Resuscitation 83 (2012) Contents lists available at ScienceDirect. Resuscitation

Transcription

1 Resuscitation 83 (2012) Contents lists available at ScienceDirect Resuscitation jo u rn al hom epage : Experimental paper Pulmonary arterial thermodilution, femoral arterial thermodilution and bioreactance cardiac output monitoring in a pediatric hemorrhagic hypovolemic shock model Yolanda Ballestero a,c, Javier Urbano a,c, Jesús López-Herce a,c,, Maria J. Solana a,c, Marta Botrán a,c, Diego Vinciguerra a,c, Jose M. Bellón b,d a Pediatric Intensive Care Department, Hospital General Universitario Gregorio Marañón, Universidad Complutense Madrid, Spain b Biomedical Investigation Foundation, Hospital General Universitario Gregorio Marañón, Universidad Complutense Madrid, Spain c Spanish Health Institute Carlos III, Maternal, Child Health and Development Network, Hospital General Universitario Gregorio Marañón, Spain d Preventive and Quality Control Service, Hospital General Universitario Gregorio Marañón, Spain a r t i c l e i n f o Article history: Received 7 April 2011 Received in revised form 14 June 2011 Accepted 28 June 2011 Keywords: Cardiac output Femoral arterial thermodilution Pulmonary arterial thermodilution Bioreactance Shock hemodynamic monitoring a b s t r a c t Aim: Bioreactance is a new non-invasive method for cardiac output measurement (NICOM). There are no studies that have analysed the utility of this technique in a pediatric animal model of hemorrhagic shock. Methods: A prospective study was performed using 9 immature Maryland pigs weighing 9 to 12 kg was performed. A Swan-Ganz catheter, a PiCCO catheter and 4 dual surface electrodes were placed at the four corners of the anterior thoracic body surface. Shock was induced by withdrawing a blood volume of 30 ml/kg, and then after, 30 ml/kg of Normal saline was administered. Seven simultaneous measurements of cardiac index (CI) were made by pulmonary artery thermodilution (PATD), Femoral artery thermodilution (), and NICOM before, during, and after hypovolaemia and during and after volume expansion. Results: The mean difference (bias) of differences (limits of agreement) between PATD and was 0.84 ( ) L/min/1.77 m 2, between PATD and NICOM was 1.95 ( ) L/min/1.77 m 2, and between and NICOM was 1.06 ( ) L/min/1.77 m 2. A moderate correlation was found between PATD and (r = 0.43; P = 0.01), but no correlation was found between bioreactance and either PATD or. Hypovolemia and volume expansion produced important significant differences in CI as measured by PATD and, while the changes with bioreactance were small and non significant. Conclusions: PATD and measurements showed similar responses to hypovolemic shock and volume expansion. Bioreactance persistently underestimates the CI and is not significantly altered by either inducing hemorrhagic shock, or later, through volume expansion. Bioreactance is not a suitable method for monitoring the CI in pediatric hemorrhagic shock Elsevier Ireland Ltd. All rights reserved. 1. Introduction Hemorrhagic shock is one of the leading causes of mortality in patients with trauma. Knowledge of the cardiac output (CO) can Abbreviations: CI, cardiac index; CO, cardiac output; CVP, central venous pressure; ECG, electrocardiography;, femoral artery thermodilution; HR, heart rate; MAP, mean arterial pressure; NICOM, cardiac output by bioreactance; PATD, pulmonary artery thermodilution; SVRI, systemic vascular resistance index; SVV, stroke volume variation. A Spanish translated version of the summary of this article appears as Appendix in the final online version at doi: /j.resuscitation Corresponding author at: Pediatric Intensive Care Department, Hospital General Universitario Gregorio Marañón, Dr Castelo 47, Madrid, Spain. Tel.: ; fax: address: pielvi@hotmail.com (J. López-Herce). be useful for managing patients in hemorrhagic shock. Although there are numerous methods for measuring the CO, the majority have many limitations for use in clinical practice in critically ill pediatric patients. The difficulty of vascular catheterisation, with a higher incidence of complications, and the lack of validation of the methods in pediatric patients are the principal difficulties when using invasive methods to measure CO in children. 1,2 The gold standard method for measuring the CO is the Fick method, but this is not used in clinical practice due to its technical difficulties. 1 The most widely used method in adult patients is pulmonary artery thermodilution (PATD) using a Swan-Ganz catheter. The major drawback with this method is the need to insert a catheter into the pulmonary artery. In children, insertion of a Swan-Ganz catheter is a highly complex procedure, and the risk of complications is higher than in adults; this technique is therefore used infrequently. 1, /$ see front matter 2011 Elsevier Ireland Ltd. All rights reserved. doi: /j.resuscitation

