Haemodynamic monitoring is typically used in critically ill patients to assess and optimize cardiac function in order to achieve and maintain adequate tissue perfusion. The use of the pulmonary artery catheter (PAC) for this task has been challenged in the last few years and there is an ongoing debate regarding its impact on outcome. Conflicting findings have been published  and some of the findings may be better explained by patient selection, lack of specific indications and differences in the treatment protocols (or the lack of it) than by the type of monitoring, that is, the PAC itself .
Based on this development, there is an increasing clinical acceptance of minimally invasive haemodynamic monitoring. Different techniques are commercially available and in the recent years they have proved adequate in replacing the PAC under certain clinical conditions. Moreover, easier handling of these techniques (compared with the PAC) may result in a widespread, early application to larger patient populations at risk for haemodynamic instability. In order to avoid inadequate tissue perfusion in these situations, ‘early goal-directed haemodynamic therapy’ is a promising concept [3,4]. However, for early use in daily clinical practice, the diversity of minimally invasive haemodynamic monitoring demands knowledge of the different techniques, the different parameters provided by the devices and their clinical validity.
The aim of this article is, therefore, to review the most widely used minimally invasive cardiac output monitoring techniques, emphasize the new parameters available for preload assessment and propose a modular stepwise monitoring concept.
Minimally invasive cardiac output monitoring techniques
The minimally invasive haemodynamic monitoring techniques discussed in this review are pulse wave analysis, Doppler measurement techniques and partial carbon dioxide rebreathing using the applied Fick's principle. Based on practical considerations, this selection does not include the bioimpedance and dye dilution technique. Both techniques may be of limited use in today's clinical practice as a result of technical aspects [5,6]. Specific features (such as invasiveness, technical details of cardiac output assessment, additional variables and limitations) of each presented technique are summarized in Table 1. In order to assess these minimally invasive haemodynamic monitoring techniques as compared with the clinical standard – the PAC – basic statistical knowledge with respect to validation studies is required and this information is summarized in a separate section.
Pulse wave analysis
Pulse wave analysis, also referred to as ‘pulse pressure analysis’ or ‘pulse contour analysis’, is based on the principle that the stroke volume can be tracked continuously by analysis of the arterial pressure waveform. Because the arterial pressure waveform is the result of an interaction between stroke volume and the vascular structure, resistance, compliance and characteristic impedance have to be considered. The first successful algorithm was developed by Wesseling et al.  more than 30 years ago. He used a computer-based algorithm in order to monitor stroke volume and calculations were performed on the hypothesis that the area under the curve of the systolic part of the arterial pressure waveform is proportional to the stroke volume. For this calculation, calibration using thermodilution was needed. However, different mathematical models are used today in order to assess stroke volume/cardiac output in the commercially available pulse wave analysis devices (PiCCO plus, Pulsion Medical Systems, Munich, Germany; PulseCO, LiDCO Ltd, London, UK and FloTrac/Vigileo, Edwards LifeSciences, Irvine, USA).
The first pulse contour device for cardiac output measurement based on the Wesseling algorithm in clinical practice was the PiCCO system. A dedicated thermistor-tipped catheter, usually introduced via the femoral artery, has to be used to track stroke volume on a beat-by-beat basis after calibration by transpulmonary thermodilution. After observing inaccurate measurements as a result of variations in systemic vascular resistance, the PiCCO algorithm has been modified in order to better address the individual patient's aortic compliance using more information related to the systolic but also the diastolic arterial waveform . Different studies in a variety of clinical settings have been performed in the last years validating the PiCCOplus system against intermittent pulmonary artery thermodilution [9,10]. The algorithm now appears to be more reliable in situations of rapid haemodynamic changes . However, despite these improvements, frequent recalibrations for accurate measurements may be required [12,13].
