Article

PDF
Access to the PDF text
Advertising


Free Article !

Archives of cardiovascular diseases
Volume 110, n° 1
pages 7-13 (janvier 2017)
Doi : 10.1016/j.acvd.2016.04.009
Received : 31 January 2016 ;  accepted : 29 April 2016
Three-dimensional transoesophageal echocardiography for cardiac output in critically ill patients: A pilot study of ultrasound versus the thermodilution method
Intérêt de l’échocardiographie trans-œsophagienne tridimensionnelle pour évaluer le débit cardiaque chez les patients admis en réanimation : étude pilote
 

Nadjib Hammoudi a, , Guillaume Hékimian b, Florent Laveau a, Marc Achkar a, Richard Isnard a, Alain Combes b
a Université Paris 6, Département de Cardiologie, Institut de Cardiologie (AP–HP), CHU Pitié-Salpêtrière, Institute of Cardiometabolism and Nutrition (ICAN), Inserm UMRS 1166, ACTION Study Group, 75013 Paris, France 
b Université Paris 6, Département de Réanimation Médicale, Institut de Cardiologie (AP–HP), CHU Pitié-Salpêtrière, Institute of Cardiometabolism and Nutrition (ICAN), 75013 Paris, France 

Corresponding author at: Institut de Cardiologie, Groupe Hospitalier Pitié-Salpêtrière, Assistance publique–Hôpitaux de Paris, 47–83, boulevard de l’Hôpital, 75651 Paris cedex 13, France.
Summary
Background

Three-dimensional transoesophageal echocardiography (3D-TOE) is a new noninvasive tool for quantitative assessment of left ventricular (LV) volumes and ejection fraction.

Aim

The objective of this pilot study was to evaluate the feasibility and accuracy of 3D-TOE for the estimation of cardiac output (CO), using transpulmonary thermodilution with the Pulse index Contour Continuous Cardiac Output (PiCCO) system as the reference method, in intensive care unit (ICU) patients.

Methods

Fifteen ICU patients on mechanical ventilation prospectively underwent PiCCO catheter implantation and 3D-TOE. 3D-TOE LV end-diastolic and end-systolic volumes were determined using semi-automated software. CO was calculated as the product of LV stroke volume (end-diastolic volumeend-systolic volume) multiplied by heart rate. CO was also determined invasively by transpulmonary thermodilution as the reference method.

Results

Among 30 haemodynamic evaluations, 29 (97%) LV 3D-TOE datasets were suitable for CO calculation. The mean 3D-TOE image acquisition and post-processing times were 46 and 155seconds, respectively. There was a correlation (r =0.78; P <0.0001) between PiCCO and 3D-TOE CO. Compared with PiCCO, the 3D-TOE CO mean bias was 0.38L/min, with limits of agreement of −1.97 to 2.74L/min.

Conclusions

Noninvasive estimation of CO by 3D-TOE is feasible in ICU patients. This new semi-automated modality is an additional promising tool for noninvasive haemodynamic assessment of ICU patients. However, the wide limits of agreement with thermodilution observed in this pilot study require further investigation in larger cohorts of patients.

The full text of this article is available in PDF format.
Résumé
Contexte

L’échocardiographie trans-œsophagienne tridimensionelle (ETO-3D) est une nouvelle modalité non invasive d’évaluation des volumes et de la fraction d’éjection du ventricule gauche (VG).

Objectif

Évaluer la faisabilité et la performance de l’ETO-3D comparativement à la thermodilution transpulmonaire par méthode PiCCO pour la mesure du débit cardiaque (DC).

Méthodes

Dans cette étude pilote, 15 patients sous ventilation mécanique admis en réanimation et bénéficiant d’un monitorage hémodynamique invasif par le système PiCCO ont été prospectivement évalués par ETO-3D. Les volumes télé-diastolique et télé-systolique du VG ont été mesurés en utilisant un logiciel semi-automatique spécifique. Le DC a ensuite été calculé en multipliant le volume d’éjection systolique du VG (volume télé-diastoliquevolume télé-systolique) par la fréquence cardiaque. Le DC a également été mesuré de façon invasive par thermodilution transpulmonaire.

Résultats

Parmi les 30 évaluations hémodynamiques effectuées, 29 (97 %) acquisitions ETO-3D étaient exploitables. Les temps moyens nécessaires pour l’acquisition et l’analyse des données ETO-3D étaient respectivement de 46 et 155 secondes. Les mesures de DC effectuées par ETO-3D et par méthode invasive étaient corrélées (r =0,78 ; p <0,0001). Le biais moyen entre les 2 méthodes de mesure était de 0,38L/min, les limites d’agrément étaient de −1,97 à 2,74L/min.