2 126 Y. Ballestero et al. / Resuscitation 83 (2012) Less invasive methods have been developed in recent years, such as femoral artery thermodilution () using the PiCCO method. 4,5 This method has been validated in adults, 6,7 but there is still limited experience in children or in pediatric animal models. 8,9 Other methods, such as pulse power analysis (LiDCO ) and pulse contour cardiac output (Flo Trac, Vigileo ), have been validated in adults but not in pediatric patients. 10 Thoracic bioreactance is a non-invasive method that analyses the variations in voltage in each beat in response to the application of a high-frequency, trans-thoracic current (NICOM TM, Cheetah Medical Inc., Wilmington, Delaware, USA). Preclinical and clinical studies have already validated the basic principles of this technique. 11 Some studies in critically ill adult patients have shown that bioreactance has a good correlation and concordance with other methods for measuring CO. 12,13 However, only one study has analysed the utility of this technique in children, specifically in those without hemodynamic instability. 14 No studies have evaluated this method in pediatric shock animal models of shock. The objective of this study was to analyse the correlation and concordance between transpulmonary thermodilution, femoral thermodilution, and bioreactance in a pediatric animal model of hypovolaemic shock. 2. Methods A prospective study was performed in 9 Maryland minipigs, isogenic for three loci of the major histocompatibility complex, weighing 10.5 ± 1.5 kg (range 9 12 kg). The study was approved by the Institutional Ethical Animal Investigation Committee Institutional Ethics and the research adhered to the guidelines for the care and use of laboratory animals. The animals were premedicated with ketamine 15 mg/kg and atropine 0.02 mg/kg. ECG monitoring and transcutaneous pulse oximetry were initiated and the animals were then anesthetised with propofol 3 mg/kg, fentanyl 5 g/kg, and atracurium 2.5 mg/kg. After endotracheal intubation they were connected to mechanical ventilation in volume control mode. Ventilation was adjusted to maintain the PaO 2 above 13.3 kpa and the PaCO 2 between 4.7 and 5.3 kpa. Positive end-expiratory pressure ventilation was not used. Anesthesia was maintained with a continuous infusion of propofol (2 mg/kg/h), fentanyl (10 g/kg/h), and atracurium (2.5 mg/kg/h). A 5.5F Swan-Ganz catheter (Edwards Lifesciences, Irvine, California, USA) was inserted via the femoral vein by cutdown to measure cardiac output and central venous pressure. A 3F to 5F Pulsiocath thermodilution catheter (Pulsion Medical System, Munich, Germany) was inserted into the femoral artery by cutdown for continuous cardiac output, blood pressure, and temperature measurement and for blood gas sampling. A 5.5F central venous catheter was inserted into the external jugular vein for volume withdrawal and replacement. The NICOM TM apparatus (Cheetah Medical, Wilmington, Delaware, USA) was used to monitor the CO by thoracic bioreactance. Four self-adhesive dressings, each with two electrodes, were used. Two of them (superior dressings) were applied to the right and left midclavicular lines. The two inferior dressings were applied cm lower in a right midclavicular line. For each measurement of the CI, 5 ml of 0.9% normal saline at a temperature below 8 C was administered via the atrial channel of the Swan-Ganz catheter, and the CI was measured simultaneously by the two methods. Two consecutive measurements were taken and the mean of the two was recorded. The following parameters were recorded at each measurement of the CI: heart rate (HR), mean arterial pressure (MAP), central venous pressure (CVP), cardiac