Department of Surgery and Bioengineering, Thoracic and Cardiovascular Unit, University of Siena, Siena, Italy
* Corresponding author: Department of Surgery and Bioengineering, Thoracic and Cardiovascular Unit, University of Siena, Viale Bracci 14, 53100 Siena, Italy. E-mail: scolletta{at}unisi.it
Accepted for publication April 4, 2005.
![]() |
Abstract |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Methods. We investigated PRAM by comparing its cardiac output (PRAM-CO) with paired measurements obtained by electromagnetic flowmetry (EM-CO) and by standard thermodilution (ThD-CO) during various haemodynamic states in a swine model. Nine pigs were monitored with a pulmonary artery catheter and a femoral artery catheter at baseline, in a hyperdynamic state produced by administration of dobutamine and in a hypodynamic state induced by progressive exsanguination. BlandAltman analysis was used.
Results. One hundred and eight paired cardiac output values over a range of EM-CO of 1.810.4 litre min1 resulted. We found close agreement between the techniques. Mean bias between EM-CO and PRAM-CO was 0.03 litre min1 (precision 0.58 litre min1). The 95% limits of agreement were 0.61 to +0.55 litre min1. Similar results between ThD-CO and PRAM-CO were found.
Conclusions. In a porcine model we have demonstrated accuracy of PRAM during various haemodynamic states. PRAM is a reliable tool to detect changes in cardiac output in pigs and has ability as a basic research tool.
Keywords: heart, cardiac output ; measurement techniques, cardiac output ; measurement techniques, pulse contour method ; measurement techniques, thermodilution ; model, pig
![]() |
Introduction |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
![]() |
Materials and methods |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Measurement of cardiac output by electromagnetic flowmetry
After median sternotomy, the probe of an electromagnetic flow meter (EFM) (model BL 613; Biotronex Laboratory, Chester, MD, USA) was placed around the ascending aorta for continuous measurement of CO (EM-CO). Calibration of the probe was checked in vitro at the end of each experiment. Values of EM-CO were recorded when the iced glucose solution was injected for ThD-CO measurement.
Measurement of cardiac output by thermodilution
Cardiac output was measured by ThD using the mean of triplicate injections of 10 ml of iced 5% glucose injected into the right atrium as fast as possible by the same investigator during apnoea at end-expiration. Any ThD-CO measurement significantly different (more than ±10%) from its pair was excluded, and a repeat measurement was obtained.
Measurement of cardiac output by PRAM
A standard arterial catheter was inserted into the femoral artery. A Baxter Truwave PX-600F transducer (Baxter-Edwards, Irvine, CA, USA) was connected to the monitoring system (Hewlett Packard, Andover, MA, USA) for continuous recording of the systemic arterial pressure waves and subsequent computation of CO. The pressure signals were acquired at 1000 Hz by means of an analoguedigital multifunction card (DAQ Card-700; National Instruments Corporation, Austin, TX, USA) working on the tension signals with 12 bits from 2.5 to 2.5 volts. All signals were recorded on a personal computer (TravelMate 507-DX; Acer, Taipei Hsien, Taiwan, ROC). The pressure signal was filtered at 25 Hz to avoid resonance effects caused by the cathetertransducer system without degrading the pressure wave amplitude. PRAM-CO values were displayed on a dedicated personal computer at each time point and recorded. For each determination of ThD-CO, a corresponding value for PRAM was made by averaging the individual beats over the approximate time needed for ThD-CO measurements. PRAM provided arterial pressure and beat-by-beat CO values continuously throughout the study.
