Cardiovascular Research 46 (2000) 82–89
www.elsevier.com / locate / cardiores
www.elsevier.nl / locate / cardiores
Comparison of intravenous and pulmonary artery injections of hypertonic
saline for the assessment of conductance catheter parallel conductance
Paul Steendijk*, Jan Baan
Department of Cardiology, Leiden University Medical Center, P.O. Box 9600, 2300 RC Leiden, The Netherlands
Received 11 November 1999; accepted 6 January 2000
Abstract
Objective: The conductance catheter provides a continuous measure of left ventricular volume. Conversion of raw data to calibrated
absolute volume requires assessment of parallel conductance. Conventionally, parallel conductance is determined by injecting a small
bolus hypertonic saline into the pulmonary artery and analyzing the signal obtained during passage of the bolus through the left ventricle.
However, in some cases, a pulmonary artery catheter is not practicable. Therefore, we investigated whether intravenous hypertonic saline
injections yield reliable parallel conductance estimates. Methods: In 13 anesthetized sheep (3365 kg) parallel conductance was obtained
by pulmonary artery and by intravenous injections. Measurements (triplicate) were done at baseline, during dobutamine and pacing, and
repeated after embolization of the right coronary artery in order to assess the effects of enlarged right ventricular volumes. We used a
multiple linear regression model to determine the relation between parallel conductance obtained by the two methods and to quantify the
effects of dobutamine, pacing, and embolization. Results: The two methods show an excellent correlation with a systematic
overestimation for intravenous injection. The mean parallel conductance obtained by pulmonary artery injection was 0.69060.009 ohm 21
whereas intravenous injection yielded 0.73960.015 ohm 21 . Interanimal variability was 0.138 ohm 21 . The difference between the two
methods was relatively small, but highly significant (10.04960.012 ohm 21 , P,0.001). Embolization resulted in significantly higher
values (10.14160.017 ohm 21 , P,0.001), but dobutamine and pacing did not significantly affect parallel conductance (10.02160.016
ohm 21 , NS). There was no interaction between these interventions and the injection method, indicating that the relation between parallel
conductances obtained by the two methods was maintained in all conditions. Conclusion: Parallel conductance obtained by intravenous
injection was significantly higher (17%) than by pulmonary artery injection. However, the relation between the two methods is highly
linear with an excellent correlation and is not affected by large hemodynamic changes. The systematic difference between the two
methods is likely due to increased conductivity of blood in the right ventricle which is present with intravenous injection but not with
pulmonary artery injection. Determination of parallel conductance by intravenous injection is a good alternative for conventional
pulmonary artery injection and may be applied in studies where pulmonary artery injection is problematic. This may include studies in
very small animals or studies in patients prone to arrhythmias or with cardiac anomalies such as pulmonary artery stenosis. In addition,
intravenous injection could be used in biventricular studies to obtain right and left ventricular parallel conductances from a single saline
injection.  2000 Elsevier Science B.V. All rights reserved.
Keywords: Blood flow; Contractile function; Hemodynamics; Ventricular function
1. Introduction
The conductance catheter method provides a continuous
on-line measurement of left ventricular volume by means
of a multielectrode catheter positioned in the left ventricle.
In combination with simultaneous measurement of left
*Corresponding author. Tel.: 131-71-526-3903; fax: 131-71-5266809.
E-mail address: [email protected] (P. Steendijk)
ventricular pressure through a sensor on the same catheter
this instrument enables quantification of ventricular function by means of pressure–volume relations. Such relations
have proven to be particularly useful because they provide
indices of systolic and diastolic ventricular function that
are relatively independent of loading conditions and as
such mainly reflect intrinsic myocardial properties.
The conductance method is based on the continuous
Time for primary review 28 days.
0008-6363 / 00 / $ – see front matter  2000 Elsevier Science B.V. All rights reserved.
PII: S0008-6363( 00 )00012-2
P. Steendijk, J. Baan / Cardiovascular Research 46 (2000) 82 – 89
measurement of the electrical conductance of the blood in
the left ventricle. This signal is converted to a volume
signal on the basis of a stacked-cylinder model and by
taking into account the specific conductivity of blood and
the catheter electrode spacing. However, the conductive
tissues and fluids surrounding the left ventricular cavity
(myocardial wall, blood in the right ventricle, lung, etc.)
also contribute to the measured conductance and introduce
an offset in the relation between true left ventricular
volume and conductance-derived volume. Therefore, to
obtain an absolute volume signal this parallel conductance
needs to be determined and subtracted from the raw
conductance signal.