index (CI) by the three methods, stroke volume variation (SVV), and systemic vascular resistance index (SVRI). Fig. 1. Bland and Altman plot of the cardiac index measured by pulmonary artery thermodilution (PATD) and by femoral artery thermodilution (). Bias between the two methods of 0.84 L/min/m 2 (limits of agreement, L/min/m 2 ). A baseline measurement of the hemodynamic parameters and CI was performed (measurement 1). Syringes of 50 ml were then used to withdraw a blood volume of 30 ml/kg over 30 minutes through the central venous line in order to induce acute hypovolemia. Further measurements of CI were performed 15 min after starting to induce hypovolemia (measurement 2), at the completion of inducing hypovolemia (measurement 3), and then 30 min after completing the induction of hypovolemia (measurement 4). Normal saline (30 ml/kg) was then administered over a period of 30 min using a standard IVAC 571 pump infusion system (model C50056, IVAC Corporation, San Diego, California, USA). Further measurements of cardiac output were made 15 min after starting the infusion (measurement 5), at the end of the infusion (measurement 6), and 30 min after the end of the infusion (measurement 7). On completion of the experiment, animals were sacrificed by the administration of sedative overdose and the intravenous injection of potassium chloride. The SPSS version 16.0 software was used for statistical analysis. An analysis was performed of the bias, correlation, and concordance between the measurements taken by pulmonary and femoral arterial thermodilution, and by bioreactance. The Bland and Altman method 15 was used to compare the results of the different measurement techniques, calculating the mean (bias) ± standard deviation (as a measure of precision) of the differences between the values obtained with each method. The differences between each pair of values were plotted over the average for each pair. Repeated measure ANOVA was used to analyse the changes in the parameters. Differences were considered significant at a P value less than Results On combined analysis of all the measurements of CI in the 9 animals, the mean CI was 4.3 ± 1.5 ml/min/1.73 m 2 measured by PATD, 3.5 ± 0.7 ml/min/1.73 m 2 by, and 2.6 ± 1.1 ml/min/1.73 m 2 by bioreactance; there were statistically significant differences between the three methods (P < 0.001). The CI measured by PATD was 24% higher than that measured by, and the mean of the differences (bias) between the two methods was 0.84 L/min/m 2 (limits of agreement, L/min/m 2 ) (Fig. 1). The mean CI measured by PATD was 40% higher than that measured by bioreactance, with a bias of 1.95 L/min/m 2 (limits of agreement, L/min/m 2 ) (Fig. 2). The mean CI measured by was 25% higher than that measured by bioreactance, with a bias of 1.06 L/min/m 2 (limits of agreement, ) (Fig. 3). The variance of measurements around the trend line (precision) was 19.4% ± 30.7% between PATD and, 45.1% ± 43.2% between PATD and bioreactance, and 30.3% ± 35.2% between and

3 Y. Ballestero et al. / Resuscitation 83 (2012) Infusion L/min/1.73m Hypovolemia PATD Fig. 2. Bland and Altman plot of the cardiac index measured by pulmonary artery thermodilution (PATD) and by bioreactance (BR). Bias of 1.95 L/min/m 2 (limits of agreement, L/min/m 2 ) Fig. 5. Changes in the systemic vascular resistance index (SVRI dyn cm 5 /m 2 ) measured by PATD and. Fig. 3. Bland and Altman plot of the cardiac index measured by femoral artery thermodilution () and by bioreactance (BR). Bias of 1.06 L/min/m 2 (limits of agreement, ). bioreactance. There was a moderate correlation between PATD and (r = 0.43; P = 0.01), but no correlation between PATD and bioreactance (r = 0.018; P = 0.9) or between and bioreactance (r = 0.169; P = 0.22). We also have calculated the agreement in each of the seven moments of measuring. The results are similar (data not showed). Fig. 4 shows the changes in CI measured at each of the study time points for the three methods. All three methods showed a fall in the CI during hypovolemia and a rise with volume expansion. The changes detected by PATD and were very large, with significant differences both for hypovolemia and for volume expansion (P = 0.001). However, the changes with bioreactance were small, and the differences were not significant