........................................................................

CHAPTER 10

V O L U M E

T H I R T Y

S E V E N

HEMODYNAMIC MONITORING: FROM CENTRAL VENOUS PRESSURE TO PULSE CONTOUR ANALYSIS

ANDREW LEIBOWITZ, M.D.
ASSOCIATE PROFESSOR OF ANESTHESIOLOGY AND SURGERY MOUNT SINAI SCHOOL OF MEDICINE NEW YORK, NEW YORK

EDITOR: MEG A. ROSENBLATT, M.D.

ASSOCIATE EDITORS: JOHN F. BUTTERWORTH IV, M.D. JEFFREY B. GROSS, M.D.

The American Society of Anesthesiologists, Inc.

........................................................................

The ASA Refresher Courses in Anesthesiology CME Program

Subscribers to ASA Refresher Courses in Anesthesiology are eligible to earn AMA PRA Category 1 Credit(s)t. Please visit www.asa-refresher-cme.asahq.org or see page iv at the beginning of this volume for complete details.

Accreditation and Designation Statement

The American Society of Anesthesiologists is accredited by the Accreditation Council for Continuing Medical Education to provide continuing medical education for physicians. The American Society of Anesthesiologists designates this educational activity for a maximum of 1 AMA PRA Category 1 Credit(s)t. Physicians should only claim credit commensurate with the extent of their participation in the activity.

Author Disclosure Information

Dr. Leibowitz has disclosed that he has no nancial interests in or signicant relationship with any commercial companies pertaining to this educational activity.

c 2009

The American Society of Anesthesiologists, Inc.

ISSN 0363-471X ISBN 978-1-6054-7424-3 An educational service to the profession under the auspices of The American Society of Anesthesiologists, Inc. Published for The Society by Lippincott Williams & Wilkins 530 Walnut Street Philadelphia, Pennsylvania 19106-3621 Library of Congress Catalog Number 74-18961. www.asa-refresher.com

PERMISSION TO PHOTOCOPY ARTICLES: This publication is protected by copyright. Permission to reproduce copies of articles for noncommercial use must be obtained from the Copyright Clearance Center, 222 Rosewood Dr., Danvers, MA 01923; (978) 750-8400, FAX: (978) 750-4470, www.copyright.com.

Hemodynamic Monitoring: From Central Venous Pressure to Pulse Contour Analysis

Andrew Leibowitz, M.D. Associate Professor of Anesthesiology and Surgery Mount Sinai School of Medicine New York, New York

This Refresher Course reviews the physiologic underpinnings, clinical application, and effectiveness of monitors that are used by anesthesiologists and intensivists in their daily practice. Only those monitors that are available for use by the average practitioner are discussed; thus, transesophageal echocardiography is not covered in detail because of the signicant additional training and certication required. However, it should be noted that there is an active movement toward recognition of limited scope or goal-directed echocardiography, sometimes also colloquially referred to as echo-lite, which primarily focuses on cardiac lling, ventricular wall motion, and the detection of pericardial effusion, performed by physicians who are less than fully trained and lack certication from the American Board of Echocardiography. Hemodynamic monitoring has gradually shifted from invasive intravascular monitoring toward less invasive devices. I review, in that order, techniques and devices including central venous pressure (CVP), pulmonary artery catheter (PAC), arterial pulse pressure variation, and the proprietary technologies of NICO, PiCCO2, LiDCO, FloTrac/Vigileo, and Doppler (e.g., CardioQ; Deltex Medical, West Sussex, United Kingdom). My inclusion or omission of any proprietary device should not be interpreted as my personal recommendation or disapproval. The included devices are representative and none of them has clearly distinguished itself as markedly superior to the others.

The Basic Science of Clinical Management

Central Venous Pressure
Measurement of CVP has been commonplace for more than 50 years since the pioneering work of Aubaniac,1 Seldinger,2 and Wilson et al.3 CVP is simply an intravascular pressure ideally measured in the superior vena cava. It is a false but widely held belief that CVP can be used to estimate intravascular volume status. Furthermore, even physicians willing to admit that this is not true usually adhere to an equally false belief that the trend of the CVP over time correlates to the trend in intravascular volume status. It is critical to understand that both of these beliefs are physiologically illogical and, through studies, have been thoroughly dismissed. From a physiologic standpoint, it makes sense that volume status cannot be inferred either from a single measurement of CVP or from its change over time. The venous system contains 70% of the blood volume and is an extremely compliant container, the main function of which is to return blood to the heart. Most of the blood contained within the venous system exists within the system at a transmural
Copyright 2009 American Society of Anesthesiologists, Inc.

