Insight into a patient’s hemodynamic status is very important to keep circulatory homeostasis as physiological as possible, especially during the perioperative period. A cardiac surgery setting is challenging for all parties involved – the anaesthesiologist, the surgeon and, of course, the patient.

The pulmonary artery catheter (PAC), popularly called the “Swan Ganz” catheter, in honour of its inceptors, has been the gold-standard of hemodynamic status insight since 1970. (1-3) However, novel hemodynamic monitors have been emerging over the last 20 or so years, which are less invasive than the PAC – even a PAC obituary was already written in 2013, which states PAC’s cousin, the continuous cardiac output PAC, as the sole surviving family member. (4)

One such monitor is the transoesophageal Doppler (in this case Deltex CardioQ™ ODM; Oesophageal Doppler Monitor, Deltex Medical, Chichester, UK). The premise is to place an ultrasonic probe into the oesophagus, at a depth that corresponds to the level of thoracic vertebrae 5 and 6 (Th5-Th6), that is 40-45cm from the nasal septum, or 35-40cm from the incisors (the probe can be placed trans-nasally or trans-orally). The tip of the probe is ideally angled at 45° towards the descending aorta (the desired position is achieved by gently rotating the probe while in the oesophagus to obtain the characteristic signal on the monitor), and blood velocity is obtained by means of continuous-wave Doppler signal emission and reception (velocity is calculated via a frequency shift equation), which is then presented as a continuous velocity/time graphical interpretation on a proprietary monitor screen. (5) Flow is calculated by using an equation for the descending aorta cross-sectional area (CSA), derived from demographic nomograms, and equals to a product of descending aorta blood velocity and descending aorta CSA. (6) Inherent transoesophageal Doppler hemodynamic variables, such as flow time (FT), peak velocity (PV), stroke distance (SD), flow time to peak (FTp) and mean acceleration (MA) are all derived from the Doppler probe measured descending aorta blood velocity using proprietary equations and algorithms. (7)

This study was done to compare hemodynamic measurements obtained via thermodilution (PAC) and a transoesophageal Doppler probe (TEDP) in patients who underwent elective cardiac bypass grafting surgery.


This is a prospective non-randomized observational clinical study that took place from February to April 2018 at the Department of Anaesthesiology and intensive care of cardiac surgical patients at the Clinical Hospital Centre Zagreb.

Data was collected from 17 patients, of whom 14 patients (13 male, aged 678 years and 1 female, aged 58 years) were selected for statistical analysis. Data from the other 3 patients was not collected according to study defined parameters (measurement timing mismatches, clerical errors due to study run-in period), and was not included in the statistical analysis.

All patients had signed an informed standard comprehensive hospital and informed anaesthesia and periprocedural monitoring consent form. Exclusion criteria were moderate and severe valvular disease, permanent atrial fibrillation, ejection fraction lower than 30%, aortic disease, upper gastrointestinal pathology and allergy to lidocaine.

The objective of the study was to compare two methods of hemodynamic monitoring, namely pulmonary artery catheter (PAC) thermodilution (make) and the transoesophageal Doppler probe (TEDP) (Deltex ODM) method, in patients who underwent elective “on-pump” cardiac bypass surgery.

Of course, other modalities of hemodynamic monitoring were standardly used, including invasive arterial pressure via intraarterial cannula (placed in left radial artery) and central venous pressure via central venous catheter (placed through right internal jugular vein).

Morphine was given intramuscularly as a premedication 30 minutes before proceeding to the operating theatre. Induction was performed using midazolam, sufentanil, etomidate and rocuronium. Subsequently, the central venous catheter (CVC) and PAC were placed through right internal jugular vein access, and TEDP was gently inserted into the oesophagus; water soluble lidocaine gel was used to facilitate optimal Doppler probe adherence to oesophageal mucosa. Maintenance of the anaesthesia was achieved with sevoflurane and a continuous infusion of sufentanil and rocuronium. During cardiopulmonary bypass, while the lungs were not ventilated, sevoflurane was administered via a vaporizer system (make) built into the cardiopulmonary bypass apparatus gas delivery. The depth of anaesthesia was continuously appraised using the patient state index monitor (SEDLine™, Masimo Corporation, Irvine, CA, USA), and the index was maintained between 30 and 50.

Following surgery, in the intensive care unit (ICU), patients were kept mechanically ventilated using synchronized intermittent mechanical ventilation (SIMV) and sedated with propofol infusion to ensure optimal conditions for hemodynamic measurements in the course of a two-hour period. After that, sedation was withdrawn and patients regained consciousness spontaneously. Postoperative analgesia was provided through a continuous sufentanil infusion for the first 24 hours.

The idea of the study was to simultaneously measure hemodynamic parameters via PAC and TEDP at four distinctive temporal points: after induction/prior to incision (time 1, T1), after chest closure/prior to leaving the operating theatre (time 2, T2), at the ICU arrival (time 3, T3) and one hour after admission to the ICU (time 4, T4).

Respective patient hemodynamics were treated by the gold standard – the PAC – to avoid clinical approach dichotomy, since this study mainly observed the feasibility of hemodynamic measurements using TEDP in cardiac bypass surgery patients, and its interchangeability with PAC, rather than outcome.

Hemodynamic variables measured in the study are listed in table 1.

Method Measured variables

Table 1. Measured hemodynamic variables. CO – cardiac output, CI – cardiac index, SV – stroke volume, SVI – stroke volume index, SD – stroke distance, MD – minute distance, PV – peak velocity, MA – mean acceleration, FT – flow time, FTc – corrected flow time, FTp – flow time to peak, PVV – pulse volume variation, SVV – stroke volume variation, SVR – systemic vascular resistance, PVR – pulmonary vascular resistance, LVSWI – left ventricular work index, RVSWI – right ventricular stroke work index, PAWP – pulmonary artery wedge pressure, HR – heart rate, IBP – invasive blood pressure, CVP – central venous pressure.

PAC measurements were obtained by 3 to 5 consecutive thermodilutions using 10cm3 of room temperature 0.9% NaCl, then averaged and calculated by Datex-Ohmeda (make) inherent monitor software, according to patient specific demographics.

TEDP measurements were done by inserting the TEDP probe after induction and then slightly adjusting it manually until the optimal signal and characteristic Doppler derived descending aorta blood flow velocity curve shape was shown on the proprietary monitor (Deltex CardioQ™ ODM monitor), and an auditory highly-pitched “whip” sound was heard through the monitor’s speaker for additional confirmation, as by the manufacturer instructions. TEDP remained in place until after the T4 measurement, then it was removed while the patients were still sedated.

Central venous pressure was obtained through an 18G lumen of a tri-luminal central venous catheter.

Measurements were done by two independent investigators, one for each method, without knowledge of the other investigator’s results. The TEDP method was always done first; PAC and CVC measurements were obtained immediately afterwards.

Inherent TEDP variables of cardiac output, stroke volume, preload, afterload and contractility were correlated to respective PAC variables that reflect the same, and to CVP, as a study variable not measured by TEDP nor by PAC, but included in thermodilution calculations.

Therefore, PAC derived cardiac output (CO) and stroke volume (SV) were correlated to their TEDP measured respective variables. Pulmonary artery wedge pressure (PAWP) and CVP were correlated to TEDP flow time variables – flow time (FT) and corrected flow time (FTc), in turn, as the indicators of preload. The contractility indicators respective correlations were done by measuring up left-ventricular work index (LVSWI) against TEDP obtained flow time to peak (FTp), peak velocity (PV), mean acceleration (MA) and stroke distance (SD) values. Afterload assessment was correlated between PAC derived systemic vascular resistance (SVR) and TEDP measured PV and FTc values.

Correlation calculations were made using the Microsoft Office Excel CORREL (Pearson correlation) function. Selected datasets with previously mentioned correlations of measure of cardiac output, cardiac index, stroke volume, preload, contractility and afterload between the two methods, plus CVP measured via CVC, were put through linear regression analysis. This was done for datasets at each respective time-frame (T1 through T4), as well as for all datasets combined. P values <0.05 were considered statistically significant.


The patients’ demographics, preoperative ejection fraction, cross-clamp time and number of coronary artery bypasses grafted are listed in table 2.

AGE (yrs) HEIGHT (cm) WEIGHT (kg) EF (%) XCLt (min) CABG (No.)
668 1767 8615 4913 5619 2,50,5

Table 2. Patients’ demographics data. EF – ejection fraction, XCLt – cross-clamp time, CABG – coronary artery bypass graft.

Out of 56 hemodynamic datasets recorded, four sets of 48 pairs of data were analysed for correlation in this paper. There were no medical complications nor technical difficulties experienced during the placement and removal of PAC and TEDP.

Tables 3 through 6, each presenting a specific hemodynamic time-frame snapshot (T1, T2, T3 and T4), show PAC and TEDP variables, as well as CVP, and their cross-referenced correlations. Linear regression analysis results of selected inherent TEDP and PAC variables of interest (cardiac output, stroke volume, preload, afterload and contractility indicators) are presented in table 7.

Table 3. Cross-referenced correlations at time T1. TEDP variables are topside horizontal; PAC and central venous catheter variables are noted on the left.

Table 4. Cross-referenced correlations at time T2. TEDP variables are topside horizontal; PAC and central venous catheter variables are noted on the left.

Table 5. Cross-referenced correlations at time T3. TEDP variables are topside horizontal; PAC and central venous catheter variables are noted on the left.

Table 6. Cross-referenced correlations at time T4. TEDP variables are topside horizontal; PAC and central venous catheter variables are noted on the left.

Table 7. Linear regression analysis. TEDP values were independent variables, while PAC values were dependent variables.

Table 8.
Cumulative measurements (time-frames T1 through T4 combined) linear regression analysis.

On a segmented time-frame dataset analysis (T1 through T4), linear regression showed little or no correlation between PAC and TEDP hemodynamic measurements. There was a slight correlation of SV measured by two methods (rT2= 0,53, p=0,053; rT3= 0,54, p=0,047; rT4= 0,53, p=0,052), and sporadically good correlation of TEDP measured PV and PAC derived LVSWI (rT2= 0,82, p=p<0,001; rT4= 0,56, p=0,036). Of all measured values, TEDP measured SD had somewhat consistent correlations to PAC derived LVSWI (rT2= 0,55, p=0,041; rT3= 0,66, p=0,010; rT4= 0,79, p=0,001).

Linear regression analysis of all the measurements combined into continuous arrays (table 8) showed poor to no correlations at all, however, some correlations were noted as follows: COTEDP-PAC (rT1-T4= 0,48, p<0,001), SVTEDP-PAC (rT1-T4= 0,47, p<0,001), PVTEDP to LVSWIPAC (rT1-T4= 0,45, p=0,001) and SDTEDP to LVSWIPAC (rT1-T4= 0,46, p<0,001).


The aim of this study was to observe for possible linear correlations between hemodynamic measurements obtained by using both PAC and TEDP methods in patients who underwent “on-pump” cardiac bypass surgery.

Looking at the data obtained at respective time-frames, there were no significant correlations between measured variables in T1 at all.

Of note are singular positive correlations at time-frame T2 of TEDP obtained PV and MA to PAC measured CO and SV – these are as follows: CO and PV (r=0,60), CO and MA (r=0,59), SV and PV (r=0,78), SV and MA (r=0,64).

Time-frames T2 through T4 show poor, although observable, correlations between PAC and TEDP measured values for SV (rT2= 0,53; rT3= 0,54; rT4= 0,53), LVSWI to PV (rT2= 0,82; rT4= 0,56,) and for LVSWI to SD (rT2= 0,55; rT3= 0,66; rT4= 0,79).

Taking all data into consideration in a continuous manner, the correlations are further lowered: COTEDP-PAC (rT1-T4= 0,48, p<0,001), SVTEDP-PAC (rT1-T4= 0,47, p<0,001), PVTEDP to LVSWIPAC (rT1-T4= 0,45, p=0,001) and SDTEDP to LVSWIPAC (rT1-T4= 0,46, p<0,001). Possible explanation for this may be an absence of noteworthy linear correlations at time-frame T1. By combining these measurements, the overall respective correlations were diminished.

On the other hand, it is difficult to explain the data obtained at time-frame T1. These measurements took place at a period right after induction and prior to incision – a period where hemodynamic instability is readily augmented by full expression of the effects of anaesthetics. During that period, patients are usually hemodynamically stabilized with the use of vasoactive drugs (noradrenaline, nitroglycerine), sometimes with additional administration of sufentanil and by changing the degree of sevoflurane exposure (guided by the patient state index monitor), not so much by augmenting intravascular volume, since cardiopulmonary bypass machine circuitry is primed with 1500mL of crystalloid/colloid solutions that would immediately lower haematocrit once cardiopulmonary bypass had started. PAC measurements were done right after TEDP measurements, which were obtained while maintaining a strong signal of the Doppler probe (confirmed by visual and auditory cues from the TEDP monitoring system). Speculatively, it is possible that pronounced peripheral vasoactivity (either spontaneous or from drugs given) in that period contributed to measured data dispersal during the minimal time lapse between measurements, since both methods require some time to produce the results (PAC – minimally 3 thermodilutions, TEDP – position verification, repositioning/reorienting the probe and data collection).

Respective correlations obtained at time-frames T2 through T4 remain weak and sporadic, although a pattern of consistency is observed for PAC derived LVSWI and TEDP measured SD (see above).

The literature on transoesophageal doppler and pulmonary artery catheter methods of obtaining hemodynamic measurements is still being accumulated, even though TEDP has been in use for almost two decades now, and the PAC method for almost half a century. A number of comparative studies have been made over the last 20 years. Data on TEDP versus PAC use and their interchangeability is still being scrutinized. Studies vary in size, observing usually ten to 40 participants, and they also vary in conclusions. (8-13)

Madan et al. found FTc a better indicator of preload to PAWP, stating the usefulness of TEDP at least as the one of PAC. (8)

Di Corte et al. found a correlation between FTc and left-ventricular end-diastolic area measured by transoesophageal echocardiography in short axis, suggesting FTc as a more reliable predictor of preload compared to PAC. (11)

Knirsch et al. have compared pediatric CardioQ™ monitoring to thermodilution and found that paediatric CardioQ™ does not reliably represent cardiac output values compared to thermodilution method. (12)

Su et al. found good correlation between TEDP measured CO and PAC measured continuous CO, however, when comparing the measurements of TEDP CO with PAC measured CO using repeated bolus technique, they observed a poorer correlation (value of 0.406) with both bias and standard deviation of bias being high. (13)

Sharma et al. compared transoesophageal Doppler monitor data to PAC hemodynamic measurements in off-pump cardiac surgery patients, concluding that it cannot be relied on as a sole method for monitoring cardiac output and derived hemodynamic variables in patients after coronary artery bypass surgery. (14)


The number of patients and measurements in this study are insufficient to generalize the results. However, it is clear there is poor correlation between observed hemodynamic variables. The interchangeability of TEDP and PAC methods for hemodynamic monitoring of cardiac surgery patients cannot be ascertained by data obtained in this study.

Further clinical studies are required to investigate the use of TEDP and its proprietary measurements value and use regarding PAC in the perioperative setting of cardiac surgery.


  1. Swan HJC, Ganz W, Forrester J et al. Catheterization of the heart in man with use of a flow directed balloon tipped catheter. N Engl J Med. 1970;283:447-451.
  2. Connors AFJ, McCaffree DR, Gray BA. Evaluation of right heart catheterization in the critically ill patient without acute myocardial infarction. N Engl J Med. 1983;308:263-67.
  3. Streisand JB, Clark NJ, Pace NL. Pulmonary arterial catheterization before anesthesia in patients undergoing cardiac surgery. Placement of the pulmonary arterial catheter before anesthesia for cardiac surgery: Safe, intelligent, and appropriate use of invasive hemodynamic monitoring. J Clin Monit. 1985;1:193-197.
  4. Marik PE. Obituary: pulmonary artery catheter 1970 to 2013. Ann Intensive Care. 2013;3:38.
  5. Singer M. Oesophageal Doppler. Curr Opin Crit Care. 2009;15:244-48.
  6. Payen D. Oesophageal Doppler monitoring: History, physical principles, and clinical applications. Int Proc J. 1994;1:3-9.
  7. Lowe GD, Chamberlain BM, Philpot EJ, Willshire RJ. Oesophageal Doppler Monitor (ODM) guided individualised goal directed fluid management (iGDFM) in surgery – a technical review. Chichester (UK): Deltex Medical; 2013. Issue No. 5.
  8. Madan AK, UyBarreta VV, Aliabadi-Wahle S, Jasperson R, Hartz RS, Flint LM, Steinberg SM. Esophageal Doppler ultrasound monitor versus pulmonary artery catheter in the hemodynamic management of critically ill surgical patients. J Trauma. 1999;46(4):607-11.
  9. Hullet B, Gibbs N, Weightman W, Thackray M, Newman M. A comparison of CardioQ and thermodilution cardiac output during off-pump coronary artery surgery. J Cardiothorac Vasc Anesth. 2003;17(6):728-32.
  10. Bein B, Worthmann F, Tonner PH et al. Compasiron of Esophageal Doppler, Pulse Contoru Analysis, and Real-Time Pulmonary Artery Thermodilution for the Continuous Meaurement of Cardiac Output. J Cardiothorac Vasc Anesth. 2004;18(2):185-189.
  11. DiCorte CJ, Latham P, Greilich PE, Cooley MV, Grayburn PA, Jessen ME. Esophageal Doppler Monitor Determinations of Cardiac Output and Preload During Cardiac Operations. Ann Thorac Surg. 2000;69:1782-6.
  12. Knirsch W, Kretschmar O, Tomaske M et al. Comparison of cardiac output measurement usinf the CardioQ™ oesophageal Doppler with cardiac output measurement using thermodilution technique in children during heart catheterization. Anaesthesia. 2008;63:851-855.
  13. Su NY, Huang CJ, Tsai P, Hsu YW, Hung YC, Cheng CR. Cardiac output measurement during cardiac surgery: esophageal Doppler versus pulmonary artery catheter. Acta Anaesthesiol Sin. 2002;40(3):127-33.
  14. Sharma J, Bhise M, Singh A, Mehta Y, Trehan N. Hemodynamic Measurements After Cardiac Surgery: Transesophageal Doppler Versus Pulmonary Artery Catheter. J Cardiothorac Vasc Anesth. 2005;19(6):746-750.
  15. Schober P, Loer SA, Schwarte LA. Haemodynamic monitoring of critically ill patients with transoesophageal Doppler technology. Neth J Crit Care. 2010;16(6):388-94.
  16. Morris C. Oesophageal Doppler monitoring, doubt and equipoise: evidence based medicine means change. Anaesthesia. 2013 Jul;68(7):684-8. doi: 10.1111/anae.12306. Epub 2013 May 15. PubMed PMID: 23672246.

Corresponding author:
Višnja Ivančan
Zagreb University Clinical Hospital Centre – KBC Zagreb, Clinic of Anaesthesiology, Reanimatology and Intensive Care, Department of Anaesthesiology and Intensive Care for Cardiac Surgery Patients
Kišpatićeva 12
10 000 Zagreb

Creative Commons LicenseThis work is licensed under a Creative Commons Attribution 4.0 International License