for hypovolemia or volume expansion (P = 0.181). The CI measured by PATD was significantly higher than that measured by bioreactance at all study time points. The same situation was found on comparing PATD with, though the differences were statistically significant only with maximum hypovolemia (P = 0.038). The CI measured by was higher than that measured by bioreactance at all times during the study, with significant differences existing at baseline (P = 0.027) and after volume expansion (P = 0.017). The value for SVRI, on combined analysis of all measurements, was found to be lower using PATD (1567 ± 504 dyn cm 5 /m 2 ) than when using (1894 ± 520 dyn cm 5 /m 2 ) (P < 0.001). The bias between the two measurements was 318 dyn cm 5 /m 2 (limits of agreement, 1576 to 940 dyn cm 5 /m 2 ). However, there were no significant differences on comparing the two methods at each one of the study time points (Fig. 5). Although the SVRI increased with hypovolaemia and decreased with volume expansion (Fig. 5), the changes did not reach statistical significance with either of the two methods (PATD, P = 0.13; P = 0.31). Fig. 6 shows the changes in the SVV. The value for SVV measured by, using all measurements, was 18.2% ± 5.2%, compared to 10.1% ± 3.8% on using bioreactance (P < 0.001). The bias between the two measurements was 8.0% (limits of agreement, %). The differences between the two methods were statistically significant at each one of the study time points (Fig. 6). With both methods, the SVV values increased with hypovolemia and fell with volume expansion, with statistically significant differences between the baseline values and the values with hypovolemia and volume expansion (, P = 0.001; bioreactance, P = 0.006). L/min/1.73 m2 Hypovolemia Infusion 6,00 5,5 5,00 4,7 4,00 3,8 3,9 3,7 3,5 3,7 3,9 3,00 2,8 2,9 3,6 2,00 2,2 2,1 1,9 1,9 2,3 2,4 2,2 1,00 Basal 15 m 30 min 60 min 75 min 90 min 120 min PATD BR % 21,0 18,0 15,0 12,0 9,0 6,0 Hypovolemia Infusion 19,0 18,2 14,4 10,2 10,7 8,7 19,0 18,2 12,5 10,5 16,2 15,4 10,5 8,6 Basal 15 min 30 min 60 min 75 min 90 min 120 min BR Fig. 4. Changes in the CI measured by pulmonary artery thermodilution (PATD), femoral artery thermodilution () and bioreactance (BR). Fig. 6. Changes in the stroke volume variation (SVV %)) measured by femoral artery thermodilution () and Bioreactance (BR).

4 128 Y. Ballestero et al. / Resuscitation 83 (2012) Discussion In critically ill adult patients, the risk-benefit relationship of routine hemodynamic monitoring using PATD has been questioned because of the invasive nature of thermodilution-derived CO. Indeed, the use of PATD has declined in recent years. Many patients are therefore managed without PATD, though there are situations in which such information can be useful. This has led to great efforts to develop methods that measure CI less invasively. Methods based on arterial pulse contour analysis, 16 transoesophageal Doppler echocardiography, 17 impedance cardiography, and carbon dioxide breath analysis 18 are available. All of these techniques except arterial pulse contour analysis are completely or relatively non-invasive; however, only impedance cardiography is suitable for continuous monitoring. Recently, the feasibility and accuracy of a completely non-invasive technique of CI estimation based on bioreactance has been demonstrated in animals and adult patients in multiple hospital settings, including catheterization laboratories and medical and surgical cardiac intensive care units. 13 Our study is the first to have investigated the validity of bioreactance through its correlation and concordance with PATD and in a pediatric animal model of hemorrhagic hypovolemic shock. We considered 20% as the limit of agreement because that is the approximate variability of the reference method Critchley and Critchley suggested that a limit of agreement of ± 30% was acceptable for CI measurements. 22 However, that recommendation was based on limits of agreement around a central value derived from the average of the two methods tested, with neither of them being considered as the reference method. The true difference between the two methods tested could therefore be greater than 30%. A number of earlier studies in adults have found a good correlation and acceptable concordance between the CI measured by bioreactance and by other methods. 