Parallel conductance can be determined by injecting a
small bolus of hypertonic saline through a balloon-flotation
catheter in the pulmonary artery. The highly conductive
saline transiently changes the conductivity of the blood,
practically without affecting parallel conductance. The
contribution of parallel conductance to the total conductance signal can be derived from a registration of the
conductance signal during the passage of the bolus through
the left ventricle [1]. The rationale to inject the hypertonic
saline in the pulmonary artery is that in this way the blood
in the right ventricle, which is part of parallel conductance,
is not affected by the hypertonic saline. In most studies
pulmonary artery injection, generally performed using the
distal port of a standard Swan–Ganz catheter, is feasible.
However, recently several new applications of the conductance method have been published where pulmonary artery
injections are either not possible or not practicable. In
these situations one may need to inject the hypertonic
saline into a peripheral vein or vena cava. One such
application is the use of the conductance catheter in very
small animals such as mice [2] where the pulmonary artery
injection is technically difficult. Furthermore, several
centers are currently performing biventricular pressure–
volume studies using conductance catheters in both the left
and the right ventricle [3]. Parallel conductance for the
right ventricle is generally obtained by hypertonic saline
injection in the vena cava and analyzing the data acquired
during passage of the bolus through the right ventricle. If
the same injection could also be used to determine the
parallel conductance for the left ventricle (using the data
acquired during the subsequent passage of the bolus
through the left ventricle), this would reduce the number of
hypertonic saline injections by half. Finally, some patients
react to the direct injection of hypertonic saline in the
pulmonary artery by coughing, which causes a disturbance
of hemodynamics that generally renders the data unusable
for analysis. This problem potentially might be resolved by
vena cava injection.
Therefore, in this study we compared inferior vena cava
and pulmonary artery injections of hypertonic saline for
the determination of parallel conductance. The study was
performed in 13 anesthetized sheep. Since the volume of
the right ventricle is likely to be the main determinant of a
83
possible difference between the two injection strategies we
performed measurements before and after embolization of
the right coronary artery which should cause a substantial
dilatation of the right ventricle. Furthermore, the influence
of inotropic and chronotropic changes were studied by
pacing and dobutamine infusion.
2. Methods
2.1. Animals
The study was approved by the animal research committee of the University of Leiden. The investigation conforms with the Guide for the Care and Use of Laboratory
Animals published by the US National Institutes of Health
(NIH Publication No. 85-23, revised 1996). Thirteen sheep
(body weight 33.464.9 kg, age 3–6 months) were premedicated with 40 mg / kg ketamine i.m. and 0.05 mg / kg
atropine i.m. The animals were intubated and ventilated
with a Servo 900B (Siemens Elema, Sweden). Upon
ventilation atracurium (0.25 mg / kg i.v.) was administered
to achieve adequate muscle relaxation. General anesthesia
was maintained with continuous i.v. infusion of ketamine
(4–10 mg / kg / h) supplemented with xylazine (1 mg / kg
i.m.). PaO 2 , PaCO 2 , and pH were checked every 30 min and
kept within normal ranges by adjusting FiO 2 and tidal
volume as necessary.
2.2. Conductance catheter
The conductance catheter technique has been described
in detail previously [1]. Briefly, a catheter with ten
electrodes is positioned along the longitudinal axis of the
left ventricle (see Fig. 1). The electrode distance is chosen
such that with electrode 1 within the apex, electrode 9 is
situated just above the aortic valve. Through the two most
proximal and the two most distal electrodes two 20 kHz
currents (current ratio 1:0.25) opposite in polarity are
applied, creating a dual electric field in the ventricular
cavity [4,5]. The six interposed electrodes are used to
measure the conductances of five intraventricular segments. Total left ventricular conductance, G(t), is calculated as the sum of these five segmental conductances.
Time-varying total left ventricular volume is calculated as
V(t) 5 (1 /a ) ? (L 2 /sb ) ? (G(t) 2 G p )
(1)
where a is the slope factor (see below), sb is the specific
conductivity of the blood measured from a blood sample
using a special cuvette, L is the catheter electrode spacing
and G p is the parallel conductance (see below). The
conductance catheter used in this study also incorporates a
solid state pressure sensor.