119

120

LEIBOWITZ

FIG. 1. Representation of the venous volume and dynamics that result in central venous pressure (CVP). The tub represents the blood volume. The stressed volume (Vs) is the volume above the hole where the pressure is measured that will result in a positive pressure; the unstressed volume (Vu) is the volume less than the hole that will exist at a pressure of zero. The measured pressure at the hole will result from a combination of the inow from the smaller arterial volume contained in the high-resistance arterial vessels, the relationship between Vs and Vu, that is, changed by moving the hole up and down, and the venous resistance (VenR). Reprinted with permission from Gelman.5

pressure of zero, whereas circulatory physiology dictates that a certain positive CVP be maintained for blood to return to the heart and homeostasis to exist (Fig. 1).4 Therefore, even large changes in the content of this highly compliant system will not consistently result in a net change in pressure measured in the superior vena cava. The physiologic premises dismissing the relationship between CVP and volume have been clinically demonstrated in ve studies using different methods of determining blood volume; the inability of the CVP to predict the potential of the cardiac output (CO) or blood pressure to increase with uid administration has been the subject of 19 other studies. All of these studies can now be reviewed very succinctly in a systematic analysis that was just published by Marik et al.5 In addition, in one recently published study by Kumar et al.,6 in volunteers given a 3-l uid challenge, one third of them failed to demonstrate a rise in their CVP, which is very clear evidence that the trend over time in CVP measurement is not a reliable estimate of what is happening to intravascular volume in even the healthiest patients in the most stable situation possible. Therefore, in patients with cardiac depression from anesthesia or sepsis, those with ongoing blood loss, or those with systemic vasodilation, it is certain that no useful relationship between CVP and blood volume exists.

Pulmonary Artery Catheter

The ow-directed, balloon-tipped PACs ability to measure CO and pressure in the pulmonary artery represented a great physiologic advance. These easily obtained measurements combined with the wedge (or pulmonary artery occlusion) pressure, which should equal the left atrial and left ventricular pressure at end diastole, and measurement of mixed venous oxygen saturation led to a widespread acceptance of this technique within only a few years of its clinical introduction.7 Problems with this technique were originally thought to be technical in nature and related to the

HEMODYNAMIC MONITORING

121

potential failure of the wedge pressure to accurately estimate the left atrial pressure.8 Later, widespread errors in PAC use, including, most importantly, user errors in obtaining and interpreting the wedge pressure, were demonstrated.9 More troubling, in the past 7 years, several high-quality retrospective studies and prospective randomized trials (in all of which user error was nearly eliminated) have failed to demonstrate an improvement in patient outcome associated with PAC use10--16 (see Table 1). Several of these investigations suggest a worse outcome when a PAC is used, even in highly competent hands using sophisticated management algorithms! Two of these trials10,11 are particularly pertinent to anesthesiologists. The Canadian Multicenter Trial10 reported that there was no difference in outcome in high-risk, elderly, American Society of Anesthesiologists 3 and 4 patients (n 1,994) randomized to management with PACs and an optimization lling pressure/oxygen delivery regimen versus standard care. Moreover, the PAC group experienced a small but statistically signicant increased total complication rate related to PAC insertion. Polanczyk et al.11 used propensity scoring to prospectively compare nonrandomized patients at identical risk who were managed with and without a PAC. This matched pairs analysis yielded 215 pairs of patients for investigation. Remarkably, patients managed with a PAC had signicantly more congestive heart failure and other adverse events than their non-PAC-managed controls.

Arterial Pulse Pressure Variation

Pulsus paradoxus, the excessive decrease in systolic pressure seen with spontaneous ventilation in the presence of cardiac tamponade, has long been recognized as an adverse clinical sign. More recently, the effect of positive pressure ventilation on the arterial pulse pressure has been described17 and the physiologic basis extensively reviewed.18,19 Understanding the three main physiologic premises
TABLE 1. Clinical Summary of Selected Recent Large Investigations Comparing Management With Versus Without a PAC Number of Patients Enrolled 1,994

Reference 10 11 12 13 14 15 16

Type Prospective multicenter Observational cohort single center Prospective multicenter Prospective single center Prospective Multicenter Prospective multicenter Prospective multicenter

Patient Group Perioperative Perioperative General ICU General ICU Decompensated Heart Failure ARDS ARDS

Signicant Outcome Differences

More adverse events in PAC group related to insertion 4,059 total; Increased heart failure and 215 matched noncardiac events in PAC group pairs 1,041 None 201 433 676 981 Increased renal insufciency and thrombocytopenia in PAC group Increased infections in PAC group None Increased related complications and blood transfusions in PAC group

ARDS adult respiratory distress syndrome; ICU intensivecare unit; PAC pulmonary artery catheter.