13 Marqué et al. compared the CI measured by bioreactance and by PATD in 29 adults in the postoperative period of cardiac surgery, finding a good correlation between the two methods (0.77), with a bias of 0.01 ± Those authors found that the relative error was less than 30% in 94% of the patients and less than 20% in 79%. 12 Squara et al, in a study performed on 119 adults, found a correlation of 0.82 between bioreactance and PATD, with a bias of 0.16 ± 0.52; the relative error was less than 20% in 85% of the patients. 11 Raval et al. found similar results on comparing bioreactance with PATD, with a correlation of 0.78 and a bias of 0.09 ± 1.19%. 13 However, our results in a pediatric animal model show that there is no correlation or concordance between the CI measured by bioreactance and that measured by PATD or, as the mean of the differences between bioreactance and TD was greater than 30%. The CI values measured by bioreactance were significantly lower than those measured by PATD or. In addition, in contrast to the findings with PATD and, the CI value measured by bioreactance did not show significant changes with hypovolaemic shock or volume expansion. A previous study validated the bioreactance in adults pigs. 23 We think that bioreactance does not work well in little pigs and children because the separation between the left and right thorax electrodes probably is not sufficient to measure an accurate CI, although we put the electrodes according to the recommendations of the manufacturer. Our first experience in little children confirms this hypothesis. 14 Our results show that the CI measured by has an acceptable correlation with PATD, and although have differences between the two methods they did not differ significantly. However the concordance between the two methods was not good. In a previous study performed by our group on an experimental model of pediatric hemorrhagic hypovolemic shock, the CI measured by PATD was 2.2 ± 0.9 L/min and by was 1.9 ± 0.8 L/min (difference not statistically significant), with a mean of the differences of 0.2 ± 0.3 L/min and a good correlation between the two methods (r = 0.88). 2 McLuckie et al. 9 studied 9 children undergoing cardiac surgery. In that study, the CO measured by was 0.2 L/min higher than by PATD, with a 95% confidence interval for relative bias of 0.04 to 0.34 L/min/m 2. Marx et al. 8 found a mean CO of 1.3 ± 0.5 with and 1.2 ± 0.5 with PATD. measurement of the CO was consistently slightly higher than PATD measurement. The mean difference was 0.12 L/min (9.6%), with a 95% confidence interval of 0.1 to 0.14 L/min. The SVRI measured by PATD and by showed considerable discordance, with a very large bias between the two methods. The SVRI increased with hypovolemia and decreased with volume expansion, although the differences were not significant with either of the two methods. In contrast, the SVV measured both by and by bioreactance increased significantly with hypovolemia and decreased with volume expansion; these may therefore be good parameters for the measurement of blood volume, as found in other studies We conclude that bioreactance is not a useful method for measurement of the CI in a pediatric model of hemorrhagic hypovolemic shock. This device should improve its algorithm for infants and children. is a less invasive method than PATD, with which it has an acceptable correlation although the concordance is not good, and it adequately detects changes in the CI caused by hypovolemia or volume expansion. Conflicts of interest statement All authors declare no conflict of interest. Acknowledgements To Mercedes Adrados and Natalia Sánchez of the Department of Experimental Medicine and Surgery of the Gregorio Marañón University General Hospital for their collaboration in performing the experiments. This study was supported by a research grant from the Spanish Health Institute Carlos III PI05/0807 and the grant N. RD08/0072: Maternal, Child Health and Development Network) within the framework of the VI National I+D+i Research Program ( ). References 1. Murdoch IA, Marsh MJ, Morrison G. In: Vincent J-L, editor. Measurement of cardiac output in children year book of intensive care and emergency medicine. Berlin: Springer; p Rúperez M, López-Herce