Until now no study has established which continuous cardiac output (CCO) monitoring system can deliver reliable and reproducible haemodynamic data essential for early detection of haemodynamically unstable patients during off-pump cardiac surgery.
Monitoring techniques available to the clinician include right heart catheterization , transoesophageal Doppler cardiac output (CO) monitoring , non-invasive cardiac output (NICO, Novametrix®) , aortic transit-time ultrasound , impedance cardiography , echocardiography  and the PICCO .
Repeated thermodilution cardiac output (TDCO) can be used before and during heart displacement to obtain CO measurements in off-pump cardiac surgery. However, because invasive CO monitoring can lead to serious complications, pulmonary artery pressure catheters (PACs) have not yet been fully accepted as a part of routine clinical management [8,9]. Concerns about limitations of TDCO as a standard monitoring system during off-pump cardiac surgery have also been expressed by Grow and colleagues  who state that an ideal solution would be a reliable, rapidly responding CCO monitoring system. Schulz and colleagues  conclude that intraoperative use of CCO measurements during open heart surgery is clinically indicated. However, the choice of a time-interval and the number of intermittent bolus thermodilution measurements to obtain reliable results for a trend of CO has not yet been subject of any international consensus. Therefore, it is justified to use CCO with a fixed time interval and standardized measurement software for CO measurements in order to compare data between institutions. A PAC for CCO measurements (PAC-CCO) offers another parameter, namely the SvO2, which is commonly monitored during off-pump coronary artery bypass (OPCAB) surgery and may correlate with CO allowing detection of sudden haemodynamic changes .
The NICO-monitoring system is based on the Fick-principle and on partial rebreathing of carbon dioxide (CO2)  and makes non-invasive CCO measurements possible. This system has already been explained in previous publications [4,12]. Strengths of the NICO are that it is less invasive, not operator-dependent and delivers continued CO measurements without delay. A downside of this technique is that it can only be applied in intubated and ventilated patients. NICO measures the pulmonary capillary blood flow (PCBF), which corresponds to the non-shunted CO. As the shunt fraction differs between patients, it is necessary to estimate it in each individual case. Shunt variations are estimated on the bases of the SPO2 and FiO2 variations. In order to obtain the CO, PCBF has to be multiplied with the shunt fraction.
In PAC-CCO measurements a pseudo-random binary sequence of heat created by a 10 cm thermal filament is transferred into the surrounding blood. The change of temperature is detected further downstream in the pulmonary artery and integrated into the CCO software with the input sequence to produce a thermodilution curve and calculate CO [13–15]. Until now, NICO has not been compared to PAC-CCO during off-pump cardiac surgery. It is also unknown which of both CO-measurement systems pick up more rapidly acute haemodynamic changes.
The aim of this prospective study was to compare CCO measurements of the NICO system with CCO of the PAC during OPCAB surgery.
Material and methods
After the hospital Ethics Committee (University Hospital of Brest, France) agreed to the study protocol, 22 patients enrolled for off-pump coronary surgery were informed and consented prior to the study. Preoperative data of patients are summarized in Table 1. An absolute inclusion criterion was the insertion of a PAC for CCO and SvO2 measurements. Exclusion criteria were the presence of valvular disease (tricuspid insufficiency), aortic stenosis, intracardiac shunt, or ascending aortic aneurysm and non-sinus rhythm.
Prior to anaesthetic induction, a radial arterial catheter was inserted under local anaesthesia. After preoxygenation (FiO2 = 1.0), anaesthesia was induced with etomidate (0.2 mg kg−1), followed by TIVA with propofol (0.7 to 2 μg mL−1, Diprifusor® Astra Zeneca laboratory) and remifentanil (0.2 to 0.5 μg kg−1 min−1, Glaxo Wellcome Laboratory). Prior to laryngoscopy, a single bolus of cisatracurium (0.15 mg kg−1, Glaxo Wellcome laboratory) was given . The NICO-system was immediately connected after endotracheal intubation. A rebreathing circuit was placed between the endotracheal intubation tube and the ventilator circuit measuring end-tidal CO2 (etCO2) and CO2 production (VCO2) before and during a cycle of partial rebreathing. The patient was ventilated with O2 in air and a tidal volume (TV) of 10 mL kg−1. Respiratory rate was set between 10 and 15 min−1 to maintain etCO2 between 32 and 38 mmHg.
