Anesthesia & Analgesia:
Technology, Computing, and Simulation: Research Reports
Noninvasive Cardiac Output Measurement in Heart Failure Subjects on Circulatory Support
Phillips, Rob MPhil*; Lichtenthal, Peter MD†; Sloniger, Julie MS†; Burstow, Darryl MD*; West, Malcolm MD*; Copeland, Jack MD‡
From the *Department of Medicine, The University of Queensland, Brisbane, Australia; †Department of Anesthesiology, and ‡Department of Surgery, The University of Arizona College of Medicine, Tucson, Arizona.
Accepted for publication August 31, 2008.
Rob Phillips and Malcolm West occupy positions on the Uscom Medical Advisory Board and Julie Sloniger is employed by Uscom Ltd. in a technical capacity.
Rob Phillips: I am the founder of Uscom Ltd. and a postgraduate student at the University of Queensland. I took no part in the data acquisition and assure the editorial staff that my involvement and scientific contribution to this article have not been influenced by this association.
Malcolm West: I am a Professor of Cardiology at the University of Queensland and a member of the Medical Advisory Board of Uscom Ltd. I own no shares in the company and took no part in the data acquisition and assure the editorial staff that my involvement and scientific contribution to this paper have not been influenced by this association.
Julie Sloniger: I am a physiology major at the University of Arizona and an employee of Uscom Ltd. I acquired the Uscom data in a totally blinded fashion. I can assure the editorial staff that my involvement and scientific contribution relating to the analysis and interpretation of data have not been influenced by this association.
Address correspondence and reprint requests to Rob Phillips, MPhil, P.O. Box J241, Coffs Harbour, 2450, New South Wales, Australia. Address e-mail to firstname.lastname@example.org.
BACKGROUND: Pulmonary artery catheter (PAC) thermodilution is commonly used in the perioperative cardiac surgical intensive care unit for measurement and management of central hemodynamics despite questions about effectiveness, difficulty of use, and safety. USCOM is a noninvasive continuous wave Doppler device for direct measurement of cardiac output (CO) and is an alternative to PAC. USCOM validation has predominantly been in the cardiac surgical intensive care unit against PAC, despite the recognized limitations in reliability of the method. We compared USCOM CO measurements with the CardioWest, an orthotopic total artificial heart (TAH), in heart failure (HF) subjects during controlled interventions.
METHOD: CO, stroke volume (SV), and heart rate (HR) were measured in a blinded fashion using the CardioWest and the USCOM device in TAH HF patients. Five-hundred eight paired measures from 18 examinations of seven subjects were acquired as flow was varied by the CardioWest controller. Bland-Altman analysis was used to compare agreement.
RESULTS: Mean values and standard deviations (±sd) for CO, SV, and HR by CardioWest and USCOM were 7.33 ± 0.46 and 7.34 ± 0.51 L/min, 56.2 ± 3.8 and 56.6 ± 3.8 mL, and 131 ± 3 and 130 ± 4 bpm, respectively. CO ranged from 5.2 to 9.3 L/min. The mean differences between methods for CO, SV, and HR were −0.01 ± 0.23 L/min, −0.34 ± 1.97 mL, and 0.9 ± 2.3 bpm, respectively, with mean percentage differences of −0.3%, −0.6%, and 0.7%. The percentage limits of agreement for CO, SV, and HR were 6.4%, 7.1%, and 3.6%.
DISCUSSION: USCOM is a feasible and accurate method for noninvasive measurement and monitoring of CO in TAH HF patients and may have a wider application in diagnosis and management of cardiovascular disease.
Heart failure (HF) is an increasingly common and serious condition characterized by impaired circulation, which may, at end stage, require cardiac transplantation. In the acute presentation, during bridge to transplantation circulatory support and in the postcardiac surgical intensive care unit (ICU), cardiac output (CO) measurement using thermodilution via a pulmonary artery catheter (PAC) is commonly used to guide therapy. The administration of fluids, inotropes, and vasoactive therapies is commonly guided by PAC in this clinical setting despite unproven benefit, associated morbidity and mortality and difficulty of use.1–5 An accurate, noninvasive alternative method for assessment of CO in HF subjects may provide safe and effective patient management. Although numerous noninvasive devices have been introduced into clinical practice, such as bioimpedance, they have not been widely adopted because of unproven accuracy and reliability in the critical care environment. The Ultrasonic Cardiac Output Monitor (USCOM) (Uscom Ltd etc) is a two-dimensional (2D) independent continuous wave Doppler (CW) device designed for simple and accurate serial assessments of right and left-sided CO, and has been validated against flowprobes in animals6 and in the postcardiac surgical setting against PAC.7–10 The CardioWest (Syncardia, Tucson, AZ) is an Food and Drug Administration approved, biventricular, orthotopic total artificial heart (TAH) driven by an external pneumatic pulsatile pump, which generates a simulated normal flow to the TAH and the native circulation in intractable HF patients as a bridge to transplant.11 The CardioWest pneumatic pump displays beat-to-beat and averaged CO parameters measured by an internal flowmeter on a computer control panel with output variables that can be controlled to optimize hemodynamics. Although flowprobe comparisons can only be made in animals and PAC is a questionable reference method,12–15 the CardioWest is another method of validating the USCOM device in humans.