Conclusions

L’évaluation non invasive du DC par ETO-3D est faisable. Cette nouvelle modalité ultrasonore est un outil prometteur pour l’évaluation hémodynamique des patients admis en réanimation. Les limites d’agrément relativement larges observées dans cette étude pilote comparativement à la theromdilution nécessitent toutefois d’être évaluer sur de plus larges populations de patients.

The full text of this article is available in PDF format.

Keywords : 3D, TOE, Cardiac output, Ultrasound, Intensive care unit

Mots clés : 3D, ETO, Ultrasons, Débit cardiaque, Réanimation

Abbreviations : 2D, 3D, CI, CO, ICU, LV, TOE


Background

Cardiac output (CO) measurement is an important variable required for the haemodynamic management of critically ill mechanically ventilated patients. Thermodilution by the Swan-Ganz catheter is still considered as the reference method in the clinical setting [1]. More recently, transpulmonary thermodilution was established as an alternative to the pulmonary catheter [2]. However, this method is also invasive, and requires artery and central venous catheterization, which can lead to severe complications [1].

Transoesophageal echocardiography (TOE) is increasingly used for the haemodynamic management of patients admitted to the intensive care unit (ICU) [3]. While the transthoracic examination of mechanically ventilated patients may be challenging, TOE is safe, and provides accurate heart imaging in most of these patients [4]. TOE training is an essential part of advanced critical care echocardiography learning [3].

Recently, a new generation of TOE probes has been introduced, allowing real-time three-dimensional (3D) imaging of the heart [5]. Currently, 3D-TOE is available for routine practice in several echocardiography machines, integrating both traditional modalities and 3D imaging. In addition, image acquisition has been simplified, and recent software that uses a semi-automated approach has dramatically reduced the 3D dataset off-line post-processing time [5].

The accuracy of 3D-ultrasound for left ventricular (LV) systolic and diastolic volume measurements and for LV ejection fraction calculation is superior to a conventional two-dimensional (2D) approach when using magnetic resonance imaging as the gold standard method [5, 6, 7]. However, the value of this technology for the estimation of CO in the ICU setting has been poorly investigated. This new semi-automated ultrasound modality, offering the advantage of providing the most reliable ultrasound quantitative assessment of LV ejection fraction, may have a suitable role in CO assessment, especially in cases where the conventional ultrasound Doppler method is not applicable (e.g. beam misalignment).

The main objective of this pilot study was to evaluate the feasibility and accuracy of 3D-TOE for the estimation of CO, using transpulmonary thermodilution as the reference method, in ICU mechanically ventilated patients.

Methods
Patients

Fifteen ICU patients on mechanical ventilation who had received a Pulse index Contour Continuous Cardiac Output (PiCCO; PULSION Medical Systems, Munich, Germany) haemodynamic monitoring device were prospectively included in the study. Exclusion criteria were: age<18 years; non-sinus rhythm; contraindication for TOE [8]; tricuspid, aortic or mitral valve regurgitation>2/4; extracorporeal membrane oxygenation support; and mechanical mitral valve prosthesis.

In addition to haemodynamic measurements, the following data were recorded: age; sex; simplified acute physiology score (SAPS II) [9]; and primary reason for ICU admission.

All patients were on continuous intravenous sedation combined with neuromuscular blocking agents during haemodynamic and TOE measurements.

The study was approved by the institutional committee for human research at our institution. Written informed consent was obtained from an appropriate designee for each patient before participation.

Invasive CO measurements

A 5F thermistor-tipped catheter was placed in the femoral artery and connected to the PiCCO system. From a central venous catheter positioned in the internal jugular or subclavian vein, 20mL of cold saline solution (<8°) were injected using the distal lumen, and CO was calculated with the transpulmonary thermodilution method [1]; the mean of three consecutive CO measurements was used. PiCCO measurements were performed independently by an intensivist who was unaware of the echocardiography results.

TOE

First, the ultrasound oesophageal probe was introduced. Immediately after PiCCO measurements, TOE examination was acquired in all patients by two experienced physicians (>1000 examinations) using the IE33 system (Philips Medical Systems, Andover, MA, USA), equipped with a 3D-TOE probe (X7-2t). The images were transferred to a workstation equipped with QLAB software version 8.1 for post-processing. All the examinations were analysed off-line, blinded to invasive measurements. All projections were obtained according to echocardiography guidelines [5, 8]. All measurements were averaged over three cardiac cycles. Ectopic and post-ectopic beats were disregarded.

