Cardiac Output Measurement During Infrarenal Aortic Surgery: Echo-Esophageal Doppler Versus Thermodilution Catheter

Cardiac Output Measurement During Infrarenal Aortic Surgery: Echo-Esophageal Doppler Versus Thermodilution Catheter Aurélie Lafanechère, MD,* Pierre Albaladejo, MD, PhD,* Mathieu Raux, MD,* Thomas Geeraerts, MD,* Rémi Bocquet, MD,* Anne Wernet, MD,* Yves Castier, MD,† and Jean Marty, MD* Objective: Aortic surgery is associated with various hemodynamic and cardiac output modifications. These disorders may be partly caused by blood flow redistribution between supra-aortic and descending aorta regions during clamping and unclamping. A new echo-esophageal Doppler (Hemosonic 100; Arrow, Reading, PA) calculates cardiac output from a simultaneous measurement of blood flow velocity and diameter of the descending aorta. This calculation may be affected by blood redistribution during aortic clamping. The aim of this study was to compare cardiac output measured by echo-esophageal Doppler and by bolus thermodilution catheter during infrarenal aortic surgery. Design: Prospective, observational study. Setting: University hospital, single institution. Participants: Twenty-two adult patients. Interventions: Infrarenal aortic surgery. Measurements and Main Results: Cardiac outputs monitored by both devices were highly correlated during the

whole surgical procedure (r2 ranging from 0.54 to 0.76). Bland and Altman analysis showed absence of significant bias before and after clamping (ranging from 0.1 ⴞ 0.73 L/min to 0.18 ⴞ 1 L/min, p > 0.05) and a significant bias of 0.5 ⴞ 1.05 L/min (p < 0.05) during aortic clamping. Limits of agreement did not differ significantly during the whole surgical procedure (ranging from ⴚ1.36/2.19 to ⴚ2.23/2.49). During clamping and unclamping, changes in cardiac output obtained by both methods were positively correlated (r2 ⴝ 0.7). Conclusions: Bias between both methods was clinically acceptable, and limits of agreement were not significantly modified by aortic clamping. However, larger studies including homogenous aortic pathologies are necessary to validate this method during infrarenal aortic surgery. © 2006 Elsevier Inc. All rights reserved.

A

cross-clamping, the reliability of the Echo-ED may be questioned in this context. The aim of this study was to compare CO measurements provided by the new Echo-ED method with the bolus thermodilution method during aortic surgery with infrarenal aortic cross-clamping.

ORTIC SURGERY is associated with significant changes in cardiac output (CO) and systemic vascular resistance after aortic cross-clamping and release.1 Hemodynamic monitoring during surgery usually relies on invasive techniques such as thermodilution CO, which remains the reference method.2 However, 2 noninvasive esophageal Doppler technologies have emerged: the esophageal Doppler monitor (EDM, CardioQ; Deltex Medical, Inc, Irving, TX) and the echo-esophageal Doppler (Echo-ED, Hemosonic100; Arrow, Reading, PA). Both methods measure blood flow velocity in the descending thoracic aorta using a pulsed Doppler flow meter. In the EDM monitor, aortic diameter is extrapolated using a normogram based on patient’s age, height, and weight; whereas in the Echo-ED monitor, an M echo mode transducer measures thoracic aortic diameter (DiamAo).3,4 Aortic blood flow (ABF) is determined by both devices from ABF velocity and aortic diameter. CO is calculated by both devices using a constant relationship between blood flow to the upper part of the body and the descending aorta. Echo-ED is a more recent device, and CO monitoring has been evaluated in only 4 studies during cardiac surgery.5-8 In a recent review concerning esophageal Doppler technologies, the authors reported an insufficient number of studies to assess the validity of Echo-ED.9 Because blood flow distribution could be modified by aortic