Continuous cardiac output measurement by the PulseCO technique via a standard peripheral arterial line is usually classified as a pulse contour method, although strictly speaking it is not a pulse contour device . This technique relies on a pulse power analysis, which is based on the principle of mass/power conservation in a system and the assumption that following the correction for compliance and calibration, there is a linear relationship between net power and net flow in the vascular system. With PulseCO, the entire pulse wave (the systolic and the diastolic part of the pressure curve) is analysed. Thus, the problem of wave reflections in the vascular system is considered, but a so-called ‘autocorrelation’ is needed to determine ‘change of power’ caused by the heart. Similar to harmonic waveform analysis, autocorrelation is a mathematical function analysing repetitive signals in cycles over time (i.e. the stroke volume and thus cardiac output). The system is calibrated using transpulmonary lithium indicator dilution (LiDCO)  and as long as no major haemodynamic changes (alterations in vascular compliance and resistance) occur, reliable continuous cardiac output measurements have been demonstrated in clinical studies [16,17]. In contrast to thermodilution, the LiDCO is not sensitive to blood temperature changes, but electrolyte and haematocrit changes may have a negative effect.
In FloTrac/Vigileo system, continuous cardiac output assessment requires a FloTrac sensor, a specific transducer, attached to an existing standard arterial line. In contrast to the PiCCOplus and the PulseCO system, this device does not require external calibration. For cardiac output measurement, the SD of pulse pressure sampled during a time window of 20 s is correlated to the ‘normal’ stroke volume based on underlying demographic data. Impedance is also derived from these data, whereas vascular compliance and resistance are determined using arterial waveform analysis. With the initial software versions of the algorithm (1.01–1.03), adaptation of the vascular status was performed every 10 min. Based on the results of early validation studies, a major modification of the algorithm was a reduction in this time window to 1 min (software versions 1.06 and higher) . Studies using these software versions showed improved cardiac output assessment [19–21]. Further software modifications addressing the issue of limited accuracy during hyperdynamic situations are currently being tested.
As a result of the general principles used, all pulse wave analysis devices share the need for an optimal arterial signal quality for valid cardiac output assessment. Moreover, arrhythmias and the use of an intraaortic balloon pump preclude reliable measurements. It has to be emphasized that with the PiCCOplus system, the insertion of the femoral catheter is contraindicated in patients with severe atheromatosis.
Flow measurement by pulsed-wave Doppler across a cardiac valve or in the left ventricular outflow tract and the assessment of the cross-sectional area at the site of the flow quantification (the aortic valve) allow cardiac output measurement by transoesophageal echocardiography (TOE). A Doppler beam orientation strictly parallel to the blood flow and an unchanged cross-sectional area over time are needed for optimal measurements. TOE measurements are time-consuming and require a high level of operator skill and knowledge. Limited reliability is typically related to technical TOE problems (image quality, variations in orifice area determination, excessive Doppler beam angle) and considerable interobserver variability [22,23].
Transoesophageal Doppler flow probes
Doppler flow measurements for cardiac output determination can also be performed in the descending aorta, in which adequate characteristic flow signals can be obtained as a result of the close proximity to the oesophagus and the aorta. Different Doppler probes are available (ODM II, Abbott, Maidenhead, UK; CardioQ/Medicina TECO, Deltex Medical Ltd, Chichester, UK; HemoSonic100, Arrow, Reading, USA) and cardiac output is typically calculated from measured aortic blood flow and aortic cross-sectional area obtained either from nomograms or by M-mode ultrasound quantification (HemoSonic 100). Cardiac output is assessed assuming a constant partition between caudal and cephalic blood supply, as brachiocephalic flow cannot be measured. The probes are smaller than conventional TOE probes and steep learning curves for probe positioning have been observed. However, validation studies in the last few years have revealed inconsistent results [24,25]. Limited accuracy may result from signal detection failure, the assumption of fixed regional blood flow or the use of nomograms to determine aortic cross-sectional area . The HemoSonic 100 device was developed to eliminate the use of nomograms by echocardiographic aortic assessment, but optimal adjustment of both Doppler and ultrasonic signal at the same time may be challenging . Therefore, the value of this minimally invasive technique may be limited in clinical practice. However, Doppler devices may be used in specific situations by skilled observers. Based on the ability to reliably track stroke volume changes over time, early goal-directed therapy in a perioperative setting may be a typical indication .