The aim of our study was to compare agreement of USCOM hemodynamic measures with those of the computer-controlled CardioWest in TAH HF subjects and assess the feasibility of USCOM use in this difficult study group.
After institutional review board approval, 508 paired measures were made using USCOM at 18 examinations performed on seven informed and consented, unintubated HF subjects treated as a bridge to transplant with CardioWest TAHs. The CardioWest control panel displays continuous updated instantaneous and averaged values for right and left-sided CO, stroke volume (SV), and heart rate (HR) and can be regulated to increase or decrease output. CardioWest values were recorded serially while contemporaneous USCOM Doppler flow profiles were acquired and stored to the device hard drive. The CO was varied by adjustment of the CardioWest controls to create a range of outputs. USCOM measurements were made of the right or left-sided output by interrogation of either the transpulmonary flow across the pulmonary valve via the parasternal acoustic access, or the transaortic flow across the aortic valve via the suprasternal or apical acoustic access. Left or right-sided examination was chosen after consideration of patient comfort and ease of acoustic access, with left-sided measures more frequently performed (77% of all measures) because of the frequent interference of postsurgical sternotomy and wound dressings. Subjects were positioned comfortably supine and a 2.2 MHz USCOM transducer positioned at the selected transcutaneous acoustic access and orientated for signal acquisition along the line of arterial flow. Doppler flow profiles were optimized for velocity and intensity, and the real-time USCOM signal acquired and stored to the hard drive for later measurement. The Doppler method depends on calculation of flow volume from the equation CO = vti × CSA × HR, where vti is the velocity time integral of the Doppler flow profile and represents the stroke distance or the distance a red blood cell travels per systolic stroke. The flow cross-sectional area (CSA) is determined from 2D ultrasonic measurement of the flow diameter using CSA = πr2. However, the USCOM device is 2D independent, with CSA calculated from a proprietary anthropometric algorithm for application in native hearts. However, as the TAH have fitted valves of known CSA, these values were directly entered into the USCOM software to calculate flow volumes.
The USCOM and CardioWest signals were acquired by separate operators blinded to the comparison values. CardioWest values were recorded by the first operator blinded to the USCOM measures, whereas the USCOM Doppler flow profiles were acquired directly to the USCOM hard drive by the second observer blinded to the CardioWest values. Measurement and recording of CO, SV, and HR values from the digitally stored Doppler flow profiles was made at a later time ensuring blinding to the CardioWest measures. The USCOM Doppler flow profiles were measured using proprietary digital auto signal tracing software, FlowTracer™, using the peak signal contour tracing method recommended by the American Society of Echocardiography Doppler conventions for CW signals.16 FlowTracer is a digital edge detect software that traces the Doppler flow profiles real-time generating beat-to-beat outputs of multiple hemodynamic variables. FlowTracer signal contouring removes a source of subjectivity and observer bias in the study and intra and interobserver variability in clinical use. Each stroke value was traced from the recorded Doppler signal and an averaged value of the clear consecutive beats displayed on the USCOM screen, usually 5–8 beats depending on the HR, was saved to the USCOM hard drive for comparison with the CardioWest average output values. The measured values by both methods were then retrieved and merged for analysis.
Mean values ± standard deviations (sd) and agreement using the Critchley modification of the Bland-Altman method17 for determining bias and precision were used to assess results using SPSS statistical analysis software (SPSS 11(Mac), Chicago, IL).
Signals were feasible in all subjects and at each examination resulting in 508 paired measures from 18 examinations in seven subjects for analysis. Mean CO by CardioWest was 7.33 ± 0.46 L/min (ranging from 6.4 to 8.8) and by USCOM 7.34 ± 0.51 L/min (range 5.18–9.33). Mean SV by CardioWest was 56.2 ± 3.8 mL (ranging from 48.9 to 64.7) and by USCOM 56.6 ± 3.8 mL (range 44.2–71.3 cm3). Mean HR by CardioWest was 131 ± 3 (ranging from 115 to 136) and by USCOM 130 ± 4 bpm (range 115–142). Mean values for all variables are summarized in Table 1. Mean differences among 18 paired CO, SV, and HR measures by CardioWest and USCOM were −0.01 ± 0.23 L/min (limits of agreement [LOA] −0.47 to 0.45), −0.34 ± 1.97 cm3 (LOA −4.28 to 3.6) and 1 ± 2 bpm (LOA −3.80 to 5.56) respectively (Table 1). Mean percentage differences among paired measures for CO, SV, and HR were −0.3%, −0.6%, and 0.7% (Fig. 1), whereas the % LOA for CO, SV, and HR was 6.4%, 7.1%, and 3.6%, respectively, suggesting no clinically significant difference between USCOM and CardioWest measures. Bland-Altman scatter plots of mean difference between measures of CO, SV, and HR are presented in Figures 2, 3, and 4.
All parameters measured by both methods showed agreement by bias and precision (LOA) analysis, and mean percentage differences between measures were not clinically significant.
