Cardiac output is one of the most significant variables measured in critically ill patients with hemodynamic instability. The bolus thermodilution technique is widely accepted as the “gold standard” for measuring cardiac output.1 – 4 However, the bolus thermodilution technique provides only intermittent information on cardiac output, and the procedure can be time consuming, especially when repeated measurements are required. For clinicians to manage critically ill patients whose hemodynamic status changes rapidly, a simple and reliable continuous cardiac output measurement technique would be useful.
Various technologies have been tested for continuous or semicontinuous measurement of cardiac output.5 – 11 For the semicontinuous measurement of cardiac output, automated intermittent measurements of cardiac output can be achieved by the use of thermal pulses through a heating filament attached to a pulmonary artery catheter.6,12 For the continuous measurement of cardiac output and blood flow velocity in various arteries, an intravascular ultrasound Doppler technique has been explored. However, the accuracy of the conventional Doppler-based technique depends on the angle between the ultrasound beam and the blood flow, i.e., Doppler angle.8 Errors in angle estimation lead to inaccurate measurements of blood flow velocity and cardiac output.
To overcome the angle dependency of the conventional Doppler-based blood flow measurement, we used 2 orthogonally positioned Doppler transducers to measure blood flow velocity in the main pulmonary artery. The 2 orthogonally positioned Doppler transducers allow trigonometric correction for differences in the angle of blood flow between each transducer, thereby making it unnecessary to define the angle between the ultrasound beam and blood flow for accurate Doppler-based blood flow velocity measurement.13 We adopted this method for continuous cardiac output measurement by developing a dual Doppler-equipped pulmonary artery catheter. In this study, we tested the accuracy of the Doppler-equipped pulmonary artery catheter by comparing its cardiac output measurements with those done by conventional techniques in animals.
Doppler-Equipped Pulmonary Artery Catheter
A Doppler-equipped pulmonary artery catheter was produced by mounting 2 ultrasound Doppler transducers on the distal tip of a conventional thermodilution pulmonary artery catheter (7.5 Fr) (Fig. 1A). Each ultrasound transducer crystal was approximately 1 × 1 × 1 mm3. Two transducers were positioned at an angle of 90° to transmit 2 ultrasound beams with a known angle relationship (Fig. 1B). The Doppler transducers were fixed against the pulmonary artery catheter with the angle of 45° (Fig. 1B). Each Doppler transducer measures blood flow velocity at 18 sampling points between 0 and 18 mm from the transducer at a pulse repetition frequency of 40 kHz. The velocity range that can be measured by this device is 0 to 76.5 cm/s, well above the flow velocity that is observed in the main pulmonary artery of humans.
The true main pulmona ry artery blood flow velocity was determined from 2 velocity components. From the 2 velocity components, the true velocity can be calculated independent of the angle of incidence.13 As is shown in Figure 1B, the true main pulmonary blood flow velocity (Vpulm) is detected as Vtransducer1 for transducer 1 and Vtransducer2 for transducer 2. Because 2 transducers were positioned at a fixed angle of 90°, using trigonometric functions, Vtransducer1 and Vtransducer2 can be presented as the following, using an angle of direction of pulmonary blood flow (α) against transducer 1:
Using a Pythagorean trigonometric identity,
. Therefore, the true pulmonary blood flow velocity (Vpulm) can be presented as
Thus, the measurement of the true pulmonary blood flow velocity is independent of the main pulmonary blood flow angle (α). The same concept of the angle-independent measurement of the blood flow that uses 2 Doppler transducers with a fixed angle can be extended to the system with 2 transducers with a fixed angle other than 90° (see the Appendix).
Each Doppler transducer transmits an ultrasound frequency of 20 MHz at a pulse repetition frequency of 40 kHz. Mean Doppler shift frequency was obtained by using a phase differential technique (modified autocorrelation technique). The phase differential technique allows real-time blood flow velocity measurements at 36 sampling points at the sampling rate at 40 kHz. This algorithm is a derivative from the autocorrelation technique that has been widely used to provide real-time 2-dimensional color display of blood flow information in the conventional 2-dimensional color Doppler echocardiogram. The phase differential technique does not use the frequency spectrum analysis of the flow velocity, such as a fast Fourier transform, thereby resulting in rapid calculations.
