Advanced cardiovascular monitoring

Transcription

1 Advanced cardiovascular monitoring Eric de Waal

2 Advanced cardiovascular monitoring E.E.C. de Waal ISBN: Printed by Gildeprint Drukkerijen, Enschede, Netherlands E.E.C. de Waal, Cover image: écluse de Négra près de Villefranche-de-Lauragais, Canal du Midi, Haute-Garonne, France. Photo by "Pinpin". Cover layout : M. van Ginkel. No part of this thesis may be reproduced in any form, by print, microfilm, or any other means, without written permission of the author.

3 Ter nagedachtenis aan mijn vader, Voor mijn moeder, Voor Marjan, Voor Renske, Marieke, Robin, Martijn en Marcel

4

5 Advanced cardiovascular monitoring Geavanceerde cardiovasculaire monitoring (met een samenvatting in het Nederlands) Proefschrift ter verkrijging van de graad van doctor aan de Universiteit Utrecht op gezag van de rector magnificus, prof. dr. J.C. Stoof, ingevolge het besluit van het college voor promoties in het openbaar te verdedigen op woensdag 24 juni 2009 des middags te uur door Eric Eugène Celina de Waal geboren op 20 mei 1961 te Nijmegen

6 Promotoren: Prof. dr. C.J. Kalkman Prof. dr. W.F. Buhre Financial support by Hemologic BV and Thoratec Corporation for the publication of this thesis is gratefully acknowledged.

7 Table of contents List of abbreviations 9 Chapter 1 General introduction 11 Chapter 2 The pulmonary artery catheter in anaesthesia and intensive care 17 Chapter 3 Cardiac output monitoring 43 Chapter 4 Validation of a new arterial pulse contour based cardiac output device 59 Chapter 5 Haemodynamic changes during low-pressure CO 2 pneumoperitoneum 75 in young children Chapter 6 Assessment of stroke volume index with three different bioimpedance 89 algorithms: lack of agreement compared to thermodilution Chapter 7 Monitoring of cardiac preload 99 Chapter 8 Dynamic preload indicators fail to predict fluid responsiveness in open 123 chest conditions Chapter 9 Stroke volume variation obtained with FloTrac/Vigileo fails to predict 139 fluid responsiveness in coronary artery bypass graft patients Chapter 10 General discussion and conclusions 149 Chapter 11 Appendix 161 Samenvatting (summary in Dutch) List of publications Dankwoord (acknowledgements in Dutch) Curriculum vitae

8

9 List of Abbreviations ABF Aortic Blood Flow Ao Aorta ARDS Acute Respiratory Distress Syndrome ASA American Society of Anesthesiologists BSA Body Surface Area CABG Coronary Artery Bypass Grafting CCO Continuous Cardiac Output CI Cardiac Index CO Cardiac Output CO 2 Carbon Dioxide CSA Cross Sectional Area CVP Central Venous Pressure DO 2 Oxygen Delivery P Change in Pressure V Change in Volume DPAP Diastolic Pulmonary Artery Pressure GEDV Global End-diastolic Volume GEDVI Global End-diastolic Volume Index HR Heart Rate IAP Intraabdominal Pressure ITBV Intrathoracic Blood Volume ITBVI Intrathoracic Blood Volume Index LAP Left Atrial Pressure LiDCO Lithium Dilution Cardiac Output LVEF Left Ventricular Ejection Fraction LVEDA Left Ventricular End-diastolic Area LVEDP Left Ventricular End-diastolic Pressure LVEDV Left Ventricular End-diastolic Volume MAP Mean Arterial Pressure MPAP Mean Pulmonary Artery Pressure PAC Pulmonary Artery Catheter PAOP Pulmonary Artery Occlusion Pressure PAP Pulmonary Artery Pressure 9

10 PCCO PCWP PEEP PIP PPV PVR RA RAP RSVT RV RVEDV RVEF SCO ScvO 2 SPAP SPV SvO 2 SV SVI SVV TEE TPCO TPTD VigileoCO VO 2 Pulse Contour Cardiac Output Pulmonary Capillary Wedge Pressure Positive End-expiratory Pressure Peak Inspiratory Pressure Pulse Pressure Variation Pulmonary Vascular Resistance Right Atrium Right Atrial Pressure Respiratory Systolic Variation Test Right Ventricle Right Ventricular End-diastolic Volume Right Ventricular Ejection Fraction Stat Cardiac Output Central Venous Oxygen Saturation Systolic Pulmonary Artery Pressure Systolic Pressure Variation Mixed Venous Oxygen Saturation Stroke Volume Stroke Volume Index Stroke Volume Variation Transesophageal Echocardiography Transpulmonary thermodilution Cardiac Output Transpulmonary Thermodilution Cardiac Output obtained with Vigileo Oxygen Consumption 10

11 Chapter 1 Introduction

12

13 Introduction Introduction The primary goal of the circulation is to supply adequate amounts of oxygen to meet the metabolic demands of the peripheral tissues. Adequate oxygen delivery depends on cardiac output (CO), oxygenation and oxygen carriers, i.e. hemoglobin. The monitoring of these parameters in the operating room and the intensive care is of outstanding importance, in particular in high risk patients [1]. Whereas most patients can be monitored adequately with non-invasive parameters, such as heart rate, blood pressure and oxygen saturation, an increasing number of patients are admitted with co-morbidity requiring additional physiologic monitoring. Moreover, in hemodynamic unstable patients, more sophisticated, even invasive monitoring may be necessary, not only to assess status of the circulation, but also to detect the response to different therapeutic interventions, provided that the chosen parameters are valid. Whereas hemoglobin and parameters of oxygenation can easily be measured, CO critically depends on cardiac preload, both being major targets of hemodynamic monitoring. 1 The Frank-Starling curve (figure 1.1) [2] describes the curvilinear relationship between cardiac preload and CO. The ventricle may operate in the preload-dependent or preload-independent portion of the Frank-Starling curve. Under normal physiologic conditions, both ventricles operate on the ascending part of the Frank-Starling curve, resulting in an increase in stroke volume after a given increase in preload. A further increase in preload does not necessarily result in an appropriate increase in stroke volume, especially when the left ventricle operates in the non-dependent part of the Frank-Starling curve. This is often the case in patients with preexisting heart failure or insufficiency which is likely to decompensate during the perioperative period. Furthermore, rather than the assessment of static preload parameters (e.g. central venous pressure, pulmonary capillary wedge pressure, left ventricular end-diastolic area), knowledge whether a volume challenge will improve hemodynamics in an individual patient is of major importance (fluid responsiveness). 13

14 Chapter 1 Figure 1.1 Frank-Starling curve. (Source (modified): Chest 2003; 124: ). Until recently, the Pulmonary Artery Catheter (PAC) was the only device for advanced monitoring of the circulation. However, recently, more sophisticated and considerable less invasive monitoring techniques have been developed (e.g. PiCCO, Pulsion, Munich, Germany; Vigileo, Edwards Lifesciences, Irvine, CA, USA) and their incremental utility in the plethora of existing cardiovascular monitoring systems remains to be determined. Therefore we investigated the validity and reliability of these CO and preload monitoring techniques. In chapter 2, the pro s and con s of CO monitoring with the pulmonary artery technique are described. In chapter 3, we describe the advantages and disadvantages of existing CO monitoring techniques in adults, including their reliability compared to the gold standard thermodilution CO technique. In chapter 4, the accuracy and reliability of a relatively new arterial pulse contour technique (Vigileo ) in adult CABG patients was studied. In chapter 5, we tested the ability of thoracic electrical bioimpedance to track changes in CO in small children. In chapter 6, the reliability of various noninvasive electrical bioimpedance algorithms to estimate CO was investigated compared to a gold standard with known accuracy. As already pointed out, CO critically depends on preload. In the second part of this thesis, we focused on different preload indicators. In chapter 7, an overview of different static (e.g. central venous pressure, pulmonary capillary wedge pressure, left ventricular end-diastolic area) and dynamic (e.g. pulse pressure variation, stroke volume variation) cardiac preload parameters and their ability to predict fluid responsiveness in clinical practice is given. Chapter 8 reports the ability of dynamic preload parameters obtained with PiCCO to predict fluid responsiveness during open-chest conditions. In chapter 9, we investigated if SVV obtained with Vigileo provide adequate functional hemodynamic monitoring. Finally, in chapter 10, we discuss the 14

15 Introduction validity and reliability of these innovative techniques and try to position them among the existing cardiovascular monitoring systems. 1 15

16 Chapter 1 References 1. Pinsky MR. Hemodynamic evaluation and monitoring in the ICU. Chest 2007; 132: Tokuda Y, Song MH, Mabuchi N, et al. Right ventricular end-diastolic volume in the postoperative care of cardiac surgery patients. A marker of the hemodynamic response to a fluid challenge. Circ J 2007; 71:

17 Chapter 2 Pulmonary artery catheter in anaesthesiology and intensive care EEC de Waal MD 1, 2, L de Rossi MD 3, WF Buhre MD 2 1 Division of Intensive Care Medicine, 2 Division of Perioperative and Emergency Medicine, Department of Anaesthesiology, University Medical Center Utrecht, Netherlands; 3 Klinik für Anästhesiology, Universitätsklinikum der RWTH Aachen, Germany. Anaesthesist 2006; 55: (translated)

18 Chapter 2 Abstract The indication for the use of the pulmonary artery catheter (PAC) in high risk patients is still a matter of discussion. Observational studies suggested that the use of the PAC does not result in decreased mortality but may even lead to an increase in mortality and morbidity. Due to this controversy, a number of randomized controlled trials have been performed throughout the past years in patients suffering from sepsis, ARDS, congestive heart failure, multi-organ failure and those undergoing high-risk non-cardiac surgery. The majority of recent randomized studies failed to demonstrate any benefit of the PAC with respect to reducing mortality and morbidity. There was, however, no association between use of the PAC and an increase in morbidity and/or mortality. This review gives an overview of clinical parameters obtained by the current generation of PACs, alternatives to the PAC and recent studies on the use of the PAC in clinical practice. 18

19 Pulmonary artery thermodilution cardiac output Introduction Few monitoring techniques in recent times have been open to so much controversy or have been studied so closely as the pulmonary artery catheter (PAC) and its use in clinical medicine [20]. Fueling this discussion is a number of new randomized trials and meta-analyses [21, 30, 38, 39, 42, and 43] in which the influence of the use of the PAC on morbidity and mortality in critically ill patients has been investigated. Even after more than 30 years of clinical use and more than 1.5 million applications per year, there are still open questions about indications, objective parameters and complications associated with the PAC [34]. Many recent papers and reviews give conflicting opinions which are not evidence based but instead rely on expert opinion or results from underpowered non-randomized studies. In this overview, we present the current position of the use of the PAC in the treatment of critically ill patients. After reading this article, the reader should be able to recognize the value of obtained parameters, be informed about alternative techniques and methods and understand the current indications of the use of PAC in clinical routine. 2 History and current use The PAC was introduced into clinical practice in 1970 by Swan and Ganz [47] and was initially used mainly for cardiac patients. Relatively quickly, the use was expanded to the field of anesthesia and intensive care. The PAC was the first available bedside monitoring technique which measured cardiac output (CO) and pressures in the pulmonary circulation (systolic pulmonary artery pressure: SPAP, diastolic pulmonary artery pressure: DPAP, mean pulmonary artery pressure: MPAP and the pulmonary artery occlusive pressure: PAOP). The PAC also made the continuous measurement of mixed-venous oxygen saturation (SvO 2 ) possible by either blood gas analysis or by using fiber optic techniques. Because of the bedside availability of cardiovascular monitoring, the use of the PAC became increasingly popular both in medical as well as in perioperative intensive care practice in the 1970s and 1980s. In 1996 Connors et al. published a multicentre cohort study in which they showed that the use of the PAC in a heterogeneous patient population did not result in a reduction in morbidity and mortality rate, but instead resulted in an increase in mortality in the same patient group [12] (figure 2.1). This study led to a renewed debate about the benefits and risks of the PAC in different categories of patients. In recent years, a number of randomized clinical trials with a sufficient number of patients in different groups reflecting different types of patients have been carried out, in which the influence of the PAC on morbidity and mortality rates has been investigated [21, 30, 39, 42]. The assessment of the benefits of the PAC is, however, complicated by the fact that a specific parameter is usually linked with a specific therapeutic action. It is therefore not the measured parameter that 19

20 Chapter 2 affects the outcome of the patient, but the therapeutic approach undertaken as a result of the measured parameter. Unfortunately, there is no generally accepted therapeutic approach to certain values of parameters measured with the PAC which leaves certain conclusions open to discussion. Figure 2.1 Influence of right heart catheterization on mortality in the intensive care unit [12]. No RHC = without right heart catheterization, RHC = with right heart catheterization. (Source: JAMA 1996; 276: ). In the current guidelines of the American Society of Anesthesiology (ASA), the use of the PAC is recommended in a number of clinical conditions [1]. Frequent among these conditions are states which cause a significant deterioration in cardiac function; for example, cardiogenic shock due to acute myocardial infarction or the perioperative period in patients with severely impaired cardiac function. These recommendations are not supported by randomized clinical studies, but almost entirely by case reports, non-randomized trials, small studies and expert opinions (table 2.1). Because of this confusion, the use of the PAC in clinical practice is becoming increasingly abandoned. However, attention is directed more and more on the risks associated with the use of the PAC. The aim of this review is to identify groups of patients in which the use of the PAC does not influence morbidity and mortality, to identify which group of patients will benefit or harm, and to identify therapeutic concepts that confirm the use of the PAC as valuable. In recent years, alternative less invasive techniques than the PAC have been developed to determine CO. These include the transpulmonary indicator dilution techniques [8, 9, 23], esophageal Doppler [46] and transesophageal echocardiography (TEE) [23]. The importance of each technique is evaluated in 20

21 Pulmonary artery thermodilution cardiac output association with the relevant PAC measurements and possible alternatives for the use of the PAC are presented. Indication Myocardial infarction with: Hypotension/Cardiogenic Shock Mechanical Complication Right Ventricle Infarction Answer Yes Yes Yes Degree of recommendation Heart Failure Uncertain D E E E 2 Pulmonary Hypertension Uncertain E Shock or hemodynamic instability Uncertain E Cardiac Surgery: Low Risk High risk Peripheral vascular surgery: Reduction of the complication rate Reduction of morbidity Aortic Surgery: Low Risk High risk No Uncertain Yes Uncertain Uncertain Yes Geriatric Patients for Surgery No E Neurosurgery Uncertain E Preeclampsia Not a routine E Trauma Yes E Sepsis/Septic Shock Uncertain D Supranormal Oxygen Delivery: SIRS High Risk Surgery Uncertain Uncertain Lung Failure Uncertain E Critically ill paediatric patients Yes E C C D D B E B C Table 2.1 Pulmonary Artery Catheter Consensus Conference 1997: Consensus Statement Does management with pulmonary artery catheters improves patient outcomes? A = Supported by at least two degree-i studies; B = Supported by a level-i-based study, C = Supported only by grade-ii studies; D = Supported by least one level III study; E = Supported by evidence IV or V. Evidence Grade: I = large randomized studies with clear results; II = small randomized trials with uncertain results; III = not randomized, simultaneous controls; IV = not randomized, historical controls and expert opinion, V = case studies, uncontrolled studies and expert opinion. Benefits and Risks In the United States, about 5 million central venous catheters and 1 million PAC s are used every year. Complications of the PAC may be caused by injury during the venous puncture, the insertion 21

22 Chapter 2 of the PAC into the pulmonary artery or during the time course of the PAC in the pulmonary circulation. A recently published "Closed Claims Analysis" of the ASA [16], in which the relevant medico legal complications of the central venous puncture and the introduction of central venous catheters were investigated, showed that establishment of a central venous line or a PAC is associated with a significant risk for morbidity and mortality [16]. The relevant complications for central venous cannulation include Seldinger embolus and air embolism, hemato- and pneumothorax and accidental arterial puncture. Rare complications are the damage of surrounding structures such as the thoracic duct, phrenic nerve or the brachial plexus. During the introduction of the PAC, and passage of the PAC through the right atrium and right ventricle into the pulmonary artery tree, transient arrhythmias may occur in approximately 35 % of patients, which rarely require a therapeutic intervention (table 2.2, table 2.3). Symptomatic ventricular tachycardias or ventricular fibrillation occur very rarely. Because of the severity of these complications and the need for immediate treatment, a defibrillator should always be available during the insertion of the PAC. During the time course that the PAC is used, a series of complications may occur, including local thrombosis and infections at the insertion site, in addition to systemic infection or endocarditis due to catheter infections. Further complications include the possibility of a small lung infarction due to temporary closure of a branch in the pulmonary artery tree or accidental longer-than-necessary inflation of the PAC balloon [42]. The most serious complication of the PAC is pulmonary artery rupture, which is associated with a high mortality risk (41-70%) [34]. The risk of pulmonary artery rupture is % which increases with increasing age, preexisting pulmonary hypertension or an inadequately positioned (inflated) balloon (table 2.2) [34]. Risks and complications Difficult venous puncture (haematoma, pneumothorax, arterial puncture) Arrhythmias Damage of heart valves Lung infarction and Thrombosis Pulmonary Artery Rupture Infection Time of occurrence Puncture Introduction PAC in situ PAC in situ Introduction PAC in situ Table 2.2 Complications of the Pulmonary Artery Catheter. The PAC is a highly invasive monitoring instrument which should be reserved for a limited number of clinical indications and should only be used for the shortest possible period of time. The 22

23 Pulmonary artery thermodilution cardiac output complication rate associated with the use of the PAC is presumed to be higher than that for other methods of CO measurement, such as the transpulmonary indicator dilution technique or echocardiography. It is therefore reasonable to consider whether the required information obtained from a PAC can be obtained with other, less invasive methods of monitoring. For example, the determination of CO is not an ultimate indication for the use of a PAC, as CO may also be determined with other less invasive techniques. 2 Complications Frequency (%) Central Venous Access Arterial puncture Bleeding at the injection site Postoperative Neuropathy Pneumothorax Air embolism Catheterization Light arrhythmias Threatening arrhythmias Mild Tricuspid insufficiency Right Bundle Branch Block Complete Heart Block PAC in situ Pulmonary Artery Rupture Positive culture (catheter tip) Catheter-associated sepsis Thrombophlebitis Pulmonary Infarction Thrombus Valvular vegetations or Endocarditis Death caused by PAC ,3-1, Incidence in the majority of studies (%) < > > Table 2.3 Incidence of complications of the Pulmonary Artery Catheter. (Source: Anesthesiology; 99: ). Measured parameters Cardiac filling pressures. For more than 20 years, the cardiac filling pressures i.e. central venous pressure (CVP) and pulmonary artery occlusion pressure (PAOP) were used to estimate the filling status of the cardiovascular system. The CVP was used as an estimate for the filling status of the venous system up to the right atrium, whereas the PAOP was used as an estimate of the left ventricular end-diastolic pressure (LVEDP), and ultimately the left ventricular end-diastolic volume (LVEDV). The aim of measuring CVP, PAOP and CO is optimization of the filling of the heart, and therefore to reach the CVP and PAOP with which the best CO is achieved. A series of clinical and experimental studies have shown that the postulated relationship between cardiac filling pressures and filling volumes in practice does not exist [10, 24]. Kumar et al. have 23

24 Chapter 2 investigated the relationship of CVP, PAOP, right and left ventricular volumes and the CO in healthy volunteers [24]. In order to vary the cardiovascular filling, 3 liter of isotonic saline was infused in healthy volunteers [24]. Simultaneous measurements of cardiac filling pressures, CO and cardiac volumes were performed before and after fluid challenge. It was shown that no positive correlation exists between changes in CVP or PAOP and changes in CO [24]. The authors conclude therefore that the relationship between cardiac filling pressures and CO in healthy volunteers was limited. In critically ill patients, however, changes in intrathoracic pressures and myocardial function exert additional influence on the relationship between filling pressures and volume status [10]. Clinical studies demonstrated that changes in CVP and PAOP do not necessarily indicate changes in right and left heart volume [10]. Figure 2.2 shows several factors influencing the determination of CVP and PAOP in detail. PA PAP PV LAP LA Ao RA CVP PCWP LVEDP RV LV PVR Mitral valve failure CVP PAP PCWP LAP LVEDP LVEDV Tricuspid valve dysfunction Increased Alveoloar Pressure Pulmonary Fibrosis LV diastolic dysfunction Figure 2.2 Representation of factors influencing the determination of CVP and PAOP. In summary, from relative and absolute changes in CVP and PAOP, only very limited conclusions can be derived regarding the cardiac filling state and myocardial function. It is not recommended to use cardiac filling pressures to dictate the fluid therapy. In order to estimate cardiac preload and the filling of the heart, other volumetric measurement techniques are more suitable [7]. Cardiac output (CO). Due to the introduction of the PAC, it was finally possible to determine CO at the bedside by using a thermodilution technique [9]. In contrast to previous methods of measuring CO using dyes, radioactively labeled erythrocytes or albumin, the thermodilution 24

25 Pulmonary artery thermodilution cardiac output technique offers the possibility to repeat measurements of CO as often as needed. The CO measurement is based on the Stewart-Hamilton principle. A bolus of indicator (usually ice-cold saline solution) is injected via the central venous line and the resulting thermodilution curve is measured downstream in the pulmonary artery. From the area under the indicator dilution curve and the injected volume, CO can be calculated. Usually, three to five bolus injections are used and the average value is calculated. Modern instruments allow the use of indicator at room temperature. CO obtained with the PAC is still the clinically accepted reference of CO determination, against which all new techniques are tested and compared. Yeldermann presented, in the beginning of the 1990s, a method by which a seemingly continuous determination of the CO (CCO = continuous cardiac output) is possible [5]. In this device, a small spiral attached to the outside of the PAC emits heat pulses, which can be detected downstream as temperature changes. Under normal conditions, the results of intermittent and CCO measurements correlate well. In patients with fluctuations in body temperature, e.g. hypothermic cardiopulmonary bypass, this continuous method however, is error-prone as the signal to noise ratio is restricted [5]. The determination of CO by pulmonary (PAC) and transpulmonary thermodilution (TPTD) is of comparable accuracy [9]. Some authors will therefore favor TPTD as the standard procedure for determination of CO in clinical routine, as it is solely reliant on the insertion of an arterial thermodilution catheter which is almost always necessary for routine monitoring in most critically ill patients [36]. In this context, it is to draw attention to the current work of Reuter and Goetz, in which the methodological and practical aspects of CO monitoring are determined in detail [36]. 2 Right ventricular ejection fraction and right ventricular end-diastolic volume (RVEF and RVEDV). A special feature of the PAC is the determination of right ventricular ejection fraction (RVEF) and right ventricular end-diastolic volume (RVEDV), which can be determined with a special thermistor catheter ( fast response"). The determination of the RVEF may be obtained intermittently by the thermodilution technique, as well as continuously [23]. A number of studies showed a good correlation of RVEF obtained with transesophageal 2-dimensional echocardiography and RVEF obtained with the PAC. However, there was a significant difference between the RVEF obtained with PAC when compared to the RVEF obtained with 3-D echocardiography [14]. The RVEF obtained with the catheter technique is approximately 15% lower than that obtained by the 3-D echocardiography [14]. Despite these differences in absolute values, the determination of the trend of RVEF is relatively valid, but the correlation at high heart rates (more than 100 beats min -1 ) is significantly worse than at normal heart rates. The thermodilution technique overestimates the RVEDV significantly [14], and the determination of 25

