Goal-directed therapy as an attempt to optimize cardiac function using intravenous fluid or inotropic support is guided by the mixed venous oxygen saturation and cardiac output (CO) in critically ill patients . This treatment has been shown to improve the outcome of surgical and septic patients [2,3]. Recent standard techniques to assess CO are pulmonary artery thermodilution or transthoracic thermodilution measurement [4,5]. However, these techniques require ICU settings, are cost and time-consuming and lead to relevant morbidity . Noninvasive assessment of CO could be an alternative to avoid missing early haemodynamic patterns or if invasive monitoring devices are not available or contraindicated. Furthermore, noninvasive measurements could support the decision of whether invasive techniques are necessary under special conditions.
Measurement of thoracic electrical bioimpedance (TEB) to estimate CO has been controversially debated before [7–9]. A recent algorithm developed by Bernstein and Osypka interprets the maximum rate of change in TEB as the ohmic equivalent of aortic blood flow acceleration. The change in electrical conductivity caused by the orientation of erythrocytes depending on the cardiac function cycle is termed electrical velocimetry . Several studies validated this new measurement technique. Comparisons between electrical velocimetry-derived CO (electrical velocimetry CO) and invasive assessment using transoesophageal Doppler echocardiography or thermodilution techniques showed a clinically acceptable agreement between the data assessed [10–13]. The agreement between electrical velocimetry CO and the pulmonary artery thermodilution value was high even in haemodynamically unstable patients . An observational trial found comparable electrical velocimetry CO and thermodilution values in 50 critically ill patients with a high percentage error but clinically acceptable agreement . In contrast to this, in another trial on cardiac surgery, the compared techniques disagreed considerably .
To our knowledge, electrical velocimetry CO has never been investigated in septic patients before. The aim of the present study was to compare electrical velocimetry and transthoracic thermodilution CO estimation in septic patients after major general surgery.
The study protocol was approved by the ethics committee at Charité University Medicine Berlin. Inclusion criteria were ongoing severe systemic inflammatory response syndrome (SIRS) or sepsis as defined by the International Sepsis Definition Conference 2001 with manifest haemodynamic instability and therefore an indication for advanced cardiac monitoring. Patients who underwent major pulmonary resection or aortic bypass grafting were excluded because transthoracic thermodilution measurement is potentially not reliable. For the same reason, patients with manifest atrial fibrillation were excluded.
All patients treated on the surgical ICU of the Department of General, Visceral, Vascular and Thoracic Surgery, Charité University Medicine Berlin Campus Mitte, who met the inclusion criteria, were assessed, and written informed consent was obtained before the measurements from every patient or his/her legal guardian.
Patients were considered haemodynamically unstable if relevant (dobutamine >2 μg kg−1 min−1) or increasing inotropic support or any vasopressors were required to keep the mean arterial blood pressure (MAP) greater than 60 mmHg or if there were strong clinical hints (e.g. good responsiveness on fluid challenges or a clearly increasing MAP after leg raising) for a low cardiac preload caused by fluid losses.
CO measurements by thoracic electrical bioimpedance/electrical velocimetry (Aesculon, Osypka Medical, Berlin, Germany) and transthoracic thermodilution (PiCCO, Pulsion Medical Systems AG, Munich, Germany) were obtained immediately after insertion of the thermodilution catheter. This catheter was regularly inserted into the femoral artery. In two cases with inguinal soft-tissue infection or known high-grade femoral atherosclerosis, an axillary access was chosen. A central venous catheter was placed in the superior vena cava in all patients. The correct position of the tip was assessed using intracardiac ECG during the insertion and an additional radiograph after admission to the ICU. The measurement of CO by the transpulmonary thermodilution technique was based on the Stewart–Hamilton equation. The temperature profile was measured in a peripheral artery after injection of cold indicator solution into the right atrium and passing the pulmonary circulation. The detailed technique is described elsewhere and its comparability against the pulmonary artery thermodilution technique using a heated pulmonary artery catheter has been previously demonstrated [16–18]. To calibrate the pulse wave analysis, three sequential thermodilution measurements with 15 ml iced saline solution were performed, averaged and used for comparison with the simultaneously obtained bioimpedance data in this study.
The measurements of CO using transthoracic electrical bioimpedance/electrical velocimetry (electrical velocimetry CO) were performed according to the instructions of the manufacturer. The approach of electrical velocimetry interprets the maximum rate of change in TEB as the ohmic equivalent of mean aortic blood flow acceleration. The change in electrical conductivity caused by alignment of erythrocytes at the opening of the aortic valve and the volumetric changes in the major thoracic blood vessels depending on the cardiac function cycle is termed electrical velocimetry. The detailed algorithm is reported elsewhere . Two standard ECG electrodes were placed on one side of the neck along the carotid artery. Another two electrodes were fixed in the left mid-axillary line on the level of the xiphoid. The mean of 10 successive CO values was displayed on the monitor as electrical velocimetry CO.
Data collection and analysis
Statistical analysis was performed with SPSS 15.0 and SAS 9.13 for Windows. Following analysis for normal distribution, Bland–Altman statistics were calculated on the raw and relative data . The unknown true value of CO was estimated by calculating the mean of thermodilution CO and electrical velocimetry CO. For the latter calculation, bias was calculated as the mean difference between thermodilution CO and electrical velocimetry CO. The upper and lower limits of agreement were calculated as bias ± 1.96 SD and defined the expected range including 95% of the differences between both methods. According to Critchley and Critchley , the percentage error was calculated as 2SDbias/mean thermodilution CO. A percentage error of less than 30% was considered to be clinically acceptable agreement.
Simultaneous measurements of thermodilution CO and electrical velocimetry CO were obtained in 30 critically ill patients during their treatment on a surgical ICU over a 10-month period. Sepsis mostly resulted from peritonitis caused by perforation of gastrointestinal organs [n = 10 (33%)], leakage of colonic anastomoses [7 (23%)], ileus [6 (20%)], acute respiratory distress syndrome (ARDS) after acute necrotizing pancreatitis [1 (3%)] and blunt thoracic trauma [1 (3%)] or severe soft-tissue infection [2 (7%)]. Two (7%) patients who developed haemodynamical instability after large multivisceral operations combined with hyperthermic intraperitoneal chemotherapy for peritoneal carcinosis were enrolled in this study. One patient with unstable SIRS after oesophagectomy was also included. In these three patients, the infection was clinically obvious but could not be proven microbiologically. The in-hospital mortality was 7% (Table 1). The extravascular lung water index (ELWI) was increased in 11 (37%) patients. At the same time, in 13 (43%) patients there was an elevated intrathoracic blood volume index, indicating an increased cardiac preload and possible hyperhydration. The severity of illness was documented using the Acute Physiology and Chronic Health Evaluation score-II (APACHE-II), Sequential Organ Failure Assessment score (SOFA) and Simplified Acute Physiology score-II (SAPS-II) before every measurement (Table 2). During all measurements, patients were calm or sedated and were in the horizontal position. Twenty-nine (97%) patients had not been given curare but were sedated and received mechanical ventilation support. We used a pressure-supported ventilation mode with a moderate positive end-expiratory pressure (PEEP; 5–10 mmHg). There were no complications (i.e. thromboembolism) caused by the monitoring devices.
Thermodilution CO and electrical velocimetry CO was −0.3 l min−1 (−58.9 to 49.4% of the mean value) with a precision of ±1.9 l min−1. The lower and the upper limits of agreement were −4.1 and 3.5 l min−1, respectively (Fig. 1). The percentage error between both techniques was 54%.
The CO in septic patients obtained with the transthoracic thermodilution technique and with the electrical velocimetry was compared in the present study, which took place in a typical surgical ICU setting with different underlying reasons for septic conditions.
The results demonstrate a poor correlation between values of both techniques. The analysis of agreement according to the method described by Bland and Altman  leads to a small bias of −0.3 l min−1 with lower and upper limits of agreement of −4.1 and 3.5 l min−1. The percentage error between both measuring techniques found in the current study was 54%, which is considerably higher than the clinically acceptable percentage error of up to 30% as defined by Critchley and Critchley . Consequently, the interchangeable use of transthoracic thermodilution and electrical velocimetry measurement of CO in critically ill patients cannot be recommended until further research with larger sample sizes is available.
Estimation of stroke volume and CO is essential for goal-directed treatment of haemodynamically unstable patients. Extended monitoring may be of special importance after major surgery and in patients with evolving sepsis [1,5]. Very early attempts to resuscitate and optimize tissue oxygenation can limit tissue hypoxia or reverse the progression to organ dysfunction and can lead to a relevant improvement in outcome [2,3,20]. Benefits of goal-directed therapy in patients with ongoing severe sepsis are not proven but have been supported by several authors .
The ideal monitoring device would be reliable in a wide spectrum of patients' conditions, validated, without any adverse effects or causes of morbidity, easy to use and inexpensive. Clearly, this tool is still missing.
The use of highly invasive techniques for assessment of cardiac function is debated controversially because of their uncertain risk–benefit ratio [6,22]. Less or noninvasive measurement techniques are explored increasingly to minimize the risks related to the intervention. A noninvasive monitoring technique estimating CO or stroke volume may be an acceptable alternative when invasive monitoring is not available .
Estimating CO from the change in TEB seems to be a very fast and easy method. Historically, the agreement between TEB and thermodilution in critically ill or septic patients was poor [7–9]. Susceptibility to distortions of the underlying equation algorithms may be responsible for inconsistent agreements . These basic equation models were modified and the technique of electrical velocimetry was developed by Bernstein to increase their reliability . In recent studies using electrical velocimetry for measurement of cardiac function, the agreement between the compared techniques varied considerably. In a porcine model, a good correlation but unacceptable agreement with transthoracic thermodilution was found, but special anatomical conditions in the animal model used may have had a relevant impact on TEB values . Two studies measuring CO during and immediately after cardiac surgery [11,15] found only a poor agreement or even contradictions between electrical velocimetry and the pulmonary artery thermodilution technique. The number of patients was small in both studies (n = 16 and n = 29, respectively). Good agreement was achieved only immediately after induction of anaesthesia . In contrast to these studies, Suttner et al.  found a close correlation and acceptable agreement between electrical velocimetry and pulmonary artery thermodilution during cardiac surgery in 74 patients and supported the results of the study by Bernstein and Lemmens . Schmidt et al.  compared electrical velocimetry and transoesophageal Doppler echocardiography for estimation of CO during coronary artery surgery. They found a good agreement after induction of anaesthesia and even stated both methods are interchangeable. The comparison of electrical velocimetry and CO derived by pulmonary artery catheter showed results contrary to the comparison with Fick's principle in children with congenital heart disease [26,27]. Only one published clinical study compared electrical velocimetry with the transthoracic thermodilution technique (PiCCO). In this trial, Zoremba et al.  stated that both techniques were comparable with clinically sufficient accuracy in 25 postsurgical ICU patients. Two further trials comparing electrical velocimetry and PiCCO are as yet available only as congress abstracts that do not allow a detailed evaluation. However, both studies found good agreement between the measuring methods [28,29].
As yet, to our knowledge, there are no published investigations validating electrical velocimetry in septic postoperative patients.
There are several limitations to the current study that possibly explain the lack of agreement between thermodilution CO and electrical velocimetry CO. First, the sample size is quite small and the study included a heterogeneous group of patients with different causes of sepsis and different haemodynamic conditions. However, this population reflects the common situation on a surgical ICU.
Second, TEB is remarkably influenced by the fluid content of the lung. Severe oedema or effusion may lead to distorted measurements . In fact, in the present study, 11 patients (37%) had an increased ELWI. An analysis of subgroups with normal or elevated ELWI did not reveal different results for agreement of electrical velocimetry CO and thermodilution CO. Percentage errors were 53 and 54%, respectively, and the absolute values of ELWI did not correlate with the difference in the mean CO assessed with both methods. Although the electrical velocimetry equation focuses on the aortic blood flow, changes in conductivity of the lungs may influence values. Also major thoracic surgery such as Ivor–Lewis oesophagectomy and mechanical ventilation with PEEP might have a relevant impact on TEB values . The special conditions of septic patients with fluid shifts, disordered ELWI and frequent capillary leak syndrome may explain the incongruent results of our trial and of the three studies comparing thermodilution CO and electrical velocimetry CO mentioned above. Unfortunately, the detailed characteristics of enrolled patients are not published in these studies [14,28,29].
Further, variables such as positioning of electrodes, poor electrode skin contact and extreme heart rates that may interfere with the accuracy of TEB were minimized in this study by following the instructions for use of the electrical velocimetry device exactly. All measurements were carried out in calm or sedated patients with exclusion of interference by muscular movement. There were no problems in placing electrodes and no interference from shivering or electrical artefacts were noticed.