Doppler method

From a 120° mid-oesophageal two-dimensional long-axis zoom of the aortic valve, LV outflow tract diameter was measured. Using pulsed-wave Doppler mode, the LV outflow tract time-velocity integral was recorded from the transgastric view. Guided both by the 2D image and colour Doppler, the Doppler sound beam was placed as parallel as possible to the LV outflow tract flow. CO was calculated as recommended (Doppler CO) [10].

3D method

From a 2D LV mid-oesophageal four-chamber view at 0°, the real-time biplane imaging modality (X-plane mode) was used to obtain a simultaneous visualization of two orthogonal LV views, and therefore to adjust the probe position and acoustic window. During a temporary interruption of ventilator support at end-expiration (5–10seconds, depending on the patient's heart rate), a 3D LV full-volume dataset was then reconstructed, using the R-wave gated method over four to six consecutive cardiac cycles, and acquired. Sector size and depth were adjusted to achieve optimal visualization of the LV at the highest possible frame rate (mean value 28±5 frames/s). During post-processing, the full-volume LV dataset was organized off-line into four-chamber, two-chamber and short-axis views. Mitral annular and apical points were placed manually on these images in end-diastole and end-systole. LV endocardial borders were then detected automatically by the software, and adjusted manually if needed. The software then used sequence analysis to track the endocardium in all frames, and to determine LV end-diastolic volume, LV end-systolic volume and LV ejection fraction. CO was calculated as the product of LV stroke volume (end-diastolic volumeend-systolic volume) multiplied by heart rate (Figure 1).



Figure 1


Figure 1. 

Post-processing of a three-dimensional transoesophageal echocardiography left ventricular (LV) dataset. The dedicated software semi-automatically determines LV end-diastolic volume (EDV), LV end-systolic volume (ESV) and LV ejection fraction (LVEF). Cardiac output (CO) was calculated as the product of LV stroke volume (SV) (EDVESV) multiplied by heart rate.

Zoom

Statistical analysis

Continuous variables are presented as medians with interquartile ranges. Qualitative variables are presented as counts and percentages. COs obtained from PiCCO and from 3D-TOE were compared using Pearson's correlation coefficient and the Bland-Altman method. In addition, the percentage error (2 standard deviations/mean CO) for TOE estimation of CO with PiCCO-derived measurement as the gold standard was calculated.

Intraobserver and interobserver variabilities for the measurement of CO were assessed in a subset of 10 patients. The coefficient of variation and intraclass correlation coefficient were determined.

MedCalc Statistical Software version 14.12.0 (MedCalc Software, Ostend, Belgium) was used for calculations.

Results

The clinical characteristics of the patients are listed in Table 1.

Thirty haemodynamic evaluations were obtained in 15 patients. At least 12hours separated two investigations in patients undergoing multiple evaluations.

3D-TOE estimation of CO was feasible in 29/30 (97%) cases. Abnormal LV segmental contraction was observed in nine (31%) cases. The mean 3D-TOE image acquisition and post-processing times were 46±19 and 155±35seconds, respectively. The COs derived from PiCCO and 3D-TOE were correlated (r =0.78; P <0.0001); the coefficient of correlation between PiCCO and Doppler CO was 0.72 (P <0.0001) (Figure 2).



Figure 2


Figure 2. 

Linear correlation between Pulse index Contour Continuous Cardiac Output (PiCCO) and three-dimensional transoesophageal echocardiography (3D-TOE) cardiac output (CO) (left), and between PiCCO and Doppler CO (right).

Zoom

Compared with PiCCO, the 3D-TOE CO mean bias was 0.38L/min (95% confidence interval [CI]: −0.07 to 0.84L/min), with a standard deviation of the difference between paired measurements of 1.2L/min, giving limits of agreement of −1.97 to 2.74L/min and a percentage error of 44%. Compared with PiCCO, the Doppler CO mean bias was 0.48L/min (95% CI: −0.07 to 1.03L/min), with a standard deviation of the difference between paired measurements of 1.45L/min, giving limits of agreement of −2.37 to 3.33L/min and a percentage error of 53% (Figure 3).



Figure 3


Figure 3. 