J, García C, Sánchez C, García E, Vigil D. Comparison between cardiac output measured by the pulmonary arterial thermodilution technique and that measured by the femoral arterial thermodilution technique in a pediatric animal model. Pediatr Cardiol 2004;25: Wippermann CF, Huth RG, Schmidt FX, Thul J, Betancor M, Schranz D. Continuous measurement of cardiac output by the Fick principle in infants and children: comparison with the thermodilution method. Intensive Care Med 1996;22: Gratz I, Kraidin J, Jacobi AG, decastro NG, Spagna P, Larijani GE. Continuous noninvasive cardiac output as estimated from the pulse contour curve. J Clin Monit 1992;8: Combes A, Berneau JB, Luyt CE, Trouillet JL. Estimation of left ventricular systolic function by single transpulmonary thermodilution. Intensive Care Med 2004;30: Sakka SG, Reinhart K, Meier-Hellmann A. Comparison of pulmonary artery and arterial thermodilution cardiac output in critically ill patients. Intensive Care Med 1999;25: Zollner C, Haller M, Weis M, et al. Beat-to-beat measurement of cardiac output by intravascular pulse contour analysis: a prospective criterion standard study in patients after cardiac surgery. J Cardiothorac Vasc Anesth 2000;14: Marx G, Sumpelmann R, Schuerholz T, et al. Cardiac output measurement by arterial thermodilution in piglets. Anesth Analg 2000;90: McLuckie A, Murdoch IA, Marsh, Anderson D. A comparison of pulmonary and femoral artery thermodilution cardiac indices in paediatric intensive care patients. Acta Paediatr 1996;85: Mathews L, Singh KRK. Cardiac output monitoring. Ann Card Anaesth 2008;11:56 68.

5 Y. Ballestero et al. / Resuscitation 83 (2012) Squara P, Denjean D, Estagnasie P, Brusset A, Dib JC, Dubois C. Noninvasive cardiac output monitoring (NICOM): a clinical validation. Intensive Care Med 2007;33: Marqué S, Cariou A, Chiche JD, Squara P. Comparison between Flotrac-Vigileo and Bioreactance, a totally non-invasive method for cardiac output monitoring. Crit Care 2009;13:R Raval NY, Squara P, Cleman M, Yalamanchili K, Winklmaier M, Burkhoff D. Multicenter evaluation of non-invasive cardiac output measurement by bioreactance technique. J Clin Monit Comput 2008;22: Ballestero Y, López-Herce J, Urbano J, et al. Measurement of cardiac output in children by bioreactance. Pediatr Cardiol 2011;32: Bland JM, Altman DG. Statistical methods for assessing agreement between two methods of clinical measurement. Lancet 1986;1: Gratz I, Kraidin J, Jacobi AG, decastro NG, Spagna P, Larijani GE. Continuous noninvasive cardiac output as estimated from the pulse contour curve. J Clin Monit 1992;8: 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: Rocco M, Spadetta G, Morelli A, et al. A comparative evaluation of thermodilution and partial CO 2 rebreathing techniques for cardiac output assessment in critically ill patients during assisted ventilation. Intensive Care Med 2004;30: Hillis LD, Firth BG, Winniford MD. Analysis of factors affecting the variability of Fick versus indicator dilution measurements of cardiac output. Am J Cardiol 1985;56: Rubini A, Del Monte D, Catena V, et al. Cardiac output measurement by the thermodilution method: an in vitro test of accuracy of three commercially available automatic cardiac output computers. Intensive Care Med 1995;21: Le Tulzo Y, Belghith M, Seguin P, et al. Reproducibility of thermodilution cardiac output determination in critically ill patients: comparison between bolus and continuous method. J Clin Monit 1996;12: 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: Keren H, Burkhoff D, Squara P. Evaluation of a noninvasive continuous cardiac output monitoring system based on thoracic bioreactance. Am J Physiol Heart Circ Physiol 2007;293:H Berkenstadt H, Friedman Z, Preisman S, Keidan I, Livingstone D, Perel A. Pulse pressure and stroke volume variations during severe haemorrhage in ventilated dogs. Br J Anaesth 2005;94: Marik PE, Cavallazzi R, Vasu T, Hirani A. Dynamic changes in arterial waveform derived variables and fluid responsiveness in mechanically ventilated patients: a systematic review of the literature. Crit Care Med 2009;37: Su NY, Huang CJ, Tsai P, Hsu YW, Hung YC, Cheng CR. Cardiac output measurement during cardiac surgery: esophageal doppler versus pulmonary artery catheter. Acta Anaesthesiol Sin 2002;40: