Oesophageal Doppler monitoring overestimates cardiac output during lumbar epidural anaesthesia

H. A. Leather and P. F. Wouters

Department of Anaesthesiology, University Hospitals, Katholieke Universiteit Leuven, U. Z. Gasthuisberg, Herestraat 49, B-3000 Leuven, Belgium*Corresponding author

Accepted for publication: February 5, 2001


    Abstract
 Top
 Abstract
 Introduction
 Methods
 Results
 Discussion
 References
 
Oesophageal Doppler monitoring (ODM) has been advocated as a non-invasive means of measuring cardiac output (CO). However, its reliance upon blood flow measurement in the descending aorta to estimate CO is susceptible to error if blood flow is redistributed between the upper and lower body. We hypothesize that lumbar epidural anesthesia (LEA), which causes blood flow redistribution, causes errors in CO estimates. We compared ODM with thermodilution (TD) measurements in fourteen patients under general anaesthesia for radical prostatectomy, who had received an epidural catheter at the intervertebral level L2–L3. Coupled measurements of CO by means of the TD and ODM techniques were performed at baseline (general anaesthetic only) and after epidural administration of 10 ml of 0.25% bupivacaine. The two methods were compared using Bland-Altman analysis: before LEA there was a bias of –0.89 litre min–1 with limits of agreement ranging between –2.67 and +0.88 litre min–1. Following lumbar sympathetic block, bias became positive (+0.55 litre min–1) and limits of agreement increased to –3.21 and +4.30 litre min–1. ODM measured a greater increase in CO after LEA ({Delta}=+1.71 (1.19) litre min–1 (mean (SD)) compared with TD ({Delta}=+0.51 (0.70) litre min–1). We conclude that following LEA, measurements with the Oesophageal Doppler Monitor II overestimate CO and show unacceptably high variability. Blood flow redistribution may limit the value of ODM.

Br J Anaesth 2001; 86: 794–7

Keywords: measurement techniques, cardiac output; measurement techniques, oesophageal Doppler; anaesthetic techniques, epidural


    Introduction
 Top
 Abstract
 Introduction
 Methods
 Results
 Discussion
 References
 
The accuracy of oesophageal Doppler monitoring (ODM)1 of cardiac output (CO) is controversial. Some authors have claimed it to be at least as accurate as thermodilution (TD).2 3 Others have shown that ODM is less accurate than the TD technique, which is the standard clinical method.4 5 The developers of commercially available ODM equipment agree that it gives less accurate absolute measurements than TD, and that the clinician using it should concentrate on trends in CO rather than absolute values.6 Several of the validation publications of the ODM machines show bias plots of relative change but not absolute values.79 Theoretically, ODM should be a relatively non-invasive, cheap and easy means of measuring CO. The technique is easier to master, less expensive and less invasive than the use of a TD catheter, but, training is required to best use the ODM technique.10 11 Variations in cross-sectional area of the descending aorta and the angle between the probe and the descending aorta may not be taken into account.12 13

The ODM II machine (Abbot Laboratories, North Chicago, IL, USA) uses a nomogram to estimate CO from Doppler flow measurements in the descending aorta. To calculate CO, the oesophageal Doppler machine has to take into account the blood flow that it does not measure (i.e. flow to the upper body), and assumes that 30% of CO goes to the upper body. Variations in blood flow distribution may, therefore, lead to errors in CO estimates. Situations involving blood flow redistribution where ODM is inaccurate include haemorrhage13 14 and aortic cross-clamping.12 Lumbar epidural anaesthesia (LEA) increases blood flow to the lower body by sympathetic nerve block, and reduces flow to the upper body by compensatory vasoconstriction caused by increased sympathetic nerve activity.1519 The ODM II machine could overestimate CO in these circumstances.

We measured the accuracy of CO estimation using ODM during LEA as an example of a clinical setting involving redistribution of blood flow between the upper and lower body.