After induction of anaesthesia a central venous line (AVA HF, Edwards Lifesciences) and a PAC (Swan-Ganz CCOmbo CCO/SvO2, Edwards Lifesciences) were inserted. Arterial and mixed venous blood gas samples were taken in order to calibrate the NICO for shunt estimation and the SvO2.
The following data were collected every 3 min with the time-weighted averaging mode of the PAC-CCO and the full NICO time cycling mode: CO, SvO2 and core temperature. Measurements extended into the postoperative period and were discontinued when the patient was extubated.
Influence of temperature: The NICO-system is based on the Fick principle and on rebreathing of CO2. The VCO2, which is measured by the NICO, depends on the patient's temperature. We verified whether core temperature had an impact on the difference between CO measurements of both methods.
Correlation between both techniques: Data were compared by the Bland–Altman method to calculate the degree of agreement and to analyse if a significant difference existed between the two methods of CO measurements . The mean difference between two measurements is called the bias. The bias can determine a systematic error of a method (over- and under-estimation). The standard deviation of all bias measurements is referred to as precision (=1 SD of bias readings). Limits of agreement were defined as: bias ±2 SD. Percentage error was the ratio of two times the SD to mean CO.
Correlation between CO of NICO and CCO: The 4372 points of comparison were recorded (2603 perioperatively and 1769 postoperatively). Bland–Altman analyses (Figs 1 and 2) show that perioperative and postoperative biases are comparable. Perioperatively, the NICO underestimated CO, while postoperatively CO was overestimated. The limits of agreement were larger during OPCAB surgery compared to the postoperative period (−3.1; +2.5 vs. −1.4; +2.2 L min−1).
Perioperatively, CO measured with PAC-CCO varied from 0.5 to 7.5 L min−1 (mean CO 3.6 L min−1) and with the NICO from 0.5 to 8.4 L min−1 (mean CO 3.9 L min−1). Postoperatively, the CO of PAC-CCO varied from 2.5 to 7.7 L min−1 (mean CO 4.5 L min−1) and that of the NICO from 2.3 to 8.4 L min−1 (mean CO 4.9 L min−1). The average CO of the PAC and the NICO was 3.7 L min−1 perioperatively and 4.7 L min−1 postoperatively.
Table 1 shows characteristics of the patients, and Table 2 the great interindividual variability between bias, limits of agreement and correlation coefficient.
Influence of temperature: Temperatures varied between 33.4 and 38.8°C. Core temperature had no influence on differences of CO measured by both techniques.
The main findings of this study were:
- The peri- and postoperative biases of the NICO and the PAC-CCO were similar.
- The NICO slightly underestimated CO intraoperatively while postoperatively compared to the PAC the CO was overestimated.
- The limits of agreement were larger during OPCAB compared to the postoperative period.
- Interindividual variability of measurement points varied greatly.
- Core temperature between 33.4 and 38.8°C did not influence the CO measurements in either method.
- The NICO detected more rapidly acute haemodynamic changes than PAC-CCO.
NICO has mainly been compared to PAC measurements in intensive care units and in the setting of cardiac surgery. However, until now, no study is available which compares the NICO and PAC-CCO during OPCAB surgery.
Large differences in etCO2 and PaCO2 will cause NICO to underestimate CO. This might explain the negative bias in this present study. However, ventilating conditions in our study were similar to Botero and colleagues  where all patients were mechanically ventilated with a tidal volume of 10 mL kg−1 and a respiratory rate of 8 to 12 per minute adjusted to achieve an end-tidal CO2 between 35 and 40 mmHg without positive end-expiratory pressure (PEEP). In the same paper, linear multiple regression analyses showed that the setting most affecting the discrepancy between NICO and TDCO was minute ventilation. Botero and colleagues also found that accuracy of the CO2 rebreathing technique is neither affected by volume-controlled ventilation (VCV), pressure-controlled ventilation (PCV) nor by spontaneous breathing, PEEP or FiO2 .
Tachibana and colleagues  observed that linear regression slopes for NICO and TDCO were almost identical when inspired VT is set at 12 mL kg−1 and respiratory rate is set at ventilation of 0.13–0.14 L min−1 kg−1. Tachibana and colleagues also write that if normocapnia is maintained by adjusting respiratory rate, the accuracy of the NICO technique can be maintained at small VT. However, at small tidal volume, the rebreathing system underreports CO, compared with the conventional thermodilution technique. Intraoperatively in our study tidal volume was 10 mL kg−1 and most likely not the reason for the underestimation of CO by the NICO.