From the Departments of *Anaesthesiology and Intensive Care, and †Vascular and Thoracic Surgery, University Hospital Beaujon, Clichy, France. Address reprint requests to Pierre Albaladejo, MD, PhD, Service d’Anesthésie-Réanimation SAMU-SMUR 94 du Professeur Jean MARTY, Hôpital Henri Mondor, 51 Avenue du Maréchal de Lattre de Tassigny, 94000 Créteil, France. E-mail: pierre.albaladejo@ hmn.aphp.fr © 2006 Elsevier Inc. All rights reserved. 1053-0770/06/2001-0006$32.00/0 doi:10.1053/j.jvca.2005.07.029 26

KEY WORDS: aortic surgery, cardiac output monitoring, esophageal Doppler, thermodilution catheter

METHODS After approval of the local ethics committee and after written informed consent, patients scheduled for aortic surgery with infrarenal aortic cross-clamping via an open transabdominal approach using a median laparotomy under general anesthesia were included in the study. Patients having a contraindication for the use of Echo-ED (eg, known or suspected esophageal ulcer, mycosis, malformation, varicose veins, or tumor) were excluded. General anesthesia was induced using intravenous thiopental (3-5 mg/kg), fentanyl (3-5 ␮g/kg), and cisatracurium (0.2 mg/kg). Anesthesia was maintained by inhalation of isoflurane (1 minimal alveolar concentration) with a nitrous oxide mixture (60% oxygen and 40% nitrous oxide). Additional doses of fentanyl, 50 ␮g, were used as necessary. None of the patients received epidural catheters or intrathecal techniques. Neuromuscular block was maintained with additional doses of cisatracurium (0.03 mg/kg) as necessary. At the end of surgery, patients were transferred to the intensive care unit before being extubated. Anesthesia was given by an independent anesthesiologist not involved in the study. After tracheal intubation, invasive arterial blood pressure was assessed by a radial artery catheter and CO was monitored by a bolus thermodilution catheter and an Echo-ED probe. The Echo-ED probe was inserted by the oral route and advanced in the esophagus to the level of the third intercostal juxtasternal space where the esophagus and aorta are in parallel. Its position was adjusted to obtain the highest Doppler velocity signal along with a simultaneous display of the aortic wall images. The probe was readjusted if necessary (loss of the aortic blood velocity or aortic wall signals). CO obtained by the Echo-ED (COED) was calculated according to the following equation incorporated in the monitor (Hemosonic 100): COED ⫽ 1.22 ABF ⫹ 0.69. This equation is derived from a regression analysis that compares 311 paired measurements of ABF and CO measured by a thermodilution catheter. These measurements were obtained in 90 adult patients from a review by Boulnois et al10 based on 3 studies.4,11,12 COED was obtained before thermodilution CO determination by an

Journal of Cardiothoracic and Vascular Anesthesia, Vol 20, No 1 (February), 2006: pp 26-30

CARDIAC OUTPUT MEASUREMENT

27

analyses were performed by standard methods. Concordance (mean bias and limits of agreement) was analyzed using a representation of Bland and Altman.13 Mean bias indicates the mean difference between COED and COTC. Limits of agreements were calculated as bias ⫾ 2 standard deviations and indicate the 95% range of fluctuations of the differences between both methods. Significance of bias was studied using a paired Student t test. A p value ⬍0.05 was considered to be statistically significant.

Table 1. Demographic and Medical Data Patients (n) Age (y, mean ⫾ 1SD) Sex (M/F) ASA score I Score II Score III Score IV Surgical procedure (n) Abdominal aortic aneurysm Aortic occlusive disease Aortic cross-clamping time (min, mean ⫾ 1 SD) Medical history (n) Hypertension Peripheral artery disease Coronary artery disease Diabetes mellitus Medications (n) Beta-adrenergic blockers Calcium channel blockers Angiotensin-converting enzyme inhibitors Diuretics