    Methods
 Top
 Abstract
 Introduction
 Methods
 Results
 Discussion
 References
 
The study was approved by the ethical committee of the University Hospitals of the Katholieke Universiteit Leuven. Informed consent was obtained from participating patients.

Fourteen male patients (mean age 65 yr (range 58–73), mean weight 80 kg (range 61–110), mean height 176 cm (range 172–184), ASA class I or II), all undergoing radical prostatectomy under general anaesthesia, were included in the study. A history of severe cardiac disease, respiratory or oesophageal pathology were exclusion criteria. The oesophageal Doppler monitor used for this study (ODM II) emits a continuous 4 MHz ultrasound signal, which is received through a 300 Hz high-pass filter. The blood flow velocity curve is displayed continuously and an on-line display of CO, heart rate and stroke volume is provided.

For the purpose of intra- and post-operative analgesia, a lumbar epidural catheter was inserted at the level of intervertebral space L2–L3 before induction of anaesthesia. Anaesthesia was induced with propofol 1.5–2 mg kg–1, vecuronium 0.1 mg kg–1 and sufentanil 0.2–0.3 µg kg–1, and maintained by means of isoflurane 1.2–1.5 MAC and 65% nitrous oxide with supplemental sufentanil. Following orotracheal intubation, the lungs were mechanically ventilated with a mixture of 35% oxygen and nitrous oxide. A 20-gauge catheter was inserted in the radial artery and a TD catheter was introduced via the right internal jugular vein.

The oesophageal probe was inserted in the standard manner by experienced personnel; optimal positioning was obtained from the appearance and sound of the Doppler signal. Patient age and weight were entered into the ODM console nomogram.

Coupled measurements of CO by TD (triplicate measurements with ice-cold saline) and ODM techniques were performed three times at fixed intervals during the general anaesthetic and then 20, 25 and 30 min after injection of 10 ml bupivacaine 0.25%. Arterial pressure, pulmonary artery pressure, central venous pressure and pulmonary capillary wedge pressure were recorded immediately after each CO determination. An i.v. infusion of 500 ml of a colloid solution (Geloplasma®) was then administered over 30 min in order to prevent severe arterial hypotension.

CO measurements with ODM and TD were compared using Bland-Altman analysis.20 Paired t-tests were used to test the significance of changes in haemodynamic parameters after epidural anaesthesia. A probability of less than 0.05 was considered statistically significant to reject the null hypothesis.


    Results
 Top
 Abstract
 Introduction
 Methods
 Results
 Discussion
 References
 
Upon awakening, all patients were free of pain and had loss of temperature sensation in the corresponding dermatomes.

The epidural anaesthesia reduced arterial pressure (mean arterial pressure from (mean (SD) 74 (12) (range 61–97) to 63 (10) (range 50–79) mm Hg) and central venous filling pressure (from 11 (3) (range 6–17) to 10 (4) (range 4–17) mm Hg) as measured at the moment of peak CO (TD method). Heart rate, pulmonary artery pressure and pulmonary capillary wedge pressure did not change (Table 1).


View this table:
[in this window]
[in a new window]
 
Table 1 Haemodynamic data before and after LEA. ABPs, d, m=systolic, diastolic and mean arterial pressure, CVP=central venous pressure, MPAP=mean pulmonary artery pressure, PCWP=pulmonary capillary wedge pressure, HR=heart rate. Data are mean (SD) [range]. Pressures are in mm Hg, heart rate is in beats min–1. *=P<0.05 vs control. Data displayed post-epidural are obtained from the measurement coinciding with peak CO as measured with TD
 
Measurements after LEA remained stable over the 10-min observation period (no significant difference between variables obtained 20 and 30 min after the onset of LEA respectively; paired t-test).

