You are on page 1of 14









The American Society of Anesthesiologists, Inc.

Leibowitz has disclosed that he has no financial interests in or significant relationship with any commercial companies pertaining to this educational activity. 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.. c 2009 The American Society of Anesthesiologists.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. Physicians should only claim credit commensurate with the extent of their participation in the activity.copyright.asa-refresher. www. 222 Rosewood Dr. Permission to reproduce copies of articles for noncommercial use must be obtained from the Copyright Clearance Center. MA 01923. PERMISSION TO PHOTOCOPY ARTICLES: This publication is protected by copyright.asa-refresher-cme. Author Disclosure Information Dr. FAX: (978) 750-4470. Pennsylvania 19106-3621 Library of Congress Catalog Number 74-18961. The American Society of Anesthesiologists designates this educational activity for a maximum of 1 AMA PRA Category 1 Credit(s)t. .com. or see page iv at the beginning of this volume for complete details. Please visit www. Inc. www. Published for The Society by Lippincott Williams & Wilkins 530 Walnut Street Philadelphia. (978) 750-8400.asahq. ISSN 0363-471X ISBN 978-1-6054-7424-3 An educational service to the profession under the auspices of The American Society of Anesthesiologists.

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. and the detection of pericardial effusion. the main function of which is to return blood to the heart. It is critical to understand that both of these beliefs are physiologically illogical and. arterial pulse pressure variation. The venous system contains 70% of the blood volume and is an extremely compliant container.g. sometimes also colloquially referred to as ‘‘echo-lite. and effectiveness of monitors that are used by anesthesiologists and intensivists in their daily practice.D. in that order. LiDCO. Furthermore. pulmonary artery catheter (PAC). PiCCO2. West Sussex. The included devices are representative and none of them has clearly distinguished itself as markedly superior to the others. From a physiologic standpoint. M. New York This Refresher Course reviews the physiologic underpinnings.Hemodynamic Monitoring: From Central Venous Pressure to Pulse Contour Analysis Andrew Leibowitz. through studies. performed by physicians who are less than ‘‘fully’’ trained and lack certification from the American Board of Echocardiography.. it makes sense that volume status cannot be inferred either from a single measurement of CVP or from its change over time. Deltex Medical. However. and Doppler (e. Inc. and the proprietary technologies of NICO. Most of the blood contained within the venous system exists within the system at a transmural Copyright Ó2009 American Society of Anesthesiologists. Hemodynamic monitoring has gradually shifted from invasive intravascular monitoring toward less invasive devices. CardioQ. It is a false but widely held belief that CVP can be used to estimate intravascular volume status. I review. thus.2 and Wilson et al.’’ which primarily focuses on cardiac filling. clinical application. FloTrac/Vigileo. 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. Only those monitors that are available for use by the average practitioner are discussed. have been thoroughly dismissed. transesophageal echocardiography is not covered in detail because of the significant additional training and certification required.3 CVP is simply an intravascular pressure ideally measured in the superior vena cava.1 Seldinger. 119 . techniques and devices including central venous pressure (CVP). ventricular wall motion. it should be noted that there is an active movement toward recognition of ‘‘limited scope’’ or ‘‘goal-directed’’ echocardiography. Associate Professor of Anesthesiology and Surgery Mount Sinai School of Medicine New York. My inclusion or omission of any proprietary device should not be interpreted as my personal recommendation or disapproval. United Kingdom).