122

LEIBOWITZ

underlies the use of this phenomena as a monitor: (1) arterial pulse pressure (systolic--diastolic pressure) is directly proportional to stroke volume and inversely related to arterial compliance20; (2) a positive pressure breath compresses the pulmonary venous system, causing an increase in left ventricular preload and an increase in stroke volume and arterial pulse pressure for a few beats; and (3) a positive pressure breath also decreases the venous return to the right heart by increasing intrathoracic pressure. This decreases the left ventricular lling; thus, after a few beats of the increased pulse pressure that immediately follow a positive pressure breath, a decrease in the pulse pressure will follow; Figure 2 demonstrates this phenomenon. For several reasons, in the presence of hypovolemia, the change in pulse pressure is greater than in normovolemia. Simply summarized, in hypovolemia, the driving force for venous return, the mean circulatory lling pressure, is reduced, the right atria and vena cava are thus more easily compressed, and the left ventricle is already operating on the steep portion of the Frank--Starling curve. Several caveats apply to the clinical application of this principle: (1) There is no universal agreement as to which exact variable should be tracked, but possibilities include: (a) systolic pressure at end apnea--minimum systolic pressure (Ddown); (b) maximum systolic pressure--minimum systolic pressure; (c) 100 (maximum systolic pressure--minimum systolic pressure)/ (maximum systolic pressure--minimum systolic pressure/2) (DPP%) (2) Increased pulse pressure variation does not necessarily mean there is a low CO or hypovolemia, but its presence (i.e., 410%) is probably the best predictor that the blood pressure will increase with the administration of intravenous uids or blood (3) The shape of the arterial pressure tracing as well as the speed and scale of the tracing combined with the ability to determine respiration will all impact the clinicians ability to use these principles successfully. The above principles were used in the design of the PiCCO, LiDCO, and FloTrac/ Vigileo devices that combine the power of continuous monitoring of DPP% or other similar variables via pulse contour analysis with features that allow for continuous CO reporting and also other derived variables.

FIG. 2. The variation of arterial pulse pressure with positive pressure breaths. See text for explanation. Reprinted with permission from Michard.18

HEMODYNAMIC MONITORING

123

NICOs
The Fick principle states that the total uptake or release of a substance by an organ is the product of the blood ow to the organ and the arteriovenous concentration difference of the substance. Physicians are most familiar with the Fick principle applied to oxygen used to calculate CO in which: V O2 CO CaO2 CvO2 where VO2 oxygen consumption; CaO2, CvO2 arterial, venous oxygen content. Clinically, this principle can be used to determine CO when oxygen consumption is determined by using a metabolic cart and an arterial sample and a mixed venous sample of blood can be obtained and their oxygen contents determined. More commonly, intensivists interested in the determination of oxygen consumption rearrange this equation to determine oxygen consumption as follows: VO2 CO CaO2 CvO2 The NICO system (Philips/Respironics) adapts this principle to CO2 to determine CO as follows: VCO2 CO CvCO2 CaCO2 where VCO2 CO2 clearance; CaCO2, CvCO2 arterial, venous CO2 content. The device is inserted between the endotracheal tube and the breathing circuit that ingeniously relies on the following principles: (1) CaCO2 can be estimated by using the end-tidal CO2 in a steady state; and (2) addition of dead space and rebreathing allow mathematical calculation of the CO as follows: VCO2 N VCO2 R CO CvCO2 N CaCO2 N CvCO2 R CaCO2 R where N normal; R rebreathing. Knowledge of algebra allows the following assumption: If a c a c a c then b d b d b d and is applied to the above equation then: CO VCO2 N VCO2 R CvCO2 N CaCO2 N CvCO2 R CaCO2 R

During rebreathing, venous CO2 should be the same as under normal conditions and CvCO2N CvCO2 R and thus this equation becomes: CO VCO2 N VCO2 R CaCO2 R CaCO2 N