Basic physical principles of PRAM
Changes in volume which occur in all arterial vessels are mostly due to radial expansion of the wall in response to blood pressure changes. This depends on various physical factors, such as the force of cardiac contraction, arterial impedance and compliance, and resistance of peripheral vessels. These variables are closely interdependent and need to be evaluated simultaneously. To this end, a variable called Z, representing the relationship between changes in pressure and changes in volume with time, is taken into account for the evaluation of SV in the various approaches to determine CO by pulse contour methods (PCMs). Pulse pressure is converted to SV by calculating the area under the pulsatile portion of the pressure wave, and Z (mm Hgxs cm3) is calculated as a factor retrospectively approximated from the results of in vitro experiments or by calibration with an independent measure of SV (i.e. ThD bolus).
At variance with other PCMs, PRAM is the practical application of a model developed completely a priori. The model did not require adjustments based on experimental data.8 The concept behind PRAM is based on the physics theory of perturbations,10 by which each physical system under the effects of a perturbing term tends to react to reacquire its own condition of stability (i.e. the situation of minimal energy required). With PRAM, the whole instead of the pulsatile systolic area under the pressure curve is measured in each cardiac cycle. At the same time, Z is obtained directly from the morphological analysis of both the pulsatile and the continuous components of the pressure waveform. The derivation of Z requires no predicted data apart from the expected mean arterial pressure.
Briefly, according to PRAM,8 Z is equal to (P/t)xK, and SV is calculated as follows (cm3):
![]() |
Study intervals
After baseline measurements had been obtained (T1, baseline 1), haemodynamic status was changed from baseline to a hyperdynamic state (2040% increase in CO monitored with EM-CO) by continuous infusion of dobutamine at the rate of 1015 µg kg1 min1 (T2, hyperdynamic), and then returned to baseline (T3, baseline 2). Next, the status was changed from baseline to a hypodynamic state by removal of 1525 ml kg1 of blood (T4, haemorrhage). After obtaining haemodynamic stability for at least 10 min, CO was measured. Three paired measurements of CO were made in each animal under each haemodynamic condition.
Statistical analysis
Paired values obtained by ThD and PRAM were divided into four groups, namely the periods T1, T2, T3 and T4. Data from all pigs were used to compute the mean and standard deviation (SD). BlandAltman analysis was used.13 Bias (mean difference between measurements) and the 95% limits of agreement (within which 95% of the difference will lie) of PRAM-CO compared with EM-CO and ThD-CO were computed. Bias and 95% limits of agreement as a percentage of the mean value of EM-CO and ThD-CO measurements were also calculated (i.e. bias %=100xbias/EM-CO; 95% limits of agreement %=100xlower limit/EM-CO100xupper limit/EM-CO). Finally, comparison of CO values among phases was performed using repeated-measures analysis of variance.
![]() |
Results |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
|
|
|
![]() |
Discussion |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
The biases of PRAM-CO and ThD-CO compared with EM-CO were negative in all conditions except haemorrhage. In a similar study, Kurita and colleagues19 compared a continuous lithium dilution technique (LiDCO) with ThD and EFM in 10 pigs during various haemodynamic states. They found that the precision values of the new technique calculated for all haemodynamic states were less than those measured using the ThD method. Moreover, ThD and LiDCO CO values were higher than those obtained by EFM. Since EFM cannot measure CO including the coronary blood flow, which constitutes about 5% of the left ventricular output in the resting condition and increases to about 15% under hypoxia,20 they concluded that ThD-CO and LiDCO exceeded the values of EM-CO.19 Our findings are in line with those of Kurita and colleagues19 and with the results later observed by the same study group again comparing the lithium dilution technique with ThD and EFM.21 In the present study, bias values in the PRAMCO/EMCO comparison, calculated for all haemodynamic states, were less than those in the ThDCO/EMCO comparison, suggesting that the variation in prediction of CO by PRAM was lower than that by ThD, and that PRAM provided better repeatability than ThD.