84
P. Steendijk, J. Baan / Cardiovascular Research 46 (2000) 82 – 89
Fig. 1. Simultaneous left ventricular (LV) and right ventricular (RV) conductance catheter signals during hypertonic saline injection in the pulmonary
artery (left panels) and during intravenous hypertonic saline injection (right panels). With intravenous injection the LV signal shows a small increase prior to
the actual entrance of the saline in the LV (LV wash-in). This initial increase reflects the passage of the saline through the RV, as evidenced by the signal
measured by a second conductance catheter which, in this particular experiment, was placed in the RV. As shown in the left panels this effect is absent
during pulmonary artery injection.
2.3. Slope factor a
After correction for parallel conductance the volume
signal derived by the conductance catheter is directly
proportional to actual ventricular volume, but generally
underestimates true volume by a fixed percentage. To
correct the underestimation the slope factor a was introduced. In practice, a is determined by comparing the
conductance-derived volume (or stroke volume) with an
independent measurement such as angiography or thermodilution. In animals such as dogs or sheep a is typically
0.8 [1,5]. In this study our main interest was to study
parallel conductance G p which is independent of a,
therefore a was not assessed and assumed to be 1.0
throughout the study.
2.4. Parallel conductance
The electric field generated by the conductance catheter
is not entirely restricted to the ventricular blood volume
but current also passes through the ventricular wall, other
cardiac chambers, and in fact to some extent through all
electrically conductive structures surrounding the heart. As
a consequence the total conductance signal is the sum of
the conductance of the blood in the left ventricle and the
‘parallel’ conductance of the surrounding structures. The
parallel conductance constitutes typically 50% of the total
signal and is assumed to be relatively constant during the
cardiac cycle [6,7]. Baan et al. [1] devised a method to
determine parallel conductance by injecting a small bolus
of hypertonic saline (2–3 ml, 10% saline) through a
balloon-flotation catheter in the pulmonary artery. This
procedure can be explained as follows. If blood conductivity in the left ventricle could be reduced to 0, the
measured total conductance would represent parallel conductance only. In practice this is not possible, but we can
transiently change conductivity (by the hypertonic saline
injection), plot measured total conductance vs. blood
conductivity and extrapolate this graph to the point where
conductivity hypothetically would be 0 and obtain parallel
conductance this way. This approach requires a beat-tobeat estimate of blood conductivity in the left ventricle
which can be derived from Eq. (1) as
sb 5 (L 2 /SV ) ? SG
(2)
where stroke volume, SV 5VED 2VES , and ‘stroke conductance’, SG 5 GED 2 GES with ED and ES, respectively,
end-diastole and end-systole (remember that a was set to
1). Eq. (2) shows that, if hemodynamics (and thus SV ) are
constant during the passage of the bolus, sb is directly
proportional to the amplitude of the conductance signal
(SG). Thus, parallel conductance can be obtained by
plotting GED vs. SG for each beat during the change in
blood conductivity, and extrapolating this relation to SG 5
0. This point corresponds to the hypothetical situation with
sb 5 0 and therefore yields parallel conductance.
With regard to this particular study it is important to
mention that in this analysis it is implicitly assumed that
parallel conductance itself is not affected by the hypertonic
saline injection. When the blood with altered conductivity
enters the coronary system it may affect parallel conductance. Therefore, preferably only data from the wash-in
phase, i.e. those beats where the conductance signal shows
P. Steendijk, J. Baan / Cardiovascular Research 46 (2000) 82 – 89
an increase (see Fig. 1), should be included in the analysis,
because for those beats the above mentioned effect will
still be relatively insignificant [8]. The same assumption is
also the rationale for the preference to inject the hypertonic
saline into the pulmonary artery, because this way the
blood in the right ventricle is not affected.
2.5. Instrumentation
An electrocardiogram was obtained using subcutaneous
needle electrodes. Self-sealing sheaths were placed in right
carotid artery and jugular vein, and left and right femoral
arteries and veins for the introduction of catheters. A 5F
thermodilution catheter (Arrow, Reading, PA, USA) was
placed in the pulmonary artery via the left femoral vein for
determination of parallel conductance either by injection
through the distal port in the pulmonary artery or through
the proximal port in vena cava superior. A 5F bipolar
pacing catheter was placed in the right atrium via the
jugular vein. A dual-field conductance catheter with 7-mm
electrode spacing (Millar, Houston, TX, USA) was inserted in the right carotid artery and its tip advanced to the
apex of the left ventricle. This catheter also incorporated a
solid state pressure transducer for measurement of highfidelity left ventricular pressure. Conductance and blood
conductivity measurements were performed using a
Leycom Sigma-5 DF signal-processor (CardioDynamics,
Zoetermeer, The Netherlands). Left ventricular pressure
was measured using a Nihon-Kohden pressure amplifier.
The sheath in the left femoral artery was used for the
introduction of a 4F multipurpose catheter for selective
catheterization and embolization of the right coronary
artery. Embolization was performed through infusion of a
total of approximately 10 000 polyvinylalcohol particles
with a diameter of 150–300 mm (Ivalon  ) [9,10]. All
catheters were placed under fluoroscopic guidance.
85
2.7. Data collection and analysis
ECG, left ventricular pressure and the five segmental
volume signals were recorded using a PC-based dataacquisition system and digitized at 12-bit accuracy and a
sample frequency of 250 Hz. All data acquired during the
hypertonic saline injections were stored on hard disk for
later analysis. Data acquisition was performed using
CONDUCT-PC (CardioDynamics, Zoetermeer, The Netherlands).
Parallel conductance was determined from the hypertonic saline injections using custom-made software. The
procedure to calculate parallel conductance requires only
roughly marking the begin- and end-points of the saline
wash-in period. Within that range a computer algorithm
automatically selects beats to be included in the calculation, based on the requirement for end-systolic, end-diastolic and stroke conductance to show a monotonic increase
during saline wash-in.
The Leycom Sigma-5 DF signal-processor requires the
user to dial in a value for the electrode spacing L and for
blood resistivity r (which equals 1 /sb ) and the analog
output of the system equals (L 2 /sb ) ? G(t) rather than the
raw conductance, G(t). Thus, the ‘parallel conductance’
calculated on the basis of these signals in fact equals
(L 2 /sb ) ? G p . Since, sb may change systematically during
the course of a study (usually there is a small gradual
increase), this may lead to a change in (L 2 /sb ) ? G p ,
unrelated to changes in the G p . Therefore, all initially
calculated values were divided by the actual L 2 /sb and
hence all reported values for parallel conductance in this
study refer to the ‘physical’ G p . Electrode spacing L was
0.7 cm in all studies. Blood conductivity sb was always
measured just prior to the hypertonic saline injections (in
this study the mean sb 5 0.0090260.00097 (ohm cm)21 ).
2.8. Statistical analysis
2.6. Protocol
Measurements were performed before and after embolization of the right coronary artery in the following
conditions: baseline, pacing (mean 141617 bpm) and
enhanced contractile state (2 mg / kg / min dobutamine). In
each condition the following measurements were performed. First, blood conductivity (sb ) was measured; next,
three consecutive hypertonic saline injections (2 ml, 10%
saline) were performed into the distal port of the thermodilution catheter to determine parallel conductance by pulmonary artery injection and, finally, three consecutive
hypertonic saline injections were performed into proximal
port of the thermodilution catheter to determine parallel
conductance by vena cava inferior injection. Since respiration may affect parallel conductance and actual end-diastolic volume all data were acquired during apnea at endexpiration.
The main purpose of this study was to compare assessment of parallel conductance by intravenous and pulmonary artery injections. Furthermore, we wanted to analyze
whether these results are affected by changes in inotropic
and chronotropic state and by geometrical changes of the
right ventricle. To statistically analyze these factors we
used the following multiple linear regression equation
[11,12]
G p 5 a 0 1 Sa iA A i 1 a M M 1 a E E 1 a C C 1 a ME ME
1 a MC MC
(3)
Using the coding as described below, the intercept a 0
yields the mean value of parallel conductance by conventional pulmonary artery injection. The n 2 1 dummy
variables A i account for between-animal differences allowing the n animals to have a different mean value (effects
86
P. Steendijk, J. Baan / Cardiovascular Research 46 (2000) 82 – 89
coding). The standard deviation of the group of animal
coefficients, a Ai , is a measure of interanimal variability of
parallel conductance. The dummy variable M codes the
method of injection: M 50 for pulmonary artery injection
and M 5 1 for intravenous injection. Consequently the
coefficient a M yields the mean difference in parallel
conductance between the two methods. Pre- and post
embolization are coded by E 5 2 1 and E 5 1 1, respectively, thus the mean effect of embolization on parallel
conductance equals 2 ? a E . The dummy variable C codes
the hemodynamic condition: C 5 2 1 for baseline and
C 5 1 for dobutamine and pacing, thus 2 ? a C gives the
mean difference between these conditions. The interaction
term a ME ME tests whether the effect of embolization on
parallel conductance (as measured by a E ) varies between
the two methods. Similarly, the interaction term a MC MC
tests whether changes in parallel conductance induced by
dobutamine and pacing (as measured by a C ) differ between
the two methods. Statistical analysis was performed using
commercial software (NCSS Statistical Software, Kaysville, UT, USA). Statistical significance was defined as P,
0.05.