All these variables can be measured by the device. To use this device, the patient must be on positive pressure ventilation with stable tidal volume and respiratory rate. The number and variety of physiologic assumptions and mathematical manipulations are the main drawbacks to using this monitor. The difference in the N and R CO2 is usually quite small (i.e., o 10 mmHg), therefore, small differences in measurement may lead to large changes in calculated

124

LEIBOWITZ

CO. In addition, shunted blood containing CO2 is eliminated from consideration in this model. These underlying physiologic limitations combined with the ability to only monitor CO alone seem to be the main reasons for NICOs relatively small market penetration. Furthermore, from a clinical and statistical point of view, it is critical to note the standard to which the new monitoring devices are held when reporting CO. PACmeasured CO is usually determined by three consecutive measurements that may differ by 15% or more. On the basis of the methods of Bland and Altman,21 and the subsequent publication of one particularly inuential meta-analysis,22 a new monitor or technique is felt to be acceptable if the CO reported is 730% of the PAC thermodilution-determined CO and it is on this basis that all of the monitors and techniques discussed in this Refresher Course are acceptable. Data comparing these devices with thermodilution-determined CO usually report the bias (the average difference between the two), precision (71 standard deviation of the bias), and limits of agreement (72 standard deviation of the bias).

Pulse Contour Devices

These devices use computer-driven algorithms to translate the arterial pressure tracing into CO and other variables.23 The PiCCO and the original LiDCO (the plus model) devices require that the CO of the device be calibrated to another method, whereas the FloTrac/Vigileo uses a more mathematically advanced algorithm and requires no calibration. A recently updated version of the LiDCO (the rapid model) also allows noncalibrated CO to be continuously reported. All these devices allow continuous monitoring of the arterial pulse pressure variation, which, even in the absence of other functionality, may be useful.

PiCCO
The PiCCO device (Pulsion Medical Systems, Munich, Germany) is widely used in Europe and was recently introduced in the United States. The device relies on several principles that are relatively new to most anesthesiologist and intensivists. The primary physiologic assumption is that the contour of the arterial pressure waveform can be analyzed and the stroke volume equals the integral of the area under the curve divided by the impedance of the aorta (z), or mathematically stated: R dP=dt SV Z This is a variation of the idea that stroke volume is equal to pulse pressure divided by arterial compliance. The determination of CO at any time will allow calculation of z after which only periodic recalibration will be required. Figure 3 illustrates this method. The device relies on the insertion both of a proprietary arterial line with a temperature sensor at the tip that must be placed in the femoral or axillary artery and a central venous line with its tip in the superior vena cava. These two catheters are used to determine CO by means of the injection of a cold uid bolus into the superior vena cava and monitoring of the temperature change in the artery; this is done via a modied Stewart--Hamilton equation and the resultant value is termed the transpulmonary thermodilution CO. After calibration, the device reports CO and stroke volume continuously. From the analysis of the pulse contour, stroke volume variance is also reported.

HEMODYNAMIC MONITORING
-T -T

125

t Calibration P [mm Hg] SV

t [s]

FIG. 3. The calibration of cardiac output with pulse contour. The transpulmonary cardiac output is obtained and used to calibrate the device so that the area under the arterial pressure curve can subsequently be used to estimate the cardiac output. t time; T temperature. Reprinted with permission from Pulsion-supplied information package.

Further physiologic assumptions primarily relying on the concept that most of the diminution in temperature of the injectate occurs within the pulmonary vascular bed also permits continuous reporting of extravascular lung water, preload (referred to as global end-diastolic volume), and afterload. The main drawback of this device is the need for both a central line and an arterial line that are in the femoral or axillary artery. In addition, rapidly changing hemodynamic conditions will warrant repeated cold water injectate to reliably obtain a properly calibrated pulse contour analysis. The manufacturer recommends recalibration at least every 8 hours. One recent study reported a prospective (nonrandomized) multicenter investigation of 331 intensive care unit patients managed with a PAC versus PiCCO.24 There was no difference in the length of stay or mortality, the two main outcomes investigated between these two groups, although the PiCCO group had a greater positive uid balance and a longer duration of mechanical balance.

