| JOURNAL HOME | CME HOME | THIS MONTH | PAST ISSUES | ETOC | COLLECTIONS |
| AUTHORS | REVIEWERS | EDITORIAL BOARD | FEEDBACK | RSS | HELP |
| ||||||||||||||
| ||||||||||||||
|
|
|
|
||||||||||||
|
||||||||||||||
Department of Anesthesiology, Yale University School of Medicine, New Haven, Connecticut; and VA Connecticut Healthcare System, West Haven, Connecticut
Address correspondence and reprint requests to Pamela E. Gray, MD, Department of Anesthesiology/186, VA Connecticut Healthcare System, 950 Campbell Ave., West Haven, CT 06516. Address e-mail to gray.pamela_e{at}west-haven.VA.gov
 |
Abstract |
|---|
 Top Abstract IntroductionMethodsResultsDiscussionReferences |
|---|
Implications: An intraoperative dobutamine infusion was used to increase aortic blood flow and demonstrate hemodynamic-induced changes in aortic valve area. These valve-area changes affect the accuracy of Doppler cardiac output determinations.
|
Introduction |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
For each CO determination, obtaining paired determinations of AVA and ascending aortic VTI incurs repeated probe manipulations. The probe must be withdrawn from the transgastric long axis (TG LAX) view, where ascending aortic flows are measured, to the midesophageal aortic valve short axis view, where the aortic valve orifice is planimetered. After manual tracings of the aortic valve orifice are completed, the probe is then advanced back to the gastric position for continued blood flow and wall motion monitoring. Because calculation of CO requires measurements of ascending aortic flow and AVA, each CO determination requires two probe manipulations and two measurements.
To simplify Doppler CO monitoring via TEE to a less labor-intensive and more clinically useful protocol, the initial AVA obtained in each patient has been used for all subsequent Doppler CO calculations (3). A potential shortcoming of this modified approach stems from the fact that AVA is not remeasured each time transvalvular flow or hemodynamics change (1). This method, therefore, does not account for alterations in AVA that may occur secondary to changes in transvalvular flow or intraaortic blood pressure. The extent to which altered hemodynamics and flow affect AVA orifice dimensions remains to be clarified.
Current understanding of the effects of hemodynamics on human AVA is mainly derived from patients with valvular aortic stenosis. In vivo planimetric studies of hemodynamic-induced changes in AVA have used dobutamine (4) or other inotropes (5) to compare AVA at baseline to AVA at the maximal drug dose, and have shown little change in AVA. In patients with aortic stenosis, additional studies using the continuity equation and a similar dobutamine protocol revealed no change in AVA (6), or mixed results (7). Extension of these results to nondiseased valves, however, would ignore the significant differences in mechanical properties between the diseased and nondiseased valve. Consequently, the degree of error introduced into CO calculations by using a single AVA measurement for all subsequent CO calculations remains unclear in the patient with a nondiseased valve. Our hypothesis is that hemodynamic-induced changes in the area of the nondiseased aortic valve represent a source of error in Doppler CO calculations and should be quantified.
Our purpose for this study was to create a controlled increase in transvalvular flow by use of a graded dobutamine infusion, measure concomitant AVA, and calculate the degree of error in CO estimations that use a fixed AVA. This would then clarify whether the degree of error is small enough to support using a fixed AVA as part of a simplified clinical protocol for Doppler CO monitoring.
|
Methods |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
Enrolled patients were examined by multiplane TEE (Sonos 1500 or 5500; Hewlett-Packard, Andover, MA) after induction of general anesthesia and tracheal intubation. Any patient with evidence of previously undiagnosed aortic valvular disease by TEE examination was eliminated from the study. Anesthetic management was at the discretion of the anesthesiology care team. After completion of a routine TEE examination, baseline hemodynamic and echocardiographic data were recorded. These data included blood pressure, heart rate, ascending aortic blood flow velocity (Vmax), and ascending aortic VTI.
Determination of Aortic Blood Flow Velocity
Continuous wave Doppler measurements of Vmax were obtained from the TG LAX, with the imaging plane at approximately 120 degrees. If this view was suboptimal, flow measurements were obtained from the deep TG LAX view. The continuous wave Doppler beam was focused at the level of the aortic valve, and its orientation adjusted to maximize amplitude and clarity of the spectral waveform and minimize the angle of incidence to blood flow to <20 degrees. All values of AVA, Vmax, and VTI were measured at suspended end expiration and represent the average from two cardiac flow cycles.