Our system records the instantaneous velocity profile over space. Instantaneous space–average velocity is obtained with the use of a multirange-gated pulsed Doppler velocipede. The pulsed Doppler velocipede maps a 1-dimensional velocity profile by establishing 18 individual range gates spanning the vessel diameter. The sample volume length for each range gate is approximately 1 mm. Instantaneous blood velocity at each of the 18 gates was calculated with the use of the phase differential technique. The instantaneous blood velocity at each of the 18 gates was calculated and stored at a sampling rate of 40 kHz. The instantaneous mean velocity over space was displayed with the velocity projections detected by transducers 1 and 2 (Fig. 2).
From a waveform of Vpulm (instantaneous main pulmonary blood flow) measured by the Doppler-equipped pulmonary artery catheter, the time velocity integral of Vpulm was continuously calculated. Cardiac output was calculated as the time velocity integral of Vpulm multiplied by the coefficient value. The coefficient value was obtained at the beginning of the experiment by dividing the cardiac output, which was derived from either the electromagnetic blood flow (EMF) probe or the thermodilution technique, by the time velocity integral of Vpulm.
After approval of the IRB, 24 mongrel dogs (19 to 30 kg) were used. Animals were anesthetized with pentobarbital (200 mg). After tracheal intubation, anesthesia was maintained with a continuous IV infusion of midazolam (1 mg/h), fentanyl (0.1 mg/h), and vecuronium (1 mg/h). Animals' lungs were mechanically ventilated with air and oxygen, with PaCO2 controlled at 35 to 45 mm Hg. Body temperature was maintained at 36°C to 37°C during the study. An arterial catheter was inserted into the right femoral artery to measure arterial blood pressure. The Doppler-equipped pulmonary artery catheter was inserted through the right femoral vein and advanced to the main pulmonary artery under the guidance of pressure and velocity waveforms. The pulmonary artery catheter was placed into the right ventricle by using pressure waveform for guidance. Then, the pulmonary artery catheter was advanced into the main pulmonary artery while visualizing the blood flow wave. After thoracotomy, an electromagnetic flowprobe (FF-130T, Nihon Kohden, Tokyo, Japan) was placed around the main pulmonary artery and connected to a flowmeter (EMF) (MFV-3200, Nihon Kohden).
We manipulated cardiac output by administering varying doses of dobutamine and propranolol in each animal to test the accuracy of the Doppler-equipped pulmonary artery catheter in a wide range of cardiac outputs (0.47 to 3.03 L/min). Cardiac output was simultaneously measured by the Doppler technique (CO-Doppler) and by EMF (CO-EMF) or the thermodilution technique (CO-Thermo).
As a primary analysis, we compared cardiac output measurements from 2 methods assuming a linear regression (CO-Doppler vs. CO-EMF, and CO-Doppler vs.CO-Thermo). To detect a 0.35 absolute error, 40 paired data points are needed for a linear correlation analysis. As a secondary analysis, the values of cardiac output measured by different techniques were compared using Bland and Altman‘s method.14 The bias was defined as the mean difference between CO-EMF and CO-Doppler values, and between CO-Thermo and CO-Doppler values. The precision was represented by the upper and lower limits of agreement. The limits of agreement were presented as the bias ± 2 standard deviations (SD), and defined the range in which 95% of the differences between the methods were expected to lie. The percentage error was calculated as the ratio of 2 SD of the bias to mean cardiac output. The percentage error was considered clinically acceptable if it was below 30%, as was previously proposed by Critchley and Critchley.15
Four of 24 dogs were excluded from the study because of the development of significant arrhythmia or death from excessive surgical bleeding. Surgical maneuvers to expose the main pulmonary artery caused uncontrollable arrhythmia in 2 animals before the insertion of the pulmonary artery catheter. Experiments were aborted in another 2 animals that suffered from continuous bleeding at the surgical site for the electromagnetic flowprobe implantation. None of the remaining 20 animals suffered from injury or perforation of the pulmonary artery or other structures by the pulmonary artery catheter.
