Transoesophageal echocardiography (TOE) has become an invaluable and widely used tool for monitoring and management of cardiovascular function in the perioperative period . In this application of TOE, the evaluation of left ventricular (LV) function and its determinants plays a central role. Reference values required for evaluation of LV function by TOE are available [2,3]. However, they were obtained in awake volunteers  or patients  and are limited to diameter measurements only. The conclusion of these two studies is that TOE values of LV function in healthy subjects do not differ from those obtained using transthoracic echocardiography. However, the non-sedated subjects of these studies possibly had an increased sympathetic tone, they were examined in the left decubitus position and were breathing spontaneously. In contrast, intraoperative TOE is usually performed in the supine position, under general anaesthesia and positive pressure ventilation of the lungs. All these factors may alter both the size and function of the LV and therefore the above conclusion may not be valid for intraoperative TOE. Moreover, the studies were performed on subjects that were younger than those who usually need perioperative TOE and the dependence of the TOE data on age and gender was not addressed. The goal of the present study was to evaluate the potential effect of these differences by defining reference values of LV dimensions and function that can be used in intraoperative TOE.
Forty-five patients who underwent non-cardiac surgery under general anaesthesia, had no clinically detectable heart disease, and who signed written informed consent were included in the study, which was approved by the hospital ethical committee. Detailed history, physical examination, routine haematology and biochemistry, 12-lead electrocardiograms and chest radiographs (in patients over 50 years) were evaluated to prevent inclusion of patients with the following conditions and findings: coronary artery disease, valvular or any other heart disease, hypertension (actual arterial pressure > 150/90 mmHg, history of hypertension or any previous treatment for hypertension), peripheral arterial or cerebrovascular disease, chronic obstructive lung disease, diabetes mellitus, obesity (> 25% of ideal body weight defined as body height in cm – 100), anaemia (haemoglobin < 12 g dL–1), absence of sinus rhythm, and intake of drugs that act on the cardiovascular system. The presence of any oesophageal or gastric pathology was considered a contraindication to TOE. The majority of the patients underwent lumbar discectomy or peripheral orthopaedic surgery. On the first postoperative day, every patient was visited by the anaesthesiologist, questioned about the possible sequelae of the TOE and informed about the TOE findings. All patients also received a written TOE report for their physician.
For premedication the patients were given midazolam 7.5 mg 1 h prior to induction of anaesthesia. They had been fasting for 8 h and not allowed to drink during the last 6 h before surgery. Anaesthesia was induced by thiopental (3–5 mg kg–1) and fentanyl (3–5 μg kg–1), intubation of the trachea was facilitated by atracurium (0.5 mg kg–1) and enflurane (0.4–1.0 Vol.%) was used for the maintenance of anaesthesia. During the TOE study, patients’ lungs were ventilated with 40% oxygen in air, tidal volume of 10–12 mL kg–1, peak inspiratory pressure 12–22 H2O and rate of 6–8 bpm. The goal of the ventilator setting was an end-tidal CO2 of 4.5–5.0 kPa and a peripheral oxygen saturation > 95%. Arterial pressure was measured by an automated oscillometric method at 3-min intervals; heart rate and rhythm were monitored continuously by a two-lead electrocardiogram (ECG). At the start of the TOE study, the patients had already been given 250 mL of Ringer’s lactated solution.
The TOE probe was inserted after endotracheal intubation and aspiration of the gastric contents. The TOE study began after stabilization of haemodynamic and respiratory variables. Biplane (HP 21363 A) or omniplane (HP21367A) 5 MHz TOE probes were used with the Sonos 2500 ultrasound system (Hewlett-Packard, Andover, MA, USA). The left ventricle was imaged in its transgastric short axis at the mid-papillary level as well as in the transgastric two-chamber view and at least 10 cardiac cycles at suspended ventilation were recorded on videotape for subsequent off-line analysis. The probe was then withdrawn to the lower or mid-oesophagus and the four- and two-chamber views of the heart obtained. Again, video recordings were made as described above.