Before LEA, TD and ODM techniques showed acceptable agreement (mean difference –0.89 litre min–1, limits of agreement –2.67 and +0.88 litre min–1; SE of the mean difference=0.95, SE of the limits of agreement=0.23 (Figs 1 and 2A)). Both ODM and TD CO increased after LEA. However, the increase measured by ODM (from 4.84 (1.26) to 6.54 (1.99) litre min–1; {Delta}=+1.71 (1.19) litre min–1) was greater than that measured using the TD technique (from 5.60 (1.16) to 6.11 (1.10) litre min–1; {Delta}=+0.51 (0.70) litre min–1; P<0.05). During LEA, in contrast to baseline circumstances, ODM overestimated CO (mean difference +0.55 litre min–1, SE of the mean difference=0.29) and displayed unacceptably wide limits of agreement (limits of agreement –3.21 and +4.30 litre min–1, SE of the limits of agreement=0.50) (Figs 1 and 2B).



View larger version (19K):
[in this window]
[in a new window]
 
Fig 1 Scattergram of CO measurements by ODM vs TD before and after LEA. COODM=CO value obtained by ODM; COTD=CO value obtained by TD.

 


View larger version (11K):
[in this window]
[in a new window]
 
Fig 2 Bias plot of CO measurements by ODM vs TD before (A) and after (B) LEA.

 

    Discussion
 Top
 Abstract
 Introduction
 Methods
 Results
 Discussion
 References
 
We observed an acceptable agreement between ODM and TD in baseline conditions, with a negative bias. This slight underestimation of CO by ODM has been reported previously4 12 but not consistently.3 5 10 Compared with several other studies, we found similar limits of agreement.35 10 12

LEA increased CO as measured by both TD and ODM, explained by a reduction in systemic vascular resistance and maintenance of an adequate preload by volume infusion.21

The main result of this study is that ODM overestimates CO during LEA. There is no theoretical reason to believe that lack of agreement between ODM and TD measurements during epidural anaesthesia should be caused by an underestimation of CO by TD. LEA causes blood pooling in the capacitance vessels of the lower limbs,16 but this should not affect the accuracy of the TD method which measures blood flow in the pulmonary circulation.

Movement artefacts are unlikely to cause error considering the type of surgery involved. Variation in cross-sectional area of the aorta has been suggested as an important potential cause of inaccuracy of ODM.12 13 Considering our patient characteristics, errors caused by aortic compliance are unlikely. A decrease in area of the descending aorta could theoretically overestimate CO. The decrease in arterial pressure which we observed could have caused a small decrease in aortic area, but it is very unlikely that it should have caused a 4-fold decrease needed if this were to be the sole cause of error.

Overestimation by ODM is probably caused by the redistribution of blood to the lower body, at the expense of the upper body segment. The fixed distribution assumed by the ODM II suggests that the machine cannot compensate for such variations in blood flow distribution. ODM has been found inaccurate in haemorrhage13 14 and aortic cross-clamping where redistribution is likely. Klotz and co-workers observed an increase in bias between TD and ODM from –0.96 to –1.51 litre min–1 during aortic cross-clamping in six patients.12 Underestimation of CO by ODM during aortic cross-clamping can be explained by blood flow redistribution from the lower to the upper body.

The new generation of ODM devices (such as the Hemosonic 100, Arrow22) measure aortic diameter. Although this may assess any possible inaccuracy of ODM caused by variations in aortic CSA, it will not solve the problem of blood flow redistribution between the upper and lower body.

We propose that the blood flow redistribution in circumstances like LEA, aortic cross-clamping, haemorrhagic and distributive shock can partially explain the inaccuracy of ODM CO measurements, and caution is necessary when using ODM to estimate CO.


    References
 Top
 Abstract
 Introduction
 Methods
 Results
 Discussion
 References
 
1 Side CD, Gosling RJ. Non-surgical assessment of cardiac function. Nature 1971; 232: 335–6[ISI][Medline]

2 Madan AK, UyBarreta VV, Aliabadi-Wahle S, et al. Esophageal Doppler ultrasound monitor versus pulmonary artery catheter in the hemodynamic management of critically ill surgical patients. J Trauma 1999; 46: 607–12[ISI][Medline]