Applied Fick's principles: partial CO2 rebreathing
Fick's principle applied to CO2 is used by the NICO system (Novametrix Medical Systems, Wallingford, USA) for cardiac output measurement. CO2 analysis is performed using a mainstream infrared and airflow sensor. CO2 production is calculated as the product of CO2 concentration and air flow during a breathing cycle and arterial CO2 content is derived from end-tidal CO2 and the corresponding dissociation curve. An intermittent partial rebreathing state in intervals of 3 min can be induced by a disposable rebreathing loop. The rebreathing cycle results in an increased end-tidal CO2 and mimics a drop in CO2 production. The differences in these values are then used to calculate cardiac output. Validation studies showed conflicting results. However, clinically acceptable cardiac output assessment is possible in intubated mechanically ventilated patients with minor lung abnormalities and fixed ventilatory settings . In contrast, variations in ventilatory modalities, mechanically assisted spontaneous breathing or use of this technique in patients with lung disorders (increased shunt fraction) resulted in decreased accuracy [29,30]. Thus, this technique may be applied in a precisely defined clinical setting to mechanically ventilated patients only.
Statistical techniques used for validation of cardiac output monitoring techniques
In the last years, the Bland–Altman analysis  has been established as the standard statistical technique for validation studies assessing new cardiac output monitoring techniques. In order to determine the agreement between a new device and intermittent thermodilution by the PAC (as standard reference technique), the differences between cardiac ouput measured by both methods (e.g. new – reference technique) are plotted against their mean values [e.g. (new + reference technique)/2]. Thus, the following information is available: mean bias – the average of all the differences, the SD around the bias and the limits of agreement – the limits within which 95% of all the points fall on either side of the bias, that is, ±1.96 times the SD around the bias. These variables can be used to describe the accuracy and precision of any given device when compared with the reference technique. The accuracy describes how close the actual measurement to the real value is and this is related to the mean bias. Precision on the other hand describes how close the values of repeated measurements are; precision is related to the limits of agreement. Whereas the concept of ‘accuracy’ is quite easy to understand, it is more difficult to interpret ‘limits of agreement’. Critchley and Critchley  proposed that the percentage error of the limits of agreement, that is, 1.96 × SD in relation to the population mean, should be used to describe the agreement and that this could be used as a cut-off to accept a new technique. They suggested that, if a clinically acceptable level of precision for a reference technique can be defined as ±20%, then, in order for a new device to be accepted, an equivalent precision, that is, ±20%, may be required. Therefore, the percentage error from the Bland–Altman plot should be less than ±28.3%, that is, the combination of the precision of two techniques. Unfortunately, this concept has been simplified subsequently by many authors. They assumed a percentage error of less than 30% to be a prerequisite for the performance of a new cardiac output measurement technique. However, this assumption may be questioned. Based on the fact that the precision of the reference technique may be – under certain conditions – less than 20%, a percentage error lower than 30% does not necessarily indicate a clinically acceptable performance of a new device . Thus, it is desirable that for future validation studies, the precision of the reference technique is assessed within the study and reported.
Parameters of preload assessment
Volume expansion is the most often performed intervention to improve the haemodynamic status of critically ill patients. Several minimally invasive monitoring devices provide different preload parameters and allow an assessment of the fluid status. These parameters are reviewed in the following section. However, as standard pressure parameters are still widely used in clinical practice, these parameters are also discussed.
