British Journal of Anaesthesia 1997; 79: 514–516
Overestimation of low cardiac output measured by thermodilution
J. P. TOURNADRE, D. CHASSARD AND R. MUCHADA
Summary
Materials and methods
We have investigated the influence of a cold water
bolus (CWB) injection on overestimation of cardiac
output (CO) in low CO states in anaesthetized
dogs. CO was measured using three methods:
(1) thermodilution (TD), (2) electromagnetic (EM)
flow meter placed on the pulmonary artery and
(3) transoesophageal echo-Doppler (OD) placed on
the descending aorta. Measurements of CO were
obtained before (steady state) and after induction
of a low CO state with thiopentone 5 mg kg91 i.v.
After CWB injection, mean CO measured by EM
and OD increased by 26% and 27%, respectively
(P0.05) during steady state, and by 85% and 75%
(P0.05) during the low CO state. This transient
increase was produced by an increase in stroke
volume, while heart rate did not change. Frank
Starling’s law may explain this variation by a
sudden increase in preload produced by CWB
injection. These results indicate that thermodilution overestimated CO during low CO states
when CWB injection was used. (Br. J. Anaesth.
1997; 79: 514–516).
After approval from our institutional Animal and
Care Committee, six dogs (weight 18 (SD 5) kg)
were anaesthetized with thiopentone 5 mg kg91,
vecuronium 0.08 mg kg91 and phenoperidine 0.015
mg kg91. The trachea was intubated and the lungs
ventilated mechanically (Bird Mark 8, Products
Corporation, Palm Springs, CA, USA) at an FIO2 of
1, ventilatory frequency 20 bpm and tidal volume 10
ml kg91. Anaesthesia was maintained with
phenoperidine 0.02 mg kg91 h91 and vecuronium
0.04 mg kg h91.
A Swan–Ganz catheter was inserted into the
pulmonary artery via the right external jugular vein
under x-ray control (7F diameter, Edwards
Laboratories, Irvine, CA, USA). Thermodilution
CO was measured using triplicate injections of 5 ml
of 5% dextrose at 2 ⬚C (cold water bolus (CWB))
during end-expiration.
A 12-mm EM flow probe (In Vivo Metric,
Healsburg, CA, USA) was placed around the pulmonary artery after dissection of the pulmonary
trunk via a left-sided thoracotomy. Pulmonary blood
flow (PBF) was measured continuously with an EM
blood flow meter (FM 501, Carolina Medical
Electronics, King, NC, USA).
A special echo-Doppler probe, described
previously,6 was used to measure aortic blood flow
(ABF) using a transoesophageal echo-Doppler (OD)
ultrasound method (Dynemo 3000, Sometec, Paris,
France). ABF, PBF, ECG and heart rate (HR) were
recorded simultaneously and continuously before,
during and after CWB injection.
Measurements were made with both continuous
methods (PBF, ABF) after 15 min of steady state
following surgical preparations and were repeated
after injection of thiopentone 5 mg kg91 i.v. to obtain
a low CO state.
Instantaneous blood flow variations were
measured by EM flow meter and OD methods
during CWB injection. These measures were
integrated by planimetry to calculate the average
stroke volume (SV) on seven cardiac cycles. These
Key words
Heart, cardiac output. Measurement techniques, cardiac
output.
Measurement
techniques,
thermodilution.
Measurement
techniques,
transoesophageal
Doppler.
Measurement techniques, flowmetry. Dog.
Several methods have been used to measure cardiac
output (CO) in humans. The thermodilution
method (TD), described initially by Fegler,1 has
been used for 20 yr since its development with the
catheter described by Swan and Ganz.2 Several
studies have demonstrated good overall agreement
between TD and other methods estimating CO.
However, several studies have shown that the TD
method overestimates CO in the low range.3–5 This
was explained by heat loss during measurement.3
In our clinical anaesthetic practice, we observed in
patients in whom CO was recorded simultaneously
by TD and transoesophageal echo-Doppler (OD),
an increase in CO measured by OD, immediately
after injection of an iced water bolus. Therefore, this
study was designed to investigate the effect of an iced
temperature injectate on CO measured by TD, OD
and by an electromagnetic (EM) flow meter placed
on the pulmonary artery of anaesthetized dogs
during normal and low CO states.
J. P. TOURNADRE*, MD, D. CHASSARD, MD, PHD, Critical Care
Division, Hôpital de l’Hôtel-Dieu, 69002 Lyon, France. R.
MUCHADA, MD, Critical Care Division, Clinique Mutualiste E.
André, 69003 Lyon, France. Accepted for publication: June 3,
1997.
*Address for correspondence: Département d’AnesthésieRéanimation, Hôpital de l’Hôtel-Dieu, 69002 Lyon, France.
Thermodilution and low cardiac output
515
were identical to cycles sampled during inscription
of the TD curve.
Data are expressed as mean (SD) and were
compared by Wilcoxon’s test between the two states
and by one-way analysis of variance (ANOVA) for
haemodynamic data obtained by the different
methods. Results were considered statistically
significant at P0.05.
Results
Results are summarized in table 1.
In all cases injection of thiopentone 5 mg kg91 produced a decrease in HR and SV. For all measurement techniques there was a decline in CO after
thiopentone (P0.05). HR remained unchanged
throughout the course of injection of the CWB using
the TD method. CO values obtained by TD were
higher than those obtained by the two other methods
(P0.05). The differences recorded with the other
methods were more important during low CO states.
Table 1 Mean (SD) pulmonary blood flow (PBF), aortic blood
flow (ABF) and cardiac output (CO) (measured by the
thermodilution method) at steady state and after injection of
thiopentone 5 mg kg91 (low CO state). *P0.05 vs before
injection of the cold water bolus (CWB); †P0.05 vs steady
state
Steady state
91
HR (beat min )
PBF (ml min91)
Before CWB
After CWB
ABF (ml min91)
Before CWB
After CWB
OD (ml min91)
Low CO state
137 (22)
120 (19)†
1583 (253)
1988 (430)*
1086 (297)†
2035 (378)*
1382 (294)
1722 (391)*
2067 (466)
947 (287)†
1682 (281)*
1847 (315)†
Figure 1 Effect of a cold water bolus (iced temperature)
injection on stroke volume measured in the descending aorta and
pulmonary artery in a dog with a low cardiac output.
CWB injection induced significant increases in PBF
and ABF (P0.05) (fig. 1). These increases were
found in the two states but were proportionally more
important during the low CO state.
Differences between the results of TD and those
of pulmonary flow meter and aortic OD obtained
after injection of CWB were not statistically significant. After CWB injection, CO, measured using EM
and OD, increased by 26% and 27%, respectively,
during steady state (P0.05), and by 85% and 75%
during low CO state (P0.05).
Discussion
Data from previous studies showed that the TD
method overestimated measurement of CO in
patients with low CO states. This could explain the
weak correlation between measurements by TD and
other methods, such as the Fick method, during low
CO conditions,7 8 dye dilution techniques9 and EM
flow meter on the pulmonary artery.10 These studies
attributed differences to heat loss under conditions
of low flow.3–5 11–13 Our study did not evaluate this
phenomenon during high output states as previous
studies found a good correlation between TD and
other methods under these conditions.2 3
In the course of the two CO states evaluated in our
study (steady and low CO state), TD overestimated
CO values compared with EM flow meter and OD
(note that these two methods are performed in two
different anatomical sites, the pulmonary artery and
descending aorta, respectively). This was more
significant in the presence of a low CO state. ABF
was lower by 13–17% compared with PBF under all
conditions. This can be explained by the fact that
measurements at this level of the aorta do not take
into account blood flow to the upper extremities,
head, bronchial tubes and cardiac areas.
The increase in aortic flow was lower than that of
pulmonary flow (75% and 85%, respectively,
P0.05) during the low CO state. This may reflect a
septal shift in response to substantial increases in
right ventricular end-diastolic volume that might
transiently decrease left ventricular end-diastolic
volume because of the restrictive effects of the
pericardium. Thus aortic flow would be lower
than pulmonary flow.
We found that CO, when measured by EM or
OD, increased in all cases after administration of
CWB injection. This transient increase was produced by an increase in SV over several cardiac
cycles, while HR remained unchanged. During rapid
injection of CWB into the right atrium, a transient
increase in CO could increase SV by increasing preload. Frank Starling’s law may explain this variation
by a sudden increase in preload after CWB injection.
This also explains why the phenomenon is more
significant at low compared with normal CO states.
The increase in SV after injection of the CWB
occurred without a change in HR. This may have
resulted from the use of thiopentone. Its effect on the
autonomic nervous system14 prevents the sinus
bradycardia caused by CWB injection described by
Harris and colleagues15 and Nishikawa and Dohi16 17
under similar conditions. Because CO:SVHR, a
516
decrease in HR may balance the increase in SV. This
may explain the good correlation between CO
determination with TD and other methods, even in
low CO states.7 18–20 However, when HR is
unchanged, especially in low CO states, the transient
increase in SV could be responsible for discrepancies
between TD and other methods of measurement of
CO.3 8 11 21
With the TD method,19 CO output is measured in
the course of the first cardiac cycles after injection of
the CWB. We found that SV increased after
injection of the CWB compared with basal values.
Thus the calculated output is overestimated. It is
remarkable to note that when we measured outputs
by EM and OD after CWB injection, we found
values close to those obtained by TD during
steady state or low CO. One can therefore estimate
that, in our study, CWB injection induced the
major part of the overestimation. The potential
human implications of these findings need to be
verified.
We used iced temperature injectates rather than
room temperature injectates, as first described by
Swan and Ganz.2 To eliminate the influence of
temperature induced by cardiac reflexes,16 17 a study
should be performed under the same conditions.
Accurate measurement of CO is an important goal
in the management of anaesthetized or critically ill
patients with compromised cardiac function. Our
findings showed that thermodilution overestimated
CO during low CO states when iced temperature
injectates were used. This suggests that other noninvasive methods are necessary to determine CO in
patients with low output states.
Acknowledgements
We thank Dr Fraser and Ilmars Lidums (Adelaide, Australia) for
help in the preparation of the manuscript.
References
1. Fegler G. Measurement of cardiac output in anaesthetized
animals by a thermodilution method. Quaterly Journal of
Experimental Physiology 1954; 39: 153–157.
2. Ganz W, Swan HJC. Measurement of blood flow by thermodilution. American Journal of Cardiology 1972; 29: 241–246.
3. Van Grondelle A, Ditchey RV, Groves BM, Wagner WW jr,
Reeves JT. Thermodilution method overestimates low
cardiac output in humans. American Journal of Physiology
1983; 245: H690–692.
British Journal of Anaesthesia
4. Hayes BE, Will JA, Zarnstorff WC, Bisgard GE. Limitations
of thermodilution cardiac output measurements in the rat.
American Journal of Physiology 1984; 246: H754–760.
5. Kissling G, Ross C, Brändle M. Validity of thermal dilution
technique for measurement of cardiac output in rats.
American Journal of Physiology 1993; 265: H1007. H1013.
6. Muchada R, Cathignol D, Lavandier B. Aortic blood flow
measurement. American Journal of Noninvasive Cardiology
1988; 2: 24–31.
7. Machecourt J, Bertrand B, Polidori C. Mesure du débit
cardiaque par méthodes invasives: méthodes et variabilité.
Annales de Cardiologie et d’Angiologie 1992; 41: 237–245.
8. Lynch J, Kaemmerer H. Comparison of a modified Fick
method with thermodilution for determining cardiac output
in critically ill patients on mechanical ventilation. Intensive
Care Medicine 1990; 16: 248–251.
9. Weisel RD, Berger RL, Hechtman HB. Measurement of
cardiac output by thermodilution. New England Journal of
Medicine 1975; 27: 682–684.
10. Renner LE, Morton MJ, Sakuma GY. Indicator amount,
temperature, and intrinsic cardiac output affect thermodilution cardiac output accuracy and reproducibility. Critical
Care Medicine 1993; 21: 586–597.
11. Wallace DC, Winslow EH. Effects of iced and room temperature injectate on cardiac output measurements in critically ill
patients with low and high cardiac outputs. Heart and Lung
1993; 22: 55–63.
12. Davies G, Hess D, Jebson P. Continuous Fick cardiac output
compared to continuous pulmonary artery electromagnetic
flow measurement in pigs. Anesthesiology 1987; 66: 805–809.
13. Maruschak GF, Potter AM, Schauble JF, Rogers MC.
Overestimation of pediatric cardiac output by thermal
indicator loss. Circulation 1982; 65: 380–383.
14. Harvey SC. Hypnotics and sedatives, the barbiturates. In:
Goodman LS, Gillman A, eds. The Pharmacological Basis of
Therapeutics. New York: MacMillian, 1980; 59–61.
15. Harris AP, Miller CF, Beattie C, Rosenfeld GI, Rogers MC.
The slowing of sinus rhythm during thermodilution cardiac
output determination and the effect of altering injectate
temperature. Anesthesiology 1985; 63: 540–541.
16. Nishikawa T, Dohi S. Slowing of heart rate during cardiac
output measurement by thermodilution. Anesthesiology 1982;
57: 538–539.
17. Nishikawa T, Dohi S. Haemodynamic changes associated
with thermodilution cardiac output determination during
myocardial ischaemia or pulmonary oedema in dogs. Acta
Anaesthesiologica Scandinavica 1992; 36: 679–683.
18. Olsson B, Pool J, Vandermoten P, Varnauskas E, Wassen R.
Validity and reproducibility of determination of cardiac
output by thermodilution in man. Cardiology 1970; 55:
136–148.
19. Ganz W, Donoso R, Marcus HS, Forrester JS, Swan HJC. A
new technique for measurement of cardiac output by thermodilution in man. American Journal of Cardiology 1971; 27:
392–439.
20. Kashtan HI, Maitland A, Salermo TA, Lichtenstein SV,
Byrick RJ. Effect of tricuspid regurgitation on thermodilution
cardiac output: studies in an animal model. Canadian Journal
of Anaesthesia 1987; 34: 246–251.
21. Woog RH, McWilliam DB. A comparison of methods of
cardiac output measurement. Anaesthesia and Intensive Care
1983; 11: 141–146.