During OPCAB surgery no changes in CO readings were observed suggesting that acute haemodynamic events recorded by PAC-CCO, NICO and SvO2 were related to changes in ventilatory settings. Perioperative ventilation was done with the same anaesthetic machine (Julian®-anaesthetic ventilator, Dragger Laboratory). However as soon as the patient was transferred to the ICU, the patient was ventilated with an Ambu-bag followed by intermittent positive pressure ventilation by an Oxylog ventilator. This could explain the discrepancy between intraoperative and postoperative results of NICO and PAC-CCO. All this implies that the NICO cannot be used for weaning a patient off the ventilator and during events in which ventilation can show important variations.
In a study by Botero and colleagues  comparing NICO to the reference method (transit-time flowmetry of the ascending aorta) before cardiopulmonary bypass, the bias of the NICO is 0.04 L min−1, the precision ±1.07 L min−1 and the limit of agreement varies between −2.1 and 2.2 (−42.2 to 48%) with an error of 44.8%. These results suggest that prior CPB the NICO in the study of Botero and colleagues is overestimating CO in contrast to our findings. However, in our study operating conditions differ from Botero and colleagues . We used PAC-CCO as a reference method and compared it to the NICO in a different surgical setting, namely OPCAB surgery.
The discrepancy of under- and overestimation between the peri- and postoperative period may also be explained by haemodynamic events during the procedure. It has been shown that the accuracy of PAC-CCO is impaired in patients with CO >10 L min−1 [19–21]. However, in our study maximal values of CO were never above 8.4 L min−1.
CI95 interval was similar to those found in other studies [4,22–30] but we cannot draw any comparative conclusions because none was done in the setting of OPCAB surgery. This also applies for the limits of agreement.
Compared to the present study, previous studies with the NICO recruited comparatively small number of patients (12 to 41 patients) [22–30], except Botero who included 68 patients . However in contrast to other studies, the numbers of measurement points in our study is higher than previously recorded [4,22–30].
Interindividual variability of CO measurements has been described with other CO monitoring systems but never with the NICO [31,32]. In our study the bias remained independent between body temperatures from 33.4 to 38.8°C. Therefore, in the operating room the use of the NICO is as reliable as PAC-CCO within the range of above body temperatures. A previous study showed good correlation between thermodilution and CCO measurements with no influence of body temperature on CO results .
Haller and colleagues  compared continuous and intermittent assessment of CO using modified PAC and concluded that both systems were not fast enough to precisely assess sudden haemodynamic changes. The latency of the PAC system might be responsible for the increased bias during periods of poor cardiac performance. Indeed, the absence of variations in PAC-CCO measurements during acute events could be due to the monitoring software: the PAC cardiac output is the average of a number of measurements and extreme values are eliminated by the algorithm of the monitor. However, the TruCCOMS system is a newer version of CCO monitoring with a response time to changes in CO, which is within seconds. Especially during OPCAB surgery, this fast response system could offer significant advantages to the anaesthetist to detect changes in CO more quickly .
Following reason can explain the delay of the NICO to recording of rapid haemodynamic changes during OPCAB surgery. The NICO assumes that CO remains constant during measurements. Hyperventilation or hypoventilation, and sudden changes in metabolism alter the mixed venous CO2 concentration. However, the time constant of metabolic changes is approximately 2–3 min, which is longer than the rebreathing time (50 s) of the NICO (TM) system. Consequently, the mixed venous CO2 concentration remains more or less constant .
Bein and colleagues  conclude that the NICO system is not suitable for CO determination during quickly changing haemodynamic conditions. However, their conclusion is in contrast to our findings. We observed that detection of acute haemodynamic changes were faster with the NICO than with PAC-CCO measurements. Also, pulmonary artery CCO measurements compared to the NICO did not detect the great majority of acute haemodynamic events. It is important to stress that Bein and colleagues never studied the rapidity of CO changes during sudden haemodynamic changes during off-pump coronary artery bypass grafting (CABG) surgery and that their conclusion is only limited to aortic reconstruction. Until now, no study has ever investigated and compared the rapidity of a CO-measurement system to detect acute haemodynamic changes during OPCAB surgery.