Measurement of Mean AVA
Echocardiographic imaging for AVA tracing used the midesophageal aortic valve short axis view, with the imaging plane at approximately 30 degrees. Adjustments in probe position and multiplane transducer angle were made to obtain an aortic valve view in which all three leaflet edges and insertion points were simultaneously visible and the aortic attachments of the leaflets were seen as fused pairs at the level of the annulus. The gain was adjusted to the lowest possible setting without loss of leaflet edge definition. The images were recorded on VHS videotape, and the valve area tracings were performed off-line. The model of an equilateral triangular aortic valve orifice, recommended by Darmon et al. (1,8) and Perrino et al. (3), was used to most closely approximate the time-averaged AVA during systole (mean systolic AVA). Planimetric determination of AVA used the software package of the Sonos 5500 or 1500, and the midsystolic frame where the orifice most closely represented an equilateral triangle. Manual tracings connected the three leaflet fusion points with straight lines to form a triangle.
Dobutamine Infusion Protocol
After baseline data were collected, a dobutamine infusion was administered. We used the protocol described by Tardif et al. (4) in their study examining flow dependence of the stenotic aortic valve. The initial infusion rate of dobutamine was 5 µg · kg-1 · min-1, and was increased by 5-µg · kg-1 · min-1 increments until either a 50% increase in Vmax or 20 µg · kg-1 · min-1 was reached. At the largest dobutamine dose, a second set of hemodynamic data, aortic flow measurements, and aortic valve echocardiographic images were collected. When data collection was complete, the dobutamine infusion was discontinued. Baseline and dobutamine valve area tracings were performed off-line by the same echocardiographer for each patient. When performing valve area determinations for the dobutamine set, this echocardiographer was blinded to the patient’s baseline AVA value.
Doppler CO (AVA · heart rate · VTI) during the dobutamine infusion was calculated two ways. First, COvariable AVA was calculated by using the AVA determined at maximal flow (AVAdobutamine). Second, COconstant AVA was calculated by using the baseline determination of AVA (AVAbaseline). The mean percentage error between the COvariable AVA and the COconstant AVA was calculated as the absolute difference between the two measurements divided by their mean.
Statistical analysis of data was determined by using the paired t-test and Wilcoxon’s signed rank test, with P values of <0.05 considered significant. Intraobserver variability of AVA determination was assessed as the mean percentage error, calculated by dividing the absolute difference between each pair of AVA determinations by their mean. Results were presented as mean ± SD. Multifactorial analysis was used to determine any independent predictors of the change in AVA. Bland-Altman bias analysis was performed to compare the matched pairs of COvariable AVA and COconstant AVA.
|
Results |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
The TG LAX view for Doppler determinations of aortic blood flow was obtained without difficulty in 16 patients (average imaging plane angle 119.4 degrees), and the deep TG LAX view was used in the remaining patient. The initial echocardiographic measurements showed an average AVAbaseline of 2.77 ± 0.22 cm2, an average Vmax of 119 ± 24.9 cm/s, and an average VTI of 23.5 ± 5.2 cm. The average multiplane angle used for AVA area measurement was 33 degrees. Baseline Doppler CO was 4.7 ± 1.7 L/min.
Hemodynamic and Doppler echocardiographic values at baseline and at maximal dobutamine infusion rate for all 17 patients are summarized in Table 1. The intraobserver variability in AVA determination was 2.0% ± 0.9%. Dobutamine infusion produced the following effects. Vmax increased 54.5% ± 19.6% (P < 0.001), with 13 of 17 patients exhibiting a >45% increase in ascending aortic flow. VTI increased 30.8% ± 25.6%. CO increased from 4.7 ± 1.7 L/min to 7.7 ± 2.0 L/min, and stroke volume increased from 65.0 ± 15.4 mL to 86.0 ± 14.3 mL, (mean 36.4% ± 27%). Increases in systolic blood pressure (SBP) (50.6% ± 34.2%), mean arterial pressure (39.5% ± 34.1%), and heart rate (18.6 ± 17.4 bpm) were also observed.
|
The alteration in AVA observed in each patient after dobutamine administration is shown in Figure 1. When COconstant AVA was compared with COvariable AVA, the average difference was 0.32 ± 0.24 L/min (range 0.03–0.97 L/min), which represents a mean percentage error of 4.2% ± 2.4% (range 0.3%–9.3%) attributable to the modified CO protocol. Figure 2 shows a Bland-Altman plot of the matched COconstant AVA and COvariable AVA measurements.
|
|
|
Discussion |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
The data gathered in this study of the human aortic valve in vivo add to previous observations in animal and in vitro studies. Variation in the area of the nondiseased aortic valve with hemodynamic changes has been suggested by investigations in open chest dogs (9) and by studies using cadaveric human aortic valves. In vitroexamination by Montarello et al. (10) of five aortic valves mounted in a pulsatile pump model demonstrated a considerable degree of flow dependence in AVA. In contrast, Sprigings et al. (11) used a similar in vitro model to examine four nondiseased aortic valves and found AVA to be flow dependent, but to a lesser degree. Other studies’ results question the flow-dependent properties of the nondiseased aortic valve, especially over a physiologic range of flows (12–16).
Our results in humans support the model that changes in hemodynamics have only a modest effect on AVA, showing approximately the same degree of variance in AVA as found in vitro by Sprigings et al. (11), as opposed to the large increases reported by Montarello et al. (10).