26 Chapter 2 both RVEF as well as RVEDV is vulnerable to factors such as heart rhythm disturbances and cardiac valvular regurgitation. To estimate cardiac preload, the determination of RVEDV is superior to the cardiac filling pressures CVP and PAOP. However, Hofer et al. showed that intrathoracic blood volume (ITBV) and left ventricular end-diastolic area (LVEDA) are superior to RVEDV as a determinant to control volume therapy [23]. Therefore, the determination of ITBV by TPTD and the determination of LVEDA by the TEE are clinical standards for the assessment of global cardiac preload [7]. The ITBV is a global parameter of cardiac filling and does not differentiate between right and left ventricular filling. However, with transthoracic and transesophageal echocardiography is a bedside determination of right and left ventricular preload possible. There are currently no studies available with a sufficient number of cases that show a positive or negative influence of RVEF or RVEDV determination on morbidity or mortality rate of patients. For this reason, the use of a PAC providing estimation of RVEF and RVEDV is reserved in most centers for specific indications, such as lung and heart transplantation, in which isolated right ventricular dysfunction is more common. Possible further indications are isolated right ventricular myocardial infarction with subsequent right heart failure. In these patients, a careful estimation of right ventricular function is of clinical significance as acute right heart failure is the leading cause of death in the initial phase. It remains however uncertain as to whether PAC guided therapy leads to an actual improvement and a reduction in mortality in such a group of patients. Mixed venous oxygen saturation. A key parameter in determining the therapy of critically ill patients in the intensive care unit is mixed venous oxygen saturation (SvO 2 ). The SvO 2 can only be determined with the PAC, because there is a complete mixture of the blood from the lower and upper part of the body in the pulmonary artery. Normal SvO 2 levels range from 65 to 80%. Lower values suggest a disturbance in organ perfusion. From a global SvO 2, one cannot derive implications about the ischemic organ, nor about the type of perfusion disorder. In recent years, central venous saturation (ScvO 2 ) has been used increasingly as an alternative for SvO 2 to adjust circulatory therapy. The determination of ScvO 2 is possible via a central venous catheter, and therefore does not require a PAC. In addition, commercially available fiber optic catheters have been developed to give a continuous measurement of ScvO 2 [35]. ScvO 2 and SvO 2 differ in absolute values as mixing of blood from the lower part of the body is not measured with the ScvO 2. Relative changes in venous oxygen saturation during hemorrhage, shock or septic circulatory conditions are mostly reflected equally. To what extent the measurement of ScvO 2 can replace SvO 2 is still controversial. Recently, Dueck et al. and Reinhart et al. published 26

27 Pulmonary artery thermodilution cardiac output studies, which suggest that ScvO 2 saturation parallels SvO 2 over a wide range of oxygen saturation [17, 35]. Pearse et al. were able to demonstrate that a ScvO 2 less than 64% is predictive for the occurrence of organ failure after major surgery [29]. It is now accepted that the early establishment of adequate organ perfusion, particularly in septic patients, is of vital importance, and with the attainment of a ScvO 2 > 65% significant as a parameter of organ perfusion. To what extent these values can be applied to other patient groups (cardiogenic shock, cardio surgical patients), has not 2 been sufficiently investigated. Rivers et al. recently showed that a treatment algorithm guided by the ScvO 2 leads to a significant improvement in survival in patients with sepsis, provided that treatment of sepsis started within 6h ( Early Goal Directed Therapy ) [40]. Therefore, ScvO 2 forms now part of the therapy algorithms in the Surviving Sepsis Campaign, [15]. However, it is essential that targets of ScvO 2 can be achieved within a defined time frame. It is essential to understand that not one separate parameter is significant, but it must be evaluated within the context of the time frame and the severity of the disease. Therapeutic algorithms, in which treatments based on SvO 2 lead to a reduction in mortality, are not incorporated in large prospective studies. Gattinoni et al. investigated in a prospective randomized study whether a treatment algorithm based on a target value of a SvO 2 greater than 70% implies a survival benefit in intensive care patients [19]. His study found no survival advantage in patients whose therapy was tailored specifically to the SvO 2 values, lacking clear definition for beginning of the disease or measurement [19]. With knowledge of CO, arterial and venous oxygen content and hemoglobin level, global oxygen delivery (DO 2 ) and oxygen consumption (VO 2 ) can be calculated. The use of the parameters DO 2 and VO 2 in clinical practice remains controversial. From a pathophysiological point of view, it is conceivable that optimizing the balance between oxygen supply and consumption improves management of cardiocirculatory function and therefore may contribute to a decreased incidence of organ failure. In the 1980s, Shoemaker et al developed a treatment algorithm where the cardiac index (greater than 4.5 liter min -1 m -2 ) and the DO 2 (greater than 600 ml min -1 m - 2) were raised to supra-normal values [45]. The importance of this therapeutic approach is now critically reviewed (see also section "clinical studies"). Systemic and pulmonary vascular resistance. With knowledge of CO and pressures in the systemic and pulmonary circulation, the systemic and pulmonary vascular resistance can be calculated: MAP CVP SVR = CO

28 Chapter 2 PAP PAOP PVR = 79.9 CO with SVR and PVR = systemic and pulmonary vascular resistance (dyne s cm -5 ), PAOP = pulmonary artery occlusion pressure (mm Hg) and CO = cardiac output (l min -1 ). The values of vascular resistance are frequently used in clinical practice to control therapy with vasoactive drugs. Especially in patients with pulmonary hypertension, the calculation of PVR before and after medical intervention is an indication for the insertion of a PAC. At the University Medical Center Utrecht, the Netherlands, a PAC is almost exclusively used in patients scheduled for heart- or lung transplantation or perioperatively in patients with known pulmonary hypertension, prone to right heart failure and total cardiac failure. In these patients, inhaled vasodilators (ino, iloprost, and epoprostenol) are used. The effectiveness of treatment with these substances is controlled with calculation of PVR and the relationship between the SVR and PVR. Whether the use of PACs in these patients reduces morbidity and mortality rate remains to be seen however, due to lack of placebo-controlled studies. The calculation of the PVR according to the above-mentioned formula is controversial. Versprille described the calculated PVR for methodical reasons as a "meaningless variable" [48]. The PAOP is used in the equation as a substitute for the left atrial pressure (LAP), but does not necessarily represent it under all conditions [48]. In experimental studies, a significant difference was found between PAOP and LAP in septic shock, as well as in induced chest trauma. Furthermore, the measurement of PAOP also depends on the position of the PAC in the pulmonary artery. In situations where the alveolar pressure exceeds the pulmonary capillary pressure, the PAOP no longer represents the effective closing pressure [48]. No other bedside monitor is currently available, allowing continuous monitoring of right ventricular function and/or pulmonary circulation. Clinical studies Until now, the benefits of the PAC for the treatment of critically ill patients remain in doubt (table 2.4, table 2.5). Contributing to this debate is a study from Connors et al. in 1996, in which the authors documented an increased mortality rate in patients with PAC monitoring [12]. This study has been controversial in its set-up as it was not prospectively randomized, included very heterogeneous patient groups and contained no PAC-based treatment guidelines [12]. There were, 28

29 Outcome PAC Controle P Therapeutic goals Therapyprotocol Number of Patients Inclusion criteria Patient group Investigator Study type ns ns ns < % 4 d 14 d idem 47.9% 5.7d 13 d Idem Mortality (28d) ICU-LOS Hospital-LOS Morbidity Except: Renal Failure Optimization of CVP and PAOP (Volume challenge until no further increase in CI); MAP > 60mmHg None Total: 201 PAC: 95 Controle: Circulatory failure 2. Oligurie 3. Vasoactive medication required 4. Mechanical ventilation General ICU Monocenter, prospective, randomised, Controlled Rhodes et al. [38], % 35% ns ns ns ns ns ns 51.3% 61% 72% 11.9 d 14.4 d idem 49.9% 59.4% 70.7% 11.6 d 14.0 d idem Mortality(14d) Mortality (28d) Mortality (90d) ICU-LOS Hospital-LOS Morbidity Optimization of the Volume status; MAP > 60 mmhg Total: 676 PAC: 335 None Controle: 341 a 1. ARDS 2. Shock Mixed ICU Multicenter (36 Centers), prospective, randomised, controled Richard et al. [39], 2003 ns ns MAP = 70 mmhg PAOP = 18 mmhg HF < 120 min -1 Hct > 27% 10 Idem 10 idem Mortality (Hospital) Hospital-LOS Morbidity Except: Pulmonary Embolism PAC (additionally): DO2I = ml -2 min -1 m CI = l min m -2 None Total: 1994 PAC: 997 Controle: 997 Surgical high risk patients (ASA III and IV), > 60 years, elective and emergency abdominal, thoracic, vascular and hip surgery Surgical ICU Multicenter, prospective, randomised, controled Sandham et al. [42], ns 66% 68% ns ns ns ns 57% 60% 11.0 d 40 d idem 60% 62% 12.1 d 34 d idem Mortality (Hospital) Mortality (ICU) Mortality (28d) ICU-LOS Hospital-LOS Morbidity at the discretion of the treating clinician Total: 1041 PAC: 519 None Controle: 522 b All patients admitted to adult ICU and identified by the treating clinician as someone who should be managed with a PAC Mixed ICU Multicenter (65 Centers), prospective, randomised, controled S. Harvey ( PAC-Man- Study ) [21], 2005 Table 2.4 Selection of currently published studies on the use of the Pulmonary artery catheter (PAC). ASA = Risk classification of the American Society of Anesthesiologists; CI = Cardiac Index; DO2I = Oxygen Delivery Index; Hct = Hematocrit; HR = Heart Rate; ICU = Intensive Care Unit; LOS = length of stay; PAC = Pulmonary Artery Catheter; PAOP = Pulmonary Artery Occlusion Pressure; MAP = Mean Artery Pressure; CVP Central Venous Pressure. a = In both groups, the use of echocardiography was allowed. b = In part, in both groups the use of other cardiac output monitoring techniques was allowed. 2 29

30 Investigators Study type Pölönen et al. [33], 2000 Monocenter, prospective, randomized, controlled Patient group Cardiac surgery patients Time of Inclusion After weaning from CPB Number of Patients Standard Goal-directed Total: 393 Protocol: 196 Standard: 197 CI > 2.5 l min Treatment goals Results -1 m -2 PAOP: mmhg MAP: mmhg Hb > 100 g l -1 Standard + SvO2 > 70 % Lact 2.0 mmol l -1 ICU-LOS Hospital-LOS Number of Patients with Organ dysfunction Goal directed 1 d 6 d 2 Standard P 1 d 7 d 11 ns < 0.05 < 0.01 Rivers et al. [40], 2001 Gan et al. [18], 2002 McKendry et al. [26], 2004 Wakeling et al. [49], 2005 Pearse et al. [30], 2005 Monocenter, prospective, randomized, controlled Monocenter, prospective, randomized, controlled Monocenter, prospective, randomized, controlled Monocenter, prospective, randomized, controlled Monocenter, prospective, randomized, controlled Severe Sepsis Septic Shock Major elective surgery Cardiac surgery patients after CPB Colorectal Surgery Major General Surgery In an emergency area After admission to ICU Intraoperatively Intraoperatively After admission to ICU Total: 263 Protocol: 130 Standard: 133 Total: 100 Protocol: 50 Standard: 50 Total: 179 Protocol: 89 Standard: 90 Total: 128 Protocol: 64 Standard: 64 Total: 122 Protocol: 62 Standard: 60 CVP 8-12 mm Hg MAP 65 mm Hg Diurese ml kg -1 h Volume replacement when: Diuresis ml kg -1 h HF > 110 min -1 SAP < 90 mmhg Standard + ScvO2 70% Volume replacement dependent from SV (Oesophageal Doppler) Hospital mortality 28d-mortality 60d-mortality APACHE II Table 2.5 Selection of recently published studies on goal directed hemodynamic therapy. CI = Cardiac Index; DO2I = Oxygen delivery indexed to BSA; CPB = Cardio-pulmonary Bypass; Hb = Hemoglobin; HR = Heart Rate; ICU = Intensive Care Unit; LOS = length of stay; MAP = Mean Arterial Pressure; PAOP = Pulmonary Artery Occlusion Pressure; SAP = Systolic Arterial Pressure; SV(I) = Stroke volume (index); CVP = Central Venous Pressure. Not Set Volume replacement dependent from CVP (Goal 12-15mmHg) Volume replacement dependent from CVP (Goal: Increase of 2 mmhg) Diuresis ml kg -1 h -2 CI > 2.5 l min -1 m SVI > 35 ml m -2 (Oesophageal Doppler) Volume replacement until no further increase in SV (Oesophageal Doppler) Volume replacement dependent from SV (Goal: 10% increase); DO2I > 600 ml min -1 m -2 (Lithiumdilution) 30.5% 33.3% 44.3% % 49.2% 56.9% <0.001 Hospital-LOS 5 7 < 0.03 ICU-LOS Hospital-LOS Hospital-LOS Patients with Gastrointestinal morbidity Patients with complications Hospital-LOS ICU-LOS 28d-mortality 60d-mortality 2.5 d 11.4 d 10 d 45.3% 44% 11 d 43 h 9.7% 11.3% 3.2 d 13.9 d 11.5 d 14.1% 68% 14 d 45 h 11.7% 15% <0.05 <

31 Pulmonary artery thermodilution cardiac output and still there are many clinicians convinced of the benefit of the PAC in the individual patient [30, 31, 41]. Many experts and research groups claim that a number of randomized multicenter studies have been recently performed in various patient groups, or are under investigation. Included were ASA III-IV patients undergoing high-risk surgical procedures [30, 42], patients with ARDS and / or shock [39], critically ill intensive care patients with high risk of mortality (PAC-MAN) [21] and patients with heart failure [43]. Cohen et al. retrospectively analyzed the impact of the PAC on the outcome in patients with acute coronary syndrome [11]. Shah et al. conducted a meta- analysis of all randomized studies between 1985 and 2005 [43]. The interpretation of the studies and metaanalyses, however, still remains difficult, as patient groups, study protocols and design of the studies vary considerably. Basically, we should distinguish between studies in which the PAC is used to guide treatment of the patient according to a well defined treatment algorithm [30, 42] and studies in which patients, with comparable serious underlying diseases, are studied and monitored with or without a PAC, but without the specific guidance of a therapeutic algorithm or limits of acceptance of parameters such as CO, PAOP, etc. [21, 39]. The interpretation of the results of the former type of study is particularly difficult, because only a disease course and monitored parameters are documented. Other influencing factors, such as doctor-patient ratio, the training of the doctors involved and their experience with the use of monitoring have not been systematically studied. Furthermore, in some studies, alternative monitoring techniques were allowed in the non- PAC-group, such as echocardiography or alternative CO-monitors. Such studies then become no longer an exclusive comparison of PAC versus control group, but an investigation of a variety of influencing factors. 2 This discussion demonstrates also the general dilemma of studies in which monitoring procedures are investigated. These studies depend on the investigated patient population, the therapeutic approach, specific treatment targets, and a different training level of personnel involved. A monitoring procedure such as the PAC can only lead to a reduction in morbidity and mortality if the collected physiological variables are taken into account and used in a therapeutic plan facilitating improved survival and decreased complication rates. Moreover, complications of the monitoring procedures are primarily assessed. The initiation of invasive monitoring and therapy at a specific moment is of particular importance in patients with rapidly progressive disease, such as sepsis [15, 30, 37, 40]. Investigations of septic patients and high-risk surgery demonstrate clearly that early therapy only leads to reduced morbidity and/or mortality if appropriate monitoring is initiated as early as possible before the onset of organ complications [40]. 31

32 Chapter 2 High-risk general surgery and intensive care In the 1980s, the research group of Shoemaker et al. investigated the influence of a supra-normal oxygen supply on morbidity and mortality of surgical patients. As previously mentioned, a retrospective analysis shows that a perioperatively increased supra-normal DO 2, in conjunction with an increase in VO 2, can have a potentially positive effect on survival and the incidence of organ failure [44]. As a consequence, these target parameters (cardiac index (CI) greater than 4.5 l min - 1 m -2, DO 2 greater than 600 ml min -1 m -2 ) were investigated in a number of different patient groups. In both surgical and intensive care patients, initial results were positive [4, 6]. Patients only then had a better prognosis when the increase of DO 2 was paralleled by an increase in VO 2, i.e. if the target organs are able to react to increased oxygen supply. If this is not the case, this therapeutic approach does not improve outcome and maybe linked to an increased mortality rate [50]. In 1995 Gattinoni et al. could not replicate a reduction in mortality shown by Shoemaker s proposed targetorientated therapy in a large randomized multicentre study [19]. Hayes et al. showed that the use of PAC-guided therapy to achieve supra-normal parameters of CO and DO 2 can even lead to an increase in mortality rate [22]. Because of the fact that in both studies not solely surgical patients were included, the results are not directly comparable with the results of Shoemaker et al, and were questioned. Sandham et al. investigated the benefits of preoperative insertion of a PAC in a group of high-risk surgical patients [42]. A total of 1994 patients undergoing major surgery were included [42]. Patients were older than 60 years and belonged to the ASA-risk groups III and IV. Nine hundred ninety seven patients received PAC-guided therapy. The desired targets were a CI of l min - 1 m -2, a DO 2 of ml min -1 m -2, a mean arterial blood pressure of 70 mm Hg and a heart rate <120 min -1. Nine hundred ninety seven patients in the control group were monitored clinically, with the option to place a central venous catheter and to use CVP for guidance of therapy. The study could not demonstrate an influence of PAC-guided therapy on 30-days, 6-months and 1-year mortality rate [42]. The incidence of organ failure (cardiac, renal) was comparable between the groups. The target values of CO and oxygen supply were only achieved in 21% of all patients preoperatively, whereas a DO 2 of 550 ml min -1 m -2 was achieved in 62.9% of patients after surgery. Patients, whose therapy was guided by a PAC, received more inotropic drugs, vasodilators, antihypertensive therapy, blood transfusions and colloid infusions. In the PAC-guided therapy group, there were 8 cases of pulmonary embolism compared to no such cases in the control group. Extrapolated to the total number of patients each year who receive a PAC, it means that about 12,000 patients per year would suffer a pulmonary embolus in association with a PAC [42]. The authors conclude that the use of a PAC-guided treatment algorithm for high-risk surgical patients 32

33 Pulmonary artery thermodilution cardiac output does not to reduce mortality, and that therefore the use of a PAC in this group of patients in this setting has no benefit (figure 2.3). Of particular importance in the interpretation of this study is that the desired CO and oxygen delivery targets were not achieved in a high number of patients in the intervention group, which makes the interpretation of the results difficult as one of the main aims of the study was not tested, i.e. the influence of intraoperative management in the intervention-group on outcome. Interestingly, in a recent study in high-risk surgical patients, in which supranormal values of CO and DO 2 were achieved in a large number of patients, a reduction of perioperative complications, without differences in mortality was observed [30]. Noteworthy is that in this study of Pearse et al, a PAC was not used, but transpulmonary lithium dilution. 2 Figure 2.3 Influence of the PAC on mortality rate in high-risk surgical patients. (Source: N Engl J Med 2003; 348: 5-14). A prospective, randomized study of 65 English intensive care units (PAC-MAN) showed a similar result as the investigation of Sandham et al. [21]. A total of 1041 intensive care patients were included, of which 519 and 522 patients were managed with and without a PAC respectively [21]. In contrast to the investigation of Sandham et al., the PAC-MAN study was exclusively performed in the intensive care setting. Both in the PAC group as well as in the control group, most patients (65% and 66%) were admitted to the ICU because of multi-organ failure. There was no significant difference in mortality between both patient groups, and was 68% in the PAC group and 66% in the control group [21]. There were no differences regarding ICU- and hospital length of stay or incidence of organ failure [21]. Therapy was not standardized and was according to the discretion of the individual doctor. 46 out of 486 patients who received a PAC had catheter-associated complications, which did not affect survival of these patients. The authors concluded that the use of 33

34 Chapter 2 a PAC in critically ill intensive care patients did not have any influence on morbidity and mortality [21]. Richard et al. investigated 676 patients, who fulfilled the standard criteria for shock mainly of septic origin and/or acute respiratory distress syndrome (ARDS) to whether the early use of a PAC within 48 hours after the onset of symptoms had an impact on morbidity and mortality [39]. The primary endpoint of this study was 28-days mortality. Secondary endpoints were 14-days and 90-days mortality, the incidence of organ failure, the incidence of dialysis and the use of vasopressor therapy. For both primary and secondary endpoints there was no difference between groups of patients monitored with a PAC and those without [39]. In summary, the results of these three studies suggest that the use of PAC does not lead to a reduction in morbidity and mortality. It remains unclear whether other high-risk patient groups such as patients scheduled for heart, lung or liver transplantation benefit from the use of a PAC. General Medical Intensive Care In 2005, two studies were published which investigated the use of PAC in medical intensive care patients. The ESCAPE study included 433 patients with severe, symptomatic heart failure assigned to receive therapy guided by clinical assessment and a PAC or clinical assessment alone [3]. The target of the study in both groups was resolution of clinical congestion, mortality rate and incidence of complications of the PAC. For patients in the PAC group additional targets were a PAOP < 15 mm Hg and a right atrial pressure of <8 mm Hg. Medications were not specified, but inotropic use was explicitly discouraged. Mean LVEF before treatment was 19% [3]. There was no significant difference in any of the endpoints between the groups. In the PAC group, significantly more complications were observed than in the control group. Four patients suffered from a catheter related infection, 2 patients experienced a catheter-related hemorrhage, and 2 patients had a pulmonary embolism. The authors concluded that the use of PAC in patients with severe heart failure did not improve outcome [3]. A second retrospective study investigated the influence of the PAC on 30-day mortality in patients with acute coronary syndrome [11]. Of 26,437 patients (from two multi centre studies: GUSTO IIb and III), 735 patients received a PAC during the initial phase of treatment [11]. Patients who received a PAC were older, suffered more frequently from diabetes mellitus, were more often ventilated and more likely to present with ST-segment elevation. Patients managed with PAC also underwent significantly more procedures: percutaneous (PCI) and surgical interventions (CABG). The use of PAC in patients with acute coronary syndrome was associated with increased 30-days 34