From the remaining 20 dogs that underwent the complete study protocol, 120 data pairs were analyzed. Seventy-two data pairs were analyzed for the comparison between CO-Doppler and CO-EMF, and 48 data pairs were analyzed for the comparison between CO-Doppler and CO-Thermo. Hemodynamic variables during cardiac output measurements are listed in Table 1. Figure 2 shows a simultaneous recording of electrocardiogram, arterial blood pressure, pulmonary artery pressure, central venous pressure, the main pulmonary blood flow velocity projections detected by transducer 1 (Vtransducer1), the main pulmonary blood flow velocity projections detected by transducer 2 (Vtransducer2), and the true main pulmonary flow velocity (Vpulm) derived from Vtransducer1 and Vtransducer2.
CO-Doppler was highly correlated with CO-EMF (y = 1.16 × −0.26, r 2 = 0.99, n = 72, P < 0.001) (Fig. 3A). When comparing the means of CO-Doppler and CO-EMF with their differences,15 CO-Doppler closely agreed withCO-EMF (Fig. 3B). The bias betweenCO-EMF andCO-Doppler was −0.02 L/min, and 95% limits of agreement were −0.32 to 0.28 L/min (Fig. 3B). The percentage error betweenCO-EMF andCO-Doppler was 16%. There was also a close correlation betweenCO-Doppler andCO-Thermo (y = 1.24 × −0.90, r 2 = 0.85, n = 48, P < 0.001) (Fig. 4A). The bias betweenCO-Thermo andCO-Doppler was 0.18 L/min, and 95% limits of agreement were −0.62 to 0.98 L/min (Fig. 4B).
This study demonstrated that continuous cardiac output measurement using the new dual Doppler-equipped pulmonary artery catheter provided acceptable accuracy in a wide range of cardiac output in animals. There have been attempts to use intravascular ultrasound techniques for the continuous measurement of cardiac output. However, the lack of defined angle between the Doppler beam and blood flow makes cardiac output measurements using the conventional intravascular ultrasound technique inaccurate and unreliable.8,16,17 In 1989, Segal et al. pioneered a single Doppler transducer-equipped pulmonary artery catheter that was designed to measure real-time pulmonary artery flow velocity.8 Unfortunately, in their single Doppler transducer design, the flow velocity measurement was dependent on the angle between pulmonary artery blood flow and the catheter. In our dual Doppler-equipped pulmonary artery catheter, an innovative design with 2 Doppler transducers at a fixed angle, the measurement of main pulmonary blood flow velocity was independent of the angle between the main pulmonary blood flow and the pulmonary artery catheter (Fig. 1B). The angle independency of this technique represents a major improvement over previous techniques.
Using the dual Doppler-equipped pulmonary artery catheter, we were able to perform continuous and accurate measurements of the blood flow velocity of the main pulmonary artery and cardiac output. Cardiac output was calculated by multiplying the time integral of the main pulmonary artery blood flow velocity by a coefficient value. The coefficient value was obtained by dividing cardiac output derived from conventional techniques by the main pulmonary artery blood flow velocity obtained from the Doppler measurements. This Doppler-equipped pulmonary artery catheter has a conventional thermodilution capability, thereby enabling the calculation of the coefficient value. Currently, we are developing a next-generation Doppler-equipped pulmonary artery catheter that will continuously measure the main pulmonary artery blood flow velocity and the main pulmonary artery cross-sectional area on a beat-to-beat basis.
In our experiments, we used a constant coefficient value that was obtained at the beginning of the experiment in each animal. Using the constant coefficient value, we found a good correlation between cardiac output measured by a Doppler-equipped pulmonary artery catheter and the cardiac output measured by other conventional techniques during the course of pharmacological manipulations of cardiac output, which often took 4 to 5 hours. Frequent calibration against the thermodilution technique or other techniques to obtain the coefficient value was not necessary. These observations may suggest that the main pulmonary artery blood flow velocity itself, without converting it to cardiac output using the coefficient value, can be used as a surrogate to cardiac output. Clinicians may be able to use continuous measurement of the main pulmonary artery blood flow velocity obtained by the Doppler-equipped pulmonary artery catheter to guide hemodynamic management of critically ill patients.
Cardiac output is commonly measured in critically ill patients using the thermodilution technique. The thermodilution technique that uses a bolus injection of fluid is often referred to as a clinical standard, but it has well-known pitfalls related to operator variation, patient pathologies and the indicator used. Intracardiac shunts can result in falsely high or low cardiac output values when either a Doppler-equipped pulmonary artery catheter or the conventional thermodilution catheter is used.