that is. the ‘‘unstressed’’ volume (Vu) is the volume less than the hole that will exist at a pressure of zero.. The tub represents the blood volume. those with ongoing blood loss. the relationship between Vs and Vu. the inability of the CVP to predict the potential of the cardiac output (CO) or blood pressure to increase with fluid administration has been the subject of 19 other studies. and measurement of mixed venous oxygen saturation led to a widespread acceptance of this technique within only a few years of its clinical introduction. Representation of the venous volume and dynamics that result in central venous pressure (CVP). changed by moving the hole up and down. it is certain that no useful relationship between CVP and blood volume exists.5 In addition. Therefore. balloon-tipped PAC’s ability to measure CO and pressure in the pulmonary artery represented a great physiologic advance. whereas circulatory physiology dictates that a certain positive CVP be maintained for blood to return to the heart and homeostasis to exist (Fig.4 Therefore. The physiologic premises dismissing the relationship between CVP and volume have been clinically demonstrated in five studies using different methods of determining blood volume. 1). which should equal the left atrial and left ventricular pressure at end diastole. in patients with cardiac depression from anesthesia or sepsis. All of these studies can now be reviewed very succinctly in a systematic analysis that was just published by Marik et al. The measured pressure at the hole will result from a combination of the inflow from the smaller arterial volume contained in the high-resistance arterial vessels. The ‘‘stressed volume’’ (Vs) is the volume above the hole where the pressure is measured that will result in a positive pressure.120 LEIBOWITZ FIG. These easily obtained measurements combined with the wedge (or pulmonary artery occlusion) pressure.7 Problems with this technique were originally thought to be technical in nature and related to the . in one recently published study by Kumar et al. Reprinted with permission from Gelman.5 pressure of zero. Pulmonary Artery Catheter The flow-directed. or those with systemic vasodilation.6 in volunteers given a 3-l fluid challenge. 1. and the venous resistance (VenR). 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. 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 excessive decrease in systolic pressure seen with spontaneous ventilation in the presence of cardiac tamponade. in the past 7 years. Increased heart failure and 215 matched noncardiac events in PAC group pairs 1.8 Later.HEMODYNAMIC MONITORING 121 potential failure of the wedge pressure to accurately estimate the left atrial pressure. The Canadian Multicenter Trial10 reported that there was no difference in outcome in high-risk.041 None 201 433 676 981 Increased renal insufficiency and thrombocytopenia in PAC group Increased infections in PAC group None Increased related complications and blood transfusions in PAC group ARDS ¼ adult respiratory distress syndrome. were demonstrated. even in highly competent hands using sophisticated management algorithms! Two of these trials10. has long been recognized as an adverse clinical sign.’’ Arterial Pulse Pressure Variation Pulsus paradoxus. This ‘‘matched pairs’’ analysis yielded 215 pairs of patients for investigation.994) randomized to management with PACs and an ‘‘optimization’’ filling pressure/oxygen delivery regimen versus standard care. PAC ¼ pulmonary artery catheter. More recently. American Society of Anesthesiologists 3 and 4 patients (n ¼ 1.11 used ‘‘propensity scoring’’ to prospectively compare nonrandomized patients at identical risk who were managed with and without a PAC.18. Clinical Summary of Selected Recent Large Investigations Comparing Management With Versus Without a PAC Number of Patients Enrolled 1. widespread errors in PAC use.19 Understanding the three main physiologic premises TABLE 1. Moreover. user errors in obtaining and interpreting the wedge pressure. Remarkably.9 More troubling. elderly. the effect of positive pressure ventilation on the arterial pulse pressure has been described17 and the physiologic basis extensively reviewed.059 total. most importantly.11 are particularly pertinent to anesthesiologists. the PAC group experienced a small but statistically significant increased total complication rate related to PAC insertion. Several of these investigations suggest a worse outcome when a PAC is used. ICU ¼ intensivecare unit.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 Significant Outcome Differences More adverse events in PAC group related to insertion 4. Polanczyk et al. . patients managed with a PAC had significantly more congestive heart failure and other adverse events than their non-PAC-managed ‘‘controls. including. 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).

(c) 100 Â (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. (2) a positive pressure breath compresses the pulmonary venous system. 410%) is probably the best predictor that the blood pressure will increase with the administration of intravenous fluids 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 clinician’s ability to use these principles successfully. a decrease in the pulse pressure will follow. This decreases the left ventricular filling. in hypovolemia.18 . Reprinted with permission from Michard. The variation of arterial pulse pressure with positive pressure breaths.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. thus. the right atria and vena cava are thus more easily compressed. but its presence (i. the driving force for venous return. Figure 2 demonstrates this phenomenon. See text for explanation. 2. the change in pulse pressure is greater than in normovolemia. LiDCO. the mean circulatory filling pressure. FIG.. and (3) a positive pressure breath also decreases the venous return to the right heart by increasing intrathoracic pressure. Several caveats apply to the clinical application of this principle: (1) There is no universal agreement as to which exact variable should be tracked. and the left ventricle is already operating on the steep portion of the Frank--Starling curve. For several reasons. (b) maximum systolic pressure--minimum systolic pressure. after a few beats of the increased pulse pressure that immediately follow a positive pressure breath.e. Simply summarized. in the presence of hypovolemia. The above principles were used in the design of the PiCCO. 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. but possibilities include: (a) systolic pressure at end apnea--minimum systolic pressure (Ddown). is reduced. causing an increase in left ventricular preload and an increase in stroke volume and arterial pulse pressure for a few beats.

