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: p.steendijk@lumc.nl (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.