Semicontinuous measurement of cardiac output can be achieved with the use of thermal pulses through a heating filament attached to a pulmonary artery catheter. Because of the time delay of cardiac output calculation, the semicontinuous measurement of cardiac output may miss rapid and transient hemodynamic changes. Other commercially available techniques for continuous cardiac output measurement include the algorithm-based analysis of a systemic arterial pressure waveform. However, the estimation of cardiac output from a systemic arterial pressure waveform relies on various assumptions, and therefore is subject to various systematic errors.18 Our device can provide a real-time measurement of cardiac output, enabling clinicians to detect acute changes in hemodynamic conditions of critically ill patients. Another potential advantage might be the display of the actual pulmonary blood flow waveform, which makes it possible to detect retrograde flow due to pulmonary regurgitation.
One of the major limitations of the Doppler-equipped pulmonary artery catheter is that its accuracy in cardiac output measurement depends on the correct placement of the Doppler transducers inside the main pulmonary artery. In this catheter, Doppler transducers are fixed at approximately 2.5 cm from the distal tip of the catheter. To keep the Doppler transducers inside the main pulmonary artery, it is critical to position the occlusion balloon at the proximal part of the right or left pulmonary artery. Such positioning can be easily obtained by intermittently inflating the occlusion balloon, a maneuver to distinguish the positioning of the pulmonary artery catheter in the main pulmonary artery and the positioning in the right or left pulmonary artery. In addition, migration of the tip of Doppler-equipped pulmonary artery catheter into the right or left pulmonary artery results in loss or severe dumping of the flow velocity waveform.
Direct facing of the Doppler transducers against the vascular wall precludes a reliable cardiac output measurement in our technique. Our preliminary experiments revealed that placing transducers at the lesser curvature of the pulmonary artery catheter significantly decreased the possibility of facing the transducer directly against the vascular wall. Therefore, we affixed the Doppler transducers to the lesser curvature of the pulmonary artery catheter. When the Doppler transducers directly faced the vascular wall, the display of this system showed a loss of laminar blood flow profile.
For this study, we modified a commercially available pulmonary artery catheter by adding 2 micro-Doppler transducers near the tip of the catheter. Because of the small size of Doppler transducers (1 × 1 × 1 mm3) used in this device, the flexibility of the pulmonary artery catheter appeared to be comparable to that of commercially available continuous cardiac output pulmonary artery catheters. None of the animals experienced perforation of the pulmonary artery or other structures. The safety of a continuous Doppler has been well established in Doppler-based fetal heart rate monitoring, which is routinely used to monitor fetal heart rates for hours and in some cases, for a few days. The acoustic power emitted by our device is less than one tenth of that of the fetal heart rate Doppler. Further studies are needed to test the safety of this device, including formal engineering studies to assess its flexibility.
Although direct comparison of the Doppler-equipped pulmonary artery catheter and the conventional thermodilution pulmonary artery catheter using 2 separate catheters may be ideal, the presence of 2 pulmonary catheters in the same animal presents technical difficulties. It also confounds measurements of cardiac output using the Doppler-equipped pulmonary artery catheter because of the ultrasound interference caused by the other catheter. Further studies are needed to indirectly compare these 2 techniques.
The newly developed Doppler-equipped pulmonary artery catheter with 2 orthogonally positioned Doppler transducers allows accurate and continuous measurement of cardiac output independent of the angle of incidence formed by the pulmonary artery catheter and the flow. The dual Doppler-equipped pulmonary artery catheter offers a promising clinical method for monitoring cardiac output continuously. Future studies in humans, especially in clinical settings, are warranted to test the accuracy and safety of this technique.