Bland-Altman analysis of the agreement between Pulse index Contour Continuous Cardiac Output (PiCCO) and three-dimensional transoesophageal echocardiography (3D-TOE) cardiac output (CO) (left), and between PiCCO and Doppler CO (right). SD: standard deviation.

Zoom

For intraobserver variability, the coefficient of variation and the intraclass correlation coefficient were, respectively, 3.9% and 0.99 (95% CI: 0.91 to 0.99) for 3D-TOE CO, and 9.3% and 0.96 (95% CI: 0.81 to 0.99) for Doppler CO. For interobserver variability, the coefficient of variation and the intraclass correlation coefficient were, respectively, 9.5% and 0.95 (95% CI: 0.76 to 0.99) for 3D-TOE CO, and 5.1% and 0.94 (95% CI: 0.71 to 0.99) for Doppler CO.

Discussion

In this pilot study, 3D-TOE appears to be a feasible method for CO measurement in ICU mechanically ventilated patients. The latest generation of 3D echocardiography equipment and software permitted the acquisition of good quality images in 97% of cases, with relatively short recording and off-line post-processing times (46 and 155seconds, respectively), making this technique suitable for routine clinical use at the bedside. This new semi-automated modality is a promising additional tool for noninvasive haemodynamic assessment of ICU patients.

Doppler-echocardiography assessment of CO can be performed using several techniques. Doppler-based CO is the most frequently used ultrasound approach in ICU patients [11]. However, this method requires two measurements obtained from two different acoustic windows. Using TOE, the LV outflow tract diameter should be measured from the mid-oesophageal long-axis view, while pulsed-wave Doppler LV outflow tract flow should be acquired from the transgastric view. This Doppler-derived CO approach is vulnerable to several measurements errors, mainly related to non-optimal alignment of the Doppler beam with aortic flow [10].

Another method relies on the 2D assessment of systolic and diastolic volumes, based on Simpson's method. However, this technique has several limitations, mainly related to the use of calculation formulae that postulate normal LV shape [7]. Indeed, this is an important issue in ICU patients, who frequently have segmental LV dysfunction, as observed in 31% of the cases included in our study.

Alternatively, accurate 3D-ultrasound LV measurement of end-diastolic and end-systolic volumes has been validated using magnetic resonance imaging as the gold standard method [6, 7]. Thus, 3D-TOE might provide more reliable semi-automated quantitative assessment of actual systolic function than the simple – and frequently used – fractional area change, visually evaluated in the transgastric short-axis view or in the four-chamber view of the cardiac chambers [12], as this monoplane 2D approach might be biased in patients with inhomogeneous systolic contraction. Moreover, the 3D-TOE LV full-volume approach without any geometrical assumption has the potential to provide accurate estimation of stroke volume and CO. Using previous generations of machines and software, the feasibility of the technique was reported for CO measurement in the operating room in stable patients just before bypass surgery [13]. However, the post-processing time to obtain CO was>400seconds in that study. Our study showed that the post-processing time was more than halved using the latest generation of software, making the technique more applicable for clinical use at the patient's bedside. To ensure that the echocardiography analyses were completely blinded from PiCCO and clinical data, the echocardiography post-processing was performed off-line in our study. Using the same software, 3D-TOE volume dataset processing is feasible on-line on the echocardiography machine.

In this pilot study, which had a limited number of observations, the percentage error and the limits of agreement between 3D-TOE and invasive measurement of CO were relatively wide. Thus, 3D-TOE could not be considered as interchangeable with invasively measured CO, and needs further evaluation in larger cohorts of patients. However, the percentage error with invasive data was smaller and the limits of agreements were narrower using 3D-TOE compared with the Doppler method. Currently, 3D-TOE should be considered as a promising, valuable, noninvasive alternative to conventional ultrasound examination, offering the advantage of providing the most reliable ultrasound quantitative assessment of LV ejection fraction.

We acknowledge several limitations to the present study. First, the current availability and cost of 3D-ultrasound technology is still an obstacle to the widespread use of the technique in ICUs worldwide. However, Doppler-echocardiography platforms, including one compact ultrasound system, which allow real-time 3D-TOE imaging, are now available, and may rapidly spread in the ICU environment. Second, sinus rhythm is required for 3D LV dataset reconstruction over several cardiac cycles. For an accurate estimation of LV stroke volume in patients with arrhythmias, the next generation of echocardiography machines should allow single-beat LV full-volume acquisition with a high frame rate. Third, the 3D data acquisitions and post-processing were performed in this study by two experts used to TOE. It would be of interest to assess in subsequent studies the feasibility of the techniques in the hands of less experienced physicians. Finally, the ability of 3D-TOE to track CO changes after a therapeutic intervention was not assessed, and should be investigated in further studies.