Large changes in haemoglobin concentration can also have a small effect on NICO measurements . Haemoglobin content and haemoglobin P50 alter only slightly in OPCAB compared to on-pump CABG surgery but miscalculations of estimated shunt fraction by the NICO device have to be taken into account if the haematocrit changes . Nevertheless, no publication has ever studied the effect of the haemoglobin concentration or haematocrit on CO measurements of the NICO or any other CO-measurement technique. Concerning the impact of haemoglobin concentration in the current study, haematocrit was measured on induction, and at any moment of haemodynamic instability like a fall in blood pressure or CO and in case of bradycardia and tachycardia. We even measured haematocrit in case of a fall of central venous pressure and also when estimated blood loss decreased the patient's haematocrit below 30%. In the latter case, we immediately corrected any decrease of haematocrit. In addition, we regularly calibrated the NICO during OPCAB as it already has been highlighted in the material and methods section.
There has not yet been any international agreement on a gold standard for CO-measurements during off-pump surgery. Obviously, the absence of this gold standard has limited our study. Nevertheless, we still consider PAC-CCO as the best reference for CO-measurements and as a gold standard for off-pump CABG surgery for reasons explained in previous publications [10,11,38].
Although the objective of this study was to compare the CCO of the NICO with the PAC-CCO, it was also observed that CO followed the changes in SvO2 and that SvO2 changed more rapidly in case of acute haemodynamic changes. In these patients, the response of the NICO to detect events of haemodynamic instability was faster than CCO measurements. The NICO needs a full cycle of 3 min to display CO as does the time-weighted averaging mode of the CCO pulmonary catheter. However, in 3 patients we observed that the SvO2 can more rapidly detect haemodynamic changes than both the NICO and the CCO pulmonary catheter.
As observed changes of SvO2 and the NICO were more or less simultaneous, the use of the NICO could be indicated in case if invasive SvO2 or CCO measurements have to be avoided. This, however, has to be confirmed by studying larger samples of patients.
Haemodynamic changes during OPCAB surgery are of different nature compared to other surgeries. Therefore, prior to extrapolate our findings with NICO and PAC-CCO to other surgical fields, further studies have to be done to confirm whether data can be transferred.
In conclusion, during OPCAB surgery NICO reliably detects CO and more rapidly than PAC-CCO. However in some patients SvO2 was the fastest parameter but larger samples of patients are needed to confirm this observation. During off-pump surgery both monitors, NICO and PAC-CCO, could be employed but NICO is quicker in order to detect rapid haemodynamic changes.
1. Sun Q, Rogiers P, Pauwels D, Vincent JL. Comparison of continuous thermodilution and bolus cardiac output measurements in septic shock. Intensive Care Med
2. 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
3. Berthelsen PG. Clinical evaluation of a partial CO2
rebreathing technique. Acta Anaesthesiol Scand
: 1175–1180; discussion 1175–1176.
4. Botero M, Kirby D, Lobato EB, Staples ED, Gravenstein N. Measurement of cardiac output before and after cardiopulmonary bypass: comparison among aortic transit-time ultrasound, thermodilution, and noninvasive partial CO2
rebreathing. J Cardiothorac Vasc Anesth
5. Shoemaker WC, Wo CC, Chan L et al
. Outcome prediction of emergency patients by noninvasive hemodynamic monitoring. Chest
6. Bettex DA, Hinselmann V, Hellermann JP, Jenni R, Schmid ER. Transoesophageal echocardiography is unreliable for cardiac output assessment after cardiac surgery compared with thermodilution. Anaesthesia
7. Rauch H, Muller M, Fleischer F, Bauer H, Martin E, Bottiger BW. Pulse contour analysis versus thermodilution in cardiac surgery patients. Acta Anaesthesiol Scand
8. 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
9. Spodick DH. The pulmonary artery catheter. Chest
10. Grow MP, Singh A, Fleming NW, Young N, Watnik M. Cardiac output monitoring during off-pump coronary artery bypass grafting. J Cardiothorac Vasc Anesth
11. Schulz K, Abel HH, Werning P. Comparison between continuous and intermittent thermodilution measurement of cardiac output during coronary artery bypass operation. Anasthesiol Intensivmed Notfallmed Schmerzther
12. Jaffe MB. Partial CO2
rebreathing cardiac output – operating principles of the NICO system. J Clin Monit Comput