35 Pulmonary artery thermodilution cardiac output mortality, even after adjustment for confounding factors such as baseline risk except in patients with cardiogenic shock. Both the ESCAPE - study as well as the retrospective analysis of Cohen et al. (11) suggest that the use of a PAC in patients with heart failure and acute coronary syndrome does not lead to a reduction of mortality. 2 Alternatives to Pulmonary Artery Catheter In recent years a number of invasive and non-invasive procedures have been developed for determination of CO [9, 36]. These include the transpulmonary dilution of an indicator. A cold or lithium bolus is injected into a central vein, and detected in the femoral or radial artery. In terms of accuracy, both procedures are comparable with the pulmonary artery thermodilution [9, 25]. Transpulmonary indicator dilution (PiCCO, LiDCO ) is less invasive in comparison with the PAC, because invasive arterial blood pressure monitoring and a central venous line belong to standard care in critically ill patients, and therefore no additional (venous) access is needed. Furthermore, the risk of puncture or infection risk with the PAC is avoided. If a cannulation of the femoral artery is not indicated, the catheter can be placed in the axillary artery. Both for the PiCCO as well as the for the LiDCO system, no large randomized studies are available to show the impact of this modality of monitoring on patient outcome. In clinical studies, CO measurements are also performed with an esophageal Doppler technique [13]. This technique intends to show a good correlation with invasive procedures [13], but the accuracy is significantly diminished with reduced or increased CO [13]. In addition, this technique is only moderately tolerated in awake or superficially sedated patients. Singer et al. showed in several randomized studies with a low number of included patients that the perioperative optimization of the CO with the esophageal Doppler technology leads to a shortened length of hospital stay [26, 46]. Another method for CO monitoring is the CO 2 rebreathing technique, a modification of the Fick principle [28]. This technique is commercially available (NICO ) and enables a non-invasive determination of pulmonary blood flow. A number of clinical studies showed that in stable conditions, there is an acceptable correlation of the CO 2 rebreathing technique and thermodilution technique in the determination of CO [27]. However, Neuhäuser et al. as well as Bein et al. showed that this correlation was unsatisfactory under specific conditions, such as after weaning from CPB or Xenon anesthesia, [2, 28]. In addition, the validity of these measurements in spontaneous breathing patients has not been sufficiently investigated. 35

36 Chapter 2 A third method for determination of CO is TEE. The blood flow velocity over the left ventricular outflow tract or in the pulmonary artery is measured using Doppler techniques. By measuring vessel diameter and heart rate, the CO can then be calculated [32]. Possible errors can occur with incorrect determination of vessel diameter and incorrect alignment of the Doppler probe with the blood flow. The insertion of the TEE probe in awake patients is only indicated in a select group of patients, but is poorly tolerated. Therefore, this procedure is not useful for continuous assessment of CO. However, TEE may give additional information about the filling state, morphology and anatomy of the heart and function of the heart valves. In addition, this technique allows detection of pathological processes of the heart (thrombi, endocarditis) and pericardium (tamponade, pericardial effusion). Therefore, TEE is the very best in patients with hemodynamic instability. The validity of CO measurements by means of TEE is subject of controversy. The accuracy is acceptable under clinical conditions; however, in situations with very low or high CO, measurements are less accurate. Similar to the CO 2 re-breathing technique is TEE-based determination of CO valuable for trend analysis of CO rather than for exact measurement. As already pointed out, the specificity of the cardiac filling pressures (CVP, PAOP) for the assessment of the filling status in surgical patients is limited. Other techniques with higher sensitivity (TEE, PiCCO, and LiDCO ) are available in clinical routine and can be used to monitor cardiac preload. However, the continuous measurement of pulmonary artery pressure can only be achieved with the PAC and should be used, in our opinion, in selected patients. In particular, patients with medically treated right ventricular failure, with mechanical support or with increased pulmonary vascular resistance may benefit from the use of a PAC. Summary The indications for the use of PACs have been significantly narrowed owing to the results of a number of publications. Measurement of CO or volume status is no longer an acceptable indication. Other modalities such as TPTD and TEE are available alternatives, and may be better able to estimate the volume status of a patient. The determination and monitoring of pulmonary artery pressures in patients with pulmonary hypertension with or without right heart failure, is still the domain of the PAC, however without a clear benefit in morbidity and mortality of these patients. Even in these patients, less invasive monitoring techniques such as TEE are increasingly used, but not without limitations. The importance of PACs in the guidance of the treatment of acute coronary syndrome with cardiogenic shock has not yet been determined, but in patients with heart failure, a survival benefit could not be demonstrated. The majority of studies do not show an increase in mortality in patients monitored with a PAC. However, it becomes clearer that the PAC is a 36

37 Pulmonary artery thermodilution cardiac output monitoring technique with specific risks. Alternative approaches have not yet been investigated in large randomized trials, and therefore, no conclusions about their impact on morbidity and mortality rate can be drawn. The use of the PAC (as well as any other monitoring modality) should therefore be restricted to indications where the measured parameters have consequences for therapy, and where an adequate 2 therapy without accurate monitoring is not feasible. 37

38 Chapter 2 References 1. Practice guidelines for pulmonary artery catheterization: an updated report by the American Society of Anesthesiologists Task Force on Pulmonary Artery Catheterization. Anesthesiology 2003; 99: Bein B, Worthmann F, Tonner PH, et al. Comparison of esophageal Doppler, pulse contour analysis, and real-time pulmonary artery thermodilution for the continuous measurement of cardiac output. J Cardiothorac Vasc Anesth 2004; 18: 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: Bishop MH, Shoemaker WC, Appel PL, et al. Prospective, randomized trial of survivor values of cardiac index, oxygen delivery, and oxygen consumption as resuscitation endpoints in severe trauma. J Trauma 1995; 38: Bottiger BW, Rauch H, Bohrer H, et al. Continuous versus intermittent cardiac output measurement in cardiac surgical patients undergoing hypothermic cardiopulmonary bypass. J Cardiothorac Vasc Anesth 1995; 9: Boyd O, Grounds RM, Bennett ED. A randomized clinical trial of the effect of deliberate perioperative increase of oxygen delivery on mortality in high-risk surgical patients. JAMA 1993; 270: Buhre W, Buhre K, Kazmaier S, et al. Assessment of cardiac preload by indicator dilution and transoesophageal echocardiography. Eur J Anaesthesiol 2001; 18: Buhre W, Kazmaier S, Sonntag H, et al. Changes in cardiac output and intrathoracic blood volume: a mathematical coupling of data? Acta Anaesthesiol Scand 2001; 45: Buhre W, Weyland A, Kazmaier S, et al. Comparison of cardiac output assessed by pulsecontour analysis and thermodilution in patients undergoing minimally invasive direct coronary artery bypass grafting. J Cardiothorac Vasc Anesth 1999; 13: Buhre W, Weyland A, Schorn B, et al. Changes in central venous pressure and pulmonary capillary wedge pressure do not indicate changes in right and left heart volume in patients undergoing coronary artery bypass surgery. Eur J Anaesthesiol 1999; 16: Cohen MG, Kelly RV, Kong DF, et al. Pulmonary artery catheterization in acute coronary syndromes: insights from the GUSTO IIb and GUSTO III trials. Am J Med 2005; 118: Connors AF Jr, Speroff T, Dawson NV, et al. The effectiveness of right heart catheterization in the initial care of critically ill patients. SUPPORT Investigators. JAMA 1996; 276:

39 Pulmonary artery thermodilution cardiac output Dark PM, Singer M. The validity of trans-esophageal Doppler ultrasonography as a measure of cardiac output in critically ill adults. Intensive Care Med 2004; 30: De Simone R, Wolf I, Mottl-Link S, et al. Intraoperative assessment of right ventricular volume and function. Eur J Cardiothorac Surg 2005; 27: Dellinger RP, Carlet JM, Masur H, et al. Surviving Sepsis Campaign guidelines for management of severe sepsis and septic shock. Crit Care Med 2004; 32: Domino KB, Bowdle TA, Posner KL, et al. Injuries and liability related to central vascular catheters: a closed claims analysis. Anesthesiology 2004; 100: Dueck MH, Klimek M, Appenrodt S, et al. Trends but not individual values of central venous oxygen saturation agree with mixed venous oxygen saturation during varying hemodynamic conditions. Anesthesiology 2005; 103: Gan TJ, Soppitt A, Maroof M, et al. Goal-directed intraoperative fluid administration reduces length of hospital stay after major surgery. Anesthesiology 2002; 97: Gattinoni L, Brazzi L, Pelosi P, et al. A trial of goal-oriented hemodynamic therapy in critically ill patients. SvO 2 Collaborative Group. N Engl J Med 1995; 333: Hall JB. Searching for evidence to support pulmonary artery catheter use in critically ill patients. JAMA 2005; 294: 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: Hayes MA, Timmins AC, Yau EH, et al. Elevation of systemic oxygen delivery in the treatment of critically ill patients. N Engl J Med 1994; 330: Hofer CK, Furrer L, Matter-Ensner S, et al. Volumetric preload measurement by thermodilution: a comparison with transoesophageal echocardiography. Br J Anaesth 2005; 94: Kumar A, Anel R, Bunnell E, et al. Preload-independent mechanisms contribute to increased stroke volume following large volume saline infusion in normal volunteers: a prospective interventional study. Crit Care 2004; 8: R Linton RA, Band DM, Haire KM. A new method of measuring cardiac output in man using lithium dilution. Br J Anaesth 1993; 71: McKendry M, McGloin H, Saberi D, et al. Randomised controlled trial assessing the impact of a nurse delivered, flow monitored protocol for optimisation of circulatory status after cardiac surgery. BMJ 2004; 329:

40 Chapter Mielck F, Buhre W, Hanekop G, et al. Comparison of continuous cardiac output measurements in patients after cardiac surgery. J Cardiothorac Vasc Anesth 2003; 17: Neuhauser C, Muller M, Brau M, et al. Partial CO 2 rebreathing technique versus thermodilution: measurement of cardiac output before and after operations with extracorporeal circulation. Anaesthesist 2002; 51: Pearse R, Dawson D, Fawcett J, et al. Changes in central venous saturation after major surgery, and association with outcome. Crit Care 2005; 9: R 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 [ISRCTN ]. Crit Care 2005; 9: R Pinsky MR, Vincent JL. Let us use the pulmonary artery catheter correctly and only when we need it. Crit Care Med 2005; 33: Poelaert J, Schmidt C, Van Aken H, et al. A comparison of transoesophageal echocardiographic Doppler across the aortic valve and the thermodilution technique for estimating cardiac output. Anaesthesia 1999; 54: Polonen P, Ruokonen E, Hippelainen M, et al. A prospective, randomized study of goaloriented hemodynamic therapy in cardiac surgical patients. Anesth Analg 2000: 90: Procaccini B, Clementi G. Pulmonary artery catheterization in 9071 cardiac surgery patients: a review of complications. Ital Heart J Suppl 2004; 5: Reinhart K, Kuhn HJ, Hartog C, et al. Continuous central venous and pulmonary artery oxygen saturation monitoring in the critically ill. Intensive Care Med 2004; 30: Reuter DA, Goetz AE. Measurement of cardiac output. Anaesthesist 2005; 54: Rhodes A, Bennett ED. Early goal-directed therapy: an evidence-based review. Crit Care Med 2004; 32: S 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: 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: Rivers E, Nguyen B, Havstad S, et al. Early goal-directed therapy in the treatment of severe sepsis and septic shock. N Engl J Med 2001; 345: Sakr Y, Vincent JL, Reinhart K, et al. Use of the pulmonary artery catheter is not associated with worse outcome in the ICU. Chest 2005; 128: Sandham JD, Hull RD, Brant RF, et al. A randomized, controlled trial of the use of 40

41 Pulmonary artery thermodilution cardiac output pulmonary-artery catheters in high-risk surgical patients. N Engl J Med 2003; 348: Shah MR, Hasselblad V, Stevenson LW, et al. Impact of the pulmonary artery catheter in critically ill patients: meta-analysis of randomized clinical trials. JAMA 2005; 294: Shoemaker WC, Appel PL, Kram HB, et al. Prospective trial of supranormal values of survivors as therapeutic goals in high-risk surgical patients. Chest 1988; 94: Shoemaker WC, Appel PL, Waxman K, et al. Clinical trial of survivors' cardiorespiratory patterns as therapeutic goals in critically ill postoperative patients. Crit Care Med 1982; 10: Sinclair S, James S, Singer M. Intraoperative intravascular volume optimisation and length of hospital stay after repair of proximal femoral fracture: randomised controlled trial. BMJ 1997; 315: Swan HJ, Ganz W, Forrester J, et al. Catheterization of the heart in man with use of a flowdirected balloon-tipped catheter. N Engl J Med 1970; 283: Versprille A. Pulmonary vascular resistance. A meaningless variable. Intensive Care Med 1984; 10: 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: Yu M, Levy MM, Smith P, et al. Effect of maximizing oxygen delivery on morbidity and mortality rates in critically ill patients: a prospective, randomized, controlled study. Crit Care Med 1993; 21:

42

43 Chapter 3 Cardiac output monitoring EEC de Waal MD 1, F Wappler MD 2, WF Buhre MD 1,2 1 Division of Perioperative and Emergency Care, University Medical Centre, Utrecht, Netherlands; 2 Division of Anaesthesia and Intensive Care Medicine, University of Witten-Herdecke, Germany Curr Opin Anaesthesiol 2009; 22: 71-78

44 Chapter 3 Abstract Purpose of review: The primary goal of hemodynamic therapy is the prevention of inadequate tissue perfusion and inadequate oxygenation. Advanced cardiovascular monitoring is a prerequisite to optimize hemodynamic treatment in critically ill patients prone to cardiocirculatory failure. The most ideal cardiac output (CO) monitor should be reliable, continuous, noninvasive, operatorindependent, cost-effective and having a fast response time. Moreover, simultaneous measurement of cardiac preload enables the diagnosis of hypo- and hypervolemia. Recent findings: During the last years, a number of significant studies in the field of CO monitoring has been published. The available CO monitoring techniques can be divided into invasive techniques, minimal invasive techniques, and noninvasive techniques. Summary: Minor invasive arterial thermodilution is the standard for the estimation of CO. Less invasive and continuous techniques such as pulse contour CO and arterial waveform analysis are preferable. The accuracy of non-calibrated pulse-contour analysis is still a matter of discussion although recent studies demonstrate acceptable accuracy compared to a standard technique. Doppler techniques are minimal invasive and offer a reasonable trend monitoring of CO. Non-invasive continuous techniques, such as bioimpedance and bioreactance are prone for further investigation. 44

45 Cardiac output monitoring Introduction The pulmonary artery catheter (PAC) was the first clinical device enabling bedside measurement of cardiac output (CO). During the past 35 years, measurement of CO became popular in several subgroups of patients, such as those undergoing cardiac surgery and those with sepsis and acute respiratory distress syndrome (ARDS). The impact of monitoring on outcome in patients during anesthesia and intensive care has, however, never been proven [1, 2]. Today, a number of less invasive or even noninvasive monitoring devices are available, representing a wide range from noninvasive bioimpedance to transpulmonary thermodilution. Validation studies [3-6] have shown that some of these techniques can replace the PAC. The ideal CO monitor should be reliable, continuous, noninvasive, operator-independent and cost-effective and should have a fast response time (beat to beat). Some of the techniques mentioned in this review fulfill most of these criteria and are now established techniques, whereas others are still under development; however, adequately powered multicenter outcome studies are still lacking [2]. 3 Invasive thermodilution techniques In the following section, we describe the available invasive CO measurement techniques (table 3.1 and table 3.2). Pulmonary artery thermodilution cardiac output. Since its introduction in 1970, pulmonary artery thermodilution has been the clinical standard in CO measurement. Measurement of thermodilution CO is based on the Stewart-Hamilton equation [2, 6]. Under optimal conditions, the coefficient of variation for repeated bolus thermodilution measurements is less than 10%. It is obvious that pulmonary artery thermodilution measures pulmonary blood flow (PBF), which is normally only insignificantly different from systemic blood flow [7]. Moreover, PBF varies with the respiratory cycle, and thus measurements should be randomly spread over the respiratory cycle. In addition to measurement of CO, modifications of the original PAC enable measurement of right ventricular function (RVEF) and right ventricular enddiastolic volume (RVEDV) [7, 8]; however, the PAC may cause relevant complications [9-12]. Several reports described intrinsic morbidity and mortality. Therefore, the use of the PAC should be restricted to highly selected patient populations [13]. The selective use of the PAC is justified only in patients with right ventricular failure and patients with increased pulmonary vascular resistance requiring vasodilator therapy [14, 15, 16*]. The use of the PAC in low-risk cardiac surgery, vascular surgery, and major abdominal, orthopedic or neurosurgical procedures cannot be 45

46 Chapter 3 CO Technique Invasiveness Intermittent versus continuous PAC BTCO +++ Intermittent PAC CCO +++ Continuous TPCO ++ Intermittent LiDCO ++ Intermittent APCO + Continuous PCCO PulseCO ++ Continuous EDCO + Continuous Limitations PAC related complications Arrhythmias Tricuspid regurgitation Intracardiac and extracardiac shunts Injected indicator volume PAC related complications Larger temperature shifts Intracardiac and extracardiac shunts Fast infusions Arrhythmias Large temperature shifts Intracardiac and extracardiac shunts Arterial signal quality Rapid changes in vascular motor tone Arterial signal quality Rapid changes in vascular motor tone IABP Arterial signal quality Rapid changes in vascular motor tone Esophageal disorder Turbulent flow Operator-dependent TTE/TEE ± Intermittent Esophageal disorder Operator-dependent PDDCO + Continuous Bioimpedance - Continuous Peripheral signal detection Allergy to indocyanin green Movement artifacts Thoracic fluid overload Abnormal thoracic anatomy Cardiac valve disease Intracardiac and extracardiac shunts Tachyarrhythmias Additional information CVP PAP PCWP SvO 2 CVP PAP PCWP SvO 2 GEDV EVLW PPV CBV SVV SV CI SVI SVV Flow time Anatomic and functional cardiac assessment Intravascular blood volume PEP LVET Table 3.1 Specific characteristics of different cardiac output monitoring techniques. APCO, arterial pulse wave cardiac output; CBV, central blood volume; CI, cardiac index; CVP, central venous pressure; EDCO, esophageal Doppler cardiac output; EVLW, extravascular lung water; GEDV, global enddiastolic volume; LVET, left ventricular ejection time; PAC BTCO, pulmonary artery catheter bolus thermodilution cardiac output; PAC CCO, pulmonary artery catheter continuous cardiac output; PAP, pulmonary artery pressure; PCCO, pulse contour continuous cardiac output; PCWP, pulmonary capillary wedge pressure; PDDCO, pulse dye densitometry cardiac output; PEP, preejection period; PPV, pulse pressure variation; PulseCO, pulse contour cardiac output; SV, stroke volume; SVI, stroke volume index; SvO 2, mixed venous oxygen 46

47 Cardiac output monitoring saturation; SVV, stroke volume variation; TEE, transesophageal echocardiography; TPCO, transpulmonary thermodilution cardiac output; TTE, transthoracic echocardiography. recommended [17]. Advocates of the PAC suggest that it is of crucial importance that physicians and nursing staff are familiar with the PAC technology, including insertion and positioning of the PAC. Use of the PAC necessitates training and education as misinterpretation of data obtained with the PAC is common. Thermal filament Continuous Cardiac Output (CCO). Continuous pulmonary CO is measured by thermodilution using a modified PAC with an embedded heating filament (Edwards Lifesciences, Irvine, California, USA) which releases small thermal pulses every seconds following a pseudorandom binary sequence [18]. The change in blood temperature is detected downstream via a rapid-response distal thermistor at the tip of this specific PAC and CO is proportional to the area under the thermodilution curve. Every seconds, a trended continuous CO (CCO) measurement is displayed, which reflects an average flow over the previous 3-6 min. To minimize the response time, this device includes a stat CO (SCO) mode, averaging CO over the last three measurements. As relatively small quantities of heat are used to calculate cardiac output, sudden changes in temperature or infusion of high quantities of cold infusate may influence the accuracy and reliability of the method [6]. Hyperthermia does not influence the accuracy of CCO monitoring, although immediately after hypothermic cardiopulmonary bypass (CPB), a relative increase in bias is reported. Recently, Bao and Wu [18] observed that the delayed response also limits the use of this technique in patients undergoing liver transplantation. 3 Transpulmonary thermodilution cardiac output. In contrast to pulmonary artery thermodilution technique, the transpulmonary (or transcardiopulmonary) approach uses arterial thermodilution for the measurement of stroke volume (SV) and CO [6, 7, 19]. A central venous catheter for the injection of an ice-cold indicator is needed, together with an arterial thermistor-tipped catheter normally placed in the femoral artery, or in the axillary or brachial artery. Calculation of CO is based on the Stewart-Hamilton equation. The validity and reliability of transpulmonary thermodilution is comparable to pulmonary artery thermodilution [20]. As transpulmonary thermodilution is less invasive compared to pulmonary artery thermodilution, the transpulmonary CO (TPCO) technique is more often used, particularly when CO monitoring is deemed necessary over a longer period of time. 47

48 Chapter 3 First Author [ref] Publication year Experimental Technique Reference Setting technique Location Population Bias Precision (2 SD) 95% CI ME (%) Chakravarthy et al. [19] PAC CCO PAC BTCO OR Off-Pump CABG to 1.43 NA Bao and Wu. [18] PAC CCO PAC BTCO OR Liver transplantation to Spöhr et al. [7] TPCO PAC CCO ICU Septic shock patients to Costa et al. [21] LiDCO PAC BTCO ICU Liver transplantation to De Waal et al. [31] VigileoCO TPCO OR and ICU CABG to Sakka et al. [35] VigileoCO TPCO ICU Sepsis to Button et al. [6] VigileoCO PAC BTCO OR and ICU Cardiac surgery: several time points 0.10 to to Chakravarthy et al. [19] VigileoCO PAC BTCO OR Off-Pump CABG to 0.81 NA Cannesson et al. [28] VigileoCO PAC BTCO ICU CABG to Breukers et al. [27] VigileoCO PAC BTCO ICU Cardiac surgery to Prasser et al. [34] VigileoCO PAC BTCO ICU Critically ill patients to Compton et al. [30*] VigileoCO TPCO ICU Hemodynamic instable pts to Mayer et al. [4] VigileoCO PAC BTCO OR and ICU CABG to Biais et al. [26] VigileoCO SCO OR and ICU Liver transplantation to Mehta et al. [25] VigileoCO PAC BTCO OR Off-Pump CABG to to Spöhr et al. [7] PCCO PAC CCO ICU Septic shock patients to Halvorsen et al. [38] PCCI PAC BTCI OR Off-Pump CABG to Chakravarthy et al. [19] PCCO PAC BTCO OR Off-Pump CABG to 2.37 NA Fakler et al. [20] PCCI TPCI ICU Costa et al. [21] Pulse CO PAC BTCO ICU Post-congenital heart surgery children Orthotopic liver transplantation to 0.85 NA to Missant et al. [23] Pulse CO PAC BTCO OR Off-Pump CABG to