3. Results
Fig. 1 shows representative examples of hypertonic
saline injections by the two methods used in this study.
The left panel shows a conventional pulmonary artery
injection with a typical gradual change in conductance
during passage of the bolus through the left ventricle. The
right panel of Fig. 1 shows an intravenous injection: Here
the passage through the left ventricle is preceded by a
passage through the right ventricle which is picked up by
the conductance catheter in the left ventricle as a small
initial increase in the signal. In this particular experiment a
second conductance catheter was placed in the right
ventricle (bottom tracings of Fig. 1). The passage of the
hypertonic saline through the right ventricle precedes the
passage through the left ventricle by about eight beats in
case of intravenous injection and is absent with pulmonary
artery injection. Although in the analysis we aim to include
only beats from the wash-in period for the left ventricle,
inevidently there will be some overlap between the right
ventricle wash-out period and the left ventricle wash-in
period which in theory should cause an artificial increase
in parallel conductance. In this case the average parallel
conductance for three repeat injections yielded
0.53360.040 ohm 21 for pulmonary artery injection and
0.55960.018 ohm 21 for intravenous injection. Given the
blood conductivity sb 5 0.00806 (ohm cm)21 and the
electrode spacing L 5 0.7 cm, the corresponding correction
volumes were 32.462.5 ml and 34.061.1 ml, respectively.
Fig. 2 graphically depicts all data acquired in this study.
In each animal at the various conditions three pulmonary
artery and three intravenous hypertonic saline injections
Fig. 2. Relation between parallel conductance (G p ) obtained by intravenous (IV) injection and by pulmonary artery (PA) injection. Each data
point represents the mean (6S.D.) of three repeat intravenous and
pulmonary artery injections. Solid line shows linear fit, dashed line the
line of identity.
were performed and the resulting mean (6S.D.) parallel
conductances for the two methods were plotted versus each
other. This figure illustrates the main findings in this study:
parallel conductance by intravenous injection has an
excellent correlation with pulmonary artery injections, but
overestimates the conventional method. The standard
deviations reflect the variability of the assessments, which
was very similar for both methods. Pulmonary artery
injection yielded a mean standard deviation of 0.043
ohm 21 , intravenous injection 0.040 ohm 21 , which represents 6.6% and 5.7% of the mean parallel conductances by
those methods, respectively.
To further analyze the data we created a Bland–Altman
plot [13] (Fig. 3). This plot confirms that intravenous
injection systematically yields higher values with a trend
(r 2 5 0.21) to greater differences at higher values of
parallel conductance. The mean difference between the two
methods was 0.04960.045 ohm 21 , which is significantly
different from zero (P,0.05) and represents 7% of the
mean parallel conductance.
To determine the effects of hemodynamic changes
induced by embolization, dobutamine and pacing on
parallel conductance by the two methods and to assess the
interanimal variability we used a multiple linear regression
model. The fitted coefficients are shown in Table 1. The
interpretation of these coefficients, as shown in Table 2,
indicates that mean parallel conductance by pulmonary
artery injection was 0.69060.009 ohm 21 whereas intravenous injection yielded 0.73960.015 ohm 21 . The
difference between the two methods was relatively small
(7%), but was highly significant: 0.04960.012 ohm 21
(P,0.001). Embolization increased parallel conductance
P. Steendijk, J. Baan / Cardiovascular Research 46 (2000) 82 – 89
87
not statistically significant, indicating that neither embolization, nor hemodynamic changes induced by dobutamine
or pacing significantly affected the relation between the
two methods.
4. Discussion
Fig. 3. Bland–Altman plot showing the mean versus the difference of
parallel conductance by the intravenous and the pulmonary artery
injection method. Solid horizontal line denotes the mean difference (bias)
and the dashed lines denote bias6S.D. See text for details.
significantly (P,0.001) by 0.14160.017 ohm 21 , but
changes in hemodynamics by dobutamine and pacing did
not induce significant changes (0.02160.016 ohm 21 , NS).