The two-dimensional measurements were performed off-line according to the recommendations of the American Society of Echocardiography [4–6]. The nomenclature of the TOE cross-sectional views follows the recently published recommendations . The end-diastolic frame was defined as the frame synchronous with the R wave of the QRS complex of the ECG; the end-systolic frame represented the frame with the smallest LV dimensions. The caliper diameter measurements and the area tracings were determined by the black–white endocardial cavity interface . Each measurement represents the mean of three to five cardiac cycles recorded in expiration
The following measurements were performed:
1 Transgastric short-axis view: end-diastolic and end-systolic antero-posterior and septo-lateral diameters (EDDap, ESDap, EDDsl, ESDsl respectively), end-diastolic and end-systolic area with cross-sectional papillary muscle areas included (EDA, ESA, respectively), total (epicardial) end-systolic area (TESA), end-diastolic and end-systolic endocardial circumference (EDC, ESC, respectively), and end-diastolic and end-systolic inferior and anterior wall thickness (EDWTi, ESWTi, EDWTa, ESWTa, respectively).
2 Transgastric two-chamber view: end-diastolic and end-systolic diameters (EDD, ESD, respectively) immediately below the mitral valve (basal level) and at the mid-papillary muscle level.
3 Four-chamber view: internal end-diastolic and end-systolic length (EDL, ESL, respectively), EDA and ESA.
4 Mid-oesophageal two-chamber view: internal EDL and ESL, total EDL (TEDL), EDA and ESA.
The two-dimensional data (diameters, lengths and areas) were normalized for body surface area (BSA) and are presented as indices (I). In addition, the following functional variables were calculated: fractional area change (FAC %)=(EDA − ESA/EDA) × 100, stroke area index (SAI)=EDAI − ESAI, fractional diameter shortening (FS %)=(EDD − ESD/EDD) × 100, wall thickening (WT %)=(ESWT − EDWT/ESWT) × 100, velocity of circumferential fibre shortening (Vcfc)=[(EDC − ESC)/(EDC × LVET)] × √RR [8,9], end-systolic (meridional) wall stress (ESWS)=1.33 × SAP × ESA/TESA − ESA [10,11], stroke-work index (SWI)=MAP × (EDAI − ESAI) , LV mass index (LVMI)=1.055 × 0.833 [(TESA × TEDL) − (ESA × EDL)]/BSA , relative wall thickness (RWT)=2 × EDWT/EDD , arterial stiffness index (ASI)=SAP − DAP/SAI , end-systolic pressure/area ratio=SAP/ESA and end-systolic wall stress/area ratio=ESWS/ESA .
One experienced investigator performed all measurements. Selected measurements were repeated after 6–10 months using the video recordings of 23 subjects. The intraobserver variability (%) was calculated as the mean absolute difference between both readings divided by the mean of all measurements multiplied by 100. The intraobserver bias and the 95% limits of agreement were calculated as the mean difference between the first and second readings ± 2 standard deviations. Pearson’s correlation coefficients were determined for the paired measurements . In addition, in 11 randomly selected patients a second experienced observer measured LV diameters and areas in the transgastric short-axis view. The interobserver variability was calculated in the same way as described above for the intraobserver variability.
The GraphPad Instat program (GraphPad Software, San Diego, CA, USA) was used to calculate the means, standard deviations and 95% confidence intervals and to test the differences between both genders by means of a Student’s t-test. The strength of the correlation between age and the echocardiographic variables were assessed by Pearson’s correlation coefficient. The same program was used for the variability calculations.
The demographic and haemodynamic data of the 45 patients (26 men and 19 women) are presented in Table 1. Men had higher mean weight, height, and BSA than women, whereas there were no differences in body mass index, arterial pressure or heart rate.
The TOE examination was completed in all patients and, on average, lasted 14 min (range 12–17 min). Haemodynamics remained stable during the TOE and no patient required pharmacological support. The postoperative controls revealed no complications or complaints that could have been attributed to the TOE.