PRAM is based on the principle that, in any given vessel, volume changes occur mainly because of radial expansion in response to variations in pressure.8 Similar approaches have been studied by several authors in the course of the last three decades,22 and have led to the development of important clinical applications (e.g. the PiCCO and LiDCO systems and the Modelflow method).2326 However, the PiCCO system does not start without a ThD-CO; moreover, even though both the LiDCO system and the Modelflow method starts pulse contour CO computations directly after connection to a radial artery pressure signal with a trending of CO, these two methods have a higher accuracy only after considering age, sex and external calibration data.2326 In contrast, PRAM can measure absolute values of SV, independently from calibration, by determining parameters able to characterize the elastic properties of the arteries from the objective analysis of the pressure wave profile.8 9
Generally, PCMs and PRAM may have advantages over PAC-derived ThD measurements.9 2226 PRAM seems to be easy to use. It provides a fast response time (beat-to-beat readout), and abrupt changes in CO resulting from blood loss, tamponade, off-pump cardiac surgery and changes in arterial resistance may be detected more quickly than with ThD. More importantly, PRAM does not require external calibration by ThD, and requires no other additional invasive procedure. Since it does not require injection of thermal solution a central line is not required, avoiding both time-consuming and potential complications due to the insertion of a central catheter.8 9 Moreover, the positioning of a pulmonary artery catheter for ThD measurements may be contraindicated in minor surgical procedures and in low-risk patients.2 3 Therefore, PRAM might be considered a practical alternative to the traditional ThD method when siting a PAC is deemed harmful or not essential for clinical management.
Despite some disadvantages of bolus ThD, this is the most commonly used clinical method of CO determination and continues to be considered as the standard for comparison with other methods. Nevertheless, the bolus ThD technique only measures CO over a variable time period of 1545 s, and it may not be so appropriate as a gold standard technique for evaluation of a new, continuous method of measuring systemic blood flow. Therefore, the thermal filament-wrapped PAC employing the ThD principle to measure CO continuously might be more appropriate for comparison of beat-by-beat techniques.2729 However, the CO displayed is updated every 30 s and reflects the average flow from the previous 34 min. Since this could generate errors in collecting CO data due to the lack of an exact event marker to synchronize continuous ThD-CO with PRAM-CO measurements, we rejected the idea of using the continuous ThD method and compared PRAM with the bolus ThD technique. With regard to ThD, the results of the present study agree with those of Romano and Pistolesi, who demonstrated an excellent correlation with ThD and PRAM in haemodynamically stable patients undergoing cardiac catheterization.8
In a recent study, our research group investigated PRAM in low-risk and haemodynamically stable patients undergoing coronary artery bypass surgery, and concluded that PRAM was accurate for real-time monitoring of CO during cardiac surgery and in the postoperative period.9 However, in that previous study we did not obtain good agreement between PRAM-CO and ThD-CO measurements after the end of extracorporeal circulation, and considered the hypothesis that our results may be accounted for by the loss of reliability of ThD after the end of ECC.4 5 9 Notoriously, variability in ThD method occurs because of differences in injection technique, fluctuations in blood temperature, electrical interference, and cyclical variations in CO with ventilation.6 7 30 Moreover, variations in pulmonary artery blood temperature transiently increase after extracorporeal circulation, and the increased thermal noise may cause significant errors in the results of ThD techniques.4 5
To date, only two studies provide an indication of the intrinsic accuracy of PRAM.8 9 However, they do not give a direct answer to the questions arising from the accuracy of the methods in measuring CO in various haemodynamic states. We performed this study on PRAM under various haemodynamic conditions in animals since it enabled us to evaluate the performance of the system over a wide range of COs that could not be tested ethically in humans.
Some interesting findings emerged from the present study. CO estimates obtained by PRAM at baseline state showed good agreement with those simultaneously obtained by EFM and conventional bolus ThD. Over the intervals studied and a wide range of blood flow values, no significant difference between PRAM and the established techniques emerged. This agreement persisted despite changes in the use of vasoactive drugs and blood volume losses (hyperdynamic and hypodynamic states, respectively). These findings confirm the reliability and accuracy of this new method in CO monitoring during various haemodynamic states. However, we selected specific clinical patterns to gain familiarity with this new method, avoiding any random factor that could have influenced animal management. Further studies will be required to assess PRAM in the setting of extreme haemodynamic conditions and of severe haemorrhage.
