Swaminathan, Madhav MD; Phillips-Bute, Barbara G. PhD; Mathew, Joseph P. MD
Department of Anesthesiology, Duke University Medical Center, Durham, North Carolina
Accepted for publication April 7, 2003.
Address correspondence and reprint requests to Madhav Swaminathan, MD, Department of Anesthesiology, Box 3094, Duke University Medical Center, Durham, NC 27710. Address e-mail to firstname.lastname@example.org.
Supported by the Cardiothoracic Division of the Department of Anesthesiology, Duke University Medical Center, Durham, NC.
Presented in part at the 76th Clinical and Scientific Congress of the International Anesthesia Research Society, San Diego, CA, March, 2002.
Echocardiographic indices of myocardial performance are useful in the diagnosis and management of a wide variety of cardiovascular diseases and serve as valuable predictors of outcome (1–5). However, the assessment of left ventricular (LV) systolic function by transesophageal echocardiography (TEE) with use of ejection fraction and other volume-based estimates of ventricular function may be difficult because of the dependency of LV filling on loading conditions and cavity foreshortening in the longitudinal view (6). Load-independent indices of LV systolic function, such as rate-corrected mean velocity of fiber shortening and myocardial performance index, involve the measurement of systolic time intervals, including the LV ejection time (LVET) (3,6)
M-mode echocardiography replaced traditional methods of assessing LVET, such as measurement of the carotid pulse contour or phonocardiograms, in the mid 1970s (7,8). However, the introduction of phased-array transducers enabled the use of Doppler and two-dimensional imaging modalities in clinical situations. Doppler echocardiography is now routinely used to estimate time intervals and is an essential component of the myocardial performance index [(isovolumetric relaxation time + isovolumetric contraction time)/ejection time], a powerful echocardiographic predictor of cardiovascular morbidity and mortality (3,4,9,10). TEE is a useful tool for the intraoperative assessment of cardiac performance. Transvalvular Doppler velocities of the aortic valve (AV) are usually obtained during surgery by positioning the TEE probe in the deep transgastric position to obtain a long-axis image of the AV and ascending aorta (deep TG LAX view) (11). However, this view involves significant probe manipulation and may be technically challenging, with inability to obtain accurate evaluation in up to 12% of patients (12). In such situations, M-mode echocardiography of the AV may be used to determine ejection-time intervals. However, M-mode echocardiography has not been compared with Doppler-derived estimation of LVET by TEE. Therefore, we tested the hypothesis that M-mode measurement of the duration of AV opening is comparable to Doppler-derived measurement of LVET by TEE.
After IRB approval was obtained, data were gathered from 31 consecutive adult patients undergoing cardiac surgery at Duke University Medical Center from July 2001 to March 2002. Patients with a preoperative or intraoperative diagnosis of AV disease (incompetence, stenosis, calcification, or prior AV surgery) were excluded from the study. Therefore, all patients included in this study had normal native AV function. Other exclusion criteria were electrical pacing or nonsinus cardiac rhythm. Demographic, anthropometric, and echocardiography data were gathered for all patients.
TEE was performed with a multiplane, phased-array TEE probe (T6210 Omniplane II transducer; Phillips Medical Systems, Andover, MA). Images were digitally acquired on a Phillips Sonos 5500 Ultrasound Imaging System (Phillips Medical Systems). All images were acquired by anesthesiologists not directly involved in the anesthetic management of the patient (MS and JPM). The electrocardiogram (ECG) was recorded simultaneously in each case. In addition to a comprehensive TEE examination according to prescribed guidelines (11), the following images were obtained specifically for study purposes:
1. Continuous-wave Doppler recording of transvalvular velocities across the AV in either the deep TG LAX at 0° or the two-chamber TG LAX at 120°. The view with the Doppler interrogation angle closest to 0° was selected (Fig. 1).
2. M-mode recording across the AV leaflets in the midesophageal long-axis view at 120° (Fig. 2).
TEE images were obtained only when hemodynamics (heart rate (HR) and systemic and pulmonary arterial pressures) did not change by >10% during the recording process. All recordings were made at end-expiration and at a recording speed of 50 cm/s and included at least 3 consecutive wave forms in a freeze-frame. Images were digitally stored for off-line analysis.
The ejection time by Doppler-derived recording was defined as the time from onset to termination of transaortic flow velocity. The ejection time by M-mode was defined as the time from opening to closure of AV leaflets (13). In all cases, the mean of three consecutive wave forms was measured and adjusted for mean HR (mHR; represented by the mean R-R interval on a simultaneously recorded ECG) by the following formula: MATH
LVET was obtained by M-mode or Doppler.