CvO2 ¼ arterial. R ¼ rebreathing. Clinically.’’ Physicians are most familiar with the Fick principle applied to oxygen used to calculate CO in which: V O2 CO ¼ CaO2 ÀCvO2 where VO2 ¼ oxygen consumption. venous CO2 should be the same as under normal conditions and CvCO2N ¼ CvCO2 R and thus this equation becomes: CO ¼ VCO2 N À VCO2 R CaCO2 R À CaCO2 N All these variables can be measured by the device. Knowledge of algebra allows the following assumption: If a c ð a À cÞ a c ¼ then ¼ ¼ b d ðb À dÞ b d and is applied to the above equation then: CO ¼ VCO2 N À VCO2 R ðCvCO2 N À CaCO2 NÞ À ðCvCO2 R À CaCO2 RÞ During rebreathing. venous oxygen content. the patient must be on positive pressure ventilation with stable tidal volume and respiratory rate.e. More commonly. small differences in measurement may lead to large changes in calculated . and (2) addition of dead space and rebreathing allow mathematical calculation of the CO as follows: VCO2 N VCO2 R ¼ CO ¼ CvCO2 N À CaCO2 N CvCO2 R À CaCO2 R where N ¼ normal. therefore. 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. CaCO2. 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.. venous CO2 content. The number and variety of physiologic assumptions and mathematical manipulations are the main drawbacks to using this monitor. intensivists interested in the determination of oxygen consumption rearrange this equation to determine oxygen consumption as follows: VO2 ¼ CO Â ðCaO2 ÀCvO2 Þ The NICO system (Philips/Respironics) adapts this principle to CO2 to determine CO as follows: VCO2 CO ¼ CvCO2 ÀCaCO2 where VCO2 ¼ CO2 clearance. The difference in the N and R CO2 is usually quite small (i.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 flow to the organ and the arteriovenous concentration difference of the substance. To use this device. CaO2. CvCO2 ¼ arterial. o 10 mmHg).

21 and the subsequent publication of one particularly influential meta-analysis. 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. precision (71 standard deviation of the bias). All these devices allow continuous monitoring of the arterial pulse pressure variation.23 The PiCCO and the original LiDCO (the ‘‘plus’’ model) devices require that the CO of the device be calibrated to another method. On the basis of the methods of Bland and Altman. In addition. From the analysis of the pulse contour. The device relies on several principles that are relatively new to most anesthesiologist and intensivists.’’ After calibration. stroke volume variance is also reported. These underlying physiologic limitations combined with the ability to only monitor CO alone seem to be the main reasons for NICO’s relatively small market penetration. or mathematically stated: R dP=dt SV ¼ Z This is a variation of the idea that stroke volume is equal to pulse pressure divided by arterial compliance. Germany) is widely used in Europe and was recently introduced in the United States. shunted blood containing CO2 is eliminated from consideration in this model. These two catheters are used to determine CO by means of the injection of a cold fluid bolus into the superior vena cava and monitoring of the temperature change in the artery. The determination of CO at any time will allow calculation of z after which only periodic recalibration will be required. . it is critical to note the standard to which the new monitoring devices are held when reporting CO.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. A recently updated version of the LiDCO (the ‘‘rapid’’ model) also allows noncalibrated CO to be continuously reported. which. 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). Munich. may be useful.’’ Data comparing these devices with thermodilution-determined CO usually report the bias (the average difference between the two). the device reports CO and stroke volume continuously. PiCCO The PiCCO device (Pulsion Medical Systems. whereas the FloTrac/Vigileo uses a more mathematically advanced algorithm and requires no calibration. this is done via a modified Stewart--Hamilton equation and the resultant value is termed the ‘‘transpulmonary thermodilution CO.124 LEIBOWITZ CO. from a clinical and statistical point of view. Figure 3 illustrates this method. PACmeasured CO is usually determined by three consecutive measurements that may differ by 15% or more. Furthermore. 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. even in the absence of other functionality.

