocardiography and cardiac
pacing. PACE 5: 607, 1982
8. Fisher DC, Sahn DJ, Friedman MJ, Larson D, Valdes-Cruz LM,
Horowitz S, Goldberg SJ, Allen HD: The mitral valve orifice
method for noninvasive two-dimensional echo Doppler determinations of cardiac output. Circulation 67: 872, 1983
9. Ormiston JA, Shah PM, Tei C, Wong M: Size and motion of the
mitral valve annulus in man. I. A two-dimensional echocardiographic method and findings in normal subjects. Circulation 64:
113, 1981
10. Hatle L, Angelsen B: Doppler ultrasound in cardiology, ed 1.
Philadelphia, 1982, Lea & Febiger, p 17
11. Hatle L, Angelsen B: Doppler ultrasound in cardiology, ed 1.
Philadelphia. 1982, Lea & Febiger, p 18
431
Pulsed Doppler echocardiographic determination of stroke volume and cardiac output:
clinical validation of two new methods using the apical window.
J F Lewis, L C Kuo, J G Nelson, M C Limacher and M A Quinones
Downloaded from http://circ.ahajournals.org/ by guest on June 14, 2017
Circulation. 1984;70:425-431
doi: 10.1161/01.CIR.70.3.425
Circulation is published by the American Heart Association, 7272 Greenville Avenue, Dallas, TX 75231
Copyright © 1984 American Heart Association, Inc. All rights reserved.
Print ISSN: 0009-7322. Online ISSN: 1524-4539
The online version of this article, along with updated information and services, is located on
the World Wide Web at:
http://circ.ahajournals.org/content/70/3/425
Permissions: Requests for permissions to reproduce figures, tables, or portions of articles originally
published in Circulation can be obtained via RightsLink, a service of the Copyright Clearance Center, not the
Editorial Office. Once the online version of the published article for which permission is being requested is
located, click Request Permissions in the middle column of the Web page under Services. Further
information about this process is available in the Permissions and Rights Question and Answer document.
Reprints: Information about reprints can be found online at:
http://www.lww.com/reprints
Subscriptions: Information about subscribing to Circulation is online at:
http://circ.ahajournals.org//subscriptions/