Static preload parameters
Preload according to Frank-Starling is defined as the end-diastolic myocardial fibre tension. Unfortunately, this cannot be assessed in a clinical setting and therefore only surrogate markers of preload can be measured. Traditionally, monitoring of central venous pressure is used as global parameter of the heart filling, that is, right and left ventricular preload, in patients without significant pulmonary diseases or impaired right ventricular function. In the presence of these disorders, pulmonary artery occlusion pressure assessed by a PAC is usually considered to be the only reliable indicator of left heart filling. Studies in the last few years, however, have shown a lack of correlation between cardiac filling pressures and stroke volume [34,35]. Transduced intrathoracic pressures interfering with the pressure preload measurements may predominantly explain these findings. Pressure is an important determinant of a compliant system (such as the heart) but this refers to transmural and not intravascular pressure. Moreover, frequently observed alterations in the cardiac valves or cardiac obstructions can introduce a measurement error. Still, cardiac filling pressures are often easily available and based on their limitations, interpretation of these parameters has to be undertaken in the clinical context. Moreover, trends are more important than absolute value and the dynamics of changes are essential.
With the limitations of pressure preload parameters, assessment of end-diastolic volume is desirable in daily clinical practice. Two thermodilution-based techniques provide volumetric preload parameters, which are calculated from cardiac output and indicator passage times. Continuous end-diastolic volume index (CEDVI) can be assessed by a modified PAC (CCOmbo CCO/SvO2/CEDV catheter 774HF75, Edwards Lifesciences) and as a result of the measurement site, the assessed volume is restricted to the right-sided heart chambers. Global end-diastolic volume index (GEDVI) and the closely related intrathoracic blood volume index (ITBVI) are determined via transpulmonary thermodilution, which is used for the calibration of the PiCCO system. The PiCCO approach assumes that the thermal indicator travels from its application (via the central venous access) through the different central compartments connected in series to the indicator detection site. Based on different indicator detection times, the different volumes can be calculated. All volumetric preload parameters have shown to be superior as preload indicators to cardiac filling pressures [34,36] and may be used to guide perioperative fluid therapy . In addition to these parameters, extravascular lung water index (EVLWI) can be determined by the PiCCO system . EVLWI may be useful for the diagnosis of pulmonary oedema and the evaluation of different ventilatory strategies in acute respiratory distress syndrome (ARDS) patients.
Echocardiographic preload assessment
Different echocardiographic approaches allow preload assessment either by measuring left ventricular end-diastolic area (LVEDA) or calculating volume (LVEDV). LVEDA is determined at the mid-papillary level in a short-axis view, whereas for the volumetric assessment different echocardiographic positions and measurements are needed for the calculation of LVEDV. These calculations are based on the Simpson algorithm and assume that the ventricle consists of the sum of small cylinders and a truncated ellipse. Conflicting results have been reported using the Simpson algorithm and, typically, LVEDA is used as surrogate of left ventricular preload . However, because this technique is highly operator-dependent , the real benefit of echocardiography is the visualization of ventricular function, wall motion abnormalities and cardiac filling as well as the ‘real-time’ guidance of fluid therapy in acute, critical haemodynamic situations. Moreover, echocardiography is of invaluable practical use as a diagnostic tool (for example, the detection of papillary muscle rupture, pericardial tamponade or dissection of the ascending aorta).