The interaction coefficients a ME and a MC were small and
Table 1
Coefficients multiple linear regression model G p 5 a 0 1 Sa iA A i 1 a M M 1
a E E 1 a C C 1 a ME ME 1 a MC MC a
Coefficients
P value
a
0
0.69060.009
,0.001
a
a
a
a
a
a
a
a
a
a
a
a
A
1
A
2
A
3
A
4
A
5
A
6
A
7
A
8
A
9
A
10
A
11
A
12
20.04760.030
20.25560.018
20.14460.030
0.12360.016
20.04660.022
0.01660.020
0.07860.018
20.23460.015
0.07360.017
0.11960.018
20.02560.018
0.16060.025
0.113
,0.001
,0.001
,0.001
0.032
0.406
,0.001
,0.001
,0.001
,0.001
0.160
,0.001
aM
aE
aC
0.04960.012
0.07160.009
0.01160.008
,0.001
,0.001
0.182
a ME
a MC
0.00560.012
0.00260.011
0.687
0.862
a
Multiple linear regression model to quantify effects on parallel
conductance of interanimal variability (A i ), method of injection (M),
embolization of the right coronary artery (E), dobutamine infusion and
pacing (C5condition), and the interactions between embolization and
injection method (ME) and between condition and method (MC). Dummy
variables code as follows; M 5 0: pulmonary artery injection; M 5 1:
intravenous injection; E 5 2 1: control (pre-embolization); E 5 1: postembolization; C 5 2 1: baseline; C 5 1: dobutamine or pacing.
Conversion of raw conductance catheter data to calibrated absolute volumes requires the assessment of parallel
conductance. Previous studies have shown that this calibration factor can be determined reliably using the hypertonic saline dilution method. The conventional procedure is to inject the hypertonic saline in the pulmonary
artery thereby avoiding that the procedure itself changes
parallel conductance through blood conductivity changes
in the right ventricle. However, in some cases pulmonary
artery injection is either impossible or not practicable.
Recently, the conductance catheter methodology has been
applied in open-chest mice [2,14] using miniaturized
conductance catheters. Although in these studies parallel
conductance was not assessed, infusion of approximately
2–4 ml hypertonic saline in these animals with a typical
stroke volume of 20 ml is likely to be sufficient for a
reliable assessment of parallel conductance. However, in
these small animals pulmonary artery injection is technically difficult and intravenous injection could provide a
relatively simple alternative. The need for this alternative
would be even greater if further improvements in conductance catheter design enable application in closed chest
mice as anticipated by the authors. Biventricular pressure–
volume studies [15–17] represent another example in
which intravenous injection is advantageous. In these
studies, where conductance catheters are placed in both the
left and the right ventricle, parallel conductances need to
be determined for both the right and the left ventricle. To
assess parallel conductance for the right ventricle the
normal procedure is to perform an intravenous saline
injection and analyze the signal from the right ventricular
catheter during passage of the bolus through the right
ventricle. Using the same injection and analyzing the left
ventricular signal during the subsequent passage of the
bolus through the left ventricle reduces the required
number of saline injections by half. This would be useful
especially in those studies where the protocol includes
several conditions that need to be calibrated and it may
avoid that the cumulated effect of repeated hypertonic
saline injections disturbs the physiological balance. Furthermore, in the catheterization laboratory some patients
appear to be very sensitive to the injection of hypertonic
saline in the pulmonary artery and react by coughing. This
causes a hemodynamic disturbance that generally precludes a reliable assessment of parallel conductance, since
the analysis requires hemodynamic stability. Presumably
intravenous injection could reduce this problem and enable
calibration in those patients. More in general, in some
P. Steendijk, J. Baan / Cardiovascular Research 46 (2000) 82 – 89
88
Table 2
Comparison of the pulmonary artery and intravenous injection methods a
Parameter (ohm 21 )
p
G by PA injection
G p by IV injection
Difference between IV and PA method
Effect of embolization
Effect of condition
Method–embolization interaction
Method–condition interaction
Interanimal variability
Calculation
Mean6S.D.
0
a
a0 1 aM
aM
2 ? aE
2 ? aC
2 ? a ME
2 ? a MC
S.D. (a Ai )
0.69060.009
0.73960.015
0.04960.012
0.14160.017
0.02160.016
0.00960.023
0.00460.022
0.138
P value
,0.001
,0.001
0.182
0.687
0.862
a
G p : parallel conductance; PA: pulmonary artery; IV: intravenous; S.D.: standard deviation. Parameters were derived from the coefficients of the
multiple linear regression model (see Table 1, and explanation in text).
cases the cardiologist may prefer intravenous injections for
practical reasons. For example, with a severe pulmonary
stenosis placement of a pulmonary artery catheter may be
problematic, or in patients susceptible for cardiac arrhythmias one may prefer to avoid right heart catheterization.