Clinical Implications
The small but significant increase in AVA seen during dobutamine infusion could result in underestimation of Doppler CO. In our study, CO calculations that did not account for this increase in AVA produced an error of 4.2% ± 2.5%. When considered in the context of current CO methods and intrinsic measurement variability in Doppler CO determination, this represents a comparatively small source of error.
The investigation of Darmon et al. (1) using the triangular AVA method for Doppler CO calculations, showed a mean repeatability coefficient for Doppler CO of 0.51 L/min. Previous examination of the intraobserver variability associated with the measurements used to calculate Doppler CO cited a mean percentage error for repeated VTI and AVA determinations of 2% ± 1.5% and 1.4% ± 1.3%, respectively (3). In comparison, the most widely used method of CO determination, thermodilution, has a cited variability of 15% (17). Therefore, our data suggest that even in the context of the marked hemodynamic changes examined in this study, changes in AVA are associated with a clinically acceptable degree of error using the modified Doppler CO protocol.
Our results have useful clinical applications for intraoperative Doppler CO monitoring. CO monitoring using TEE measurements of ascending aortic flows is in good agreement with thermodilution methods by several investigators (1–3). The development of the rotatable imaging plane of multiplane TEE has made measurement of ascending aortic flows less labor intensive, because the aortic root is often easily visualized in the TG LAX view at approximately 120 degrees (2,3), decreasing the need for the more technically challenging deep TG LAX, or apical, view.
Our results, showing a modest degree of variation of AVA in nondiseased valves, support the simplified protocol for Doppler CO monitoring which uses a single baseline measurement of AVA in all subsequent CO determinations. By substituting the baseline AVA determination, subsequent CO monitoring is accomplished in the transgastric position without major probe manipulation. Because the transgastric position is also preferred for monitoring myocardial ischemia and function, with this approach, clinicians performing intraoperative echocardiography simply select the 120-degree imaging plane for ascending aortic flow measurements and CO determination, and the 0-degree imaging plane for the short axis view images used in wall motion and ventricular function monitoring.
Limitations of Study Model
This was a study of male veterans, many with coexisting cardiovascular disease. Although patients with coexisting cardiovascular disease represent a majority of those adults in whom intraoperative CO monitoring is used, results in other patient groups could yield different results. All patients were studied under general anesthesia, and although Doppler CO monitoring is most often used in anesthetized patients, results in a nonanesthetized population might differ. This study examined AVA over a range of normal-to-high CO and arterial pressures. The degree of variation in AVA may differ in the lower ranges of CO and arterial pressure.
Our study was not designed to determine the physical factors that lead to hemodynamic-induced changes in AVA. Alterations in flow, pressure, and tissue compliance are potential etiologic factors. Because these variables are interrelated in the intact human model, we are unable to define how each of these specific variables relates to the observed changes in AVA. Higashidate et al. (9) addressed this issue using their open chest dog model. Their investigation demonstrated that AVA increased when aortic blood flow and pressure concomitantly increased, but also showed increases in AVA when a ligature was placed around the ascending aorta, and only intraaortic pressure was increased. These canine data suggest that hemodynamic-induced increases in AVA are likely multifactorial.
Limitations of Planimetric AVA
Multiplane TEE provides excellent views of the aortic valve, allowing accurate planimetric determinations of AVA (18–20). Systole produces dynamic changes in aortic valve orifice geometry. At best, the examination of AVA for CO calculation would involve instantaneous matched pairs of valve areas and flow determinations throughout systole. Because this is not feasible in the clinical setting, time-averaged systolic AVA (mean AVA), estimated from a single systolic mea-surement, is used to calculate stroke volume. For the purposes of Doppler CO determination, mean AVA is best represented by a triangular rather than a circular orifice model. The triangular model represents the shape of the aortic valve orifice during the majority of systole (8), and best approximates time-averaged systolic AVA (1,8).
In conclusion, we have shown that the mean area of the nondiseased human aortic valve increased with dobutamine infusion and that these increases introduced a modest degree of error into Doppler CO calculations. These findings suggest that a modified protocol of intraoperative Doppler CO monitoring, which uses a single measure of planimetered AVA for all subsequent CO calculations, introduces an acceptable degree of error.
|
References |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
|
|
 |
|
  K. Norozi, C. Beck, W. A. Osthaus, I. Wille, A. Wessel, and H. Bertram Electrical velocimetry for measuring cardiac output in children with congenital heart disease Br. J. Anaesth., January 1, 2008; 100(1): 88 - 94. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. Pemberton, X. Li, T. Karamlou, C. A. Sandquist, K. Thiele, I. Shen, R. M. Ungerleider, A. Kenny, and D. J. Sahn The use of live three-dimensional Doppler echocardiography in the measurement of cardiac output: An in vivo animal study J. Am. Coll. Cardiol., February 1, 2005; 45(3): 433 - 438. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