3 Valtier B, Cholley BP, Belot JP, de la Coussaye JE, Mateo J, Payen DM. Noninvasive monitoring of cardiac output in critically ill patients using transesophageal Doppler. Am J Respir Crit Care Med 1998; 158: 77–83[Abstract/Free Full Text]

4 Schmid ER, Spahn DR, Tornic M. Reliability of a new generation transesophageal Doppler device for cardiac output monitoring. Anesth Analg 1993; 77: 971–9[Abstract]

5 Keyl C, Rödig G, Lemberger P, Hobbhahn J. A comparison of the use of transoesophageal Doppler and thermodilution techniques for cardiac output determination. Eur J Anaesthesiol 1996; 13: 136–42[ISI][Medline]

6 Singer M. Esophageal Doppler monitoring of aortic blood flow: beat-by-beat cardiac output monitoring. Int Anesthesiol Clin 1993; 31: 99–125[ISI][Medline]

7 Singer M, Bennett ED. Noninvasive optimization of left ventricular filling using esophageal Doppler. Crit Care Med 1991; 19: 1132–7[ISI][Medline]

8 Singer M, Clarke J, Bennett ED. Continuous hemodynamic monitoring by esophageal Doppler. Crit Care Med 1989; 17: 447–52[ISI][Medline]

9 Singer M, Bennett D. Optimisation of positive end expiratory pressure for maximal delivery of oxygen to tissues using oesophageal Doppler ultrasonography. BMJ 1989; 298: 1350–3[ISI][Medline]

10 Lefrant JY, Bruelle P, Aya AGM, et al. Training is required to improve the reliability of esophageal Doppler to measure cardiac output in critically ill patients. Intensive Care Med 1998; 24: 347–52[ISI][Medline]

11 Krishnamurthy B, McMurray TJ, McClean E. The peri-operative use of the oesophageal Doppler monitor in patients undergoing coronary artery revascularisation. A comparison with the continuous cardiac output monitor. Anaesthesia 1997; 52: 624–9[ISI][Medline]

12 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 1995; 74: 655–60[Abstract/Free Full Text]

13 Wong DH, Watson T, Gordon I, et al. Comparison of changes in transit time ultrasound, esophageal Doppler, and thermodilution cardiac output after changes in preload, afterload, and contractility in pigs. Anesth Analg 1991; 72: 584–8[Abstract]

14 Kamal GD, Symreng T, Starr J. Inconsistent esophageal Doppler cardiac output during acute blood loss. Anesthesiology 1990; 72: 95–9[ISI][Medline]

15 Chamberlain DP, Chamberlain BDL. Changes in the skin temperature of the trunk and their relationship to sympathetic blockade during spinal anesthesia. Anesthesiology 1986; 65: 139–43[ISI][Medline]

16 Arndt JO, Höck A, Stanton-Hicks M, Stühmeier K.-D. Peridural anesthesia and the distribution of blood in supine humans. Anesthesiology 1985; 63: 616–23[ISI][Medline]

17 Taniguchi M, Kasaba T, Takasaki M. Epidural anesthesia enhances sympathetic nerve activity in the unanesthetized segments in cats. Anesth Analg 1997; 84: 391–7

18 Hopf HB, Weissbach B, Peters J. High thoracic segmental epidural anesthesia diminishes sympathetic outflow to the legs, despite restriction of sensory blockade to the upper thorax. Anesthesiology 1990; 73: 882–9[ISI][Medline]

19 Hickey NC, Wilkes MP, Howes D, Watt J, Shearman CP. The effect of epidural anaesthesia on peripheral resistance and graft flow following femorodistal reconstruction. Eur J Vasc Endovasc Surg 1995; 9: 93–6[ISI][Medline]

20 Bland JM, Altman DG. Statistical methods for assessing agreement between two methods of clinical measurement. Lancet 1986; 1: 307–10[ISI][Medline]

21 Wright PMC, Fee JPH. Cardiovascular support during combined extradural and general anaesthesia. Br J Anaesth 1992; 68: 585–9[Abstract]

22 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 1998; 26: 2066–72[ISI][Medline]