LiDCO
The original LiDCO device (LiDCO Ltd., Cambridge, United Kingdom), much like the PiCCO device, requires CO to be measured to calibrate its internal algorithm.25 It uses a peripheral injection of lithium ion 0.15 to 0.3 mMol with a 15-ml saline ush and a proprietary arterial line containing a lithium sensor to construct a dilution curve for the lithium ion and: COL=min R Lithium dose in mmol 60 R 1 PCV dLi=dt

d[Li]/dt is the area under the primary curve and PCV is the packed cell volume [Hb (g/dL)/34]. A correction for PCV is necessary because lithium is distributed in the plasma. The LiDCO device then relies on a more complex algorithm (i.e., PulseCO) than the PiCCO device. It assumes that arterial compliance changes with blood pressure and this change is similar in all humans. A few downsides to this device result from the use of lithium injections. The drug, lithium carbonate, hyponatremia, and some other drugs that contain quaternary ammonium ions (including some muscle relaxants), and hyponatremia interfere with lithium injection as a calibrating standard.

126

LEIBOWITZ

A randomized trial by Pearse et al.26 compared the use of the LiDCO device coupled with a goal-directed strategy versus conventional management and showed a reduction in the complications and hospital length of stay after major surgery in the treatment group but no difference in mortality. This was a small study of only 122 patients and the treatment group received dopexamine; therefore, it is difcult to draw any broad conclusion from this study.

FloTrac/Vigileo
The FloTrac/Vigileo system (Edwards Life Sciences, Irvine, CA) differs signicantly from the PiCCO and LiDCO devices in that no external calibration for CO is required at all and the device can be used with a standard arterial line. Instead of calibrating the device to a measured CO (transpulmonary thermodilution CO in the case of PiCCO and lithium dilution CO in the case of LiDCO), the arterial pressure waveform is analyzed in conjunction with demographic data consisting of age, height, weight, and sex. This device has recently had its algorithm updated after which bias, precision, and limits of agreement were improved.27 The underlying mathematics and physics used are difcult for the mathematically unsophisticated nonengineer to comprehend. No outcome study using the most recent version of this device has been reported.

Esophageal Doppler (e.g., CardioQ)

CO can also be determined completely noninvasively by using small esophageal echo-type probes. The technique relies on the fact that ow through a tube is equal to the cross-sectional area of the tube multiplied by the velocity--time integral of the uid. In the case of pulsatile ow, the velocity changes with time and the area under the velocity versus time curve (velocity--time integral), is used as shown in Figure 4. Doppler monitoring devices may use transthoracic or transesophageal probes. Transesophageal probes have been developed that can be easily inserted and are small and soft enough to be left in place for prolonged periods of time. For example, the CardioQ (Deltex Medical) uses an esophageal probe that is inserted until an optimal descending aortic tracing is obtained. In combination with basic demographic information, the velocity--time tracing is used to report continuous CO, stroke volume (Fig. 4), and corrected ow time, which is a velocity

Velocity

VTI = area under velocity time curve = velocity time integral

Time

Cross sectional area = r2

Stroke Volume = cross sectional area x velocity time integral

FIG. 4. How the stroke volume is calculated by esophageal Doppler devices: Stroke volume cross--sectional area velocity--time integral.

HEMODYNAMIC MONITORING

127

measurement corrected for age, peak velocity, minute distance, and heart rate. The manufacturer suggests using this device to construct a volume response optimization strategy in which the effect of successive boluses on CO, stroke volume, and velocity can be seen and uid titrated to optimize CO. Although there are several versions of esophageal CO devices that are available commercially, the CardioQ is nearly unique in that several outcome studies have been published attesting to its effectiveness. On that basis, the Center for Medicare and Medicaid Service will reimburse physicians for using the information supplied by esophageal Doppler for ventilated patients in the intensive care unit and operative patients with a need for intraoperative uid optimization. However, the majority of the patients studied were at a relatively minimal cardiac risk and the outcome improvements (i.e., length of stay) are of unknown importance because of the long length of stay in both the monitored and control groups.

Conclusion
Although CVP and measurements available from the PAC show no association with estimated intravascular volume and provide no proven outcome benet for manipulating hemodynamic values, there are newer monitoring techniques that may provide benet to patients. These monitors require a better understanding of how the circulatory system adjusts to changes in blood volume and extracellular uid volume. To date, although these devices can nearly duplicate the CO data supplied by the PAC (and do so continuously) and provide data heretofore unavailable to clinicians, these devices have not been demonstrated to change outcomes.