Gawne et al. [18] introduced a dual-frequency method to
estimate parallel conductance, however a recent study
indicates that this method does not provide a reliable
substitute for the saline dilution method [19].
So far all validation studies, i.e. those studies comparing
saline dilution calibrated conductance signals with some
independent absolute volume measurement, have used
pulmonary artery injections [1,16,20–22]. Generally these
studies show that the relation between left ventricular
volume derived by the conductance catheter and the
independent method has an offset close to zero. This
indicates that the saline dilution method using pulmonary
artery injection accurately estimates parallel conductance.
In addition, theoretically pulmonary artery injection is
preferred, but in some cases one may need to use intravenous injections for practical reasons.
In this study parallel conductances were always determined as the mean of three repeat injections. The
corresponding standard deviations are a good measure of
the repeat variability of the assessments which was about
6% for both methods. Previous studies have shown that
5–10% repeat variabilities are inherent to indicator dilution
methods (such as dye- or thermodilution) even if respiratory influences and indicator injections are carefully controlled [23–27]. Given this variability, assessment of
parallel conductance on the basis of a single injection
generally appears to be inadequate and repeat assessments
in each hemodynamic condition are recommended.
5. Conclusions
Our results indicate that there is an excellent linear
correlation between parallel conductances obtained by the
pulmonary artery and intravenous injection methods and
the relation between the two methods is not significantly
affected by substantial hemodynamic changes induced by
embolization of the right coronary artery, dobutamine
infusion and pacing. Linear regression analysis and Bland–
Altman plots show that intravenous injection systematically overestimates pulmonary artery injection by about 7%.
Presumably, the overestimation mainly reflects a slightly
increased conductivity of the blood in the right ventricle,
which is present with intravenous injection and not with
pulmonary artery injection. Embolization of the right
coronary artery did increase parallel conductance as one
would expect due to the, presumably, enlarged right
ventricle. We anticipated that the overestimation with
intravenous injection would be relatively more pronounced
after embolization due to the enlargement and a reduced
ejection fraction of the right ventricle. However, there was
no significant interactive effect between the embolization
and the injection method, indicating that the relation
between parallel conductances assessed by the two methods was not affected by embolization. Pacing and
dobutamine had only a marginal effect on parallel conductance and no interaction with the injection method was
present. Consequently, taking the pulmonary artery injection method as the de facto gold standard, parallel
conductance obtained by intravenous injection should be
corrected by a simple multiplicative factor of 0.93, and this
appears to be valid for a wide range of hemodynamic
conditions. Strictly speaking these results only apply to
closed-chest sheep and may not be directly extrapolated to
other species including man. Therefore further validation
of this method is warranted.
References
[1] Baan J, Van der Velde ET, De Bruin H, Smeenk G, Koops J, Van
Dijk AD et al. Continuous measurement of left ventricular volume
in animals and humans by conductance catheter. Circulation
1984;70:812–823.
[2] Georgakopoulos D, Christe M, Giewat M, Seidman C, Seidman J,
Kass D. The pathogenesis of familial hypertrophic cardiomyopathy:
early and evolving effects from an alpha-cardiac myosin heavy
chain missense mutation. Nature Med 1999;5(3):327–330.
[3] Szwarc RS, Ball HA. Simultaneous LV and RV volumes by
conductance catheter: effects of lung insufflation on parallel conductance. Am J Physiol 1998;275(2 Pt 2):H653–661.
P. Steendijk, J. Baan / Cardiovascular Research 46 (2000) 82 – 89
[4] Steendijk P, Van Der Velde ET, Baan J. Single and dual excitation of
the conductance–volume catheter analysed in a spheroidal mathematical model of the canine left ventricle. Eur Heart J
1992;13(Suppl E):28–34.
[5] Steendijk P, Van Der Velde ET, Baan J. Left ventricular stroke
volume by single and dual excitation of conductance catheter in
dogs. Am J Physiol 1993;264(Heart Circ Physiol 33):H2198–
H2207.
[6] Lankford ED, Kass DA, Maughan WL, Shoukas AA. Does volume
catheter parallel conductance vary during a cardiac cycle? Am J
Physiol 1990;258(Heart Circ Physiol 27):H1933–H1942.