The two-dimensional views were obtained and measurements were performed in all patients. The results are presented in Tables 2–6.
Men had larger end-diastolic and end-systolic LV area indices in transgastric short-axis, four-chamber and two-chamber views. The differences in the LV size between men and women also persisted when body height or body mass index was used instead of BSA for normalization of the data (not shown). FS %, FAC % and WT % were similar in both groups except for the FAC in the four-chamber view. In this view, women showed higher FAC values that were associated with a smaller end-systolic length. Furthermore, compared with men, women had lower LV mass indices and higher ES pressure/area as well as ESWS/ESA ratios. EDA and ESA in the mid-oesophageal two-chamber view showed an inverse correlation with age (r =–0.39 and –0.47, P =0.009 and 0.001 respectively).
Intraobserver and interobserver variabilities were assessed for LV dimensions and function in the transgastric short-axis view, the most frequently performed intraoperative on-line measurements. Results are displayed in Tables 7a and b. The intraobserver variability varied between 2.5% and 8.1%; the interobserver variability between 1.7% and 4.5%.
The objective of our study was to define normal values of LV dimensions, and systolic function in anaesthetized and ventilated supine patients during standard anaesthesia including muscle relaxation and positive pressure pulmonary ventilation. General anaesthesia is associated with a lower sympathetic tone, lower oxygen consumption and the anaesthetic agents can directly affect the heart and peripheral vessels . These anaesthesia related effects, together with the increase in intrathoracic pressure, result in lower venous return and lower cardiac output. The capability of the cardiovascular system to counteract the effects of anaesthesia and controlled ventilation, for instance by activation of the baroreflex, is reduced in the anaesthetic state . Moreover, relaxation of the respiratory muscles, particularly the diaphragm, can affect the position and geometry of the heart. Therefore, the use of reference values obtained in awake subjects could be misleading. Thus, the data of our study can be used as reference values whenever LV structure and function is evaluated in anaesthetized and ventilated subjects. In accordance with the expected cardiovascular effects of general anaesthesia and controlled ventilation, the LV dimensions were smaller than the published normal values in awake subjects. Drexler and his colleagues  used a monoplane TOE probe to measure LV diameter in 25 healthy volunteers ranging in age from 19 to 30 years. Cohen and his colleagues  examined 60 monoplane studies that were reported as normal during a 2-year period from two clinical centres; however, in their study, LV diameter measurements were performed only in 22 patients. Both of these studies concluded that the LV dimensions in their awake subjects corresponded closely to the published normal values obtained from transthoracic echocardiography [8,17–19]. Compared with the results of these two TOE studies, the LV anteroposterior and septolateral diameters in our subjects were 12–13% smaller both at end-diastole and end-systole. Furthermore, our transgastric short-axis areas were up to 40% smaller than the published corresponding transthoracic parasternal short-axis values .
In our study, men had greater LV dimensions than women. After correction for body size, the difference in all areas and long-axis diameters persisted. Separate reference values for men and women are not usually used [8,9], but according to our data they should be taken into consideration to avoid incorrect diagnosis and inappropriate treatment of hypovolaemia [20,21]. In women, the finding of a small LV end-diastolic size accompanied with near end-systolic obliteration of the cavity was frequent and not associated with hypotension or any other clinical signs of hypovolaemia. In general, the smaller LV size can be explained by redistribution of the effective blood volume caused by anaesthesia and ventilation in the presence of maintained LV pump function. In women, an increased contractility, suggested by higher end-systolic pressure or stress/area indices, may also contribute to the smaller LV size.
The indices of global LV systolic function such as FS %, FAC % and Vcfc in our subjects were not different from the corresponding normal values obtained by transthoracic echocardiography [8,9]. While men can exhibit lower LV ejection fractions than women , in our study the difference in FAC % between genders was minimal and not significant.