References
1. Aubaniac RL: Intravenous subclavicular injection [in French]. Presse Med 1952; 60: 1456--58. 2. Seldinger SI: Catheter replacement of the needle in percutaneous arteriography; a new technique. Acta Radiol 1952; 39:368--76. 3. Wilson IN, Grow JB, Demong CV, et al.: Central venous pressure in optimal blood volume maintenance. Arch Surg 1962; 85:563--78. 4. Gelman S: Venous function and central venous pressure. Anesthesiology 2008; 108: 735--48. 5. Marik PE, Baram M, Vahid B: Does central venous pressure predict uid responsiveness? A systematic review of the literature and the tale of seven mares. Chest 2008; 134: 172--8. 6. Kumar A, Anel R, Bunnell E, et al.: Pulmonary artery occlusion pressure and central venous pressure fail to predict ventricular lling volume, cardiac performance, or the response to volume infusion in normal subject. Crit Care Med 2004; 32:691--9. 7. Swan HJ, Ganz W, Forrester J, et al.: Catheterization of the heart in man with use of a owdirected balloon-tipped catheter. N Engl J Med 1970; 283:447--51. 8. Raper R, Sibbald WJ: Misled by the wedge? The Swan-Ganz catheter and left ventricular preload. Chest 1986; 89:427--34. 9. Iberti TJ, Fischer EP, Leibowitz AB, et al.: A multicenter study of physicians knowledge of the pulmonary artery catheter. Pulmonary Artery Catheter Study Group. JAMA 1990; 264:2928--32. 10. 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.

128

LEIBOWITZ

11. Polanczyk CA, Rohde LE, Goldman L, et al.: Right heart catheterization and cardiac complications in patients undergoing noncardiac surgery: An observational study. JAMA 2001; 286:309--14. 12. Harvey S, Harrison DA, Singer M, et al.: Assessment of the clinical effectiveness of pulmonary artery catheters in management of patients in intensive care (PAC-Man): A randomised controlled trial. Lancet 2005; 366:472--7. 13. Rhodes A, Cusack RJ, Newman PJ, et al.: A randomised, controlled trial of the pulmonary artery catheter in critically ill patients. Intensive Care Med 2002; 28:256--64. 14. Binanay C, Califf RM, Hasselblad V, et al.: Evaluation study of congestive heart failure and pulmonary artery catheterization effectiveness: The ESCAPE trial. JAMA 2005; 294:1625--33. 15. Richard C, Warszawski J, Anguel N, et al.: Early use of the pulmonary artery catheter and outcomes in patients with shock and acute respiratory distress syndrome: A randomized controlled trial. JAMA 2003; 290:2713--20. 16. The National Heart, Lung, and Blood Institute Acute Respiratory Distress Syndrome (ARDS) Clinical Trials Network, Wheeler AP, Bernard GR, et al.: Pulmonary artery catheter versus central venous catheter to guide treatment of acute lung injury. N Engl J Med 2006; 354:2213--416. 17. Rick JJ, Burke SS: Respirator pulsus paradox. South Med J 1978; 71:1376--8. 18. Michard F: Changes in arterial pressure during mechanical ventilation. Anesthesiology 2005; 103:419--27. 19. Perel A: Automated assessment of uid responsiveness in mechanically ventilated patients. Anesth Analg 2008;106:1031--3. 20. Chemla D, Hebert JL, Coirault C, et al.: Total arterial compliance estimated by stroke volume-to-aortic pulse pressure ration in humans. Am J heart Circ Physiol 1998; 274:500--5. 21. Bland JM, Altman DG: Statistical methods for assessing agreement between methods of clinical measurement. Lancet 1986; 8:307--10. 22. 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. 23. Chaney JC, Derdak S: Minimally invasive hemodynamic monitoring for the intensivists. Current and emerging technology. Crit Care Med 2002; 30:2338--45. 24. Uchino S, Bellomo R, Morimatsu H, et al.: Pulmonary artery catheter versus pulse countour analysis: A prospective epidemiologic study. Crit Care 2006; 10:R174. 25. Pearse RM, Ikram K, Barry J: Equipment review: An appraisal of the LiDCO plus method of measuring cardiac output. Crit Care 2004; 8:190--5. 26. Pearse R, Dawson D, Fawcett J, et al.: Early goal-directed therapy after major surgery reduces complications and duration of hospital stay. A randomized controlled trial. Crit Care 2005; 9:R687. 27. Mayer J, Bold J, Wolf MW: Cardiac output derived from arterial pressure waveform analysis in patients undergoing cardiac surgery: Validity of a second generation device. Anesth Analg 2008; 106:867--72.