1. Forrester JS, Diamond G, Chatterjee K, Swan HJ. Medical therapy of acute myocardial infarction by application of hemodynamic subsets (second of two parts). N Engl J Med 1976;295:1404–13
2. Swan HJ, Ganz W, Forrester J, Marcus H, Diamond G, Chonette D. Catheterization of the heart in man with use of a flow-directed balloon-tipped catheter. N Engl J Med 1970;283:447–51
3. Ganz W, Donoso R, Marcus HS, Forrester JS, Swan HJ. A new technique for measurement of cardiac output by thermodilution in man. Am J Cardiol 1971;27:392–6
4. Ganz W, Swan HJ. Measurement of blood flow by thermodilution. Am J Cardiol 1972;29:241–6
5. Roewer N, Bednarz F, Schulte am Esch J. Continuous measurement of intracardiac and pulmonary blood flow velocities with transesophageal pulsed Doppler echocardiography: technique and initial clinical experience. J Cardiothorac Anesth 1987;1:418–28
6. Yelderman ML, Ramsay MA, Quinn MD, Paulsen AW, McKown RC, Gillman PH. Continuous thermodilution cardiac output measurement in intensive care unit patients. J Cardiothorac Vasc Anesth 1992;6:270–4
7. Yelderman M. Continuous measurement of cardiac output with the use of stochastic system identification techniques. J Clin Monit 1990;6:322–32
8. Segal J, Pearl RG, Ford AJ Jr, Stern RA, Gehlbach SM. Instantaneous and continuous cardiac output obtained with a Doppler pulmonary artery catheter. J Am Coll Cardiol 1989;13:1382–92
9. Darmon PL, Hillel Z, Mogtader A, Mindich B, Thys D. Cardiac output by transesophageal echocardiography using continuous- wave Doppler across the aortic valve. Anesthesiology 1994;80:796–805
10. Savino JS, Troianos CA, Aukburg S, Weiss R, Reichek N. Measurement of pulmonary blood flow with transesophageal two-dimensional and Doppler echocardiography. Anesthesiology 1991;75:445–51
11. Muhiudeen IA, Kuecherer HF, Lee E, Cahalan MK, Schiller NB. Intraoperative estimation of cardiac output by transesophageal pulsed Doppler echocardiography. Anesthesiology 1991;74:9–14
12. Rodig G, Keyl C, Liebold A, Hobbhahn J. Intra-operative evaluation of a continuous versus intermittent bolus thermodilution technique of cardiac output measurement in cardiac surgical patients. Eur J Anaesthesiol 1998;15:196–201
13. Akamatsu S, Kondo Y, Dohi S. Velocity measurements with a new ultrasonic Doppler method independent of angle of incidence. J Anesth 1996;10:133–9
14. Bland JM, Altman DG. Statistical methods for assessing agreement between two methods of clinical measurement. Int J Nurs Stud 2010;47:913–6
15. Critchley LA, Critchley JA. A meta-analysis of studies using bias and precision statistics to compare cardiac output measurement techniques. J Clin Monit Comput 1999;15:85–91
16. Doucette JW, Corl PD, Payne HM, Flynn AE, Goto M, Nassi M, Segal J. Validation of a Doppler guide wire for intravascular measurement of coronary artery flow velocity. Circulation 1992;85:1899–911
17. Siegel LC, Pearl RG. Noninvasive cardiac output measurement: troubled technologies and troubled studies. Anesth Analg 1992;74:790–2
18. Biais M, Nouette-Gaulain K, Cottenceau V, Vallet A, Cochard JF, Revel P, Sztark F. Cardiac output measurement in patients undergoing liver transplantation: pulmonary artery catheter versus uncalibrated arterial pressure waveform analysis. Anesth Analg 2008;106:1480–6
True velocity was determined from 2 velocity components or vector projections of the blood velocity vector. Two Doppler shift frequencies, Δf1 and Δf2, which are linearly proportional to the 2 measured velocity components, are expressed by the following Doppler equations:
The 2 ultrasound beams (beam 1 and beam 2) with 2 different angles of incidence (α and α + θ) generate Doppler shift Δf1 and Doppler shift Δf2, respectively. From these 2 equations, we can eliminate the unknown angle α, and the true velocity can be given by the following equation:
Note that c = the velocity of sound in tissue, fc = the frequency of the emitted ultrasonic signal, v = the velocity of the blood flow, α = the unknown angle of incidence of the emitted ultrasound beam 1, and θ = the known angle between the direction of the emitted ultrasound beam 1 and that of the ultrasound beam 2. The transducers were purposefully positioned at 90° with respect to each other so that equation  may be simplified and the true velocity v can be expressed by the following equation:
© 2011 International Anesthesia Research Society