The wall thickness measured for the inferior and anterior walls of the LV closely corresponded with published normal values [8,9]. We did not measure wall thickness of the septal and lateral walls because of an unsatisfactory lateral resolution of the recorded images. The calculated wall thickening was similar for the anterior and inferior walls and was also similar to published data [23,24]. Thus, both global and regional systolic LV functions were not adversely affected by our standard anaesthetic technique, possibly because the lower preload was counterbalanced by a decrease in afterload of the LV . Moreover, the anaesthetic agents in the doses used were unlikely to impair myocardial contractility .
The finding of an inverse correlation between LV area or length in the two-chamber view and age in our study agrees with the findings of previous transthoracic studies that demonstrated a decrease in both end-systolic and end-diastolic sizes of the LV with increasing age [14,27]. In contrast to others , we did not find an increase in LV wall thickness with increasing age, an observation that may be related to the careful exclusion of patients with increased arterial pressure from our study and the limited number of elderly patients.
There is little published data on the reproducibility of intraoperative TOE measurements of LV size and function. The intra- and interobserver variability of the measurements performed in anaesthetized and ventilated patients without cardiac or other systemic disease has not been previously studied. The variability of the two-dimensional measurements of 3.5–8.1% found in our study is similar to the variability of 3–6% reported for repeated measurements from the same video recordings of the transthoracic short-axis view [28–30] as well as the transgastric short-axis view . The variability of the measurements performed in end-systole appeared to be higher than those of the measurements at end-diastole. This has been reported previously for transthoracic echocardiography and results from the same absolute deviation causing a greater percentile difference in a smaller LV . In our anaesthetized and ventilated patients, the frequent near obliteration of the LV cavity at end-systole rendered the tracing more difficult and could also have contributed to the higher variability.
In the present study, the LV dimensions and flows were assessed under general anaesthesia, muscle paralysis and controlled ventilation of the lungs. From our data it was not possible to separate the effects of these three variables on the size and function of the LV. The results may be different in anaesthetized but spontaneously breathing patients or in ventilated but not paralyzed and only sedated patients. The results may also be different after surgical incision that increases sympathetic tone and could counterbalance the effects of anaesthesia. The TOE probe is, however, in most cases inserted and LV function evaluated immediately after induction of anaesthesia before the start of surgery and the findings serve as baseline for later changes.
Because anaesthetic agents affect overall cardiovascular function, our reference values strictly speaking are valid only for the conditions of the balanced anaesthesia used in our study. However, such an anaesthetic technique is that which is most widely used in both cardiac and non-cardiac surgery. A similar limitation exists with regard to fluid management.
Cardiac dimensions are related to body dimensions and therefore the TOE values must be normalized to allow a comparison between different subjects and a definition of normal limits. We divided the dimensions by BSA because this is still the predominant method of normalization both in clinical practice and the literature. This method assumes a linear relationship between cardiac and body dimensions that is, however, not present at extremes of body weight. The use of body height  or an allometric scaling of our data  may have been more appropriate, particularly if obese subjects had been included in the study.
Often there may be no time to normalize data during the operative period. In such situations, Table 8 which presents normal uncorrected LV dimensions and function indices for men and women of average height and weight, may be helpful as a quick reference.
We have limited our assessment of observer variability to the two-dimensional measurements in the transgastric short-axis view that is commonly used for intraoperative monitoring of LV function and on which most of the therapeutic decisions are based. The reproducibility of measurements made in other views may, of course, be different.
Our hypothesis is that the published normal values of LV dimensions and function obtained by TOE in awake and spontaneously breathing subjects may not be valid for anaesthetized and ventilated patients. Indirect comparison of our data with previously published normal values led us to conclude that anaesthetized and ventilated subjects have smaller LV dimensions. On the other hand, global and regional LV systolic functions are well preserved. The presented values of LV size and function should be applied whenever LV function and its determinants are evaluated in patients under general anaesthesia and controlled ventilation.