49 Cardiac output monitoring Green. [42] NICO EDCO OR Major abdominal surgery to 5.10 NA Missant et al. [23] TEE-CO AoV PAC BTCO OR Off-Pump CABG to Baulig et al. [45] PDDCO PAC BTCO OR and ICU CABG to SVI-K TPSVI OR and ICU CABG to SVI-SB TPSVI OR and ICU CABG to De Waal et al. [47]* SVI-W TPSVI OR and ICU CABG to Gujjar et al. [48] CO-K PAC BTCO ICU Postcardiac surgery (±IABP) to Zoremba et al. [49] EV-CO PAC BTCO ICU Critically ill patients to Raval et al. [50] Bioreactance CO PAC BTCO CL Cardiac catheterization lab to 1.87 NA Table 3.2 Overview of the accuracy of different cardiac output monitoring techniques, compared with a reference technique. APCO = arterial pulse wave cardiac output; CABG = coronary artery bypass grafting; Bioreactance CO = cardiac output obtained with bioreactance; CI = confidence interval; CL = catheterization lab; CO-K = cardiac output according to Kubicek; EDCO = esophageal Doppler cardiac output; EV-CO= electrical velocimetry cardiac output; IABP = intraaortic balloon pump; ICU = intensive care unit; LiDCO = lithium dilution cardiac output; ME = mean error; NA = not available; NICO = noninvasive partial CO2 rebreathing cardiac output; OR = operating room; PAC BTCI = pulmonary artery catheter bolus thermodilution cardiac index; PAC BTCO = pulmonary artery catheter bolus thermodilution cardiac output; PAC CCO = pulmonary artery catheter continuous cardiac output; PCCI = pulse contour continuous cardiac index; PCCO = pulse contour continuous cardiac output; PDDCO = pulse dye densitometry cardiac output; PulseCO = continuous cardiac output calibrated by lithium; SCO = stat cardiac output obtained with Vigilance; SVIK = stroke volume index according to Kubicek; SVISB = stroke volume index according to Sramek- Bernstein; SVIW = stroke volume index according to Wang; TEE-CO AoV = transesophageal echocardiography-derived cardiac output at the aortic valve; TPCO = transpulmonary cardiac output; TPSVI = stroke volume index obtained with transpulmonary thermodilution. 3 49

50 Chapter 3 Lithium dilution cardiac output. An example of dye dilution CO monitoring is based on intravascular injection of a small dose (1 ml) of an isotonic lithium chloride solution (150 mmol) [21]. An advantage of the lithium indicator dilution CO technique (LiDCO) is that no central venous line is necessary because the indicator bolus can be applied via a peripheral line [22]. The resulting lithium concentration-time curve is recorded by withdrawing blood (4.5 ml min -1 ) through a special disposable sensor, attached to the patient s arterial line, which consists of a lithium-selective electrode in a flow-through cell. The voltage across the lithium-selective membrane is digitized on-line and recorded via a computer converting the voltage signal to a lithium concentration. LiDCO is calculated according to the equation: LiCl 60 LiDCO = AUC (1 PCV ) where LiCl is lithium chloride (mmol), AUC is area under the primary dilution curve and PCV is packed cell volume, which can be calculated when the actual hematocrit is known. The lithium dilution technique is of sufficient accuracy when there is constant blood flow, homogenous mixing of blood and when there is no indicator loss between site of injection and detection site [21-23]. Minimal invasive continuous arterial pulse waveform analysis At present three different commercially available devices (Vigileo, PiCCO, LiDCO) are available for CCO monitoring using the arterial pulse waveform technique. Non-calibrated arterial pressure-based cardiac output. One recently developed CO monitoring system (Vigileo, Edwards Lifesciences, Irvine, California, USA) is also based on arterial pulse contour analysis. A special blood flow sensor (FloTrac), which is connected to an arterial line (radial, brachial or femoral artery) is needed; no external calibration is necessary [4, 24, 25]. This device calculates CO on a continuous base (every 20 sec) by multiplying heart rate by calculated SV. SV is calculated using arterial pulsatility (standard deviation of the pressure wave over a 20 sec interval), according to the equation: SV = K x Pulsatility [6, 26-29, 30*, 31]. The constant K is derived from the patient s specific vascular compliance based on biometric values (gender, age, height and weight) according to the method described by Langewouters et al. [32], and waveform characteristics (skewness and kurtosis of the individual arterial pressure curve). The calibration constant K is automatically recalculated every 50

51 Cardiac output monitoring minute. Several studies [4, 25, 33-35] concerning the accuracy of VigileoCO monitoring using radial or femoral arterial lines in a variety of patients have been performed with different software versions of the device. Newer studies demonstrated clinical acceptable precision in comparison to a standard technique of known accuracy; however, the validity critically depends on the software version. Rapid changes in vascular motor tone lead to impaired accuracy of CO monitoring. This is of importance as no calibration technique is incorporated in the device; thus, the user must be aware of the potential limitations of this technique. Calibrated continuous arterial pulse contour cardiac output (PCCO). Two commercially available CO measurement systems (PiCCO, LiDCO) require calibration before measurement of pulse contour cardiac output (PCCO) based on the assumption that the systolic part of the arterial pressure waveform represents SV [36*, 37, 38]. The PiCCO system requires transpulmonary thermodilution for the calibration procedure, whereas LiDCO can be calibrated using the lithium dilution; however, recalibration after profound changes in arterial compliance (e.g. sepsis, after CPB) and/or hemodynamics is a prerequisite for adequate measurement of CO with PCCO [36*, 37, 38] mainly because of changes in vasomotor tone [39]. When these criteria are fulfilled, the accuracy of both techniques is sufficient for clinical purposes. 3 Alternative less invasive techniques During recent years, a number of less invasive or even noninvasive techniques were developed. In this section, the available techniques and their validity and reliability are described in comparison to a technique of known accuracy. Doppler ultrasound methods. Nowadays, several types of Doppler techniques are commercially available for the estimation of CO by measurement of aortic blood flow (ABF) [40-42]. An ultrasound beam directed along the ABF is reflected by moving of the red blood cells with a shift in frequency (Doppler effect), which is proportional to the blood flow velocity according to the equation: F d = 2 f 0 / C V cos Ө where F d is the change in frequency (Doppler shift), f 0 is the transmitted frequency, V is the blood flow velocity, and Ө is the angle between the direction of ultrasound beam and blood flow. CO is 51

52 Chapter 3 estimated by multiplying the blood flow velocity with the cross-sectional area (CSA) of the aorta at the insonation point. The esophageal Doppler probe is introduced orally or nasally and placed at the level of the descending aorta. This technique has some advantages over the classical suprasternal technique, the most important being a more stable probe position once the descending aorta is insonated. Three models of esophageal CO monitoring systems are commercially available and differ from each other in some important ways. Two systems use a built-in nomogram to obtain the descending aortic diameter (CardioQ, Deltex Medical, Chicester, Sussex, UK; Medicina TECO, Berkshire, UK), whereas the other system uses M-mode echocardiography for the measurement of the descending aortic diameter (HemoSonic, Arrow International, Reading, PA). By turning the esophageal Doppler probe, the best Doppler image should be achieved. ABF is calculated by multiplying ABF velocity with the CSA of the descending aorta and heart rate. Limitations of this technique are turbulent flow, negotiation of blood flow to the upper part of the body, and the angle of insonating the aorta [41, 42]. Moreover, the technique is poorly tolerated in awake, nonintubated patients, and cannot be used in patients with esophageal disorder. Once a Doppler probe is in place, transesophageal echocardiography (TEE) cannot be performed. In summary, esophageal Dopplerderived ABF is a semi-invasive approach, which enables trend monitoring of CO. The technique can easily be incorporated into clinical algorithms. The limits of agreement of this technique are increased compared with invasive techniques; however, in contrast to most other techniques, it has been demonstrated in subsets of patients that hemodynamic treatment according to Doppler-derived CO measurements leads to a decrease in perioperative morbidity and length of stay in intensive care units [43*, 44]. Doppler flow measurements obtained with transthoracic echocardiography (TTE) or TEE can also be used to estimate CO. The accuracy depends on image quality, sample site, angle of insonation, the profile of the blood flow velocity distribution across the blood flow, the velocity signal-to-noise ratio and the possibility of measuring the diameter of the vessel and the shape of the valve. Most often, measurements of blood flow velocity and CSA are performed both by TTE and TEE at the level of a cardiac valve or the right ventricular (RVOT) or left ventricular outflow tract (LVOT). The best results are usually obtained by the transaortic approach using the triangular shape assumption of aortic valve opening, and CO determination at the LVOT. In summary, Doppler echocardiography is technically demanding, time-consuming and requires a skilled operator. It is a safe, fairly reproducible and reasonably accurate method for CO measurement in selected patients, provided that the signal quality is adequate during recording. 52

53 Cardiac output monitoring Pulse Dye Densitometry. Pulse dye densitometry (PDD; DDG2001 analyzer, Nihon Kohden, Tokyo, Japan) facilitates minimally invasive, subsequent CO measurements by estimating the arterial concentration of indocyanine green (ICG) after a bolus injection [45]. ICG distributes exclusively in the intravascular space via binding to α 1 -lipoproteins. After passage through the pulmonary circulation, the amount of ICG can be detected non-invasively via a finger-tip sensor, which emits light with wavelengths of 805 and 890 nm. The ratio of ICG-concentration measured at 805 and 890 nm is used to calculate the ICG concentration-time curve. The concentration of ICG (mg/l) is computed continuously, and CO is calculated from the observed dye-densitogram. The ICG dye is nontoxic, except for rare cases of anaphylaxis and allergic reactions. ICG is cleared from the blood exclusively via the liver without undergoing either intrahepatic conjugation or enterohepatic metabolism. Usually, the ICG concentration decreases to 1% of the initial concentration after 20 min, enabling a new measurement. Several studies [45, 46] concerning the accuracy of PDD have been published with conflicting results. 3 Bioimpedance cardiography. Bioimpedance cardiography is based on the application of a high-frequency, low-alternating electrical current to the thorax (thoracic electrical bioimpedance). Changes in bioimpedance to this current are related to cardiac events and blood flow in the thorax. The conversion from changes in bioimpedance to stroke volume requires mathematical conversion [47*]. Several investigators provided algorithms based on different mathematical models [48] of the thorax and they may differ in the location and amount of the current injecting and voltage sensing electrodes. Recently electrical velocimetry was introduced as a new bioimpedance technique with a new algorithm: the Bernstein-Osypka equation (Aesculon, Osypka Medical, Berlin, Germany) [49]. Two main differences exist compared to earlier approaches. The volume of the electrically participating tissue (VEPT) was a homogenously blood filled cylinder or truncated cone. In the newest algorithm, only the intrathoracic blood volume compartment is the VEPT. A second difference is based on the conceptualization of what the Newtonian hemodynamic equivalent dz/dt max represents in the electrical domain. In the older techniques, dz/dt max represents the ohmic equivalent of the peak flow in the ascending aorta, while in the Bernstein-Osypka equation dz/dt max represents the ohmic equivalent of the peak aortic blood acceleration. The accuracy and reliability of the majority of thoracic bioimpedance devices have been evaluated with inconclusive and conflicting results, which may lead to inappropriate clinical interventions [47*]. Common cylinder and cone based models for bioimpedance stroke volume calculation are oversimplifications of the complex electrical events 53

54 Chapter 3 occurring inside the thorax during the cardiac cycle, even when only the intrathoracic blood volume is used as a model [47*]. Therefore, bioimpedance CO is not accepted as a valid and reproducible method in clinical routine at present. In whole body electrical bioimpedance, the whole body is used as a conductor for the alternating current. Although the results seem promising, this technique needs further investigation. Finally, bioreactance technology (Bioreactance, Cheetah, Medical Inc., Indianapolis, Indiana, USA) is based on the discovery that changes in aortic blood volume in the ascending aorta induce small changes in the frequency and phase of electrical signals (frequency-modulation and phasemodulation) propagating across the thorax [50*]. These small changes are highly correlated with blood flow and can thus be used to monitor CO. Conclusion Several methods for the measurement of CO are available. Pulmonary artery thermodilution, transpulmonary thermodilution and lithium dilution are valid and reliable intermittent techniques, which can be used in awake and anesthetized patients. These methods are relatively observerindependent. Pulse contour-derived CO is currently the most frequently used continuous technique for the assessment of CO. In combination with intermittent transpulmonary thermodilution, this technique enables a minor invasive approach when compared with the classical PAC technique. Arterial pressure waveform analysis is a reasonable and well received technique in several patient populations. The precision of Doppler measurements of CO is still a matter of discussion; however, clinical studies demonstrated that this technique enabled target-directed therapy resulting in fewer complications in different patient populations. Bioimpedance CO is not accepted as a valid and reproducible method in clinical routine; however, there are some promising new technologies awaiting clinical studies. 54

55 Cardiac output monitoring References and recommended reading Papers of particular interest, published within the annual period of review, have been highlighted as: * of special interest ** of outstanding interest 1. Mutoh T, Kazumata K, Ajiki M, et al. Goal-directed fluid management by bedside transpulmonary hemodynamic monitoring after subarachnoid hemorrhage. Stroke 2007; 38: Pinsky MR. Hemodynamic evaluation and monitoring in the ICU. Chest 2007; 132: Cavallaro F, Sandroni C, Antonelli M. Functional hemodynamic monitoring and dynamic indices of fluid responsiveness. Minerva Anestesiol 2008; 74: Mayer J, Boldt J, Wolf MW, et al. Cardiac output derived from arterial pressure waveform analysis in patients undergoing cardiac surgery: validity of a second generation device. Anesth Analg 2008; 106: Morgan P, Al-Subaie N, Rhodes A. Minimally invasive cardiac output monitoring. Curr Opin Crit Care 2008; 14: 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: Spöhr F, Hettrich P, Bauer H, et al. Comparison of two methods for enhanced continuous circulatory monitoring in patients with septic shock. Intensive Care Med 2007; 33: Rocca GD, Costa MG, Feltracco P, et al. Continuous right ventricular end diastolic volume and right ventricular ejection fraction during liver transplantation: a multicenter study. Liver Transpl 2008; 14: Buhre W, Rossaint R. Perioperative management and monitoring in anaesthesia. Lancet 2003; 362: Ahmed H, Kaufman D, Zenilman ME. A knot in the heart. Am Surg 2008; 74: Chen LC, Huang PH. Entrapment of a Swan-Ganz catheter. J Chin Med Assoc 2007; 70: George RB, Olufolabi AJ, Muir HA. Critical arrhythmia associated with pulmonary artery catheterization in a parturient with severe pulmonary hypertension. Can J Anaesth 2007; 54: De Waal EE, de Rossi L, Buhre W. [Pulmonary artery catheter in anaesthesiology and intensive care medicine.]. Anaesthesist 2006; 31:

56 Chapter Bremer HC, Kreisel W, Roecker K, et al. Phosphodiesterase 5 inhibitors lower both portal and pulmonary pressure in portopulmonary hypertension: a case report. J Med Case reports 2007; 1: Rex S, Busch T, Vettelschoss M, et al. Intraoperative management of severe pulmonary hypertension during cardiac surgery with inhaled iloprost. Anesthesiology 2003; 99: Rex S, Schaelte G, Metzelder S, et al. Inhaled iloprost to control pulmonary artery hypertension in patients undergoing mitral valve surgery: a prospective, randomizedcontrolled trial. Acta Anaesthesiol Scand 2008; 52: *This study demonstrates the usefulness of the PAC in a selected group of high-risk patients undergoing cardiac surgery. 17. Lyons WS. Re: Routine perioperative pulmonary artery catheterization has no effect on rate of complications in vascular surgery: a meta-analysis. Am Surg 2002; 68: Bao FP, Wu J. Continuous versus bolus cardiac output monitoring during orthotopic liver transplantation. Hepatobiliary Pancreat Dis Int 2008; 7: Chakravarthy M, Patil TA, Jayaprakash K, et al. Comparison of simultaneous estimation of cardiac output by four techniques in patients undergoing off-pump coronary artery bypass surgery A prospective observational study. Ann Card Anaesth 2007; 10: Fakler U, Pauli C, Balling G, et al. Cardiac index monitoring by pulse contour analysis and thermodilution after pediatric cardiac surgery. J Thorac Cardiovasc Surg 2007; 133: Costa MG, Della Rocca G, Chiarandini P, et al. Continuous and intermittent cardiac output measurement in hyperdynamic conditions: pulmonary artery catheter vs. lithium dilution technique. Intensive Care Med 2008; 34: Christiansen C, Hostrup A, Tonnesen E, et al. [Hemodynamic monitoring with the lithium dilution cardiac output system]. Ugeskr Laeger 2008; 170: Missant C, Rex S, Wouters PF. Accuracy of cardiac output measurements with pulse contour analysis (PulseCO ) and Doppler echocardiography during off-pump coronary artery bypass grafting. Eur J Anaesthesiol 2008; 25: Scheeren TW, Wiesenack C, Compton FD, et al. Performance of a minimally invasive cardiac output monitoring system (Flotrac/Vigileo). Br J Anaesth 2008; 101: Mehta Y, Chand RK, Sawhney R, et al. Cardiac output monitoring: comparison of a new arterial pressure waveform analysis to the bolus thermodilution technique in patients undergoing off-pump coronary artery bypass surgery. J Cardiothorac Vasc Anesth 2008; 22: Biais M, Nouette-Gaulain K, Cottenceau V, et al. Cardiac output measurement in patients undergoing liver transplantation: pulmonary artery catheter versus uncalibrated arterial 56

57 Cardiac output monitoring pressure waveform analysis. Anesth Analg 2008; 106: Breukers RM, Sepehrkhouy S, Spiegelenberg SR, et al. Cardiac output measured by a new arterial pressure waveform analysis method without calibration compared with thermodilution after cardiac surgery. J Cardiothorac Vasc Anesth 2007; 21: 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: Collange O, Xavier L, Kuntzman H, et al. FloTrac for monitoring arterial pressure and cardiac output during phaeochromocytoma surgery. Eur J Anaesthesiol 2008; 25: Compton FD, Zukunft B, Hoffmann C, et al. Performance of a minimally invasive uncalibrated cardiac output monitoring system (Flotrac /Vigileo ) in haemodynamically unstable patients. Br J Anaesth 2008; 100: *Interesting study, in which the authors compared the value of different techniques in unstable patients. 31. De Waal EE, Kalkman CJ, Rex S, et al. Validation of a new arterial pulse contour-based cardiac output device. Crit Care Med 2007; 35: Langewouters GJ, Wesseling KH, Goedhard WJ. The pressure dependent dynamic elasticity of 35 thoracic and 16 abdominal human aortas in vitro described by a five component model. J Biomechanics 1985; 18: Opdam HI, Wan L, Bellomo R. A pilot assessment of the FloTrac cardiac output monitoring system. Intensive Care Med 2007; 33: Prasser C, Bele S, Keyl C, et al. Evaluation of a new arterial pressure-based cardiac output device requiring no external calibration. BMC Anesthesiol 2007; 7: Sakka SG, Kozieras J, Thuemer O, et al. Measurement of cardiac output: a comparison between transpulmonary thermodilution and uncalibrated pulse contour analysis. Br J Anaesth 2007; 99: Piehl MD, Manning JE, McCurdy SL, et al. Pulse contour cardiac output analysis in a piglet model of severe hemorrhagic shock. Crit Care Med 2008; 36: *Interesting study demonstrating the influence of vasomotor tone on the validity of pulse contour-based CO monitoring. 37. Johansson A, Chew M. Reliability of continuous pulse contour cardiac output measurement during hemodynamic instability. J Clin Monit Comput 2007; 21: Halvorsen PS, Sokolov A, Cvancarova M, et al. Continuous cardiac output during off-pump coronary artery bypass surgery: pulse-contour analyses vs pulmonary artery thermodilution. Br J Anaesth 2007; 99:

58 Chapter Buhre W, Rex S. Is continuous really continuous? Crit Care Med 2008; 36: Monsel A, Salvat-Toussaint A, Durand P, et al. The transesophageal Doppler and hemodynamic effects of epidural anesthesia in infants anesthetized with sevoflurane and sufentanil. Anesth Analg 2007; 105: Monnet X, Chemla D, Osman D, et al. Measuring aortic diameter improves accuracy of esophageal Doppler in assessing fluid responsiveness. Crit Care Med 2007; 35: Green DW. Comparison of cardiac outputs during major surgery using the Deltex CardioQ oesophageal Doppler monitor and the Novametrix-Respironics NICO: a prospective observational study. Int J Surg 2007; 5: 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: *This clinical study demonstrates the benefit of a CO-based management of patients with respect to morbidity and mortality. 44. McFall MR, Woods WG, Wakeling HG. The use of oesophageal Doppler cardiac output measurement to optimize fluid management during colorectal surgery. Eur J Anaesthesiol 2004; 21: Baulig W, Bernhard EO, Bettex D, et al. Cardiac output measurement by pulse dye densitometry in cardiac surgery. Anaesthesia 2005; 60: Hori T, Yamamoto C, Yagi S, et al. Assessment of cardiac output in liver transplantation recipients. Hepatobiliary Pancreat Dis Int 2008; 7: De Waal EE, Konings MK, Kalkman CJ, et al. Assessment of stroke volume index with three different bioimpedance algorithms: lack of agreement compared to thermodilution. Intensive Care Med 2008; 34: *Innovative study design, as raw data were processed to compare different algorithms of CO monitoring by bioimpedance. 48. Gujjar AR, Muralidhar K, Banakal S, et al. Non-invasive Cardiac Output by Transthoracic Electrical Bioimpedance in Post-cardiac Surgery Patients: Comparison with Thermodilution Method. J Clin Monit Comput 2008; 22: Zoremba N, Bickenbach J, Krauss B, et al. Comparison of electrical velocimetry and thermodilution techniques for the measurement of cardiac output. Acta Anaesthesiol Scand 2007; 51: Raval NY, Squara P, Cleman M, et al. Multicenter evaluation of noninvasive cardiac output measurement by bioreactance technique. J Clin Monit Comput 2008; 22: *Interesting study on a new technique of non-invasive CO monitoring. 58

59 Chapter 4 Validation of a new arterial pulse contour based cardiac output device EEC de Waal MD 1, 2, CJ KalkmanMD PhD 1, S Rex MD 3, WF Buhre MD 1 1 Division of Perioperative and Emergency Care, 2 Department of Intensive Care, University Medical Center, Utrecht, Netherlands; 3 Department of Anesthesiology, University Hospital of Aachen, Germany. Crit Care Med 2007; 35:

60 Chapter 4 Abstract Objective: Evaluation of accuracy and precision of an arterial pulse contour based continuous cardiac output device (Vigileo). VigileoCO was compared with intermittent transpulmonary thermodilution (TPCO) and an established arterial pulse contour based CO (PCCO). Design: Prospective clinical study. Setting: University hospital. Patients: 22 patients undergoing coronary artery bypass graft surgery (CABG). Interventions: Defined volume load during surgery and in the postoperative period. Measurements and Main results: 184 pairs of VigileoCO and TPCO, 140 pairs of VigileoCO and PCCO, and 140 pairs of PCCO and TPCO were obtained. Measurements were performed after induction of anesthesia (T 1 ), sternotomy (T 2 ), immediately (T 3 ) and 20 min after volume challenge with 10 ml kg -1 hydroxyethyl starch 6% (T 4 ), 15 minutes after CPB (T 5 ), after re-transfusion of autologous blood (T 6 ), after arrival at the ICU (T 7 ), immediately (T 8 ) and 20 min after a second volume load with 10 ml kg -1 hydroxyethyl starch 6% (T 9 ). TPCO was used to calibrate PCCO. For pooled data, including un-calibrated PCCO data immediately after weaning from CPB (T 5 ), the correlation coefficient of TPCO vs. VigileoCO, PCCO vs. VigileoCO and TPCO vs. PCCO was 0.75, 0.60 and 0.75 respectively. Bland-Altman analysis showed a bias of 0.00, and 0.02 l min -1, the precision (=SD) was 0.87, 1.08 and 0.93 l min -1 and a mean error of 33%, 40% and 35%. When comparing calibrated PCCO values (T 2 -T 4, T 6, T 7 -T 9 ), the correlation coefficients of PCCO-VigileoCO and TPCO-PCCO were 0.72 and 0.85, bias was and 0.19 l min -1 and mean error was 33% and 27% respectively. Best correlations and the least differences between TPCO and VigileoCO were observed in post-bypass closed chest conditions and in the ICU period. Conclusions: Our results showed that VigileoCO enables clinically acceptable assessment of CO in post-bypass closed chest conditions and during stable conditions in the ICU. 60

61 Arterial pulse contour based cardiac output Introduction The measurement of cardiac output (CO) is still an important technique in the hemodynamic management of perioperative and critically ill patients. Until now, bolus pulmonary artery thermodilution using the pulmonary artery catheter (PAC) has remained the clinical reference technique of CO-monitoring [1, 2]. However, pulmonary artery catheterization is highly invasive, time consuming and associated with a considerable risk of morbidity and mortality [3, 4]. To avoid the complications of the PAC, a number of efforts have been made to develop alternative lessinvasive techniques of intermittent or continuous CO-monitoring. The ideal cardiac output monitor should be non-invasive, valid and reliable under various pathological hemodynamic conditions, operator independent, easy to use, continuous and cost-effective [5]. The PiCCO technique (Pulsion, Munich, Germany) uses a special arterial thermodilution catheter in the femoral, axillary or brachial artery and measures continuous cardiac output by analysis of the arterial pulse contour. An initial calibration and subsequent re-calibration using transpulmonary thermodilution (TPCO) is mandatory [6]. The accuracy of the PiCCO-algorithm has been proven both clinically and experimentally [6, 7]. A good agreement between TPCO and pulmonary artery thermodilution cardiac output has been demonstrated [8, 9], even during off-pump coronary artery bypass surgery [10]. Recently, the Vigileo system (Edwards Lifesciences, Irvine, CA, USA) has been introduced into clinical practice. Basically, this system includes a newly developed algorithm for arterial pulse contour analysis using a special blood flow sensor (FloTrac Sensor, Edwards Lifesciences, Irvine, CA, USA) which can be used with every arterial line. No thermodilution cardiac output is needed for calibration of this technique [11, 12]. Until now, no clinical data comparing the results of the Vigileo and the PiCCO as the reference technique of known accuracy are available. Therefore, we performed a clinical trial to investigate the accuracy and precision of this new arterial pulse contour device compared to a standard technique of known accuracy in coronary artery bypass graft (CABG) patients. 4 Materials and methods Patients. After approval by the institutional review board and written informed consent, 22 patients undergoing elective CABG-surgery were included. Patients with severely reduced LV-function (ejection fraction < 35%), intracardiac shunts, significant valvular heart disease, occlusive peripheral arterial disease or patients undergoing emergency surgery were excluded. All cardiac medications were continued until the day of surgery, except digitalis, ACE-inhibitors and diuretics. 61

62 Chapter 4 Anesthesia. Patients were premedicated with midazolam mg orally one hour before induction of anesthesia. General anesthesia was induced with sufentanil (2 µg kg -1 ) and midazolam (0.05 mg kg -1 ). Tracheal intubation was facilitated by pancuroniumbromide (0.1 mg kg -1 ). After endotracheal intubation, anesthesia was maintained with a continuous infusion of sufentanil (0.5 µg kg -1 h -1 ) and MAC Sevoflurane. All patients were mechanically ventilated with an inspired oxygen concentration of 0.4 and a positive end-expiratory pressure of 5 cm H 2 O. After induction of anesthesia, a 5-F thermistor-tipped catheter (Pulsiocath PV2015L20A, Pulsion Medical systems, Munich, Germany) was inserted into the femoral artery. A double lumen central venous line was inserted into the right internal jugular vein. Patients were in the supine position throughout the entire study period. Cardiopulmonary bypass (CPB). The CPB circuit was primed with a crystalloid-colloid mixture: 250 ml Mannitol 10%, 500 ml hydroxyethyl starch 10%, 1500 ml Ringers Lactate and 100 ml sodium hydrogen carbonate 8.4%. Before CPB, 300 IU kg -1 heparin was administered to achieve an activated clotting time (ACT) > 480 sec. After clamping of the aorta, cardiac arrest was induced using crystalloid or blood cardioplegia at the discretion of the attending surgeon. CPB was managed according to the α-stat principle with a minimal nasopharyngeal temperature of 32 C and a nonpulsatile flow of 2.0 to 2.4 l min -1 m -2. After termination of CPB, protamine was administered to antagonize the heparin effect. Transpulmonary thermodilution and pulse contour cardiac output. The arterial thermistor catheter was connected via a 3-way stopcock to the PiCCO pressure transducer, positioned at the level of the mid-axillary line for monitoring of arterial pressure, TPCO as well as pulse contour cardiac output (PCCO). Four central venous bolus injections of 15 ml cold isotonic saline were injected within 3 seconds for the measurement of TPCO at every time point. The first set of measurement was used for initial calibration of PCCO [8]. In addition, the Vigileo system was connected via the FloTrac sensor to the femoral artery catheter. Calculation of CO via the Vigileo system is based on height, weight, gender, age of the patient and arterial blood pressure [13]. The underlying algorithm is the property of the manufacturer. Study protocol. Hemodynamic measurements included recordings of heart rate (HR), mean arterial pressure (MAP), central venous pressure (CVP), TPCO, PCCO and VigileoCO, which were measured intra- and postoperatively at the following time points: after induction of anesthesia (T 1 ), after sternotomy (T 2 ), immediately after a volume load of 10 ml kg -1 hydroxyethyl starch 6% (T 3 ), 20 min after this volume load (T 4 ), 15 minutes after weaning from CPB (T 5 ), after retransfusion of autologous blood (from the extracorporeal circulation) (T 6 ), after arrival at the intensive care unit (T 7 ), immediately after a second volume load of 10 ml kg -1 hydroxyethyl starch 6% (T 8 ) and 20 min 62

63 Arterial pulse contour based cardiac output later (T 9 ). At each time point, PCCO and VigileoCO were recorded prior to the four transpulmonary thermodilution procedures, i.e. before re-calibration of the PiCCO device. After the four transpulmonary thermodilution procedures, VigileoCO was recorded again and averaged with the earlier obtained value. At T 1 and T 5, no PCCO values were obtained because of (re)-starting the system in the OR and in the intensive care. SVR was calculated according to the formula SVR = (MAP CVP) x 79.9/CO, in which transpulmonary cardiac output was used. Data Analysis and Statistics. After gathering the data for this study, we performed a power analysis. The sample size calculation was performed to limit the width of a 95% confidence interval (CI) for the mean error. Based on a mean cardiac output of 5.0 l min -1, a mean error of 30% [14] and a desired total width of the CI of no more than 20%, a sample size of 22 was needed. Statistical analysis was performed using GraphPad PRISM (version 4.0., GraphPad software Inc, San Diego, CA, USA). All data are expressed as mean ± standard deviation (SD) unless otherwise stated. Hemodynamic parameters at each time point were compared to baseline by analysis of variance for repeated measurements (ANOVA). If the ANOVA revealed a significant interaction, post-hoc analysis was performed using paired samples t-test. Pearson s correlation coefficient was used to describe the correlation between the three techniques. Bias, precision and limits of agreement (LOA) were calculated according to Bland and Altman [15]. Mean error was calculated from 2 * precision divided by mean TPCO (or PCCO) and expressed as percentage [14]. Mean TPCO was used when TPCO was the reference technique, while mean PCCO was used when PCCO was the reference technique. 4 Results Patient characteristics, patient history and home medication are given in table 4.1. The time course of HR, CVP, systemic vascular resistance (SVR), TPCO, PCCO, VigileoCO, blood and rectal temperatures are presented in table 4.2. After induction of anesthesia and before surgery, a phenylephrine bolus was administered if MAP was below 60 mmhg. From the 22 patients studied, 3 patients received low dose dopamine (mean 4 µg kg -1 min -1 ) during weaning from bypass and two patients received dopamine (2.7 and 5.7 µg kg -1 min -1 ) during postoperative measurements. Moreover, 6 patients were paced via atrial leads because of sinus bradycardia at time point T 5 and T 6, 5 patients out of that 6 were paced at T 7 and T 8, while 4 out of that five were paced at T 9. No significant differences were observed when comparing bias, precision and mean error in patients with atrial pacing compared to patients with sinus rhythm. CVP and TPCO increased significantly after volume load in the OR (T 3, T 4 ) and in the ICU (T 8, T 9 ), whereas MAP and VigileoCO increased significantly only after volume load in the ICU. After 63

64 Chapter 4 weaning from CPB, MAP was decreased compared to pre-cpb values. The decrease in MAP was accompanied by a significant decrease in TPCO (and a decrease in VigileoCO) compared to pre- CPB values. One patient died the second day after surgery, due to pericardial tamponade and subsequent cardiac failure. Mean ± SD Range Age (yr) 66 ± Weight (kg) 80 ± Height (cm) 174 ± BSA (m 2 ) 1.95 ± BMI (kg m -2 ) 26.3 ± CPB time (min) 85 ± Cross clamp time (min) 61 ± Gender 18 m : 4 f Relevant history Number of patients Diabetes 7 Hypertension 15 Myocardial infarction 11 COPD 6 Medication Number of patients β-blocker 18 Calcium blocker 9 ACE-inhibitor 10 AR-blocker 2 Nitrates 11 Diuretics 7 Table 4.1. Patient characteristics, relevant history and home medication. A total of 184 sets of CO-measurements were available for comparison of TPCO and VigileoCO. The analysis of pooled data according to Bland-Altman is presented in table % of the data were within 2SD of the bias. Bias between TPCO and VigileoCO was 0.00 l min -1, with a precision of 0.87 l min -1. The limits of agreement were l min l min -1 with a mean error of 33%. 64

65 Arterial pulse contour based cardiac output T1 T2 T3 T4 T5 T6 T7 T8 T9 HR (bpm) 59 ± 9 62 ± ± 8 61 ± 9 72 ± ± ± ± ± 12 MAP (mmhg) 63 ± ± ± ± ± 8 a 60 ± 11 b 69 ± 12 c 78 ± 13 d 80 ± 15 d CVP (mmhg) 6 ± 2 7 ± 3 10 ± 3 a 9 ± 3 7 ± 2 10 ± 3 b 8 ± 3 13 ± 3 d 11 ± 3 d TPCO (l min -1 ) 4.02 ± ± ± 1.22 a 5.71 ± 1.08 a 4.96 ± 0.92 a 5.60 ± ± ± 1.28 d 6.17 ± 1.38 d PCCO (l min -1 ) ± ± ± 1.09 a 5.79 ± 1.58 a 5.33 ± ± ± 1.32 VigileoCO (l min -1 ) 3.95 ± ± ± ± ± ± ± ± 0.98 d 6.03 ± 1.00 d SVR (dyne sec cm -5 ) 1194 ± ± ± 188 a 837 ± 205 a 721 ± 118 a 754 ± ± 259 c 868 ± ± 227 Blood temp (ºC) 35.9 ± ± ± 0.6 a 35.3 ± 0.6 a 36.4 ± 0.4 a 36.2 ± 0.4 b 35.8 ± 0.5 c 35.4 ± 0.5 d 35.5 ± 0.5 d Rectal temp (ºC) 36.2 ± ± ± 0.5 a 35.7 ± 0.5 a 36.4 ± ± ± 0.5 c 35.9 ± 0.5 d 35.7 ± 0.5 d Table 4.2 Hemodynamic data obtained at each time point. All values are expressed as mean ± SD; n = 22. T1 = after induction before skin incision, T2 = after sternotomy, T3 = immediately after a volume load of 10 ml kg -1 hydroxyethyl starch 6%, T4 = 20 min after volume load, T5 = 15 minutes after weaning from bypass, T6 = after retransfusion of autologous blood (from the extracorporeal circulation), T7 = after arrival at the intensive care unit, T8 = immediately after a second volume load of 10 ml kg -1 hydroxyethyl starch 6%, and T9 = 20 min later. HR = heart rate, MAP = mean arterial pressure, CVP = central venous pressure, SVR = systemic vascular resistance. a p < 0.05 (versus T2); b p < 0.05 (versus T5); c p < 0.05 (versus T6); d p < 0.05 (versus T7). 4 65

66 Chapter 4 The bias of PCCO and VigileoCO was l min -1, the precision being 1.08 l min -1, resulting in LOA of l min l min -1 and a mean error of 40%. The bias between TPCO and PCCO was 0.02 l min -1, the precision was 0.93 l min -1, the LOA l min l min -1 and the mean error 35%. Time r TPCO versus VigileoCO PCCO versus VigileoCO TPCO versus PCCO Bias l min -1 Precision l min -1 ME (%) T r Bias l min -1 Precision l min -1 ME (%) r Bias l min -1 Precision l min -1 ME (%) T T T T T T T T Pooled data - T5 Pooled data Table 4.3 Bland-Altman analysis and Pearson s correlation coefficient for data per time point and pooled data. n = 22. T 1 = after induction before skin incision, T 2 = after sternotomy, T 3 = immediately after a volume load of 10 ml kg -1 hydroxyethyl starch 6%, T 4 = 20 min after volume load, T 5 = 15 minutes after weaning from bypass, T 6 = after retransfusion of autologous blood (from the extracorporeal circulation), T 7 = after arrival at the intensive care unit, T 8 = immediately after a second volume load of 10 ml kg -1 hydroxyethyl starch 6%, and T 9 = 20 min later. r = Pearson s correlation coefficient. Analyses of data for every defined measurement point are given in table 4.3. The correlation coefficient between TPCO and VigileoCO varied between 0.53 and The mean error between both methods varied between 24% and 45%. The best correlations and the least differences between TPCO and VigileoCO were observed in patients in post-bypass closed chest conditions and during the post-operative ICU period. The correlation coefficient between PCCO and VigileoCO varied between 0.21 and 0.78, with a mean error between 26% and 56%. As with TPCO and VigileoCO, the best correlation between PCCO and VigileoCO was observed in the post-operative period. 66

67 Arterial pulse contour based cardiac output At T 5, after weaning from bypass, we observed a worse correlation (r =0.48) between TPCO and non-calibrated PCCO values with a mean error of 53%. In contrast, the correlation between TPCO and VigileoCO at that time point was 0.58 and the mean error was 33%, respectively. 4 Figure 4.1 Bias ± 2*SD according to Bland-Altman for pooled data excluding data obtained at T 5. If non-calibrated PCCO-values obtained at T 5 were not included in the analysis (as recommended by the manufacturer), the respective correlation coefficients of TPCO-VigileoCO, PCCO- VigileoCO and TPCO-PCCO were 0.76, 0.72 and 0.85 with a bias of 0.01 l min -1, l min -1 and 67

68 Chapter l min -1 respectively and a mean error of 33%, 33% and 27% respectively (table 4.3, figure 4.1 and figure 4.2). Figure 4.2 Pearson s correlation coefficients for pooled data excluding data obtained at T 5. 68

69 Arterial pulse contour based cardiac output Discussion In this controlled clinical trial, we studied a recently introduced, pulse-contour based continuous CO-monitor (Vigileo, Edwards Lifesciences, Irvine, CA, USA) during the perioperative time course in patients undergoing CABG-surgery. Our results suggest an acceptable bias and precision between TPCO and arterial pulse contour based VigileoCO during post-bypass closed chest conditions and in the intensive care unit. The accuracy of the Vigileo device was found to be clinically acceptable, except for pre-bypass values, when patients received bolus doses of vasopressors resulting in a sudden increase in vascular tone. The mean error of the established PCCO system (PiCCO, Version 7.0, Pulsion Medical Systems, Munich, Germany) in comparison to TPCO was below 30% at all time points except T 3, T 5 and T 6. When comparing the PCCO with VigileoCO, mean errors were generally higher compared to the comparisons of TPCO with VigileoCO alone. Immediately after weaning from CPB, there was a worse correlation between PCCO and VigileoCO as well as between PCCO and TPCO. Pulmonary artery thermodilution cardiac output is still the accepted reference technique for clinical assessment of CO [1, 2]. However, the PAC is a highly invasive monitoring technique and subsequent efforts have been made to develop alternative, minor invasive CO-monitoring devices. One of the less invasive alternatives is the use of TPCO instead of the PAC. The accuracy and precision of the TPCO technique is comparable to classical pulmonary artery thermodilution [6-8, 10]. Moreover, using TPCO in conjunction with the PiCCO system, measurement of PCCO is available [9]. Both the PiCCO system and the Vigileo system are arterial pressure curve based CO monitoring systems. Therefore, we used TPCO as the reference technique in the present study. We compared both techniques with arterial pressure curves obtained in the femoral artery. CO determination via the femoral artery seems to be superior to CO determination in the radial artery, particularly due to less damping of the femoral arterial pressure curve during the first few hours after weaning from CPB. However, the incidence of infection in femoral artery cannulation is higher (0.44% versus 0.13% for radial artery cannulation), while the main complication rate of radial artery cannulation is temporary occlusion (19.7% versus 1.18% for femoral artery cannulation) [16]. Another major problem of studying different methods for CO monitoring is the fact that regularly no real reference technique (e.g. an electromagnetic flow meter) is available in a clinical setting. Therefore, the true CO is unknown, as the reference technique in the underlying study (transpulmonary thermodilution) has an inherent bias of approximately 10-20% [6, 7]. In 1999, Critchley and Critchley performed a meta-analysis of studies comparing different techniques of cardiac output monitoring [14]. They analyzed the percentage errors of different methods against 4 69

70 Chapter 4 the clinical reference technique and used it to determine clinically acceptable limits of agreements between methods. According to this type of analysis, mean errors up to ± 30% are acceptable for clinical purposes. Therefore, we calculated not only bias, precision and limits of agreement according to Bland-Altman, but also the mean error between the three different techniques. In the present study, mean error between VigileoCO and TPCO as the reference technique, showed a mean error of 33%, which is relatively close to the upper acceptable limit of 30%. The highest difference between VigileoCO and TPCO were found after induction of anesthesia, which may be due to the fact that patients received boluses of phenylephrine in that period due to a decrease in mean arterial pressure below 60 mmhg. No patients received vasopressor support after weaning from CPB, therefore it is not likely that drug induced changes in vascular tone affected the results of the present study. However, during situations with rapidly changing cardiovascular conditions (e.g. after induction of anesthesia and sternotomy), the difference between methods was increased. Most likely, the increased difference in CO may be caused by the algorithms for calculating of CO incorporated in the Vigileo device. As far as we know, the Vigileo system calculates stroke volume (SV) using arterial pulsatility (standard deviation of the pressure wave over a 20 seconds interval), according to the equation SV = K * Pulsatility [11]. K is a constant quantifying arterial compliance and vascular resistance. K is derived from patient characteristics (gender, age, height and weight) according to the method described by Langewouters [13], and waveform characteristics (skewness and kurtosis of individual arterial pulse waves) [11]. The calibration constant K is automatically recalculated (software version 01.01) every 10 minutes. Therefore under conditions of rapidly changing vascular resistance, the response of the Vigileo device in the time domain will be probably delayed. However, the manufacturer recently provided a new software version (01.07), recalibrating the system every minute. This may improve the accuracy and precision during rapid changes in vasomotor tone. Changes of CO by a defined volume challenge were detected adequately under open (T 3, T 4 ) and closed chest conditions (T 8, T 9 ). Recently, Manecke et al. compared the Vigileo monitor with intermittent CO-measurements by pulmonary artery thermodilution (10). Over a wide range of CO ( l min -1 ), they found a close agreement (bias 0.04 l min -1, precision 0.99 l min -1 ) between both techniques. These results are in line with the present study. However, one major limitation of the study from Manecke et al. is that no standardized changes in volume status were done. Mayer and colleagues compared the VigileoCO with pulmonary artery thermodilution CO in patients undergoing CABG and/or valve repair [17]. In their study, the overall mean error was 46%, which was substantially higher than the overall mean error of 33% obtained in our study. However, the recent study of Mayer et al. differed 70

71 Arterial pulse contour based cardiac output from our study in several important points, which may explain the diverging results. First, the patient population differed considerably. Whereas Mayer et al. included a heterogeneous patient population (Patients undergoing CABG and/or valve surgery), our study population was homogeneous (CABG-patients). Second, in our study the FloTrac sensor was connected to a femoral arterial line, whereas they connected the FloTrac sensor to the radial arterial line, which may also influence the results in particular after CPB. Moreover, Mayer et al. conducted no definitive fluid challenge. It seems possible that these differences in study design may have influenced the results of the study from Mayer et al. and at least in part can explain the increased mean error comparing VigileoCO and pulmonary artery thermodilution CO. We used the PiCCO system as the reference technique, which was already studied under a variety of experimental and clinical circumstances [6-10, 18]. We obtained a good agreement between PCCO and TPCO, enabling continuous measurement of CO with the PCCO technology. The mean error of pooled data is 27% if uncalibrated data obtained at T 5 were excluded from the analysis. In contrast to Vigileo, measurement of arterial pulse contour derived cardiac output by the PiCCO is dependent on re-calibration if hemodynamic conditions change rapidly. However, some authors claimed that PCCO can be measured with acceptable agreement without re-calibration even for a longer period of time [9]. In the present study, PCCO values obtained immediately after CPB (T 5 ) without re-calibration showed a worse agreement to TPCO-values obtained by thermodilution immediately thereafter. Therefore, we conclude that re-calibration after profound changes in hemodynamics are a pre-requisite for adequate measurement of CO with the PCCO technique. The results of the present study confirmed the findings from Sander et al [18]. Sander et al. observed also a good agreement between TPCO and PCCO before CPB but found less good agreement between TPCO and PCCO after CPB. When comparing both arterial pulse contour derived techniques (VigileoCO vs. PCCO), bias, precision, limits of agreement and mean error were found to be increased compared to the results obtained for the comparison between the Vigileo device and the reference technique (TPCO). The most pronounced changes were found immediately after weaning from CPB, which at least in part may be explained by the fact that uncalibrated PCCO values were obtained. Moreover, VigileoCO after sternotomy was different from PCCO and TPCO values. During the remaining time points, both minor invasive techniques offers comparable accuracy and precision. The lowest mean error was observed during the ICU period, this holds true for both the Vigileo and the PCCO technique. A second volume load maneuver resulted in a significant increase in CO. 4 71