All images were interpreted by anesthesiologists certified in perioperative TEE (MS and JPM). Doppler-derived measurements were made by one anesthesiologist (MS) who was blinded to the M-mode measurements made by the other (JPM). Both sets of measurements were independently analyzed by a statistician not involved in anesthetic management or data collection (BGP-B).
Simple descriptive statistics (mean and standard deviation (sd)) were used to describe the study population demographics. Both sets of observations were analyzed with the paired Student’s t-test, and significance was assessed at a two-tailed P value of <0.05. The mHR-corrected M-mode-derived and Doppler-derived ejection times were normally distributed, and their joint distribution was bivariate normal. Therefore, Pearson’s correlation coefficient was calculated for evaluating the strength of the linear relationship between these two variables. A P value of <0.05 was considered to signify whether the correlation coefficient was statistically significant. To assess the degree of agreement between the two measurements, a bias analysis was performed (14), with calculation of the mean difference and sd of the differences. The range of ±2 sd of the differences estimates the 95% level of agreement between the two techniques. To further strengthen the comparison of these two techniques, the Passing-Bablok regression method was used to assess the significance of agreement. This test does not require the normal distribution of variables on both axes. The significance of the linear agreement was analyzed with the CUSUM (cumulative summation) test for linearity; P < 0.05 indicated a significant consistent difference between the two techniques.
Although the limits of agreement between different echocardiographic techniques that are clinically acceptable have not been previously defined in the setting of LVET measurement, they have been assessed in the setting of cardiac output measurement. A meta-analysis of methods used to compare cardiac output estimation found that the overall limits of agreement in studies evaluating Doppler estimates was ±65% (22%–225%) (15). According to Critchley and Critchley (15), using current reference methods, acceptance of a new technique should rely on limits of agreement of up to ±30%. We believe that this recommendation may be extrapolated to the comparison of M-mode-derived and Doppler-derived estimates of ejection time measurement. All statistical analyses were performed with Analyze-it statistical software (Version 1.63; Analyze-it Software Inc., Leeds, UK).
Demographic data from all 31 patients are shown in Table 1. Most study patients underwent coronary artery bypass graft surgery, but some underwent heart transplantation, mitral valve replacement, and transmyocardial laser revascularization. Adequate TG TEE images for Doppler interrogation could not be obtained in one patient, who was excluded from subsequent analysis. Complete echocardiographic data were therefore available in 30 patients. Descriptive statistics for ejection time measurements are shown in Table 2. There was no significant difference between mean M-mode-derived (362.03 ± 41.6 ms) and mean Doppler-derived (360.06 ± 37.0 ms) LVET measurements (P = 0.61; mean difference between measurements, 1.98 ms; 95% confidence limits of between-measurement differences, −5.97 to 9.92 ms). Pearson correlation analysis (Fig. 3) revealed a correlation coefficient of 0.86, indicating a significant correlation between M-mode-derived and Doppler-derived measurements (P < 0.0001).
The Bland-Altman analysis of bias (Fig. 4, Table 3) revealed that there was no significant bias between M-mode-derived and Doppler-derived values (observed bias, 1.98 ms; 95% confidence intervals, −5.97 to 9.92 ms). The Passing-Bablok analysis of agreement (Fig. 5) showed that the observations were close to the “perfect fit” slope of 1. The CUSUM test of linearity showed that the observed slope was not significantly different from the perfect slope (1.05 versus 1.0; P > 0.1).
We confirmed our hypothesis that M-mode measurement of the duration of AV opening is comparable to Doppler-derived measurement of LVET by TEE. There was a significant correlation between values measured by the two techniques. The Bland-Altman plot also confirmed close agreement between the two measurements. M-mode echocardiography is an acceptable alternative to Doppler imaging for measurement of LVET.
The intraoperative assessment of ventricular function is essential in the management of patients undergoing cardiac surgery. TEE is an important tool that provides comprehensive information on myocardial performance in real time. Several estimates of ventricular function are based on ejection fraction, ventricular volumes, or diastolic function analyses. Most of these estimates are dependent on loading conditions of the heart (1,16). In the setting of cardiac surgery, the rapid and large volume shifts that accompany cardiopulmonary bypass or vasoactive drug infusions may make some load-dependent echocardiographic estimations inaccurate (17). Changes in preload or afterload may change the ejection fraction without affecting myocardial contractility (6). Traditional load-independent indices of systolic ventricular function were based on the measurement of velocity of fiber shortening and preejection period/ejection time ratios (6,16). Among newer load-independent indices of ventricular performance, the myocardial performance index, a Doppler-derived estimate of global LV function incorporating systolic and diastolic time intervals, has been described and reported to be a powerful predictor of morbidity and mortality in several cardiovascular diseases (3,9,10).