Dynamic preload parameters
Stroke volume variation (SVV) and pulse pressure variation (PPV) occur due to cyclic changes in intrathoracic pressure induced by inspiration and expiration during mechanical ventilation (Fig. 1a). SVV and PPV have been recognized as an interesting concept for guiding fluid replacement therapy more than 20 years ago . Only since the introduction of the minimally invasive haemodynamic monitoring devices based on pulse wave analysis has automated, quantification of this phenomenon been possible (Fig. 1b). The method appears to visualize the individual cardiac response (changes in stroke volume) related to myocardial contractility due to diastolic volume loading. In presence of hypovolaemia, large SVVs can be observed and the preload dependence of left ventricular function is pronounced, that is, the ventricle operates on the ascending limb of the Frank-Starling curve. During volume expansion, there is a rightward shift of left ventricular function on the Frank-Starling curve, which corresponds to the observed decrease in SVV (Fig. 1c). Both parameters have been used to assess fluid responsiveness and have been shown in a number of investigations to be sensitive in predicting the ventricular response to fluid administration, that is, the detection of fluid responders [41–43]. However, alteration of vasomotor tone may influence PPV more than SVV . Moreover, it is crucial to realize that this dynamic preload assessment is only reliable in fully sedated, mechanically ventilated patients with a relatively high tidal volume. Finally, a regular heart rhythm is mandatory. In the presence of these limitations, the ‘passive leg raising concept’ can be applied instead. This concept relies on the response of stroke volume to an internal fluid shift induced by a modified Trendelenburg position and has shown to reliably predict fluid responsiveness .
Modular stepwise monitoring concept
All of the preceding minimally invasive monitoring techniques can be integrated in a modular stepwise monitoring concept (Fig. 2). Haemodynamic assessment of critically ill patients can range from invasive blood pressure measurement and central venous catheterization, including the interpretation of blood gas analysis (with the measurement of lactate levels and central venous oxygenation ) to the use of a PAC. The indications for the next step in the concept may be determined by the patient's response to an initial fluid trial, but may also be based on the patient's physical status, related cardiovascular or respiratory comorbidities and, in the perioperative setting, by the performed intervention. Moreover, limitations of minimally invasive haemodynamic monitoring systems have to be considered and in selected situations, a PAC may be mandatory. Transthoracic or transoesophageal echocardiographic evaluation of an uncertain haemodynamic situation is a cornerstone in this concept.
Different minimally invasive haemodynamic monitoring systems can be reliably used in daily practice and have replaced the PAC in many clinical settings. However, the specific properties and limitations of these systems have to be considered. Devices based on pulse wave analysis may have the largest potential for routine continuous cardiac output measurement. In addition, this technique allows the assessment of fluid responsiveness through the measurement of pulse pressure or SVV. Minimally invasive monitoring techniques may be integrated in a modular stepwise concept ranging from invasive blood pressure monitoring to the use of the PAC based on an initial fluid response, patient-related factors and limitations of all monitoring systems. Echocardiographic evaluation has to be considered a clinical standard in the assessment of a critical haemodynamic situation.
1 Harvey S, Young D, Brampton W, et al.
Pulmonary artery catheters for adult patients in intensive care. Cochrane Database Syst Rev
2 Takala J. The pulmonary artery catheter: the tool versus treatments based on the tool. Crit Care 2006; 10:162.
3 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.
4 Wakeling HG, McFall MR, Jenkins CS, et al
. Intraoperative oesophageal Doppler
guided fluid management shortens postoperative hospital stay after major bowel surgery. Br J Anaesth 2005; 95:623–642.
5 Hirschl MM, Kittler H, Woisetschlager C, et al
. Simultaneous comparison of thoracic bioimpedance and arterial pulse waveform-derived cardiac output
with thermodilution measurement. Crit Care Med 2000; 28:1798–1802.
6 Hofer CK, Buhlmann S, Klaghofer R, et al
. Pulsed dye densitometry with two different sensor types for cardiac output
measurement after cardiac surgery: a comparison with the thermodilution technique. Acta Anaesthesiol Scand 2004; 48:653–657.
7 Wesseling KH, Purschke R, Smith NT, et al
. A computer module for the continuous monitoring of cardiac output
in the operating theatre and the ICU. Acta Anaesthesiol Belg 1976; 27(suppl):327–341.
8 Godje O, Hoke K, Goetz AE, et al
. Reliability of a new algorithm for continuous cardiac output
determination by pulse-contour analysis during hemodynamic instability. Crit Care Med 2002; 30:52–58.