72 Chapter 4 Conclusion The results of the present study suggest that the new arterial pulse contour device enables assessment of cardiac output in patients undergoing CABG surgery with clinically acceptable bias and precision in the post-bypass period under closed chest conditions and in the intensive care. PiCCO and Vigileo are interchangeable in the postoperative period. The latter technique is of particular interest as no calibration is needed and the flow sensor can be used with any common arterial line. VigileoCO was found to exceed TPCO in the pre-bypass period and open chest condition, which makes both techniques not interchangeable under these circumstances. Thereby, during rapidly changing hemodynamic conditions, the accuracy and precision is not acceptable. Further refinement of the algorithm resulting in decreased response time may improve the accuracy under such hemodynamic conditions. 72

73 Arterial pulse contour based cardiac output References 1. Swan HJC, Ganz W, Forrester J, et al. Catheterization of the heart in man with use of a flow-directed balloon-tipped catheter. N Engl J Med 1970; 283: De Waal EEC, De Rossi L, Buhre W. Pulmonary artery catheter in anaesthesiology and intensive care. Anaesthesist 2006; 55: Connors AF Jr, Castele RJ, Farhat NZ, et al. Complications of right heart catheterization. A prospective autopsy study. Chest 1985: 88: Peters SG, Afessa B, Decker PA, et al. Increased risk associated with pulmonary artery catheterization in the medical intensive care unit. J Crit Care 2003; 18: Jensen L, Yakimets J, Teo KK. A review of impedance cardiography. Heart Lung 1995; 24: Goedje O, Hoeke K, Lichtwarck-Aschoff M, et al. Continuous cardiac output by femoral arterial thermodilution calibrated pulse contour analysis: comparison with pulmonary arterial thermodilution. Crit Care Med 1999; 27: 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: Buhre W, Weyland A, Kazmaier S, et al. Comparison of cardiac output assessed by pulsecontour analysis and thermodilution in patients undergoing minimally invasive direct coronary artery bypass grafting. J Cardiothor Vasc Anesth 1999; 13: Gödje 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: Halvorsen PS, Espinoza A, Lundblad R, et al. Agreement between PiCCO pulse-contour analysis, pulmonary artery thermodilution and transthoracic thermodilution during off-pump coronary artery bypass surgery. Acta Anaesthesiol Scand 2006; 50: Manecke GR. Edwards FloTrac sensor and Vigileo monitor: easy, accurate, reliable cardiac output assessment using the arterial pulse wave. Expert Rev Med Devices 2005; 2: Headley JM. Arterial pressure-based technologies: a new trend in cardiac output monitoring. Crit Care Nurs Clin N Am 2006; 18: Langewouters GJ, Wesseling KH, Goedhard WJ. The pressure dependent dynamic elasticity of 35 thoracic and 16 abdominal human aortas in vitro described in a five component model. J Biomech 1985; 18:

74 Chapter Critchley LAH, Critchley JAJH. A meta-analysis of studies using bias and precision statistics to compare cardiac output measurement techniques. J Clin Monit 1999; 15: Bland JM, Altman DG. Statistical methods for assessing agreement between two methods of clinical measurement. Lancet 1986; 1: Scheer BV, Perel A, Pfeiffer U. Clinical review: complications and risk factors of peripheral arterial catheters used for haemodynamic monitoring in anaesthesia and intensive care medicine. Crit Care 2002; 6: Mayer J, Boldt J, Schöllhorn T, et al. Semi-invasive monitoring of cardiac output by a new device using arterial pressure waveform analysis: a comparison with intermittent pulmonary artery thermodilution in patients undergoing cardiac surgery. Br J Anaesth 2007; 98: Sander M, von Heymann C, Foer A, et al. Pulse contour analysis after normothermic cardiopulmonary bypass in cardiac surgery patients. Crit Care 2005; 9:

75 Chapter 5 Haemodynamic changes during low-pressure CO 2 pneumoperitoneum in young children EEC de Waal MD, CJ Kalkman MD PhD Department of Anaesthesiology, Wilhelmina Children s Hospital, University Medical Centre, Utrecht, Netherlands, Paediatr Anaesth 2003; 13: 18-25

76 Chapter 5 Abstract Background: Both mechanical and pharmacological effects may contribute to the haemodynamic consequences of CO 2 pneumoperitoneum. The aim of the present study was to evaluate the haemodynamic effects of low-pressure pneumoperitoneum (IAP 5 mm Hg) in young children (<3 yr). Methods: Thirteen children aged 6-36 months (ASA physical status I-III) scheduled for laparoscopic fundoplication for gastroesophageal reflux were investigated in the head-up position (10º). Non-invasive thoracic electrical bioimpedance cardiac index (CI), stroke volume index (SVI), heart rate (HR), mean arterial pressure (MAP) and peak inspiratory pressure (PIP) were recorded together with P et CO 2 and P a CO 2 at five time points: before insufflation, 20, 35 and 70 min after start of CO 2 insufflation and 12 min after desufflation. During insufflation, minute ventilation was not adjusted and the intraabdominal pressure was maintained at 5 mm Hg. Results: During insufflation, P et CO 2 increased from 29 ± 4 to 37 ± 5 mm Hg (p < 0.001) and P a CO 2 increased from 31 ± 4 to 39 ± 5 mm Hg (p < 0.01). CI increased from 2.39 ± 0.86 to 2.92 ± 0.94 l min -1 m -2 (p < 0.01), HR increased from 108 ± 10 to 126 ± 22 bpm (p < 0.01), MAP increased from 52 ± 10 to 63 ± 9 (p < 0.05) and PIP increased from 16 ± 3 to 18 ± 3 cm H 2 O (p < 0.001). There were no changes in SVI and arterial oxygen saturation. Conclusion: We conclude that low-pressure CO 2 pneumoperitoneum (with IAPs not exceeding 5 mm Hg) for laparoscopic fundoplication in infants and children does not decrease their cardiac index. 76

77 Bioimpedance cardiac output in children Introduction Laparoscopic surgical procedures are increasingly used in infants and children [1]. In adults, the haemodynamic effects of pneumoperitoneum can be considerable. In particular, decreases in cardiac output (CO) and increases in systemic vascular resistance (SVR) have been reported [2-6]. Relatively few studies of the cardiovascular effects of laparoscopic surgery in children have been published. The haemodynamic changes reported during pneumoperitoneum in children include increases in heart rate (HR) and blood pressure [7-11], together with decreases in cardiac index (CI) [12, 13] and aortic blood flow [14]. These haemodynamic changes were observed during laparoscopic procedures with intraabdominal pressures above 8 mm Hg while minute ventilation was adjusted in an attempt to maintain normocapnia during insufflation. The aim of the present study was to investigate the haemodynamic effects of low-pressure pneumoperitoneum (5 mm Hg) in young children undergoing elective laparoscopic fundoplication (Thal-procedure [15]). The CO 2 pneumoperitoneum was superimposed on a baseline of moderate hypocapnia and fixed minute ventilation. Methods The protocol was approved by the Hospital Ethics Committee and written informed parental consent was obtained in all cases. 13 children (ASA physical status I III) at least 53 weeks postconception and aged < 3 years scheduled for laparoscopic fundoplication for treatment of primary or secondary gastroesophageal reflux were investigated. Children with significant cardiovascular diseases, children with clinical signs of pulmonary disease without adequate treatment and children with increased intracranial pressure or decreased intracranial compliance were excluded. In all cases, a low pressure CO 2 pneumoperitoneum (5 mm Hg) was applied. 5 Anaesthetic technique No premedication was given and general anaesthesia was induced via facemask using oxygen, nitrous oxide and halothane. When unconscious, an intravenous line was inserted and atracurium 0.5 mg kg -1 and sufentanil 0.2 µg kg -1 were given IV. Administration of nitrous oxide and halothane was discontinued and the child was manually ventilated with oxygen/air (F i O 2 0.4) and isoflurane (F i 1%). Nitrous oxide is not used during the laparoscopic guided surgery because of its potential for flammability and dilation of the bowel. After orotracheal intubation, continuous infusions of atracurium 0.5 mg kg -1 h -1 and sufentanil 0.5 µg kg -1 h -1 were started and the operating table was positioned in the head-up position (10 ). This position was maintained during the entire study period. The child was mechanically ventilated (Cicero, Dräger, Lübeck, Germany) with oxygen/air 77

78 Chapter 5 (F i O 2 0.4) and isoflurane (F i 1%). This level of isoflurane was kept constant during the study period. Tidal volume was set at 10 ml kg -1, while respiratory rate was adjusted to achieve a moderate hypocapnia (end-tidal CO 2 between 25 and 32 mm Hg) before insufflation. Thereafter, these ventilatory settings remained unchanged during the operation. An IV-infusion of glucose 2.5% NaCl 0.45% 10 mg kg -1 h -1 was started. A urinary catheter, a large bore gastric tube and a radial arterial line were inserted. Arterial pressure was measured with the transducer at mid-thoracic level. Before the start of operation, an extra bolus of sufentanil 0.1 µg kg -1 IV was given to prevent any hemodynamic response to surgical stimulation. Carbon dioxide was insufflated using an electronic laparoflator (Karl Storz Type , Tuttlingen, Germany). Insufflation was terminated at an intraabdominal pressure (IAP) of 5 mm Hg. After desufflation, the continuous infusions of atracurium and sufentanil were discontinued. The administration of isoflurane was stopped after the last measurement and the patient was weaned from ventilation and extubated when consciousness had returned. Data collection With the patient in the head-up position, the following variables were recorded before insufflation (T 0 ), 20 minutes after start of CO 2 -insufflation (T 1 ), 35 minutes after start of CO 2 -insufflation (T 2 ), 70 minutes after start of CO 2 -insufflation (T 3 ) and 12 min after desufflation of the pneumoperitoneum at the end of surgery (T 4 ): cardiac output (CO), heart rate (HR), stroke volume (SV), mean arterial pressure (MAP), end-tidal CO 2 (P et CO 2 ), partial pressure of CO 2 in arterial blood (P a CO 2 ), ph, arterial oxygen saturation (SaO 2 ) and peak inspiratory pressure (PIP). The duration of insufflation was recorded. Arterial blood gas samples were collected at T 0, T 2, T 3 and T 4. Thoracic electrical bioimpedance cardiac output CO was continuously measured by means of non-invasive thoracic electrical bioimpedance using a BoMed NCCOM3-R7 thoracic bioimpedance monitor (BoMed Medical Manufacturing Ltd., Irvine, California). In order to bypass the built-in adult algorithms for use in the neonatal and paediatric patients, a different (artificial) set of height (cm) and weight (kg) has to be utilised to produce clinically valid CO values by this bioimpedance monitor (CardioDynamics International Corporation, bulletin # 0020 Dec 30, 1993). Before placing the electrodes, the skin was cleaned with alcohol in order to achieve a skin-to-electrode impedance as low as possible. Four pairs of electrodes were positioned on each side of the neck and the chest. The two sensing electrode pairs were placed in the mid-axillary line at the level of the sternal xiphoid and on the lateral side of the neck just above the clavicles respectively. The current injecting electrodes were placed in the 78

79 Bioimpedance cardiac output in children neck of the patient just above the sensing electrodes and below the thoracic sensing electrodes. Beat-to-beat SV was determined using changes in transthoracic electrical conductivity. In order to counteract to the respiratory changes in SV and CO (CI), SV is averaged over 16 heartbeats. As alternating current is passed through the thorax, pulsatile variations in thoracic aortic blood flow and velocity alter the impedance, which is measured and recorded as changes in thoracic impedance. SV is calculated from the Bernstein-Sramek formula [16] SV = VEPT x VET x (dz/dt max )/Z 0 where VEPT (volume of electrically participating tissue) is a constant derived from the patient s height (H) and weight (W) and is corrected by a factor to account for the discrepancy between the patient s actual (a) and ideal (i) weight, according to the equation VEPT = W a /W i x [(0.17xH)³/4.2]. VET is the ventricular ejection time; dz/dt max is the maximum rate of change of impedance during systole, and Z 0 is the basal thoracic impedance. SV x HR = CO. CI and stroke volume index (SVI) were calculated, dividing CO and SV respectively by body surface area (BSA). SV and CO were determined in a stable period without irregular cardiac rhythms, tachyarrhythmias and coagulation. Statistical analysis All results are expressed as mean ± standard deviation (SD). Analysis of variance (ANOVA) for repeated measurements was performed to test for time effects. Post-hoc paired samples t-tests with Bonferroni correction were performed for the haemodynamic variables to determine if variables obtained at T 1 were significantly different from baseline and if variables obtained at T 4 were significantly different from those obtained at T 3 and baseline. P a CO 2 - and ph-results were analyzed using post-hoc paired samples t-tests with Bonferroni correction in order to determine if results obtained at T 2 were significantly different from baseline and if results obtained at T 4 were significantly different from those obtained at T 3 and baseline. Linear regression analysis was used to test whether there was a relation between P a CO 2 and CI and a relation between the increases in P a CO 2 compared to the increases in CI. A p-value < 0.05 was considered statistically significant. The SPSS program (version 9.0 for Windows NT, SPSS, and Chicago, IL) was used for all statistical analyses. 5 Results Patient characteristics are presented in table 5.1. Figure 5.1 shows the end tidal versus arterial carbon dioxide. During CO 2 -insufflation P et CO 2 increased from 29 ± 4 to 37 ± 5 mm Hg (p < 0.001) and P a CO 2 increased from 31 ± 4 to 39 ± 5 mm Hg (p = 0.003). PIP increased from 16 ± 3 to 18 ± 3 (p < 0.003). The increase of end tidal and arterial CO 2 resulted in a decrease in ph from 7.40 ±

80 Chapter 5 to 7.31 ± 0.07 (p = 0.003). Desufflation resulted in decreases in P et CO 2 and PIP. Figure 5.2 shows the haemodynamic changes induced by CO 2 insufflation and after desufflation. HR increased from 108 ± 10 to 126 ± 22 bpm (p = 0.009), MAP increased from 52 ± 10 to 63 ± 9 mm Hg (p = 0.015) and CI increased from 2.4 ± 0.9 to 2.9 ± 0.9 l min -1 m -2 (p = 0.006). Pneumoperitoneum caused no significant changes in stroke volume index. These changes persisted during the entire insufflation period. There was no significant correlation between P a CO 2 and CI, nor the increase in P a CO 2 and the increase in CI. Mean ± SD Range Age (mo) 20 ± Male/Female 5 / 8 Weight (kg) 10.1 ± Height (cm) 79 ± BSA (m 2 ) 0.48 ± Duration of insufflation (min) 123 ± Table 5.1 Patients characteristics. All values are expressed as mean ± SD; n = 13. Figure 5.1 End tidal and arterial CO 2. All values are expressed as mean ± SD. Discussion Because the abdominal wall is more pliable than that of adults [17], adequate visualization of the intraabdominal contents in children is possible at lower IAP than in adult laparoscopic surgery. A relatively low IAP up to 5 mm Hg in young children is sufficient to facilitate laparoscopic surgery. CO 2 insufflation with the operating table in the reverse-trendelenburg position while maintaining 80

81 Bioimpedance cardiac output in children minute ventilation constant resulted in significant increases in CI, HR, MAP and PIP. CI increased mainly as a result of increased HR. This suggests that mechanical factors that are known to contribute to a decrease in CO in adults (level of IAP, upward shift of diaphragm, diminished venous return) do not play such a role in small children when lower pressures are used. Therefore, it is most likely that a positive mechanical effect of lower intraabdominal pressures on venous return combined with the pharmacological effects of CO 2 absorption result in the observed haemodynamic changes in our study. 5 Figure 5.2 Haemodynamic variables before, during and after CO 2 -insufflation for laparoscopic fundoplication. All values are expressed as mean ± SD. The cardiovascular effects of pneumoperitoneum have been studied extensively in adults. Most authors reported an increase in blood pressure, systemic vascular resistance, left ventricular end systolic wall stress and end-tidal CO 2, together with a decrease in CI, SVI and fractional area shortening [2-6]. In general, the IAP in these studies in adults is 15 mm Hg or more. The haemodynamic effects of laparoscopic surgery have been studied in children, but there are several differences between studies regarding the level of IAP (ranging from 6 to 13 mm Hg): whether minute ventilation was adjusted in an attempt to maintain normocapnia, whether arterial CO 2 was measured and the technique to estimate changes in CO. In children, increases in blood pressure and 81

82 Chapter 5 HR during insufflation were noted by Tobias [7], Sfez [8] and Schäfer [9]. In these studies PIP and P et CO 2 increased together with decreases in SaO 2 and lung compliance. Table 5.2 summarizes results from studies concerning CO measured in children during CO 2 insufflation. Sakka and coworkers [12] noted a decrease in CI measured with transesophageal echocardiography in small children for the first 12 mm Hg period, while this decrease was not seen during the second increase of IAP to 12 mm Hg. Kardos [13] reported a significant reduction in CO during laparoscopy with a IAP of mm Hg in children aged 13 ± 2.3 years. A decrease in Aortic Blood Flow Index and an increase in SVRI were noted during pneumoperitoneum (IAP 10 mm Hg) by Gueugniaud [14] et al using an echo-doppler device to measure flow velocity in the descending thoracic aorta together with measurement of the aortic diameter, allowing estimation of the CO. In contrast, we observed an increase in CO during insufflation of carbon dioxide intraperitoneally using thoracic electrical bioimpedance. The validity of our CO results is dependent on the accuracy of bioimpedance CO estimation during low-pressure pneumoperitoneum. Studies by Introna [18], Braden [19] and McKinley [20] showed that thoracic electrical bioimpedance CO correlates with thermodilution CO and direct Fick-derived CO in children. Artefact-free data obtained during a stable period were used for analysis. Author Age IAP (mmhg) Baseline P et CO 2 (kpa) Ventilation adjustment P et CO 2 during insufflation P a CO 2 CO measurement technique Sakka [12] 3.5 ± 1.3 yr MV no change n.a. TTE CI Kardos [13] 13 ± 2.3 yr MV max 5.8 kpa n.a. I.C. CI Gueugniaud [14] 23 ± 5 mo constant no change n.a. Echo- Doppler ABFI de Waal 20 ± 9 mo constant I.C. CI CI Table 5.2 Studies that assessed changes in cardiac output during laparoscopy in children. I.C. = Impedance Cardiography; CI = Cardiac Index. The haemodynamic effects of CO 2 -pneumoperitoneum are influenced both by the magnitude of IAP and hypercapnia-induced neurohumoral effects. One possible explanation for the different effects on cardiac output or CO 2 -related variables noticed by Sakka, Kardos and Gueugniaud respectively and the present study is that different intraabdominal pressures were used. Insufflation of any gas results in direct mechanical effects, due to the increase in IAP, while indirect pharmacological effects on cardiorespiratory function depend on the type of insufflating gas used. 82

83 Bioimpedance cardiac output in children The concept of abdominal vascular zone conditions offers a theoretical framework to interpret the complex haemodynamic responses to steady-state increases in IAP [21]. During insufflation with an IAP less than the right atrial pressure (RAP), blood is squeezed out of the venous capacitance vessels from the splanchnic circulation, resulting in increased venous return and an increase in CO. In the present study, the intraabdominal pressure was 5 mm Hg during insufflation, and we assumed that this was probably less than the RAP. An increase in IAP above the RAP causes compression of the inferior caval vein with a decrease in venous return, a decrease in preload and a subsequent decrease in CO [22]. In neonates and infants suffering from congenital heart malformation, pneumoperitoneum may lead to a decrease in preload (venous return) and an increase in afterload, with an attending increase in pulmonary and systemic vascular resistance. As this may cause a temporary or permanent reopening of intracardial shunts, e.g. foramen ovale or ductus arteriosus, with an increased risk for systemic gas embolism and heart failure [23], these patients may be considered to be less eligible for laparoscopic surgery than otherwise healthy children [11]. Sakka [12], Kardos [13] and Gueugniaud [14] noted a decrease in CI and aortic blood flow with a constant end-tidal CO 2 together with a significant increase in SVR. These findings illustrate the hypothesis of the abdominal vascular zone conditions. An IAP of 10 mm Hg, i.e. higher than the right atrial pressure, causes a decrease in venous return, a decrease in right ventricular CO, a decrease in left ventricular preload, a decrease in left ventricular CO and a decrease in aortic blood flow. No significant change in end-tidal CO 2 was noted in the study of Gueugniaud because the expected increase in P et CO 2 caused by CO 2 absorption was counteracted by a decrease in right ventricular CO. P a CO 2 was not measured in the studies of Sakka, Kardos and Gueugniaud [12-14]. Apart from the mechanical effects of increased IAP on venous return, gas insufflation may also result in changes in pulmonary function as a result of an elevation of the diaphragm, a decrease in 5 functional residual capacity (FRC), and an increase in alveolar dead space. After insufflation of CO 2 there is absorption of CO 2 in the circulation resulting in a further increase in P a CO 2 and P et CO 2. Hsing et al [24] reported that the plateau of end-tidal CO 2 during CO 2 insufflation is reached after 4.2 ± 0.6 min in children aged 11 month - 2 years and after 6.3 ± 1.0 min in children aged 2-5 years. There is also a time lag of P et CO 2 decline from plateau to baseline after desufflation of 6.2 ± 0.5 min in younger children and 8.3 ± 0.8 min in children between 2 and 5 years of age. The best assessment of the influence of CO 2 insufflation and desufflation on haemodynamic function is made in steady state conditions, i.e. after reaching a plateau. The time points of measurement in the present study were chosen according to these results. We hyperventilated the children to moderate hypocapnia before the operation by adjusting the respiratory rate. A few studies reported the effects of hypocapnia on the circulation. Hypocapnia (34 83