United Kingdom). In addition. although the PiCCO group had a greater positive fluid balance and a longer duration of mechanical balance. . A few downsides to this device result from the use of lithium injections. One recent study reported a prospective (nonrandomized) multicenter investigation of 331 intensive care unit patients managed with a PAC versus PiCCO. lithium carbonate. preload (referred to as global end-diastolic volume). rapidly changing hemodynamic conditions will warrant repeated cold water injectate to reliably obtain a properly calibrated pulse contour analysis. hyponatremia. 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. PulseCO) than the PiCCO device.. requires CO to be measured to ‘‘calibrate’’ its internal algorithm. 3. Cambridge. and hyponatremia interfere with lithium injection as a calibrating standard.e. the two main outcomes investigated between these two groups. The LiDCO device then relies on a more complex algorithm (i. The calibration of cardiac output with pulse contour. T ¼ temperature.25 It uses a peripheral injection of lithium ion 0.3 mMol with a 15-ml saline flush and a proprietary arterial line containing a lithium sensor to construct a dilution curve for the lithium ion and: COðL=minÞ ¼ R Lithium dose in mmol  60 R ð1 À PCVÞ Â d½LiŠ=dt d[Li]/dt is the area under the primary curve and PCV is the packed cell volume [Hb (g/dL)/34]. 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. and some other drugs that contain quaternary ammonium ions (including some muscle relaxants). and afterload. much like the PiCCO device. t ¼ time. Reprinted with permission from Pulsion-supplied information package. 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. A correction for PCV is necessary because lithium is distributed in the plasma.15 to 0.HEMODYNAMIC MONITORING -∆T -∆T 125 t Calibration P [mm Hg] SV t t [s] FIG. The drug.. The manufacturer recommends recalibration at least every 8 hours. It assumes that arterial compliance changes with blood pressure and this change is similar in all humans.24 There was no difference in the length of stay or mortality. LiDCO The original LiDCO device (LiDCO Ltd.

27 The underlying mathematics and physics used are difficult for the mathematically unsophisticated nonengineer to comprehend. the velocity changes with time and the area under the velocity versus time curve (velocity--time integral). 4. 4). CardioQ) CO can also be determined completely noninvasively by using small esophageal echo-type probes.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.g. and corrected flow time. which is a velocity Velocity VTI = area under velocity time curve = velocity time integral Time Cross sectional area = πr2 Stroke Volume = cross sectional area x velocity time integral FIG.126 LEIBOWITZ A randomized trial by Pearse et al. . and limits of agreement were improved. Esophageal Doppler (e. precision. The technique relies on the fact that flow through a tube is equal to the cross-sectional area of the tube multiplied by the velocity--time integral of the fluid. stroke volume (Fig. CA) differs significantly 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.. therefore. How the stroke volume is calculated by esophageal Doppler devices: Stroke volume ¼ cross--sectional area  velocity--time integral. Irvine. 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. No outcome study using the most recent version of this device has been reported. In the case of pulsatile flow. Doppler monitoring devices may use transthoracic or transesophageal probes. it is difficult to draw any broad conclusion from this study. 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. This was a small study of only 122 patients and the treatment group received dopexamine. weight. FloTrac/Vigileo The FloTrac/Vigileo system (Edwards Life Sciences. 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. This device has recently had its algorithm updated after which bias. is used as shown in Figure 4. height. the velocity--time tracing is used to report continuous CO. and sex.

peak velocity. Vahid B: Does central venous pressure predict fluid responsiveness? A systematic review of the literature and the tale of seven mares. References 1. 4. Demong CV. Fischer EP. stroke volume. a new technique. Conclusion Although CVP and measurements available from the PAC show no association with estimated intravascular volume and provide no proven outcome benefit for manipulating hemodynamic values. the CardioQ is nearly unique in that several outcome studies have been published attesting to its effectiveness. 32:691--9. 348: 5--14. Seldinger SI: Catheter replacement of the needle in percutaneous arteriography. Although there are several versions of esophageal CO devices that are available commercially. Leibowitz AB.: Catheterization of the heart in man with use of a flowdirected balloon-tipped catheter. there are newer monitoring techniques that may provide benefit to patients. 5. 9. et al. To date. 264:2928--32.: Pulmonary artery occlusion pressure and central venous pressure fail to predict ventricular filling volume. Chest 2008. 89:427--34. 7. although these devices can nearly duplicate the CO data supplied by the PAC (and do so continuously) and provide data heretofore unavailable to clinicians. the majority of the patients studied were at a relatively minimal cardiac risk and the outcome improvements (i. 8. N Engl J Med 1970. Sandham JD. Acta Radiol 1952. controlled trial of the use of pulmonary-artery catheters in high-risk surgical patients. these devices have not been demonstrated to change outcomes. Ganz W. 60: 1456--58. Forrester J. 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 fluid optimization. These monitors require a better understanding of how the circulatory system adjusts to changes in blood volume and extracellular fluid volume. length of stay) are of unknown importance because of the long length of stay in both the monitored and control groups. Chest 1986. The manufacturer suggests using this device to construct a volume response optimization strategy in which the effect of successive boluses on CO. Presse Med 1952. 10. Grow JB. Arch Surg 1962. JAMA 1990. and velocity can be seen and fluid titrated to optimize CO. 108: 735--48. Crit Care Med 2004. . Raper R. Pulmonary Artery Catheter Study Group. or the response to volume infusion in normal subject. Kumar A.: Central venous pressure in optimal blood volume maintenance. Hull RD. Brant RF. 134: 172--8. minute distance. 39:368--76. Aubaniac RL: Intravenous subclavicular injection [in French].e. On that basis. Anesthesiology 2008. Baram M. and heart rate. Wilson IN. 2. et al. Marik PE. Swan HJ.. 283:447--51. Bunnell E. cardiac performance. 3.’’ However. 6. Gelman S: Venous function and central venous pressure. Sibbald WJ: Misled by the wedge? The Swan-Ganz catheter and left ventricular preload. et al.: A randomized.HEMODYNAMIC MONITORING 127 measurement corrected for age. N Engl J Med 2003. et al. et al.: A multicenter study of physicians’ knowledge of the pulmonary artery catheter. 85:563--78. Anel R. Iberti TJ.