An essential component of both traditional load-independent systolic function indices and the myocardial performance index is the estimation of LVET (3). Since its original description, the ejection time has been measured by Doppler velocity gradients across the AV or LV outflow tract. These Doppler images are usually acquired during TEE by maneuvering the probe in the deep TG position (deep TG LAX view) to obtain a view similar to the transthoracic apical view (11). However, this view is not technically simple to acquire consistently. In up to 12% of patients, the images from this view may preclude accurate interpretation (12). In such situations, the M-mode may be a useful imaging modality for time-interval measurements.
One of the advantages of M-mode imaging is its superior temporal resolution compared with two-dimensional echocardiography (18). This factor should make it a logical choice for time-interval measurements. Since Stefadouros and Witham (7) reported that M-mode echocardiography was an acceptable alternative to phonocardiograms and carotid pulse contour analysis, a number of studies have reported the usefulness of this imaging modality for the estimation of systolic time intervals as components of load-independent indices of LV function (19–24). In a transthoracic echocardiographic study of right ventricular systolic time intervals in 31 patients, Hsieh et al. (25) reported that Doppler-derived measurements of transpulmonary Doppler velocities were comparable to simultaneous M-mode measurements of pulmonic valve opening. However, there has been no other study comparing M-mode and Doppler echocardiographic values for LV systolic time intervals with TEE.
There are limitations with regard to our study. HR changes over time may affect the measurement of ejection time at separate time points. There are two methods of overcoming this limitation. First, both M-mode and Doppler measurements may be performed simultaneously. However, with current technology, it is not possible to simultaneously obtain two different images in separate locations with one TEE probe. Second, a HR correction may be applied to both sets of measurement to eliminate the effect of a change in HR during the study. We corrected the ejection time values in our study for HR, so we are confident that the analysis accounted for any effect of altered HR during the examination. In addition, we included only patients who were in sinus rhythm; therefore, any confounding factors known to affect ejection time, such as atrial fibrillation, electrical pacing, and supraventricular dysrhythmias, were not present in our study. Changes in ventricular function that may have affected ejection time over the short examination period, although possible, are highly unlikely, because all patients were hemodynamically stable during the TEE examination. A limitation of the use of M-mode for measurements related to AV timing occurs with aortic incompetence or early closure (as in hypertrophic cardiomyopathy), where it may not be possible to define the exact point of AV closure. This may preclude accurate interpretation of LVET. We excluded patients with AV disease to prevent such inaccuracies. In such patients, Doppler transvalvular velocity gradients may be the only method to obtain accurate ejection time measurement during TEE. Despite these limitations, we believe that our results permit confident conclusions regarding the feasibility of using M-mode echocardiography as an alternative to Doppler echocardiography for ejection-time measurement.
In summary, we found close agreement between M-mode and Doppler echocardiography for the measurement of LVET by TEE. M-mode echocardiography is an acceptable method of determining LVET in situations when transvalvular Doppler velocities cannot be obtained satisfactorily. Although a “gold standard” for ejection time measurement has not been described, the close correlation between M-mode-derived and Doppler-derived timing measurements indicates that either technique may be used with reasonable confidence for obtaining measures of ventricular performance. M-mode echocardiography is an underused imaging modality in intraoperative TEE today. This simple technique may provide useful information, especially when used in conjunction with Doppler imaging.
1. Otto C. Textbook of clinical echocardiography. 2nd ed. Philadelphia, PA: WB Saunders, 2000.
2. Poulsen SH, Jensen SE, Tei C, et al. Value of the Doppler index of myocardial performance in the early phase of acute myocardial infarction. J Am Soc Echocardiogr 2000; 13: 723–30.
3. Tei C, Ling LH, Hodge DO, et al. New index of combined systolic and diastolic myocardial performance: a simple and reproducible measure of cardiac function—a study in normals and dilated cardiomyopathy. J Cardiol 1995; 26: 357–66.
4. Tei C, Dujardin KS, Hodge DO, et al. Doppler index combining systolic and diastolic myocardial performance: clinical value in cardiac amyloidosis. J Am Coll Cardiol 1996; 28: 658–64.