84 Chapter 5 mm Hg) induced by reducing instrumental deadspace [25] or by hyperventilation (27 mm Hg) [26] in critically ill patients did not result in a decrease in CO, while decreasing the CO 2 flow in a semiclosed respiratory system without a change in ventilatory parameters caused a 19% decrease in CO [27]. CO 2 absorption has pharmacological effects via the neurohumoral axis [28]), while helium is pharmacologically inert [29]. These haemodynamic effects are modified by the underlying cardiovascular status, intravascular volume status, patient positioning and the anaesthetic technique [10]. Hypercapnia causes moderate vasodilation of arterioles in most tissues and a marked vasodilation in the brain. Indirectly, carbon dioxide activates the central nervous system. It activates the sympathoadrenal system leading to significant increase in cortisol, epinephrine, norepinephrine, renin and vasopressin. However, there are no published studies that report vasopressin concentrations in children during CO 2 pneumoperitoneum. Changes in neurohumoral status increase myocardial contractility, HR, CO and blood pressure [30-32]. At the same time, CO 2 may cause irritation of the peritoneal surface by carbonic acid, which forms as a result of the chemical reaction of CO 2 with water [33]. This may lead to perioperative stimulation of the afferent nerves conducting pain stimuli and postoperative shoulder tip pain following laparoscopy. The position of the patient during laparoscopy influences lung volume, chest wall mechanics and haemodynamic status. Movement of the patient from the supine to the reverse-trendelenburg position is associated with an increase in functional residual capacity and a decrease in venous return. Placing the patient in the Trendelenburg position leads to an increase in venous return and an increase in central venous pressure in the awake state, during anaesthesia and during laparoscopy [34]. In the present study, all measurements including baseline were performed with the table in the reverse-trendelenburg position. This would be expected to compound the negative effect of raised IAP on venous return. However, the absence of a decrease in CI in our study patients suggests that this mechanism was not of importance in our patients. We hypothesize that neurohumoral stimulation of the circulation by the absorbed CO 2 and the applicated low pressure pneumoperitoneum are the most likely reasons for the observed increase in CO. However, the design of the study does not allow us to exclude the possibility of increased heart rate due to surgical stimulation. Anaesthesia was discontinued between T 3 and T 4 and the patient was breathing spontaneously 12 minutes after desufflation of pneumoperitoneum. At time point T 4 patients were lightly anaesthetized and weaning from ventilation confounded the observed haemodynamic changes. Therefore, the data recorded at T 4 do not provide information about the effects of the release of the pneumoperitoneum. In conclusion, this study demonstrates that low-pressure carbon dioxide pneumoperitoneum with intraabdominal pressures not exceeding 5 mm Hg for laparoscopic fundoplication in infants and 84

85 Bioimpedance cardiac output in children children has different haemodynamic effects compared to studies performed with higher intraabdominal pressures. Low pressure carbon dioxide pneumoperitoneum (with IAPs not exceeding 5 mm Hg) for laparoscopic fundoplication in infants and children does not decrease their CI. Acknowledgements The authors thank Dr. T. Ionescu, Dr. N. Turner and Dr. J. de Vries for their enthusiastic support and critical reading. 5 85

86 Chapter 5 References 1. Lugo-Vicente HL. Impact of minimally invasive surgery in children. Bol Asoc Med P R 1997; 89: Joris JL, Noirot DP, Legrand MJ, et al. Hemodynamic changes during laparoscopic cholecystectomy. Anesth Analg 1993; 76: Cunningham AJ, Turner J, Rosenbaum S, et al. Transoesophageal echocardiographic assessment of haemodynamic function during laparoscopic cholecystectomy. Br J Anaesth 1993; 70: Critchley LAH, Critchley JAJH, Gin T. Haemodynamic changes in patients undergoing laparoscopic cholecystectomy: measurement by transthoracic electrical bioimpedance. Br J Anaesth 1993; 70: O Leary E, Hubbard K, Tormey W, et al. Laparoscopic cholecystectomy: haemodynamic and neuroendocrine responses after pneumoperitoneum and changes in position. Br J Anaesth 1996; 76: Branche PE, Duperret SL, Sagnard PE, et al. Left ventricular loading modifications induced by pneumoperitoneum: a time course echocardiographic study. Anesth Analg 1998; 86: Tobias JD, Holcomb III GW, Brock III JW, et al. Cardiorespiratory changes in children during laparoscopy. J Pediatr Surg 1995; 30: Sfez M, Guérard A, Desruelle P. Cardiorespiratory changes during laparoscopic fundoplication in children. Paed Anaesth 1995; 5: Schäfer R, Gerlach K, Barthel M, et al. Auswirkungen endoskopischer Operationstechniken bei Kindern auf die Beatmung. Anästhesiol Intensivmed Notfallmed Schmerzther 1997; 32: Tobias JD. Anesthetic considerations for laparoscopy in children. Semin Laparosc Surg 1998; 5: Terrier G. Anaesthesia for laparoscopic procedures in infants and children: indications, intra- and postoperative management, prevention and treatment of complications. Curr Opin Anaesthesiol 1999; 12: Sakka SG, Huetteman E, Petrat G, et al. Transoesophageal echocardiographic assessment of haemodynamic changes during laparoscopic herniorrhaphy in small children. Br J Anaesth 2000; 84: Kardos A, Vereczkey G, Pirót L, et al. Use of impedance cardiography to monitor haemodynamic changes during laparoscopy in children. Paed Anaesth 2001; 11:

87 Bioimpedance cardiac output in children 14. Gueugniaud P-Y, Abisseror M, Moussa M, et al. The hemodynamic effects of pneumoperitoneum during laparoscopic surgery in healthy infants: assessment by continuous esophageal aortic blood flow echo-doppler. Anesth Analg 1998; 86: Martinez-Frontanilla LA, Sartorelli KH, Haase GM, et al. Laparosopic Thal-fundoplication with gastrostomy in children. J Pediatr Surg 1996; 31: Bernstein DP. A new stroke volume equation for thoracic electrical bioimpedance: theory and rationale. Crit Care Med 1986; 14: Holcomb III GW. Indications for laparoscopy in children. Int Surg 1994; 79: Introna RPS, Pruett JK, Crumrine RC, et al. Use of transthoracic bioimpedance to determine cardiac output in pediatric patients. Crit Care Med 1988; 16: Braden DS, Leatherbury L, Treiber F, et al. Noninvasive assessment of cardiac output in children using impedance cardiography. Am Heart J 1990; 120: McKinley DF, Pollack MM. A comparison of thoracic bioimpedance to thermodilution cardiac output in critically ill children (Abstract). Crit Care Med 1987; 15: Takata M, Wise RA, Robotham JL. Effects of abdominal pressure on venous return: abdominal vascular zone conditions. J Appl Physiol 1990; 69: Kitano Y, Takata M, Sasaki N, et al. Influence of increased abdominal pressure on steadystate cardiac performance. J Appl Physiol 1999; 86: Sfez M. Anesthésie pour coeliochirurgie en pédiatrie. Ann Fr Anesth Réanim 1994; 13: Hsing CH, Hseu SS, Tsai SK, et al. The physiological effect of CO 2 pneumoperitoneum in pediatric laparoscopy. Acta Anaesthesiol Sin 1995; 33: Mas A, Saura P, Joseph D, et al. Effects of acute moderate changes in P a CO 2 on global hemodynamics and gastric perfusion. Crit Care Med 2000; 28: Ichai C, Levraut J, Baruch I, et al. Hypocapnia does not alter hepatic blood flow or oxygen consumption in patients with head injury. Crit Care Med 1998; 26: Gelman S, Fowler KC, Bishop SP, et al. Cardiac output distribution and regional blood flow during hypocarbia in monkeys. J Appl Physiol 1985; 58: Joris JL, Chiche J-D, Canivet J-L M, et al. Hemodynamic changes induced by laparoscopy and their endocrine correlates: effects of clonidine. J Am Coll Cardiol 1998; 32: Rademaker BM, Bannenberg JJ, Kalkman CJ, et al. Effects of pneumoperitoneum with helium on hemodynamics and oxygen transport: a comparison with carbon dioxide. J Laparoendosc Surg 1995; 5: Cullen D, Eger EI II. Cardiovascular effects of carbon dioxide in man. Anesthesiology 1974; 5 87

88 Chapter 5 41: Rasmussen JP, Dauchot PJ, DePalma RG, et al. Cardiac function and hypercarbia. Arch Surg 1978; 113: Guyton AC, Hall JE, eds. Textbook of medical physiology. Philadelphia-London-Toronto- Montreal-Sydney-Tokyo: W.B. Saunders company, 1996; Johnson PL, Sibers KS. Laparoscopy: gasless vs. CO 2 pneumoperitoneum. J Reprod Med 1997; 42: Hirvonen EA, Nuutinen LS, Vuolteenaho O. Hormonal responses and cardiac filling pressures in head-up or head-down position and pneumoperitoneum in patients undergoing operative laparoscopy. Br J Anaesth 1997; 78:

89 Chapter 6 Assessment of stroke volume index with three different bioimpedance algorithms: lack of agreement compared to thermodilution EEC de Waal MD 1, MK Konings PhD 2, CJ Kalkman MD PhD 1, WF Buhre MD 1 1 Division of Perioperative and Emergency Care, 2 Department of Medical Technology, University Medical Center, Utrecht, Netherlands Intensive Care Med 2008; 34: 735-9

90 Chapter 6 Abstract Objective: Accuracy of bioimpedance stroke volume index (SVI) is questionable, because a number of studies reported inconsistent results. It remains unclear if the algorithms alone are responsible for these findings. We analyzed the raw impedance data with three algorithms and compared bioimpedance SVI to transpulmonary thermodilution (SVI TD ). Design: Prospective observational clinical study. Setting: University hospital. Patients and participants: Twenty adult patients scheduled for coronary artery bypass grafting (CABG). Intervention: SVI TD and bioimpedance parameters were simultaneously obtained before surgery (T 1 ), after bypass (T 2 ), after sternal closure (T 3 ), at the intensive care unit (T 4 ), at normothermia (T 5 ), after extubation (T 6 ) and before discharge (T 7 ). Bioimpedance data were analyzed off-line using cylinder (Kubicek: SVI K, Wang: SVI W ) and truncated cone based algorithms (Sramek-Bernstein: SVI SB ). Measurements and results: Bias and precision between the SVI TD and SVI K, SVI SB and SVI W was 1.0 ± 10.8, 9.8 ± 11.4 and ± 8.2 ml m -2 respectively, while the mean error was abundantly above 30%. Analysis of data per time moment resulted in a mean error above 30%, except for SVI W at T 2 (28%). Conclusions: Estimation of SVI by cylinder or truncated cone based algorithms is not reliable for clinical decision making in patients undergoing CABG surgery. A more robust approach for estimating bioimpedance based SVI may exclude inconsistencies in the underlying algorithms in existing thoracic bioimpedance cardiography devices. 90

91 Bioimpedance cardiac output in adults Introduction Additional information about the cardiovascular status of critically ill patients can be obtained by measuring cardiac output (CO). Pulmonary artery thermodilution CO monitoring has remained the reference technique for three decades [1], but is invasive and associated with specific complications [2-4]. Thoracic bioimpedance cardiography, a non-invasive CO monitoring technique, exhibits many qualities of the ideal CO monitor: operator independency, continuous and cost-effective [5]. Since the late 1960s, a number of bioimpedance devices have been developed with cylinder or cone based models of a homogeneously with blood filled human thorax. Method comparison studies have demonstrated conflicting results with respect to validity and reliability [6], varying from satisfactory correlations [7-9] to poor correlations [10, 11]. Inaccuracies can result from irregular cardiac rhythms, abnormal ventilatory patterns, motion artifacts, valvular heart diseases, electrocautery, changes in hematocrit, excessive changes in body temperature, and an obese body habitus [5]. Thereby, it remains unclear if the methodology (i.e. detection of impedance signals from the thorax using a small number of electrodes) per se or limitations of the underlying algorithms are responsible for these conflicting results. We hypothesize that bioimpedance SV measured with any of three well-established bioimpedance algorithms is valid and reliable. We compared bioimpedance stroke volume index (SVI) with transpulmonary thermodilution stroke volume index (SVI TD ) as a reference of proven accuracy [12]. 6 Materials and methods After approval by the IRB and written informed consent, 20 patients scheduled for elective coronary artery bypass graft (CABG) surgery with cardiopulmonary bypass were included. Exclusion criteria were: ejection fraction < 40%, femoral arterial disease and valvular heart disease. A total intravenous anesthesia technique was used during the operation. Normocapnia was maintained during mechanical ventilation (F i O 2 0.4, PEEP 5 cm H 2 0). A 4 Fr thermodilution catheter (Pulsiocath PV2014L16) was introduced in the femoral artery and connected to a commercially available CO device (PiCCO, Pulsion, Munich, Germany. Transpulmonary thermodilution cardiac output (TPCO) was measured by quadruple injections of 15 ml ice-cold saline into the right atrium and used for transpulmonary thermodilution stroke volume calculation. After rubbing and cleaning the skin with alcohol to achieve a skin-to-electrode impedance as low as possible, two current injecting electrodes were placed on the forehead and the left hip and two voltage sensing electrodes were placed on the lateral side of the neck just above the left clavicle and 91

92 Chapter 6 in the left mid-axillary line at the level of the sternal xiphoid. A 0.3 ma alternating current (64 khz) was applied. A thoracic bioimpedance cardiograph (HL-4, Hemologic, Amersfoort, The Netherlands) was used for recording raw bioimpedance signals in the perioperative period. The first derivative of the thoracic impedance (dz/dt) and the ECG signal were displayed on the screen. Raw data were analyzed off-line over a 20 second period (LabView, E-solutions, Arnhem, The Netherlands) and used for bioimpedance SV calculation using three distinct reconstruction algorithms: Kubicek [13], Sramek-Bernstein [14] and Wang [15]. Data collected after induction before skin incision (T 1 ), after weaning from cardiopulmonary bypass (T 2 ), after sternal closure (T 3 ), after admission at the intensive care (T 4 ), after reaching normothermia (36.5ºC) (T 5 ), after extubation (T 6 ) and before discharge to the ward (T 7 ) were: heart rate (HR), mean arterial pressure (MAP), central venous pressure (CVP), bioimpedance raw data and TPCO measurements. SVI was calculated by dividing SV by BSA. Sample size calculation was performed to limit the width of a 95% confidence interval for the mean error. Based on a mean CO of 5.0 l min -1, a correlation coefficient of 0.65, a mean error of 30% [16] and a confidence interval of 95%, a sample size of 20 patients was calculated. Statistical analysis was performed using PRISM 4.0 (GraphPad software Inc, San Diego, CA) and SPSS (SPSS Inc, Chicago, IL). If the ANOVA revealed a significant interaction, post-hoc analysis was performed using students t-test with Bonferroni correction. Validity and reproducibility between bioimpedance SVI and SVI TD were tested according to Bland-Altman [17]: bias, precision (= SD of bias), limits of agreement (LOA) and mean error [15] for absolute SVIvalues, and for relative changes in SVI ( SVI). Mean error was calculated as 2*precision divided by the mean SVI TD. Pooled data and data per time moment were analyzed. A p-value < 0.05 was considered significant. Results 15 male and 5 female patients (64 ± 10 years, 79 ± 12 kg, 171 ± 8 cm (mean ± SD) and BSA range m 2 ) were included. 93 out of 140 series of SVI obtained with each technique were available for statistical analysis. 47 series of SVI data could not be used for further analysis because of failure to obtain SVI due to insufficient raw bioimpedance signals. Time course of SVI and Bland-Altman analysis for every method are shown in figure 6.1. Bias, precision, LOA and mean error between SVI TD and SVI K were 1.0 ± 10.8 ml m -2, ml m -2 and 63% respectively, while the results for SVI TD and SVI SB were 9.8 ± 11.4 ml m -2, ml m -2 and 67% and for 92

93 Bioimpedance cardiac output in adults SVI TD and SVI W ± 8.2 ml m -2, ml m -2 bioimpedance data for each algorithm at each time moment are given in table 6.1. and 48% respectively. Analysis of 6 Figure 6.1 Time course of stroke volume index (SVI), and Bland Altman analysis for pooled data for every method. Discussion In this study, three bioimpedance algorithms assessing bioimpedance SVI were compared to SVI TD during the perioperative period in CABG patients. However, significant deviations were found and accurate clinical decision making was not possible based on absolute values or changes in bioimpedance SVI. No single algorithm was superior to another. Interestingly, application of the Wang algorithm produced consistent underestimation, whereas the two other algorithms overestimated SVI. Our study differed from previous studies in several important aspects. Raw 93

94 Chapter 6 time Bias (ml m -2 ) Precision (ml m -2 ) SVIK SVISB SVIW LOA (ml m -2 ) Mean error (%) Bias (ml m -2 ) Precision (ml m -2 ) LOA (ml m -2 ) Mean error (%) Bias (ml m -2 ) Precision (ml m -2 ) LOA (ml m -2 ) Mean error (%) T , , , T , , , T , , , T , , , T , , , T , , , T , , , Table 6.1 Bias, precision, limits of agreement (LOA) and mean error of stroke volume index calculated with three bioimpedance algorithms versus transpulmonary thermodilution stroke volume index (SVITD) for each time moment. T1 = after induction before skin incision; T2 = after weaning from cardiopulmonary bypass; T3 = after sternal closure; T4 = after admission at the intensive care; T5 = after reaching normothermia (36.5ºC); T6 = after extubation; T7 = before discharge to the ward. 94

95 Bioimpedance cardiac output in adults voltage data were measured and used for off-line calculation of bioimpedance SVI on the basis of different bioimpedance algorithms, commonly used in commercially available devices. Therefore, data were obtained without using different bioimpedance devices and calculation was independent from built-in proprietary software algorithms. Measurements were performed not only in the OR, but also in the intensive care in ventilated as well as in spontaneously breathing patients. The difference of bioimpedance SVI compared to SVI TD for any of the three algorithms may be explained by the fact that the relation between the signal on the voltage sensing electrodes and the resulting SVI is based on assumptions in relation to multiple effects. Whereas SV is equal to the change of the left ventricular volume during the systole, the voltage signal measured with bioimpedance is a result of volume changes of different intrathoracic compartments during the cardiac cycle, such as the intracardiac cavities, the aorta, the superior and inferior vena cava and the pulmonary circulation on the injected current [18]. Vascular diseases (atherosclerosis) can affect the relative contribution of the aorta to the bioimpedance signal, because the volume change of the aorta during the cardiac cycle depends on aortic compliance. Moreover, a considerable anatomical variability exists between patients and within the cardiac cycle. The orientation of the central heart axis in relation to the thorax cavity varies considerably between patients, but also during the cardiac cycle. Both influence the main current density field, and hence the relative contribution of SV to the bioimpedance signal. It is questionable whether it is even possible to measure SVI reliably using thoracic bioimpedance with only one single voltage input stream, given the fact that all of three distinctly different algorithms failed to produce satisfactory agreement with SVI TD. Therefore, an increase in the number of data input streams (i.e. electrodes) may improve the validity and reliability of the technique. Consequently, suggestions have been made to optimize the measurement technique and the basic bioimpedance SV equation [19]. Recently, Spiess [8] and Sageman [9] studied a second generation thoracic bioimpedance cardiograph (BioZ System 1.52, Cardiodynamics International Corporation, San Diego, CA) in CABG patients and found a clinically acceptable correlation between pulmonary artery thermodilution and bioimpedance. However, mean error in the study of Spiess was 26% after induction of anesthesia and exceeded the clinically acceptable 30% during the other measurements [8]. In contrast, our study showed a mean error exceeding 30% with the exception of t 2 using the Wang algorithm. In conclusion, common cylinder and cone based models for bioimpedance SVI calculation are not reliable compared to SVI TD measurements in CABG patients. These models are oversimplifications of the complex electrical events occurring inside the thorax during the cardiac cycle. The problem 6 95

96 Chapter 6 of retrieving SV from voltage data may be considered as a special case of the general inverse conductivity theory [20]. There is need for a more robust mathematical approach (see appendix), including an increase in the number of voltage measurement input streams, an accurate description of the physics of current density distributions and taking into account the full spectrum of all relevant patient anatomical variabilities. 96

97 Bioimpedance cardiac output in adults References 1. Swan HJ, Ganz W, Forrester J, et al. Catheterization of the heart in man with use of a flowdirected balloon-tipped catheter. N Engl J Med 1970; 283: 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 randomized controlled trial. Lancet 2005; 366: 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: Peters SG, Afessa B, Decker PA, et al. Increased risk associated with pulmonary artery catheterization in the medical intensive care unit. J Crit Care 2003; 18: Jensen L, Yakimets J, Teo KK. A review of impedance cardiography. Heart Lung 1995; 24: Raaijmakers E, Faes TJ, Scholten RJ, et al. A meta-analysis of three decades of validating thoracic impedance cardiography. Crit Care Med 1999; 27: Thangathurai D, Charbonnet C, Roessler P, et al. Continuous intraoperative noninvasive cardiac output monitoring using a new thoracic bioimpedance device. J Cardiothorac Vasc Anesth 1997; 11: Spiess BD, Patel MA, Soltow LO, et al. Comparison of bioimpedance versus thermodilution cardiac output during cardiac surgery: evaluation of a second-generation bioimpedance device. J Cardiothorac Vasc Anesth 2001; 15: Sageman WS, Riffenburgh RH, Spiess BD. Equivalence of bioimpedance and thermodilution in measuring cardiac index after cardiac surgery. J Cardiothorac Vasc Anesth 2002; 16: Young JD, McQuillan P. Comparison of thoracic electrical bioimpedance and thermodilution for the measurement of cardiac index in patients with severe sepsis. Br J Anaesth 1993; 70: Doering L, Lum E, Dracup K, et al. Predictors of between-method differences in cardiac output measurement using thoracic electrical bioimpedance and thermodilution. Crit Care Med 1995; 23: Buhre W, Weyland A, Kazmaier S, et al. Comparison of cardiac output assessed by pulsecontour analysis and thermodilution in patients undergoing minimally invasive direct coronary artery bypass grafting. J Cardiothorac Vasc Anesth 1999; 13: Kubicek WG, Karnegis JN, Patterson RP, et al. Development and evaluation of an 6 97

98 Chapter 6 impedance cardiac output system. Aerosp Med 1966; 37: Bernstein DP. A new stroke volume equation for thoracic electrical bioimpedance: theory and rationale. Crit Care Med 1986; 14: Wang Y, Haynor DR, Kim Y. A finite-element study of the effects of electrode position on the measured impedance change in impedance cardiography. IEEE Trans Biomed Eng 2001; 48: 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: Bland JM, Altman DG. Statistical methods for assessing agreement between two methods of clinical measurement. Lancet 1986; 1: Kauppinen PK, Hyttinen JA, Malmivuo JA. Sensitivity distributions of impedance cardiography using band and spot electrodes analyzed by a three-dimensional computer model. Ann Biomed Eng 1998; 26: Raaijmakers E, Faes TJ, Goovaerts HG, et al. Thoracic geometry and its relation to electrical current distribution: consequences for electrode placement in electrical impedance cardiography. Med Biol Eng Comput 1998; 36: Konings MK, Bouma CJ, Mali WP, et al. 2D Intravascular EIT using a non-iterative, nonlinear reconstruction algorithm. Lecture Notes Comput Sci 1997; 1230:

99 Chapter 7 Monitoring of cardiac preload EEC de Waal MD 1, WF Buhre MD 1,2 1 Division of Perioperative and Emergency Medicine, University Medical Center, Utrecht, Netherlands; 2 Lehrstuhl für Anästhesiologie II, Universität Witten-Herdecke, Klinik für Anästhesiologie und operative Intensivmedizin, Klinikum Köln-Merheim, Köln, Germany In: J. Boldt (Editor), Hemodynamic monitoring, Unimed, Bremen, Germany.