[7] White PA, Chaturvedi RR, Shore D, Lincoln C, Szwarc RS, Bishop
AJ et al. Left ventricular parallel conductance during cardiac cycle
in children with congenital heart disease. Am J Physiol 1997;273(1
Pt 2):H295–302.
[8] Baan J, Van der Velde ET, Steendijk P, Koops J. Calibration and
application of the conductance catheter for ventricular volume
measurement. Automedica 1989;11:357–365.
[9] Jerzewski A, Steendijk P, Pattynama P, Leeuwenburgh B, de Roos
A, Baan J. Right ventricular systolic function and ventricular
interaction during acute embolisation of the left anterior descending
coronary artery in sheep. Cardiovasc Res 1999;43:86–95.
[10] Myhre ES, Johansen A, Bjornstad J, Piene H. The effect of
contractility and preload on matching between the canine left
ventricle and afterload. Circulation 1986;73(1):161–171.
[11] Slinker BK, Glantz SA. Multiple linear regression is a useful
alternative to traditional analyses of variance. Am J Physiol
1988;255:353–367.
[12] Glantz SA, Slinker BK. Primer of applied regression and analysis of
variance, New York: McGraw Hill, 1990.
[13] Bland JM, Altman DG. Statistical methods for assessing agreement
between two methods of clinical measurement. Lancet
1986;1(8476):307–310.
[14] Georgakopoulos D, Mitzner WA, Chen CH, Byrne BJ, Millar HD,
Hare JM et al. In vivo murine left ventricular pressure–volume
relations by miniaturized conductance micromanometry. Am J
Physiol 1998;274(4 Pt 2):H1416–1422.
[15] Pinsky MR, Perlini S, Solda PL, Pantaleo P, Calciati A, Bernardi L.
Dynamic right and left ventricular interactions in the rabbit:
simultaneous measurement of ventricular pressure–volume loops. J
Crit Care 1996;11(2):65–76.
89
[16] Amirhamzeh MM, Dean DA, Jia CX, Cabreriza SE, Yano OJ,
Burkhoff D et al. Validation of right and left ventricular conductance
and echocardiography for cardiac function studies. Ann Thoracic
Surg 1996;62(4):1104–1109.
[17] Brookes C, Ravn H, White P, Moeldrup U, Oldershaw P, Redington
A. Acute right ventricular dilatation in response to ischemia
significantly impairs left ventricular systolic performance. Circulation 1999;100(7):761–767.
[18] Gawne TJ, Gray KS, Goldstein RE. Estimating left ventricular offset
volume using dual-frequency conductance catheters. J Appl Physiol
1987;62:872–876.
[19] White PA, Brookes CI, Ravn HB, Stenbog EE, Christensen TD,
Chaturvedi RR et al. The effect of changing excitation frequency on
parallel conductance in different sized hearts. Cardiovasc Res
1998;38(3):668–675.
[20] Burkhoff D, Van Der Velde E, Kass D, Baan J, Maughan WL,
Sagawa K. Accuracy of volume measurement by conductance
catheter in isolated, ejecting canine hearts. Circulation 1985;72:440–
447.
[21] Tjon A Meeuw L, Hess OM, Nonogi H, Monrad ES, Leskosek B,
Krayenbuehl HP. Left ventricular volume determination in dogs: a
comparison between conductance technique and angiocardiography.
Eur Heart J 1988;9:1018–1026.
[22] Cassidy SC, Teitel DF. The conductance volume catheter technique
for measurement of left ventricular volume in young piglets. Pediatr
Res 1992;31:85–90.
[23] Jansen JR, Schreuder JJ, Settels JJ, Kloek JJ, Versprille A. An
adequate strategy for the thermodilution technique in patients during
mechanical ventilation. Intensive Care Med 1990;16(7):422–425.
[24] Nishikawa T, Dohi S. Errors in the measurement of cardiac output
by thermodilution. Can J Anaesthesia 1993;40(2):142–153.
[25] Jansen JR. The thermodilution method for the clinical assessment of
cardiac output. Intensive Care Med 1995;21(8):691–697.
[26] Imai T, Takahashi K, Fukura H, Morishita Y. Measurement of
cardiac output by pulse dye densitometry using indocyanine green: a
comparison with the thermodilution method. Anesthesiology
1997;87(4):816–822.
[27] Schmid ER, Schmidlin D, Tornic M, Seifert B. Continuous thermodilution cardiac output: clinical validation against a reference
technique of known accuracy. Intensive Care Med 1999;25(2):166–
172.