5. Gan TJ, Soppitt A, Maroof M, et al. Goal-directed intraoperative fluid administration reduces length of hospital stay after major surgery. Anesthesiology 2002; 97: 820–6.
6. Skiles JA, Griffin BP. Transesophageal echocardiographic (TEE) evaluation of ventricular function. Cardiol Clin 2000; 18: 681–97.
7. Stefadouros MA, Witham AC. Systolic time intervals by echocardiography. Circulation 1975; 51: 114–7.
8. Hirschfeld S, Meyer R, Schwartz DC, et al. Measurement of right and left ventricular systolic time intervals by echocardiography. Circulation 1975; 51: 304–9.
9. Tei C, Nishimura RA, Seward JB, Tajik AJ. Noninvasive Doppler-derived myocardial performance index: correlation with simultaneous measurements of cardiac catheterization measurements. J Am Soc Echocardiogr 1997; 10: 169–78.
10. Poulsen SH, Jensen SE, Nielsen JC, et al. Serial changes and prognostic implications of a Doppler-derived index of combined left ventricular systolic and diastolic myocardial performance in acute myocardial infarction. Am J Cardiol 2000; 85: 19–25.
11. Shanewise JS, Cheung AT, Aronson S, et al. ASE/SCA guidelines for performing a comprehensive intraoperative multiplane transesophageal echocardiography examination: recommendations of the American Society of Echocardiography Council for Intraoperative Echocardiography and the Society of Cardiovascular Anesthesiologists Task Force for Certification in Perioperative Transesophageal Echocardiography. Anesth Analg 1999; 89: 870–84.
12. Katz WE, Gasior TA, Quinlan JJ, Gorcsan J III. Transgastric continuous-wave Doppler to determine cardiac output. Am J Cardiol 1993; 71: 853–7.
13. Erbel R. Principles of global LV function analysis. In: Roelandt J, Sutherland G, Iliceto S, Linker D, eds. Cardiac ultrasound. New York: Churchill Livingston, 1993: 219–32.
14. Bland JM, Altman DG. Statistical methods for assessing agreement between two methods of clinical measurement. Lancet 1986; 1: 307–10.
15. Critchley LA, Critchley JA. A meta-analysis of studies using bias and precision statistics to compare cardiac output measurement techniques. J Clin Monit Comput 1999; 15: 85–91.
16. Kuecherer HF, Foster E. Hemodynamics by transesophageal echocardiography. Cardiol Clin 1993; 11: 475–87.
17. Garcia MJ, Thomas JD, Klein AL. New Doppler echocardiographic applications for the study of diastolic function. J Am Coll Cardiol 1998; 32: 865–75.
18. Edelman S. Understanding ultrasound physics. 2nd ed. College Station, TX: Tops Printing Inc, 1994.
19. Lemne C, Lindvall K, Georgiades A, et al. Structural cardiac changes in relation to 24-h ambulatory blood pressure levels in borderline hypertension. J Intern Med 1995; 238: 49–57.
20. Pennestri F, Biasucci LM, Rinelli G, et al. Abnormal intraventricular flow patterns in left ventricular dysfunction determined by color Doppler study. Am Heart J 1992; 124: 966–74.
21. Eriksson SV, Offstad J, Kjekshus J. M-mode echocardiography in patients with severe congestive heart failure: a subgroup analysis in the Cooperative North Scandinavian Enalapril Survival Study (CONSENSUS). Drugs 1990; 39: 43–8,discussion 53–4.
22. Roman MJ, Devereaux RB, Cody RJ. Ability of left ventricular stress-shortening relations, end-systolic stress/volume ratio and indirect indexes to detect severe contractile failure in ischemic or idiopathic dilated cardiomyopathy. Am J Cardiol 1989; 64: 1338–43.
23. Nitta M, Nakamura T, Hultgren HN, et al. Progression of aortic stenosis in adult men: detection by noninvasive methods. Chest 1987; 92: 40–3.
24. Danielsen R, Nordrehaug JE, Lien E, Vik-Mo H. Subclinical left ventricular abnormalities in young subjects with long-term type 1 diabetes mellitus detected by digitized M-mode echocardiography. Am J Cardiol 1987; 60: 143–6.
25. Hsieh KS, Sanders SP, Colan SD, et al. Right ventricular systolic time intervals: comparison of echocardiographic and Doppler-derived values. Am Heart J 1986; 112: 103–7.