ASI arterial stiffness index BSA body surface area DAP diastolic arterial pressure ECG electrocardiogram EDA end-diastolic area EDAI end-diastolic area index EDC end-diastolic circumference EDD end-diastolic diameter EDDap end-diastolic antero-posterior diameter EDDsl end-diastolic septolateral diameter EDL end-diastolic length EDWT end-diastolic wall thickness EDWTa end-diastolic anterior wall thickness EDWTi end-diastolic inferior wall thickness ESA end-systolic area ESAI end-systolic area index ESC end-systolic circumference ESD end-systolic diameter ESDap end-systolic antero-posterior diameter ESDsl end-systolic septolateral diameter ESL end-systolic length ESWS end-systolic wall stress ESWTa end-systolic anterior wall thickness ESWTi end-systolic inferior wall thickness FAC percentage fractional area change FS percentage fractional diameter shortening I indices (dimension/BSA) LV left ventricular LVET left ventricular end tidal LVMI left ventricular mass index MAP mean arterial pressure RWT relative wall thickness SAI stroke area index SAP systolic arterial pressure SWI stroke work index TEDL total end-diastolic length TESA total epicardial end-systolic area TOE transoesophageal echocardiography WT percentage wall thickening
1 Practice Guidelines for Perioperative Transesophageal Echocardiography. Anesthesiology
1996; 84: 986–1006.
2 Drexler M, Erbel R, Mueller U, Wittlich N, Mohr-Kahaly S, Meyer J. Measurement of intracardiac dimensions and structures in normal young adult subjects by transesophageal echocardiography. Am J Cardiol
1990; 65: 1491–1496.
3 Cohen GI, White M, Sochowski RA et al. Reference values
for normal adult transesophageal echocardiographic measurements. J Am Soc Echocardiogr
1995; 8: 221–230.
4 American Society of Echocardiography Committee on Standards Subcommittee on Quantitation of Two- Dimensional Echocardiograms. Recommendations for quantitation of the left ventricle by two-dimensional echocardiography. J Am Soc Echocardiogr
1989; 2: 358–367.
5 Schiller NB. Two-dimensional echocardiographic determination of left ventricular volume, systolic function, and mass. Circulation
1991; 84 (Suppl. I): I-280–287.
6 Devereux RB, Roman MJ. Evaluation of cardiac and vascular structure and function by echocardiography and other noninvasive techniques. In: Laragh JH, Brenner BM, eds. Hypertension: Pathophysiology, Diagnosis, and Management
, 2nd Edn. New York: Raven Press, 1995: 1169–1177.
7 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 and the Society of Cardiovascular Anesthesiologists Task Force for Certification in Perioperative Transesophageal Echocardiography. Anesth Analg
1999; 89: 870–884.
8 Feigenbaum H. Echocardiography
, 4th Edn. Philadelphia: Lea & Febiger, 1986.
9 Weyman AE. Principles and Practice of Echocardiography
, 2nd Edn. Philadelphia: Lea & Febiger, 1994: 1289–1298.
10 Douglas PS, Reichek N, Plappert T, Muhammad A, St John Sutton MG. Comparison of echocardiographic methods for assessment of left ventricular shortening and wall stress. J Am Coll Cardiol
1987; 9: 945–951.
11 Greim C, Roewer N, Meissner C, Bause H, Schulte am Esch J. Abschätzung akuter linksventrikulärer Nachlaständerungen. Anaesthesist
1995; 44: 108–115.
12 Atkins BZ, Silvestry SC, Davis JW, Kisslo JA, Glower Jr DD. Means for load variation during echocardiographic assessment of the Frank-Starling relationship. J Am Soc Echocardiogr
1999; 12: 792–800.
13 Slotwiner DJ, Devereux RB, Schwartz JE et al.
Relation of age to left ventricular function in clinically normal adults. Am J Cardiol
1998; 82: 621–626.