100

101 Cardiac preload monitoring Introduction Cardiac output (CO) is coupled on adequate cardiac preload and optimization of preload is one of the major targets of hemodynamic management. The most common reason for arterial hypotension is intravascular hypovolemia, and therefore, adequate fluid management is of major importance in the care of hemodynamically unstable patients. Monitoring of cardiac preload should fulfill two criteria. First, hemodynamic monitoring should alert the physician to impeding cardiovascular instability; it is routinely used in this manner in the operating room during high risk surgery. Second, it should obtain specific information to the disease process, which may facilitate diagnosis and treatment and enables monitoring of the response to therapy. From a physiological point of view, the Frank-Starling [1] curve describes the curvilinear relationship between cardiac preload and left ventricular CO (figure 7.1). Under normal physiologic conditions, both ventricles operate on the ascending part of the Frank-Starling curve [2], resulting in a concomitant increase in stroke volume (SV) when preload increases. However, a further increase in preload does not necessarily result in an appropriate increase in SV if the left ventricle operates already in the steep part of the Frank-Starling curve. In particular, in patients with pre-existing heart failure, an increase in preload can result in a significant decrease in CO (Overfilling). 7 Figure 7.1 Frank-Starling curve. (Source (modified): Chest 2003; 124: ). However, measurement of cardiac preload in clinical practice is not easy to obtain, as no continuous, user-independent, on-line measurement technique of left ventricular end-diastolic volume (LVEDV) is available until now. Therefore, surrogate preload parameters for LVEDV are commonly used in clinical practice. These preload parameters can be subdivided into static and 101

102 Chapter 7 Static preload indicators Volumetric preload indicators Dynamic preload indicators Abbreviation CVP PAOP LVEDAI FTc DT GEDVI/ITBV RVEDVI ITBVI SVI SOI down SPV PPV SVV VPV RSVT Vpeak ABF Name of parameter central venous pressure pulmonary artery occlusion pressure left ventricular enddiastolic area flow time corrected for heart rate deceleration time of the early rapid filling phase global end-diastolic volume index right ventricular enddiastolic volume index intrathoracic blood volume index stroke volume index variation stroke output index variation Difference between SBP during short apnoea and its minimal value during one mechanical breath Systolic Pressure Variation Pulse Pressure Variation Stroke Volume Variation Ventilation-induced Plethysmographic Variation Respiratory Systolic Variation Test Variation in peak flow velocity at the aorta valve Variation in Aortic Blood Flow in the descending aorta Measurement Measured directly with PAC or CV-line Measured directly with PAC (Off-line) by planimetry by means of echocardiography machine, indexed to BSA Measured directly with ODM TTE/TEE Transpulmonary thermodilutionpicco (Off-line) by planimetry by means of echocardiography, indexed to BSA, or measured directly with PAC Estimated by PiCCO Measured with ED Measured with ED Measured (off-line) by analysis of arterial BP line Measured (off-line) by analysis of arterial BP line Pulse contour analysis Pulse contour analysis Measured (off-line) by analysis of plethysmography curve Measured and calculated off-line from the recording of the response of arterial BP to respiratory maneuver Measured with TTE/TEE and calculated off-line Measured with ED and calculated off-line by analysis of Aortic Blood Flow at the Aortic valve Formula for calculation Measured directly Measured directly or off-line Measured directly during transpulmonary thermodilution Measured directly during transpulmonary thermodilution SVI = SVI post FC - SVI pre FC FTc SVI FTc SVI SOI = ( ) postfc ( ) prefc Difference between minimal and maximal values of systolic BP during mechanical breath Difference between minimal and maximal values of pulse pressure during one mechanical breath related to the average between the values Difference between minimal and maximal values of stroke volume during one mechanical breath related to the average between the values Difference between minimal and maximal values of pulse amplitude of the plethysmography waveform during mechanical breath Difference between minimal and maximal values of ABF during mechanical breath 102

103 Cardiac preload monitoring Table 7.1 Clinically used cardiac preload parameters. dynamic preload parameters. Static parameters are either pressure based, (e.g. central venous pressure (CVP) and pulmonary artery occlusion pressure (PAOP)) or volume based (e.g. intrathoracic blood volume (ITBV), global end-diastolic volume (GEDV) and left ventricular enddiastolic area (LVEDA)). In contrast to static preload parameters, dynamic preload parameters are usually based on beat to beat analysis of arterial pressure and flow. Table 7.1 gives an overview of currently used cardiac preload parameters. In clinical practice, the main interest is if the patient is normo-, hypo- or even hypervolemic (figure 7.2) and if the individual patient responds with an increase in CO to fluid administration (Fluid responsiveness) [3]. Thus, parameters describing fluid responsiveness should be preferred. Factors affecting cardiac preload are: vasomotor tone, mechanical ventilation and the course of the underlying disease. Moreover, sympathetic tone and the activation of humoral or cellular mechanisms can rapidly result in a change in preload. The individual patient condition and the severity of underlying disease are also of major importance. In the presence of pre-existing cardiac disease, in particular heart failure, the effects of fluid management are often different from the expected normal course. SV Hypovolemia Normovolemia Hypervolemia 7 LVEDV Figure 7.2 Frank-Starling curve. During hypervolemic conditions, any further increase in LVEDV results in a decrease in stroke volume. SV = stroke volume; LVEDV = left ventricular end-diastolic volume. (Source: N Engl J Med 1977; 296: ). Therefore, the ideal monitor of cardiac preload should be a fast response system with on-line data acquisition. As invasive monitoring technologies can lead to specific complications, minor or noninvasive techniques should be preferred. For this purpose, dynamic preload indices like stroke 103

104 Chapter 7 volume variation (SVV) and pulse pressure variation (PPV) are increasingly being used [4]. Recent studies suggest that these indices are reliable predictors of fluid responsiveness in different patient populations. Directly measured parameter Normal Range Cardiac output (CO) l min -1 Mixed venous oxygen saturation (SvO 2 ) 75 % Pulmonary artery pressures (PAP) Pulmonary Artery Occlusion Pressure (PAOP) Systolic Diastolic Mean mmhg 4-12 mmhg 9-19 mmhg 4-12 mmhg Blood temperature C Calculated parameter Formula Normal Range (MAP-CVP)/CO Wood units (mmhg l -1 min -1 ) Systemic Vascular Resistance (SVR) [(MAP-CVP)/CO] 80 (MPAP-PAOP)/CO dynes sec cm Wood units (mmhg l -1 min -1 ) Pulmonary Vascular Resistance (PVR) [(MPAP-PAOP)/CO] dynes sec cm -5 Stroke Volume (SV) CO/HR ml Arterial Oxygen Content (CaO 2 ) PaO SaO 2 Hb ml dl -1 Mixed Venous Oxygen Content (CvO 2 ) PvO SvO 2 Hb ml dl -1 Arteriovenous Oxygen Difference (avdo 2 ) CaO 2 -CvO ml dl -1 Oxygen Delivery (DO 2 ) CO CaO ml min -1 Oxygen Consumption (VO 2 ) CO avdo ml min -1 Oxygen Extraction Rate (ER-O 2 ) VO 2 /DO % Table 7.2 Parameters measured and calculated with the PAC. Static preload parameters Intermittent measures of preload not influenced by cyclic changes during the respiratory cycle are described as static preload parameters. Since the introduction of the pulmonary artery catheter (PAC) [5], hemodynamic monitoring in anesthesia and intensive care medicine was based on the PAC for more than two decades. However, pulmonary artery catheterization is highly invasive, time consuming and associated with a considerable risk of morbidity and mortality [6, 7]. The PAC enables not only the measurement of thermodilution CO, but also the assessment of mixed venous 104

105 Cardiac preload monitoring oxygen saturation (SvO 2 ), blood temperature and CVP and PAOP (table 7.2) [8]. Assessment of cardiac preload was primarily based on the measurement of CVP and PAOP. PA PAP PV LAP LA Ao RA CVP PCWP LVEDP RV LV PVR Mitral valve failure CVP PAP PCWP LAP LVEDP LVEDV Tricuspid valve dysfunction Increased Alveoloar Pressure Pulmonary Fibrosis LV diastolic dysfunction Figure 7.3 Factors influencing the relationship between left ventricular end-diastolic volume (LVEDV) and pulmonary artery occlusion pressure (PAOP). CVP and PAOP. Although CVP and PAOP are recommended to be used to guide fluid therapy in critically ill patients [9, 10], they fail to correlate with either ventricular end-diastolic volume (EDV) or changes in SV after fluid substitution. Thereby, changes in filling pressures are not necessarily indicative for changes in cardiac preload [4, 11-14]. A few clinical studies confirmed that CVP does not predict fluid responsiveness adequately [4, 12, 13, 15]. However, many clinicians still use PAOP as an estimate for left ventricular (LV) preload (i.e. LVEDV), assuming that an increase in PAOP reflects an increase in LV filling. However, the relationship between PAOP as an indirect measure of LVEDP and LVEDV is mainly determined by the compliance of the LV, which is typically decreased in the majority of critically ill patients, because of ventricular hypertrophy, distension of the left ventricle, myocardial ischemia, sepsis, pericardial effusion/tamponade, and the need for catecholamine support. Moreover, a variety of other conditions can affect the relationship between LVEDV and PAOP (figure 7.3). PAOP can only be measured intramural and does not reflect transmural pressure, hereby disregarding the influence of juxtacardiac pressures. In mechanically ventilated patients, this invalidates the use of PAOP as a parameter of cardiac preload. In the presence of PEEP, end-expiratory pressure may have a significant influence on the measured PAOP, which may lead to an overestimation of LV filling 7 105

106 Chapter 7 pressure. Consequently, the majority of clinical studies showed that PAOP was not able to adequately reflect cardiac preload and fails to predict the individual patient s response to a fluid challenge [3, 16, 17]. Oesophageal Doppler preload parameters. The Oesophageal Doppler Monitor is a non-invasive bedside monitor using an oesophageal Doppler probe, measuring aortic blood flow (ABF) and flow time corrected for heart rate (FTc). The ability of FTc as a preload parameter has been studied in comparison to other static preload indices, such as PAOP. In two recent studies in patients with acute circulatory failure, FTc failed to predict the fluid response [18, 19] probably because FTc is inversely related to afterload, which is influenced by the use of vasoactive drugs, e.g. norepinephrin. Recently, it was demonstrated that FTc predicted fluid responsiveness adequately, but should be used only in conjunction with other clinical information [20]. However, the technique is minor invasive and enables on-line assessment of SV. As SV is the primary target parameter of fluid-optimization, a number of clinicians use oesophageal Doppler monitoring as a fast and easy to use method for the assessment of CO in patients with intermediate risk, usually before initiating invasive techniques. Echocardiography. Echocardiography is used either via the transthoracic or the transoesophageal approach. Both transthoracic (TTE) and transesophageal echocardiography (TEE) enables the measurement of a number of static and dynamic preload parameters (table 7.1). LVEDA can be obtained by tracing the endocardial border including the papillary muscles in the short axis crosssectional view of the left ventricle (figure 7.4). In several studies, the value of LVEDA to predict fluid responsiveness has been studied in different subsets of patients. However, these studies are inconsistent and report either a poor predictive value [21, 22], a moderate predictive value [23-25] or even accurate predictive value [26]. From a theoretical point of view, the measurement of LVEDA should reflect preload dependency with high accuracy compared to other techniques. Conflicting results might be explained by the method used to measure LVEDA, which do not adequately reflect the geometry of the left ventricle. In particular, in patients with disturbed geometry of the left ventricular cavity, the measurement of LVEDA is difficult. However, in patients scheduled for routine TEE, e.g. cardiac surgical patients, LVEDA is a commonly used technique for the assessment of cardiac preload. In particular, echocardiography enables direct visualization of the therapeutic efforts in these patients. Besides cardiac preload, the morphology and pathology of the entire heart can be studied by means of echocardiography; therefore cardiac 106

107 Cardiac preload monitoring echocardiography is the method of choice for monitoring of patients with acute hemodynamic decompensation of unknown origin in the OR and the ICU. Figure 7.4 Endocardial border detection of LVEDA. LVEDA in this patient was 25 cm 2. Volumetric preload parameters In the past years, a number of techniques have been developed which enables monitoring of volumetric parameters. These parameters include: right ventricular end-diastolic volume (RVEDV), global end-diastolic volume (GEDV) and intrathoracic blood volume (ITBV). RVEDV can be calculated from simultaneous measurements of right ventricular ejection fraction and CO via a fastresponse thermistor PAC as follows: ( CO/ HR) RVEDV = RVEF 7 with CO = cardiac output, HR = Heart Rate and RVEF = Right Ventricular Ejection Fraction. Other techniques for the estimation of RVEDV are cardiac scintigraphy and echocardiography (4- chamber view TTE or TEE). However, several studies demonstrate that RVEDV is not as sensitive as GEDV or ITBV in predicting cardiac preload and volume responsiveness [28]. Transpulmonary thermodilution enables the assessment of intrathoracic blood volume (ITBV) and global end diastolic volume (GEDV). ITBV comprises the volume of all four cardiac chambers and the pulmonary circulation whereas GEDV only comprises the cardiac volumes in end-diastole. From a physiological point of view, ITBV is a surrogate parameter for the central blood volume 107

108 Chapter 7 which serves as the fluid reservoir for the left ventricle. So, at least in theory a severe hypovolemia should be reflected by a decrease in ITBV and GEDV. In clinical studies, it has been demonstrated that both ITBV and GEDV are more sensitive than CVP and PAOP and comparable to echocardiographically determined LVEDA [28-34]. Figure 7.5 Systolic pressure variation (SPV) after one mechanical breath followed by an end-expiratory pause. Reference line permits the measurement of Up and Down. SAP = Systemic Arterial Pressure; AP = Airway Pressure. (Source: Intensive Care Med 2003; 29: ). Dynamic preload parameters Systolic pressure variation, pulse pressure variation and stroke volume variation. Preload parameters influenced by cyclic changes in the respiratory cycle are the most popular dynamic preload parameters. These parameters, such as pulse pressure variation (PPV) and stroke volume variation (SVV) are derived from changes in the arterial pressure curve reflecting respiratory variations of SV. The changes in alveolar and pleural pressure during mechanical ventilation (i) result in cyclic alterations of ventricular pre- and afterload, (ii) affect ventricular interdependence, and (iii) are directly transmitted to the thoracic aorta [35]. For the majority of these mechanisms, it appears fundamental that changes in intrathoracic pressures during mechanical ventilation are transmitted to cardiovascular structures within the thorax. During mechanical ventilation, positive airway pressures in the thorax causes intermittent changes in biventricular preload. A decrease in preload and an increase in afterload of the right ventricle (RV) result in a decrease in RV stroke volume during inspiration, which in turn induces a subsequent decrease in LV preload, SV and systolic blood pressure, reaching its nadir during expiration ( Down) (figure 7.5). Left ventricular SV usually increases slightly during inspiration as a consequence of increased LV filling, better LV 108

109 Cardiac preload monitoring compliance due to a decrease in RV dimensions and, decreased LV afterload. This increase in LV stroke volume causes a concomitant increase in systolic blood pressure ( Up). Thus, respiratory variations in arterial pressure reflect, in large part, the corresponding variations in left ventricular SV. The pressure variation is exaggerated if the LV is working on the steep portion of the Frank-Starling curve. As hypovolemic ventricles (i.e. ventricles that will benefit from preload augmentation) operate on the ascending part of the Frank-Starling curve (figure 7.1), a given preload change (as imposed by ventilation) will induce a significant change in SV, SVV, systolic pressure variation (SPV) and ultimately in PPV. Accordingly, PPV and SVV have been demonstrated to adequately predict fluid responsiveness in different patient populations, including neurosurgical, cardiac surgical and critically ill patients [16, 17, 36-38]. One of the main limitations of PPV and SVV is that both parameters may only be used in mechanically ventilated patients with stable heart rhythm. Both SVV and PPV are at least in part dependent on tidal volume. Tidal volume > 8 ml/kg increased the validity of the measurements [39, 40]. SPV is defined as the difference between maximal and minimal values of systolic blood pressure during one ventilatory cycle: ( SP SPV (%) = ( SP max max SPmin ) 100(%) + SP ) / 2 min with SP max = maximal value of systolic pressure and SP min = minimal value of systolic pressure. Using the systolic pressure at end expiration as a reference point, SPV is further divided in two components: an increase ( Up) and a decrease ( Down) in systolic pressure versus baseline (figure 7.5). It was demonstrated recently that Down and SPV are sensitive parameters for cardiac preload and that both parameters can predict fluid responsiveness [38]. PPV describes the ventilation-induced variation of beat-to-beat pulse pressure (the difference between systolic and diastolic arterial pressure) from the mean value during the respiratory cycle and is calculated [10, 19, 21-23] as 7 ( PP PPV (%) = ( PP max max PPmin ) 100(%) + PP ) / 2 min with PP max = maximal value of pulse pressure and PP min = minimal value of pulse pressure. Accordingly, SVV represents the variation of pulse-contour derived beat-to-beat stroke volume from the mean value during the respiratory cycle and is calculated as [16-18, 21-23]: 109

110 Chapter 7 SVV (%) ( SV = ( SV max min SV mean ) ) 100(%) with SV max = maximal value of stroke volume and SV min = minimal value of stroke volume. A number of commercially available devices (PiCCO Pulsion, Munich, Germany, FloTrac /Vigileo Edwards Lifesciences, Irvine, CA, USA) enable monitoring of SVV [41]. The algorithm used in the PiCCO - system is based on a continuously sliding time window of 30 secs to calculate the SV mean. The time window is divided into four 7.5-sec periods; within each period the highest (SV max )) and the lowest (SV min ) SV were determined and the average of the four 7.5 sec intervals were used to calculate SVV [42]. Recent studies suggest that these indices are the most reliable predictors of fluid responsiveness in different patient populations [1, 15, 35, 36, 42-47] although one study is published yet concerning the value of SVV obtained with FloTrac /Vigileo in predicting fluid responsiveness in cardiac surgery patients [48]. Fluid responsiveness in patients with reduced left ventricular function A number of patients suffer from reduced left ventricular function. Therefore it is of clinical importance that measurement of preload with different techniques is valid in this specific patient group. Reuter [49] investigated the ability of static (CVP, PAOP and left ventricular end-diastolic area index) and dynamic preload parameters (SVV) of predicting fluid responsiveness in patients with a LVEF < 35%. The authors concluded that SVV predicts fluid responsiveness accurately. However, patients with a LVEF < 35% are known to be particularly afterload sensitive, and cyclic changes in ventricular afterload induced by mechanical ventilation are an important determinant of variations in SV and pulse pressure. Hence, there is a theoretical concern that an increase in SVV and PPV may erroneously indicate fluid responsiveness in patients with depressed left ventricular function. Fluid responsiveness in spontaneously breathing patients Many studies concerning cardiac preload and fluid responsiveness have been performed in patients receiving mechanical ventilation. These patients are generally sedated and sometimes paralysed to avoid any spontaneous respiratory movements. However, spontaneous respiration usually affects dynamic preload parameters when compared to mechanical ventilation. During spontaneous inspiration, the intrathoracic pressure is negative [50, 51]. Respiratory changes in alveolar and pleural pressures are lower during spontaneous breathing compared to mechanical ventilation. 110

111 Cardiac preload monitoring Active expiratory movements alter the cyclic changes in alveolar pressure. Active expiratory abdominal muscle contractions flushes blood from the abdominal compartment into the thorax, increasing the right ventricular preload and a subsequent increase in left ventricular preload occurs. Active expiration also induces a decrease in left ventricular afterload. The respiratory rate is variable in patients with spontaneous respiration and normally increased compared to ventilated patients. Therefore, the number of cardiac beats per respiratory cycle may be reduced. Spontaneous breathing, even in awake patients may be more sensitive to various stimuli (e.g. anxiety, pain) resulting in immediate changes in cardiac output. Heenen et al. showed that the respiratory variation in CVP, calculated as CVP end-expiration CVP end-inspiration failed to predict the response to volume expansion [51] in spontaneous breathing patients. In patients on pressure support ventilation, heartlung interactions are more complex, because the fluctuation in pleural pressures may be insufficient to change the cardiac preload even in fluid responsive patients. As already pointed out, the duration of the respiratory cycle and the tidal volumes are not fixed, which potentially hampers the use of any dynamic marker of fluid responsiveness. For this reasons, SVV does not predict fluid responsiveness in septic shock patients on pressure support ventilation [52]. Future preload parameters In routine surgery in patients at intermediate risk, the widespread use of invasive arterial catheters and transesophageal echocardiography cannot be advocated. Therefore, non-invasive predictors of fluid responsiveness, such as pulse oximetry waveform (PPVsat) [53-55] and the infrared photoplethysmography waveform (PPVfina, Finapres ) are evaluated [56]. The plethysmographic waveform is generated by blood volume changes in both arterial and venous vessels. The amplitude depends on intravascular pulse pressure, as well as on distensibility of the vascular wall. The Ventilation-induced Plethysmographic Variation (VPV) is defined as [53-55] (figure 7.6): ( POP VPV (%) = ( POP max max POPmin ) 100(%) + POP ) / 2 min 7 with POP max = maximal amplitude of the plethysmographic waveform and POP min = minimal amplitude of the plethysmographic waveform. Using a servo-controlled pressure cuff device (Finapress, Ohmeda monitoring systems, Englewood, CO, UA) applied to the middle finger, it is possible to measure non-invasively systolic and diastolic blood pressure and pulse pressure. VPV obtained during mechanical ventilation may be a simple and non-invasive technique in predicting fluid responsiveness [56-58]. 111

112 Chapter 7 Figure 7.6 Simultaneous recordings of ECG, invasive blood pressure, pulse oximetry plethysmography and respiration. Source Crit Care 2005; 9: R Echocardiographic monitoring is increasingly being used in the operating theatre, as well as in the intensive care unit, not only for anatomical reasons, but also for physiologic measurements. Peak flow velocity (Vpeak) measured at the base of the aortic valve with TTE or TEE is another newly developed dynamic preload parameter. Simultaneous recording of the airway pressure curve and beat-to-beat peak flow velocity over a single respiratory cycle [22] offers the calculation of Vpeak, the relative change in Vpeak caused by respiration. V peak (%) ( V = ( V peakmax peakmax V + V peakmin peakmin ) 100(%) ) / 2 with V peakmax = maximal peak velocity during the respiratory cycle, and V peakmin = minimal peak velocity during the respiratory cycle. 112