9 Della Rocca G, Costa MG, Coccia C, et al
. Cardiac output
monitoring: aortic transpulmonary thermodilution and pulse contour analysis agree with standard thermodilution methods in patients undergoing lung transplantation. Can J Anaesth 2003; 50:707–711.
10 Halvorsen PS, Espinoza A, Lundblad R, et al
. Agreement between PiCCO
pulse-contour analysis, pulmonal artery thermodilution and transthoracic thermodilution during off-pump coronary artery by-pass surgery. Acta Anaesthesiol Scand 2006; 50:1050–1057.
11 Felbinger TW, Reuter DA, Eltzschig HK, et al
. Cardiac index measurements during rapid preload changes: a comparison of pulmonary artery thermodilution with arterial pulse contour analysis. J Clin Anesth 2005; 17:241–248.
12 Bein B, Meybohm P, Cavus E, et al
. The reliability of pulse contour-derived cardiac output
during hemorrhage and after vasopressor administration. Anesth Analg 2007; 105:107–113.
13 Hamzaoui O, Monnet X, Richard C, et al
. Effects of changes in vascular tone on the agreement between pulse contour and transpulmonary thermodilution cardiac output
measurements within an up to 6-h calibration-free period. Crit Care Med 2008; 36:434–440.
14 Jonas MM, Tanser SJ. Lithium dilution measurement of cardiac output
and arterial pulse waveform analysis: an indicator dilution calibrated beat-by-beat system for continuous estimation of cardiac output
. Curr Opin Crit Care 2002; 8:257–261.
15 Cecconi M, Dawson D, Grounds RM, Rhodes A. Lithium dilution cardiac output
measurement in the critically ill patient: determination of precision of the technique. Intensive Care Med 2009; 35:498–504.
16 Cecconi M, Fawcett J, Grounds RM, Rhodes A. A prospective study to evaluate the accuracy of pulse power analysis to monitor cardiac output
in critically ill patients. BMC Anesthesiol 2008; 8:3.
17 Yamashita K, Nishiyama T, Yokoyama T, et al
. Cardiac output
by PulseCO is not interchangeable with thermodilution in patients undergoing OPCAB. Can J Anaesth 2005; 52:530–534.
18 Mayer J, Boldt J, Poland R, et al
. Continuous arterial pressure waveform-based cardiac output
using the FloTrac
/Vigileo: a review and meta-analysis. J Cardiothorac Vasc Anesth 2009; 23:401–406.
19 Button D, Weibel L, Reuthebuch O, et al
. Clinical evaluation of the FloTrac
/Vigileo system and two established continuous cardiac output
monitoring devices in patients undergoing cardiac surgery. Br J Anaesth 2007; 99:329–336.
20 Cannesson M, Attof Y, Rosamel P, et al
. Comparison of FloTrac cardiac output
monitoring system in patients undergoing coronary artery bypass grafting with pulmonary artery cardiac output
measurements. Eur J Anaesthesiol 2007; 24:832–839.
21 Senn A, Button D, Zollinger A, Hofer CK. Assessment of cardiac output
changes using a modified FloTrac
/Vigileo algorithm in cardiac surgery patients. Crit Care 2009; 13:R32.
22 Bettex DA, Hinselmann V, Hellermann JP, et al
. Transoesophageal echocardiography is unreliable for cardiac output
assessment after cardiac surgery compared with thermodilution. Anaesthesia 2004; 59:1184–1192.
23 Schmidlin D, Bettex D, Bernard E, et al
. Transoesophageal echocardiography in cardiac and vascular surgery: implications and observer variability. Br J Anaesth 2001; 86:497–505.
24 Hullett B, Gibbs N, Weightman W, et al
. A comparison of CardioQ and thermodilution cardiac output
during off-pump coronary artery surgery. J Cardiothorac Vasc Anesth 2003; 17:728–732.