Conclusions

Noninvasive estimation of CO by 3D-TOE is feasible in ICU mechanically ventilated patients. This new semi-automated modality appears to be a promising, valuable, noninvasive alternative to conventional ultrasound examination, offering the advantage of providing the most reliable ultrasound quantitative assessment of LV ejection fraction. This new technology needs to be investigated in a larger population of ICU patients.

Sources of funding

None.

Disclosure of interest

N. H. Consulting/advisory activities for the company Philips.

The other authors declare that they have no competing interest.

References

Litton E., Morgan M. The PiCCO monitor: a review Anaesth Intensive Care 2012 ;  40 : 393-409
Chakravarthy M., Patil T.A., Jayaprakash K., Kalligudd P., Prabhakumar D., Jawali V. Comparison of simultaneous estimation of cardiac output by four techniques in patients undergoing off-pump coronary artery bypass surgery – a prospective observational study Ann Card Anaesth 2007 ;  10 : 121-126 [cross-ref]
Expert Round Table on Echocardiography in ICU International consensus statement on training standards for advanced critical care echocardiography Intensive Care Med 2014 ;  40 : 654-666
Mayo P., Mekontso Dessap A., Vieillard-Baron A. Erratum to: myths about critical care echocardiography: the ten false beliefs that intensivists should understand Intensive Care Med 2015 ;  41 : 573
Lang R.M., Badano L.P., Tsang W., and al. EAE/ASE recommendations for image acquisition and display using three-dimensional echocardiography J Am Soc Echocardiogr 2012 ;  25 : 3-46 [inter-ref]
Jacobs L.D., Salgo I.S., Goonewardena S., and al. Rapid on-line quantification of left ventricular volume from real-time three-dimensional echocardiographic data Eur Heart J 2006 ;  27 : 460-468
Lang R.M., Badano L.P., Mor-Avi V., and al. Recommendations for cardiac chamber quantification by echocardiography in adults: an update from the American Society of Echocardiography and the European Association of Cardiovascular Imaging Eur Heart J Cardiovasc Imaging 2015 ;  16 : 233-270 [cross-ref]
Hahn R.T., Abraham T., Adams M.S., and al. Guidelines for performing a comprehensive transesophageal echocardiographic examination: recommendations from the American Society of Echocardiography and the Society of Cardiovascular Anesthesiologists J Am Soc Echocardiogr 2013 ;  26 : 921-964 [inter-ref]
Le Gall J.R., Lemeshow S., Saulnier F. A new Simplified Acute Physiology Score (SAPS II) based on a European/North American multicenter study JAMA 1993 ;  270 : 2957-2963 [cross-ref]
Quinones M.A., Otto C.M., Stoddard M., and al. Recommendations for quantification of Doppler-echocardiography: a report from the Doppler Quantification Task Force of the Nomenclature and Standards Committee of the American Society of Echocardiography J Am Soc Echocardiogr 2002 ;  15 : 167-184 [cross-ref]
Thiele R.H., Bartels K., Gan T.J. Cardiac output monitoring: a contemporary assessment and review Crit Care Med 2015 ;  43 : 177-185 [cross-ref]
Vieillard-Baron A., Charron C., Chergui K., Peyrouset O., Jardin F. Bedside echocardiographic evaluation of hemodynamics in sepsis: is a qualitative evaluation sufficient? Intensive Care Med 2006 ;  32 : 1547-1552 [cross-ref]
Culp W.C., Ball T.R., Burnett C.J. Validation and feasibility of intraoperative three-dimensional transesophageal echocardiographic cardiac output Anesth Analg 2007 ;  105 : 1219-1223 [cross-ref]



© 2016  Published by Elsevier Masson SAS.
EM-CONSULTE.COM is registrered at the CNIL, déclaration n° 1286925.
As per the Law relating to information storage and personal integrity, you have the right to oppose (art 26 of that law), access (art 34 of that law) and rectify (art 36 of that law) your personal data. You may thus request that your data, should it be inaccurate, incomplete, unclear, outdated, not be used or stored, be corrected, clarified, updated or deleted.
Personal information regarding our website's visitors, including their identity, is confidential.
The owners of this website hereby guarantee to respect the legal confidentiality conditions, applicable in France, and not to disclose this data to third parties.
Close
Article Outline