14 Carabello BA. Ratio of end-systolic stress to end-diastolic volume: is it a useful clinical tool? J Am Coll Cardiol
1989; 14: 496–498.
15 Altman DG. Practical Statistics for Medical Research
. London: Chapman & Hall, 1991: 396–403.
16 Hoeft A, Buhre W. Anaesthesia and the cardiovascular system. In: Priebe H-J, Skarvan K, eds. Cardiovascular Physiology
. London: BMJ Publishing Group, 1995: 272–306.
17 Schnittger I, Gordon EP, Fitzgerald PJ, Popp RL. Standardized intracardiac measurements of two-dimensional echocardiography. J Am Coll Cardiol
1983; 5: 934–938.
18 Erbel R, Henkel B, Ostländer C, Clas W, Brennecke R, Meyer J. Normalwerte für zweidimensionale Echokardiographie. Dtsch Med Wochenschr
1985; 110: 123–128.
19 Triulzi M, Gillan LD, Gentile F, Nevell JB, Weyman BS, Weyman AE. Normal adult cross-sectional echocardiographic values: linear dimensions and chamber areas. Echocardiography
1984; 1: 403–426.
20 Leung JM, Levine EH. Left ventricular end-systolic cavity obliteration as an estimate of intraoperative hypovolemia. Anesthesiology
1994; 81: 1102–1109.
21 Fontes ML, Bellows W, Ngo L, Mangano DT, McSPI Research Group. Assessment of ventricular function in critically ill patients: Limitations of pulmonary artery catheterisation. J Cardiothorac Vasc Anesth
1999; 13: 521–527.
22 Wong ND, Gardin JM, Kurosaki T et al.
Echocardiographic left ventricular systolic function and volumes in young adults: Distribution and factors influencing variability. Am Heart J
1995; 129: 571–577.
23 Haendchen RV, Wyatt HL, Maurer J et al.
Quantitation of regional cardiac function by two-dimensional echocardiography. I. Patterns of contraction in the normal left ventricle. Circulation
1983; 67: 1234–1245.
24 Konstadt SN, Abrahams HP, Nejat M, Reich DL. Are wall-thickening measurements reproducible? Anesth Analg
1994; 78: 619–623.
25 Dahlgren G, Settergren G, Ribeiro A, Brodin L-A. Changes in left ventricular diameter during intravenous induction of anesthesia. J Cardiothorac Vasc Anesth
1993; 7: 399–401.
26 Rathod R, Jacobs HK, Kramer NE, Rao LK, Salem MR, Towne WD. Echocardiographic assessment of ventricular performance following induction with two anesthetics. Anesthesiology
1978; 49: 86–90.
27 Wandt B, Bojö L, Hatle L, Wranne B. Left ventricular contraction pattern changes with age in normal adults. J Am Soc Echocardiogr
1998; 11: 857–863.
28 Himelman RB, Cassidy MM, Landzberg JS, Schiller NB. Reproducibility of quantitative two-dimensional echocardiography. Am Heart J
1988; 115: 425–431.
29 Kücherer HF, Kee LL, Modin G, Cheitlin MD, Schiller NB. Echocardiography in serial evaluation of left ventricular systolic and diastolic function: importance of image acquisition, quantitation, and physiological variability in clinical and investigational applications. J Am Soc Echocardiogr
1991; 4: 203–214.
30 Otterstad JE, Froeland G, St John Sutton M, Holme I. Accuracy and reproducibility of biplane two-dimensional echocardiographic measurements of left ventricular dimensions and function. Eur Heart J
1997; 18: 507–513.
31 Cahalan MK, Ionescu P, Melton HE, Adler S, Kee LL, Schiller NB. Automated real-time analysis of intraoperative transesophageal echocardiograms. Anesthesiology
1993; 78: 477–485.
32 Batterham AM, George KP, Mullineaux DR. Allometric scaling of left ventricular mass by body dimensions in males and females. Med Sci Sports Exerc
1997; 29: 181–186.