25 Moxon D, Pinder M, van Heerden PV, Parsons RW. Clinical evaluation of the HemoSonic monitor in cardiac surgical patients in the ICU. Anaesth Intensive Care 2003; 31:408–411.
26 Leather HA, Wouters PF. Oesophageal Doppler
monitoring overestimates cardiac output
during lumbar epidural anaesthesia. Br J Anaesth 2001; 86:794–797.
27 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 2003; 17:217–220.
28 Gueret G, Kiss G, Rossignol B, et al
. Cardiac output
measurements in off-pump coronary surgery: comparison between NICO
and the Swan-Ganz catheter. Eur J Anaesthesiol 2006; 23:848–854.
29 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 2004; 30:82–87.
30 Tachibana K, Imanaka H, Takeuchi M, et al
. Noninvasive cardiac output
measurement using partial carbon dioxide rebreathing is less accurate at settings of reduced minute ventilation and when spontaneous breathing is present. Anesthesiology 2003; 98:830–837.
31 Bland JM, Altman DG. Statistical methods for assessing agreement between two methods of clinical measurement. Lancet 1986; 1:307–310.
32 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.
33 Cecconi M, Rhodes A, Poloniecki J, et al
. Bench-to-bedside review: the importance of the precision of the reference technique in method comparison studies – with specific reference to the measurement of cardiac output
. Crit Care 2009; 13:201.
34 Hofer CK, Furrer L, Matter-Ensner S, et al
. Volumetric preload measurement
by thermodilution: a comparison with transoesophageal echocardiography. Br J Anaesth 2005; 94:748–755.
35 Kumar A, Anel R, Bunnell E, et al
. Pulmonary artery occlusion pressure and central venous pressure fail to predict ventricular filling volume, cardiac performance, or the response to volume infusion in normal subjects. Crit Care Med 2004; 32:691–699.
36 Della Rocca G, Costa GM, Coccia C, et al
. Preload index: pulmonary artery occlusion pressure versus intrathoracic blood volume monitoring during lung transplantation. Anesth Analg 2002; 95:835–843.
37 Goepfert MS, Reuter DA, Akyol D, et al
. Goal-directed fluid management reduces vasopressor and catecholamine use in cardiac surgery patients. Intensive Care Med 2007; 33:96–103.
38 Sakka SG, Klein M, Reinhart K, Meier-Hellmann A. Prognostic value of extravascular lung water in critically ill patients. Chest 2002; 122:2080–2086.
39 Hofer CK, Ganter MT, Rist A, et al
. The accuracy of preload assessment by different transesophageal echocardiographic techniques in patients undergoing cardiac surgery. J Cardiothorac Vasc Anesth 2008; 22:236–242.
40 Michard F, Teboul JL. Using heart-lung interactions to assess fluid responsiveness during mechanical ventilation. Crit Care 2000; 4:282–289.
41 Cannesson M, Musard H, Desebbe O, et al
. The ability of stroke volume variations obtained with Vigileo/FloTrac
system to monitor fluid responsiveness in mechanically ventilated patients. Anesth Analg 2009; 108:513–517.
42 Hofer CK, Senn A, Weibel L, Zollinger A. Assessment of stroke volume variation for prediction of fluid responsiveness using the modified FloTrac
and PiCCOplus system. Crit Care 2008; 12:R82.
43 Marx G, Cope T, McCrossan L, et al
. Assessing fluid responsiveness by stroke volume variation in mechanically ventilated patients with severe sepsis. Eur J Anaesthesiol 2004; 21:132–138.
44 Michard F, Teboul JL. Predicting fluid responsiveness in ICU patients: a critical analysis of the evidence. Chest 2002; 121:2000–2008.
45 Monnet X, Teboul JL. Passive leg raising. Intensive Care Med 2008; 34:659–663.
46 Marx G, Reinhart K. Venous oximetry. Curr Opin Crit Care 2006; 12:263–268.