Some disadvantages of PRAM remain to be addressed. (i) Several factors could affect the accuracy of CO measurements based on the analysis of arterial waveforms,31 such as stenosis of the arterial tree, arterial pathology in the proximal segments, etc., which need to be further investigated. Moreover, damped waveforms and inadequate pulse detection (e.g. severe arrhythmias, catheter dislodgement) may influence the precision of the pressure wave analysis.31 (ii) The additional data regarding the loading conditions of the heart (e.g. wedge pressure) represent important information that allows accurate interpretation (and therapy) of low CO states.9 The absence of preload measures may be a disadvantage of many pulse contour methods. In our opinion, this is the penalty we have to accept for being less invasive.
In summary, under the conditions studied, we have demonstrated that PRAM gives results comparable to CO determined by electromagnetic flowmetry, and by conventional ThD during various haemodynamic states in a swine model. Moreover, this is the first study to investigate the accuracy of PRAM in a porcine model and to assess its value as a research tool.
![]() |
References |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
2 Patel C, Laboy V, Venus B, Mathru M, Wier D. Acute complications of pulmonary artery catheter insertion in critically ill patients. Crit Care Med 1986; 14: 1957[ISI][Medline]
3 Connors AF Jr, Speroff T, Dawson NV, et al. The effectiveness of right heart catheterisation in the initial care of critically ill patients. SUPPORT investigators. JAMA 1996; 276: 88997[Abstract]
4 Bazaral MG, Petre J, Novoa R. Errors in thermodilution cardiac output measurements caused by rapid pulmonary artery temperature decreases after cardiopulmonary by-pass. Anesthesiology 1992; 77: 317[ISI][Medline]
5 Latson TW, Whitthen CW, O'Flaherty D. Ventilation, Thermal noise, and errors in cardiac output measurements after cardiopulmonary bypass. Anesthesiology 1993; 79: 123343[ISI][Medline]
6 Nishikawa T, Dohi S. Errors in the measurements of cardiac output by thermodilution. Can J Anaesth 1993; 40: 14253[Abstract]
7 Dhingra VK, Fenwick JC, Walley KR, Chittock DR, Ronco JJ. Lack of agreement between thermodilution and Fick cardiac output in critically ill patients. Chest 2002; 122: 9907
8 Romano SM, Pistolesi M. Assessment of cardiac output from systemic arterial pressure in humans. Crit Care Med 2002; 30: 183441[CrossRef][ISI][Medline]
9 Giomarelli P, Biagioli B, Scolletta S. Cardiac output monitoring by pressure recording analytical method in cardiac surgery. Eur J Cardiothorac Surg 2004; 26: 51520
10 Messiah A. Change of representation and perturbation treatment of a part of the Hamiltonian. In: Quantum Mechanics, Vol. 2. Amsterdam: North-Holland Physics Publishing, 1985; 72260
11 Burton AC. Physical principles of circulatory phenomena: the physical equilibria of the heart and blood vessels. In: Hamilton WF, eds. Handbook of Physiology, Vol. 1. Washington: American Physiological Society, 1961; 85106
12 Guyton AC. Textbook of Physiology. Philadelphia: Saunders, 1991; 151
13 Bland JM, Altman DG. Statistical methods for assessing agreement between two methods of clinical measurement. Lancet 1986; 1: 30710[CrossRef][ISI][Medline]
14 Wallace AW, Salahieh A, Lawrence A, Spector K, Owens C, Alonso D. Endotracheal cardiac output monitor. Anesthesiology 2000; 92: 17889[CrossRef][ISI][Medline]