Further limitations for the clinical use are related to the measuring techniques. In severe sepsis, rapid changes in intravascular and extravascular fluid can be expected . Information about cardiac preload volumes or at least filling pressures is necessary for adequate goal-directed clinical treatment. These data cannot be assessed with electrical velocimetry. Important data for the treatment of septic patients are not available without further monitoring devices.
Furthermore, the transthoracic thermodilution technique may overestimate CO, especially in low-output conditions caused by heart failure . Therefore, the use of only one technique as a gold standard has been criticized . Values may vary depending on temperature of injectate, speed of injection, respiratory cycle and position of the central venous catheter or arterial thermistor. To reduce these influences, in the current study, the tip of the central venous catheter was positioned in all cases in the superior vena cava and all measurements were made meticulously by a well experienced investigator at the same time of the respiratory cycle. Nevertheless, the transthoracic thermodilution technique for estimating CO has proven its reliability against the transcardiopulmonary thermodilution technique as a second standard technique for assessment of cardiac function .
In conclusion, transthoracic bioimpedance/electrical velocimetry is a simple to use and completely noninvasive technique for measuring CO. In contrast to previously published trials, the agreement between measuring CO using thermodilution and electrical velocimetry technique was poor in the present study, representing the typical setting of a surgical ICU. The percentage error of 54% was clinically unacceptably high. Therefore, electrical velocimetry and PiCCO should not be used interchangeably in patients with persistent sepsis, until further studies with larger sample sizes are available.
The present work was supported by Osypka Medical GmbH, Berlin, by providing the electrical velocimetry monitoring system.
1 Pearse R, Dawson D, Fawcett J, et al
. Early goal-directed therapy after major surgery reduces complications and duration of hospital stay. A randomised, controlled trial [ISRCTN38797445]. Crit Care 2005; 9:R687–R693.
2 Lobo SM, Lobo FR, Polachini CA, et al
. Prospective, randomized trial comparing fluids and dobutamine optimization of oxygen delivery in high-risk surgical patients [ISRCTN42445141]. Crit Care 2006; 10:R72.
3 Rivers E, Nguyen B, Havstad S, et al
. Early goal-directed therapy in the treatment of severe sepsis
and septic shock. N Engl J Med 2001; 345:1368–1377.
4 Marik PE, Baram M. Noninvasive hemodynamic monitoring in the intensive care unit. Crit Care Clin 2007; 23:383–400.
5 Ahrens T. Hemodynamics in sepsis
. AACN Adv Crit Care 2006; 17:435–445.
6 Sandham JD, Hull RD, Brant RF, et al
. A randomized, controlled trial of the use of pulmonary-artery catheters in high-risk surgical patients. N Engl J Med 2003; 348:5–14.
7 Critchley LA, Critchley JA. A meta-analysis of studies using bias and precision statistics to compare cardiac output
measurement techniques. J Clin Monit Comput 1999; 15:85–91.
8 Genoni M, Pelosi P, Romand JA, et al
. Determination of cardiac output
during mechanical ventilation by electrical bioimpedance or thermodilution
in patients with acute lung injury: effects of positive end-expiratory pressure. Crit Care Med 1998; 26:1441–1445.
9 Imhoff M, Lehner JH, Lohlein D. Noninvasive whole-body electrical bioimpedance cardiac output
and invasive thermodilution cardiac output
in high-risk surgical patients. Crit Care Med 2000; 28:2812–2818.
10 Bernstein DP, Lemmens HJ. Stroke volume equation for impedance cardiography
. Med Biol Eng Comput 2005; 43:443–450.
11 Mekis D, Kamenik M, Starc V, Jeretin S. Cardiac output
measurements with electrical velocimetry in patients undergoing CABG surgery: a comparison with intermittent thermodilution
. Eur J Anaesthesiol 2008; 25:237–242.
12 Schmidt C, Theilmeier G, Van Aken H, et al
. Comparison of electrical velocimetry and transoesophageal Doppler echocardiography for measuring stroke volume and cardiac output
. Br J Anaesth 2005; 95:603–610.
13 Suttner S, Schollhorn T, Boldt J, et al
. Noninvasive assessment of cardiac output
using thoracic electrical bioimpedance in hemodynamically stable and unstable patients after cardiac surgery: a comparison with pulmonary artery thermodilution
. Intensive Care Med 2006; 32:2053–2058.
14 Zoremba N, Bickenbach J, Krauss B, et al
. Comparison of electrical velocimetry and thermodilution
techniques for the measurement of cardiac output
. Acta Anaesthesiol Scand 2007; 51:1314–1319.
15 Heringlake M, Handke U, Hanke T, et al
. Lack of agreement between thermodilution
and electrical velocimetry cardiac output
measurements. Intensive Care Med 2007; 33:2168–2172.
16 Della RG, Costa MG, Pompei L, et al
. Continuous and intermittent cardiac output
measurement: pulmonary artery catheter versus aortic transpulmonary technique. Br J Anaesth 2002; 88:350–356.
17 Sakka SG, Reinhart K, Meier-Hellmann A. Comparison of pulmonary artery and arterial thermodilution cardiac output
in critically ill patients. Intensive Care Med 1999; 25:843–846.
18 Sakka SG, Reinhart K, Wegscheider K, Meier-Hellmann A. Is the placement of a pulmonary artery catheter still justified solely for the measurement of cardiac output
? J Cardiothorac Vasc Anesth 2000; 14:119–124.
19 Bland J, Altman D. Statistical methods for assessing agreement between two methods of clinical measurement. Lancet 1986; 1:307–310.
20 Levy MM, Macias WL, Vincent JL, et al
. Early changes in organ function predict eventual survival in severe sepsis
. Crit Care Med 2005; 33:2194–2201.
21 Hadian M, Pinsky MR. Functional hemodynamic monitoring. Curr Opin Crit Care 2007; 13:318–323.
22 Hadian M, Pinsky MR. Evidence-based review of the use of the pulmonary artery catheter: impact data and complications. Crit Care 2006; 10(Suppl. 3):8.
23 Shoemaker WC, Belzberg H, Wo CC, et al
. Multicenter study of noninvasive monitoring systems as alternatives to invasive monitoring of acutely ill emergency patients. Chest 1998; 114:1643–1652.
24 de Waal EE, Konings MK, Kalkman CJ, Buhre WF. Assessment of stroke volume index with three different bioimpedance algorithms: lack of agreement compared to thermodilution
. Intensive Care Med 2008; 34:735–739.
25 Osthaus WA, Huber D, Beck C, et al
. Comparison of electrical velocimetry and transpulmonary thermodilution
for measuring cardiac output
in piglets. Paediatr Anaesth 2007; 17:749–755.
26 Norozi K, Beck C, Osthaus WA, et al
. Electrical velocimetry for measuring cardiac output
in children with congenital heart disease. Br J Anaesth 2008; 100:88–94.
27 Tomaske M, Knirsch W, Kretschmar O, et al
. Cardiac output
measurement in children: comparison of Aesculon cardiac output
monitor and thermodilution
. Br J Anaesth 2008; 100:517–520.
28 Lindner P, Mellert F, Haushofer M, et al
. Noninvasive cardiac output
measurement in cardiovascular ICU-patients – comparison of conventional impedance cardiography
, electrical velocimetry and thermodilution
methods. Thorac Cardiovasc Surg 2007; 55:A64, doi: 10.1055/s-2007-967607.
29 Faltlhauser A, Weber T, Thomas A. Electrical velocimetry – clinical validation of a new noninvasive CO-monitor in postoperative non cardiac surgical patients [abstract].: 234-S. Critical Care Medicine. Society of Critical Care Medicine 35th Critical Care Congress: San Francisco, California, USA; 7–11 January 2006; vol. 33.
30 Critchley LA, Critchley JA. Lung fluid and impedance cardiography
. Anaesthesia 1998; 53:369–372.
31 Marx G. Fluid therapy in sepsis
with capillary leakage. Eur J Anaesthesiol 2003; 20:429–442.
32 van Grondelle A, Ditchey RV, Groves BM, et al
method overestimates low cardiac output
in humans. Am J Physiol 1983; 245:H690–H692.
33 Yung GL, Fedullo PF, Kinninger K, et al
. Comparison of impedance cardiography
to direct Fick and thermodilution cardiac output
determination in pulmonary arterial hypertension. Congest Heart Fail 2004; 10:7–10.