13. Yelderman ML, Ramsay MA, Quinn MD, Paulsen AW, McKown RC, Gillman PH. Continuous thermodilution cardiac output measurement in intensive care unit patients. J Cardiothorac Vasc Anesth
14. Yelderman M. Continuous measurement of cardiac output with the use of stochastic system identification techniques. J Clin Monit
15. Thrush D, Downs JB, Smith RA. Continuous thermodilution cardiac output: agreement with Fick and bolus thermodilution methods. J Cardiothorac Vasc Anesth
16. Gueret G, Rossignol B, Kiss G et al
. Is muscle relaxant necessary for cardiac surgery? Anesth Analg
17. Bland JM, Altman DG. Statistical methods for assessing agreement between 2 methods of clinical measurement. Lancet
18. Tachibana K, Imanaka H, Miyano H, Takeuchi M, Kumon K, Nishimura M. Effect of ventilatory settings on accuracy of cardiac output measurement using partial CO2
19. Burchell SA, Yu M, Takiguchi SA, Ohta RM, Myers SA. Evaluation of a continuous cardiac output and mixed venous oxygen saturation catheter in critically ill surgical patients. Crit Care Med
20. Zollner C, Polasek J, Kilger E et al
. Evaluation of a new continuous thermodilution cardiac output monitor in cardiac surgical patients: a prospective criterion standard study. Crit Care Med
21. Jacquet L, Hanique G, Glorieux D, Matte P, Goenen M. Analysis of the accuracy of continuous thermodilution cardiac output measurement. Comparison with intermittent thermodilution and Fick cardiac output measurement. Intensive Care Med
22. Mielck F, Buhre W, Hanekop G, Tirilomis T, Hilgers R, Sonntag H. Comparison of continuous cardiac output measurements in patients after cardiac surgery. J Cardiothorac Vasc Anesth
23. Rocco M, Spadetta G, Morelli A et al
. A comparative evaluation of thermodilution and partial CO2
rebreathing techniques for cardiac output assessment in critically ill patients during assisted ventilation. Intensive Care Med
24. Murias GE, Villagra A, Vatua S et al
. Evaluation of a noninvasive method for cardiac output measurement in critical care patients. Intensive Care Med
25. de Abreu MG, Geiger S, Winkler T et al
. Evaluation of a new device for noninvasive measurement of nonshunted pulmonary capillary blood flow in patients with acute lung injury. Intensive Care Med
26. van Heerden PV, Baker S, Lim SI, Weidman C, Bulsara M. Clinical evaluation of the non-invasive cardiac output (NICO) monitor in the intensive care unit. Anaesth Intensive Care
27. Odenstedt H, Stenqvist O, Lundin S. Clinical evaluation of a partial CO2
rebreathing technique for cardiac output monitoring in critically ill patients. Acta Anaesthesiol Scand
28. Nilsson LB, Eldrup N, Berthelsen PG. Lack of agreement between thermodilution and carbon dioxide-rebreathing cardiac output. Acta Anaesthesiol Scand
29. Binder JC, Parkin WG. Non-invasive cardiac output determination: comparison of a new partial-rebreathing technique with thermodilution. Anaesth Intensive Care
30. Rocco M, Spadetta G, Morelli A et al
. A comparative evaluation of thermodilution and partial CO2
rebreathing techniques for cardiac output assessment in critically ill patients during assisted ventilation. Intensive Care Med
31. Dhingra VK, Fenwick JC, Walley KR, Chittock DR, Ronco JJ. Lack of agreement between thermodilution and fick cardiac output in critically ill patients. Chest
32. Schrijen F, Henriquez A, Renondo J, Poincelot F, Pichene M. Inter- and intrasubject variability of the thermodilution measurement of right ventricular ejection fraction and volume in patients with chronic obstructive pulmonary disease. Cardiovasc Res
33. Haller M, Zollner C, Briegel J, Forst H. Evaluation of a new continuous thermodilution cardiac output monitor in critically ill patients: a prospective criterion standard study. Crit Care Med
34. Padua G, Canestrelli G, Pala G, Sechi D, Spanu MC. Original insight into continuous cardiac output monitoring: “TruCCOMS” correlation with other methods. Minerva Anestesiol
35. Fujii S, Kikura M, Takada T, Katoh S, Aoyama N, Sato S. A noninvasive partial carbon dioxide rebreathing technique for measurement of pulmonary capillary blood flow is also a useful oxygenation monitor during one-lung ventilation. J Clin Anesth
36. Bein B, Hanne P, Hanss R et al
. Effect of xenon anaesthesia on accuracy of cardiac output measurement using partial CO2
37. Haryadi DG, Orr JA, Kuck K, McJames S, Westenskow DR. Partial CO2
rebreathing indirect Fick technique for non-invasive measurement of cardiac output. J Clin Monit Comput
38. Zink W, Noll J, Rauch H et al
. Continuous assessment of right ventricular ejection fraction: new pulmonary artery catheter vs. transoesophageal echocardiography. Anaesthesia