15 Wong DH, Watson T, Gordon I, et al. Comparison of changes in transit ultrasound, esophageal doppler, and thermodilution cardiac output after changes in preload, afterload, and contractility in pigs. Anesth Analg 1991; 72: 5848[Abstract]
16 Aadahl P, Aakhus S, Saether OD, Stomholm T, Myhre HO. Cardiac output measurement during cross-clamping of the descending thoracic aorta in pigs. A comparison between transit-time ultrasound, thermodilution and pulsed Doppler ultrasound. Acta Anaesthesiol Scand 2000; 44: 1805[CrossRef][ISI][Medline]
17 Gregoretti S, Henderson CT, Bradley EL Jr. Simultaneous cardiac output measurements by transtracheal Doppler, electromagnetic flow meter, and thermodilution during various hemodynamic states in pigs. Anesth Analg 1991; 73: 4559[Abstract]
18 Broomhead CJ, Wright SJ, Kiff KM, Withington PS. Validation of thoracic electrical bioimpedance as a porcine research. Br J Anaesth 1997; 78: 3235
19 Kurita T, Morita K, Kato S, Kikura M, Horie M, Ikeda K. Comparison of the accuracy of the lithium dilution technique with the thermodilution technique for measurement of cardiac output. Br J Anaesth 1997; 79: 7705
20 Kuipers JR, Sidi D, Heymann MA, Rudolph AM. Comparison of methods of measuring cardiac output in newborn lambs. Pediatr Res 1982; 16: 5948[Abstract]
21 Kurita T, Morita K, Kato S, et al. Lithium dilution cardiac output measurements using a peripheral injection site comparison with central injection technique and thermodilution. J Clin Monit Comput 1999; 15: 27985[CrossRef][ISI][Medline]
22 Linton NWF, Linton RAF. Estimation of changes in cardiac output from the arterial blood pressure waveform in the upper limb. Br J Anaesth 2001; 86: 48696
23 Godje O, Hoke K, Goetz AE, et al. Reliability of a new algorithm for continuous cardiac output determination by pulse-contour analysis during hemodynamic instability. Crit Care Med 2002; 30: 528[CrossRef][ISI][Medline]
24 Jonas MM, Tanser SJ. Lithium dilution measurement of cardiac output and arterial pulse waveform analysis: an indicator dilution calibrated beat-by-beat system for continuous estimation of cardiac output. Curr Opin Crit Care 2002; 8: 25761[CrossRef][Medline]
25 Wesseling KH, Jansen JRC, Settels JJ, Schreuder JJ. Computation of aortic flow pressure in humans using a nonlinear, three-element model. J Appl Physiol 1993; 74: 256673[Abstract]
26 Hamilton TT, Huber LM, Jessen ME. PulseCO: a less-invasive method to monitoring cardiac output from arterial pressure after cardiac surgery. Ann Thorac Surg 2002; 74: 140812[CrossRef]
27 Mihaljevic T, von Segesser LK, Tonz M, et al. Continuous versus bolus thermodilution cardiac output measurements: a comparative study. Crit Care Med 1995; 23: 9449[CrossRef][ISI][Medline]
28 Bottiger BW, Soder M, Rauch H, et al. Semi-continuous versus injectate cardiac output measurement in intensive care patients after cardiac surgery. Intensive Care Med 1996; 22: 3128[CrossRef][ISI][Medline]
29 Rauch H, Muller M, Fleischer F, Bauer H, Martin E, Bottiger BW. Pulse contour analysis versus thermodilution in cardiac surgery patients. Acta Anaesthsiol Scand 2002; 46: 4249[CrossRef][ISI][Medline]
30 van Grondelle A, Ditchey RV, Groves BM, Wagner WW Jr, Reeves JT. Thermodilution method overestimate low cardiac output in humans. Am J Physiol 1983; 245: 6902
31 van Lieshout JJ, Wesseling KH. Continuous cardiac output by pulse contour analysis? Br J Anaesth 2001; 86: 4679