22 65 ⫾ 8 18/4 0 17 4 1

RESULTS

7 15 63 ⫾ 30

Twenty-two patients were included in the study. Demographic and clinical data are presented in Table 1. Seven patients underwent surgery for abdominal aortic aneurysm and 15 for aortic occlusive disease. Ten patients received a bolus injection of ephedrine (3-12 mg), 6 of them immediately after induction and 8 of them immediately after unclamping. One patient required a dobutamine infusion (7.5 ␮g/kg/min) after unclamping. Isoflurane was used after aortic clamping to control blood pressure. No other vasodilators were used in these patients. Hemodynamic data are presented in Table 2. COED and COTC decreased significantly during aortic cross-clamping and increased significantly after unclamping the aorta when compared with the preceding surgical period. A significant decrease in DiamAo was observed immediately after unclamping. COED and COTC were positively correlated during each period of the surgical procedure: (A) r2 ⫽ 0.76, (B) r2 ⫽ 0.54, (C) r2 ⫽ 0.56, (D) r2 ⫽ 0.6, (E) r2 ⫽ 0.64, and (F) r2 ⫽ 0.67. Figure 1 (Bland and Altman plot) shows agreement between COTD and COED during 3 different surgical periods: before aortic clamping (B), before unclamping (D), and at the end of surgery (F). Other periods showed similar results. Mean bias and limits of agreement obtained during the 6 periods of surgery are shown in Table 3. The mean difference between COED and COTC (mean bias) was not statistically significant before clamping (0.1 ⫾ 0.73 L/min [A] and 0.13 ⫾ 1.18 L/min [B], p ⬎ 0.05) and after unclamping (0.18 ⫾ 1 L/min [E] and 0.15 ⫾ 1 L/min [F], p ⬎ 0.05), but became statistically significant during aortic clamping (0.54 ⫾ 1.05 L/min [C], p ⬍ 0.05). The limits of agreement (mean bias ⫾ 2 standard deviations) did not differ significantly during the whole surgery.

12 16 8 6 10 7 10 5

Abbreviations: M, male; F, female.

anesthesiologist experienced with this device not involved in the anesthesia management. Thermodilution cardiac output (COTC) was measured by a pulmonary artery catheter (Edwards Lifescience, Irvine, CA) inserted via the internal jugular vein. The retained value was the mean of 3 consecutive bolus injections (NaCl 0.9 %, 10 mL at room temperature) throughout the whole respiratory cycle. COTC was obtained immediately after COED by the anesthesiologist involved in the anesthesia management who was not blinded to the COED value. Heart rate; systolic, diastolic, and mean arterial pressure; mean pulmonary artery pressure; pulmonary artery occlusion pressure; DiamAo; ABF; COED; and COTC were recorded during the following 6 periods: after induction of anesthesia (A), before clamping (B), 10 minutes after clamping (C), before unclamping (D), 10 minutes after unclamping (E), and at the end of surgery (F). Standard statistical analysis was performed using Statview software (SAS, Cary, NC). Values are expressed as mean ⫾ 1 standard deviation. Hemodynamic data and DiamAo were analyzed with a 2-way analysis of variance for repeated measurements. Simple regression

Table 2. Hemodynamic Parameters During the 6 Surgical Periods Before Clamping A

HR (beats/min) SAP (mmHg) MAP (mmHg) DAP (mmHg) MPAP (mmHg) PAOP (mmHg) COTC (L/min) ABF (L/min) COED (L/min) DiamAo (mm)

69 ⫾ 16 106 ⫾ 24 75 ⫾ 18 57 ⫾ 14 27 ⫾ 14 14 ⫾ 6 4.6 ⫾ 1.5 3.2 ⫾ 1.2 4.5 ⫾ 1.5 26.3 ⫾ 4

During Clamping

After Unclamping

B

C

D

E

F

69 ⫾ 20 113 ⫾ 27 77 ⫾ 20 61 ⫾ 17 25 ⫾ 16 12 ⫾ 5 4.9 ⫾ 1.6 3.4 ⫾ 1.3 4.8 ⫾ 1.6 26.9 ⫾ 5

66 ⫾ 20 108 ⫾ 25 72 ⫾ 17 57 ⫾ 15 23 ⫾ 13 12 ⫾ 5 4.0 ⫾ 1.3* 2.4 ⫾ 1.3* 3.6 ⫾ 1.6* 26.5 ⫾ 4

67 ⫾ 18 110 ⫾ 23 75 ⫾ 15 56 ⫾ 12 25 ⫾ 16 11 ⫾ 4 4.3 ⫾ 1.4 2.5 ⫾ 1.3 3.7 ⫾ 1.6 26.7 ⫾ 4

70 ⫾ 18 101 ⫾ 26 69 ⫾ 17 55 ⫾ 19 27 ⫾ 12 12 ⫾ 5 4.7 ⫾ 1.4* 3.2 ⫾ 1.4* 4.6 ⫾ 1.6* 25.9 ⫾ 4*

70 ⫾ 16 104 ⫾ 20 72 ⫾ 14 56 ⫾ 16 26 ⫾ 14 11 ⫾ 6 5.0 ⫾ 1.3 3.4 ⫾ 1.4 4.9 ⫾ 1.6 26 ⫾ 4

Abbreviations: SAP, systolic arterial pressure; MAP, mean arterial pressure; DAP, diastolic arterial pressure; MPAP, mean pulmonary artery pressure; PAOP, pulmonary artery occlusion pressure. *p ⬍ 0.05 versus previous period; values are mean ⫾ 1 standard deviation.

28

LAFANECHÈRE ET AL

Fig 2. Plot comparing percent change in COTC and COED during clamping (B v C) and unclamping aorta (D v E).

Figure 2 represents the percent change in COTC during aortic clamping (B v C) and unclamping (D v E) compared with the corresponding percent change in COED. Changes in COED were positively correlated to changes in COTC (r2 ⫽ 0.7). Twenty percent of COED and COTC increases or decreases were in opposite directions during clamping and unclamping. On 5 occasions, COED changed less than 10% when COTD changed greater than 20%, and on 2 occasions COED changed more than 40% when COTD changed less than 10%. DISCUSSION

Fig 1. Bland and Altman plot comparing COED and COTC before clamping (B), before unclamping (D), and at the end of surgery (F). The solid line represents the mean difference (mean bias) between COED and COTC, and the dotted line defines the 2 standard deviation limits of agreement.

It was hypothesized that aortic clamping and the resulting change in blood flow distribution could be a source of error in measuring cardiac output by Echo-ED. The present study shows that during aortic surgery with infrarenal aortic crossclamping, COs measured by thermodilution and by Echo-ED are highly correlated even during clamping. A significant bias between the 2 methods is observed only during aortic clamping, but limits of agreement do not differ during the whole surgical procedure. Changes in COED were positively correlated to changes in COTC during clamping and unclamping. This study is the first to evaluate the reliability of CO measurements obtained by the Echo-ED method during aortic surgery. Two previous studies compared CO monitored by

Table 3. Bland and Altman Analysis for Cardiac Output Measurement by Echo-Esophageal Doppler and Thermodilution Catheter

(A) After induction of anesthesia (B) Before clamping (C) 10 min after clamping (D) Before unclamping (E) 10 min after unclamping (F) At the end of surgery Abbreviation: CI, confidence interval.

Mean Bias (95% CI)

Lower Limits of Agreement (95% CI)

Upper Limits of Agreement (95% CI)

0.10 (⫺0.21/0.41) 0.13 (⫺0.39/0.65) 0.43 (⫺0,02/0,88) 0.54 (0.45/0.99) 0.18 (⫺0.25/0.61) 0.15 (⫺0.28/0.58)

⫺1,36 (⫺1.92/⫺1.07) ⫺2.23 (⫺3.11/⫺1.33) ⫺1.65 (⫺2.44/⫺0.86) ⫺1.56 (⫺2.35/⫺0.77) ⫺1.82 (⫺2.56/⫺1.08) ⫺1.85 (⫺2.59/⫺1.11)

2.19 (1.63/2.75) 2.49 (1.6/3.38) 2.51 (1.72/3.3) 2.64 (1.85/3.43) 2.18 (1.44/2.92) 2.15 (1.41/2.89)

CARDIAC OUTPUT MEASUREMENT

29

esophageal Doppler with COTC during aortic surgery.14,15 However, in both studies, the aortic diameter was obtained from a normogram. In the present study, the Echo-ED device allowed a continuous and valid measurement of the aortic diameter.4 This measurement is of interest considering possible modifications in thoracic aortic diameter related to infrarenal cross-clamping. These patients’ aortic diameter showed a decrease of less than 1 mm during unclamping. The authors considered this small variation as not clinically relevant. This absence of aortic dilatation during clamping may be because of the rigidity of the aortic wall in atherosclerotic patients or to the low range of blood pressure. In fact, large blood pressure variations have been implicated in modifying aortic diameter.15 During aortic clamping, a statistically significant and positive mean bias was observed between COTC and COED (0.54 ⫾ 1.05 L/min, p ⬍ 0.05). This suggests a role of blood flow redistribution. However, this bias was not as high as previously reported during aortic surgery. During aortic cross-clamping, Perrino et al15 found a mean bias of 0.7 L/min, and Klotz et al14 observed an increased bias of 1.51 L/min between both methods. As suggested by the authors, in these patients a significant “aortic dilatation induced by clamp” may occur. Another explanation could be a greater blood flow redistribution during aortic clamping in these patients. Indeed, in the present study, 16 of 22 patients presented with a past medical history of peripheral artery disease. Blood flow distribution may have been modified in these patients before aortic cross-clamping, which may have minimized the effects of the aortic crossclamping. Echo-ED has been previously compared with bolus thermodilution in only 3 studies.6-8 Su et al8 compared both methods among 12 patients during cardiac surgery (coronary artery bypass grafting). Jaeggi et al6 and Moxon et al7 included 22 and 13 patients who had recently undergone cardiac surgery. In these studies, the mean bias ⫾ standard deviation ranged from 0.23 ⫾ 1.06 L/min to 0.23 ⫾ 0.8 L/min between both devices. In the present study, the mean bias ⫾ standard deviation ranged from 0.10 ⫾ 0.73 L/min to 0.5 ⫾ 1.05 L/min. This is similar to previous results observed in cardiac surgery. Although judgment of bias and limits of agreement are subjective and not standardized, limits of agreement between Echo-ED and thermodilution are wide. Several factors could explain this result. First, the intraobserver variability of the bolus thermodilution method has led some authors to question its validity as a reference method.16 In the study by Su et al8 including 24 patients undergoing coronary artery bypass graft surgery, COED compared with COTC obtained by continuous thermodilution (12 patients) showed a mean bias of 0.05 ⫾

0.49 L/min, whereas COED compared with bolus COTC (12 patients) showed a higher mean bias of 0.11 ⫾ 1.12 L/min. As reported by Dark and Singer,9 continuous thermodilution has a better repeatability and may be considered as a more appropriate “gold standard.” Second, although Echo-ED offers continuous CO monitoring, frequent probe repositioning was necessary in this study. Thus, despite experienced investigators, an interobserver and intraobserver variability in COED, not evaluated in this study, may have contributed to the magnitude of agreement. Third, COED is calculated from a regression analysis by Boulnois and Pechoux10 based on 3 studies.4,11,12 Because the characteristics of the patients from these studies were not clearly defined, the validity of this equation may be questioned. Two previous studies showed that ABF measured by Echo-ED tracks COTC changes.4,17 In the present work, 20% of COED and COTC variations during clamping and unclamping were in opposite directions. On 5 occasions, COED changed less than 10% when COTC changed more than 20%, and on 2 occasions COED changed more than 40% when COTC changed less than 10%. However, the COED intraobserver variability needs to be tested in order to define a threshold value for clinically relevant COED changes. Testing this variability is an important point because a training period and frequent probe repositioning are required to obtain a good signal. It would also help to evaluate the impact of Echo-ED on decision-making and to define Echo-EDs place during aortic surgery. In fact, when compared with thermodilution, this device does not measure venous pressures. The Doppler waveform offers an estimation of preload through the flow time and cardiac function through the maximum acceleration. However, further studies are required to determine the ability of those parameters to diagnose hypovolemia, cardiac insufficiency, or vasoplegic states. In summary, during aortic surgery, this study shows a clinically acceptable bias between COED and COTC, and the limits of agreement, although wide, were not significantly modified by infrarenal aortic clamping. Changes in COED were positively correlated to changes in COTC during clamping and unclamping. However, this study has several limitations. The method used as a gold standard is not a continuous determination of CO. The study was not blinded, and the small number of patients with aortic occlusive disease may have been insufficient to show a relevant difference between methods. Larger blinded studies with homogenous populations of patients are necessary to define COED intraobserver and interobserver variability and determine its validity during infrarenal aortic surgery.

REFERENCES 1. Gelman S: The pathophysiology of aortic cross-clamping and unclamping. Anesthesiology 82:1026-1960, 1995 2. Taylor RW, Ahrens T, Viejo A, et al: Pulmonary Artery Catheter Consensus Conference: Consensus statement. Crit Care Med 25:910925, 1997 3. Laupland KB, Bands CJ: Utility of esophageal Doppler as a minimally invasive hemodynamic monitor: A review. Can J Anaesth 49:393-401, 2002

4. Cariou A, Monchi M, Joly LM, et al: Noninvasive cardiac output monitoring by aortic blood flow determination: Evaluation of the Sometec Dynemo-3000 system. Crit Care Med 26:2066-2072, 1998 5. Bein B, Worthmann F, Tonner PH, et al: Comparison of esophageal Doppler, pulse contour analysis, and real-time pulmonary artery thermodilution for the continuous measurement of cardiac output. J Cardiothorac Vasc Anesth 18:185-192, 2004

30

6. Jaeggi P, Hofer CK, Klaghofer R, et al: Measurement of cardiac output after cardiac surgery by a new transesophageal Doppler device. J Cardiothorac Vasc Anesth 17:217-220, 2003 7. Moxon D, Pinder M, van Heerden PV, et al: Clinical evaluation of the HemoSonic monitor in cardiac surgical patients in the ICU. Anaesth Intensive Care 31:408-411, 2003 8. Su N, Huang C, Tsai P, et al: Cardiac output measurement during cardiac surgery: Esophageal Doppler versus pulmonary artery catheter. Acta Anaesthesiol Sin 40:127-133, 2002 9. Dark PM, Singer M: The validity of transesophageal Doppler ultrasonography as a measure of cardiac output in critically ill adults. Intensive Care Med 2004 30:2060-2066, 2004 10. Boulnois JL, Pechoux T: Noninvasive cardiac output monitoring by aortic blood flow measurement with the Dynemo 3000. J Clin Monit Comput 16:127-140, 2000 11. Bernardin G, Tiger F, Fouche R, et al: Continuous noninvasive measurement of aortic blood flow in critically ill patients with a new esophageal echo-Doppler system. J Crit Care 13:177-183, 1998

LAFANECHÈRE ET AL

12. Dummler R, Emmerich M, Klein G, et al: Semi-invasive cardiac output measurement using a combined transesophageal ultrasound device. Early experiences. Anaesthesist 49:207-210, 2000 13. Bland JM, Altman DG: Statistical methods for assessing agreement between two methods of clinical measurement. Lancet 1:307-310, 1986 14. Klotz KF, Klingsiek S, Singer M, et al: Continuous measurement of cardiac output during aortic cross-clamping by the oesophageal Doppler monitor ODM 1. Br J Anaesth 74:655-660, 1995 15. Perrino AC Jr, Fleming J, LaMantia KR: Transesophageal Doppler cardiac output monitoring: performance during aortic reconstructive surgery. Anesth Analg 73:705-710, 1991 16. Taylor SH, Silke B: Is the measurement of cardiac output useful in clinical practice? Br J Anaesth 60:90S-98S, 1988 17. Odenstedt H, Aneman A, Oi Y, et al: Descending aortic blood flow and cardiac output: A clinical and experimental study of continuous oesophageal echo-Doppler flowmetry. Acta Anaesthesiol Scand 45:180-187, 2001