Rick JJ. Chemla D. 103:419--27. Cusack RJ. 12. 17. Chaney JC. 16.: Pulmonary artery catheter versus central venous catheter to guide treatment of acute lung injury. et al. Morimatsu H. 23. Warszawski J. Crit Care 2004. Singer M. 30:2338--45. 294:1625--33. et al. Bellomo R. 14. et al. Coirault C. Lung. Anguel N. Anesth Analg 2008.: Pulmonary artery catheter versus pulse countour analysis: A prospective epidemiologic study. Polanczyk CA. 15:85--91. J Clin Monit Comput 1999. 8:190--5. 27. N Engl J Med 2006. Hebert JL. Rhodes A. 20.106:1031--3. Newman PJ. 10:R174. 19. Anesthesiology 2005. Perel A: Automated assessment of fluid responsiveness in mechanically ventilated patients. Mayer J. Rohde LE. Bold J. Crit Care 2005. and Blood Institute Acute Respiratory Distress Syndrome (ARDS) Clinical Trials Network. Bland JM. JAMA 2003. 286:309--14.: Total arterial compliance estimated by stroke volume-to-aortic pulse pressure ration in humans. Hasselblad V. Lancet 2005.: A randomised. The National Heart. et al. et al. A randomized controlled trial. Michard F: Changes in arterial pressure during mechanical ventilation. Burke SS: Respirator pulsus paradox. Califf RM.: Evaluation study of congestive heart failure and pulmonary artery catheterization effectiveness: The ESCAPE trial. Am J heart Circ Physiol 1998. Wheeler AP. South Med J 1978.: Assessment of the clinical effectiveness of pulmonary artery catheters in management of patients in intensive care (PAC-Man): A randomised controlled trial. Crit Care 2006. Barry J: Equipment review: An appraisal of the LiDCO plus method of measuring cardiac output. 15. 366:472--7. JAMA 2005. controlled trial of the pulmonary artery catheter in critically ill patients.: Early goal-directed therapy after major surgery reduces complications and duration of hospital stay. et al. Wolf MW: Cardiac output derived from arterial pressure waveform analysis in patients undergoing cardiac surgery: Validity of a second generation device. Crit Care Med 2002. et al. 71:1376--8. Bernard GR. Dawson D. 21. 28:256--64. Anesth Analg 2008. Harvey S. Altman DG: Statistical methods for assessing agreement between methods of clinical measurement. 25. 13.: Early use of the pulmonary artery catheter and outcomes in patients with shock and acute respiratory distress syndrome: A randomized controlled trial. 274:500--5. JAMA 2001. Ikram K.: Right heart catheterization and cardiac complications in patients undergoing noncardiac surgery: An observational study. Critchley LA. Derdak S: Minimally invasive hemodynamic monitoring for the intensivists. Binanay C. Current and emerging technology. Fawcett J. Goldman L. 8:307--10. 18. Critchley JA: A meta-analysis of studies using bias and precision statistics to compare cardiac output measurement techniques. et al. Richard C. Uchino S. 106:867--72.128 LEIBOWITZ 11. 290:2713--20. 22. Pearse R. 24. Intensive Care Med 2002. . Harrison DA. 9:R687. et al. 26. 354:2213--416. Pearse RM. Lancet 1986.