These results suggest that the USCOM device accurately measures cardiac function compared with the CardioWest flowmeter over a range of controlled outputs (5.2–9.3 L/min) in TAH HF patients. These measurements were possible at each examination and in all patients despite postcardiac surgical patients often being difficult to insonate and to acquire reliable echocardiographic signals.
The USCOM device has previously been validated against PAC in ICU postcardiac surgical patients7–10; however, the accuracy of PAC is uncertain so the use of a controlled circulation like the CardioWest to further test the accuracy of USCOM in the ICU is useful. Critchley et al.17 suggested a 30% LOA for acceptable clinical agreement between CO measurement methods in comparisons with PAC. However, Critchley et al. acknowledged that PAC has intrinsic variability in the order of 20 or more % suggesting that PAC may be a flawed reference standard for evaluating new CO measurement methods.
The ESCAPE trial, a 26 center randomized, controlled trial found no indication for routine use of PAC to adjust therapy during hospitalization for long-term HF and recommended that future trials should test noninvasive assessments with specific treatment strategies to better tailor therapy for both survival time and survival quality.1 However, the clinical ineffectiveness of PAC may be related either to the inaccuracy of the measurements or the ineffectiveness of the therapeutic protocols based on this information. This study confirmed the accuracy of USCOM measurements of circulation in TAH HF subjects, and further research should now be directed to develop and test therapeutic protocols to determine whether outcomes benefit can be proven using USCOM-guided management of circulation.
Importantly, the comparison of USCOM against the CardioWest circulatory model allows some assessment of the reliability of PAC as a reference standard for CO studies. Comparisons of precision between PAC and USCOM CO measurements in postcardiac surgical patients have been performed in four studies with a reported %LOA of 40%, 29%, 35% and 23%, and 15%, respectively, for a mean of 28%.7–10 This study of USCOM against CardioWest, a more stable and accurate measure of CO, reported a %LOA of 6.4%. Previous studies in animals of USCOM against transit time flowprobes, a true flow “gold standard” with accuracy of ±2%, also demonstrated high levels of precision with %LOAs of 13% and 5.5%.6,18 That USCOM shows higher precision when compared with more accurate methods of flow measurement, such as CardioWest and flowprobes than PAC, infers that USCOM may be a more accurate measure of hemodynamics than PAC, particularly where the flow CSA is already known. A more accurate method for measurement of CO may allow the development of improved circulatory management protocols with outcomes benefit in HF patients.
An accurate, noninvasive CO measurement device would not only significantly reduce the ICU infection risk but could also be used to extend the reach of ICU management to presurgical assessment, post-ICU monitoring, and postdischarge review, as well as to the outpatient setting for assessment and management of chronic cardiovascular disease, thus limiting acute ICU admissions.
In this study, averaged USCOM values of CO, SV, and HR were compared with the averaged CardioWest values; however, USCOM provides real-time, beat-to-beat measures of hemodynamic variables allowing calculation of SV variability (SVV). This beat-to-beat variability of SV is a diagnostic and prognostic indicator in HF and a guide to intravascular fluid volume optimization, thus adding value to the hemodynamic examination.19–21 Many invasive monitoring methods average SV measures from an extended time domain (up to 2 min for PAC) to generate a rolling CO readout which masks SVV. Unlike PAC, the USCOM device makes no measure of filling pressures. Although clinicians have traditionally used filling pressures to guide fluid administration, the measurement of filling pressure has neither been uniformly accurate4 nor resulted in management which has improved outcomes.2,21,22 Hofer et al.22 found that SVV was almost twice as predictive of fluid responsiveness as static measures of central venous pressure and pulmonary capillary wedge pressure. Michard and Teboul23 in a retrospective meta-analysis of fluid responsiveness studies found right arterial pressure accurately predicted fluid responsiveness in only 40% of studies and right ventricular end-diastolic volume in 33% of studies. In the same review, respiratory changes in pulse pressure and aortic blood velocity had positive and negative predictive values 94% and 91%, and 96% and 100%, respectively. Whereas Monnet and Teboul,24 in their review of the literature assessing predictability of fluid responsiveness, conclude that “static measures of cardiac preload are not appropriate to assess preload reserve.” The concept of preload-recruitable SV reserve is a more physiologic approach to fluid management and requires the use of an accurate beat-to-beat monitor of SV combined with endogenous fluid challenges, such as passive leg lifts, hepatic compression, or observations of respiratory changes in SV. In this study, USCOM accurately measured SV in TAH HF subjects and therefore has the potential to predict fluid responsiveness; however, further study will be required to develop this application.
Although signal acquisition was feasible in all subjects at each examination in this study, it is possible that there may be difficulty in signal acquisition in some subjects immediately after thoracotomy when both hemothorax and pneumothorax may limit acoustic transmission. We have noted that postural adjustment, usually rolling the patient to the left lateral decubitus position allows blood to settle to the dependent thorax and air to aggregate in the raised hemithorax, creating an acoustic access through the mediastinum.
A limitation of this study is that stable and controlled circulation was tested, and so results may not be duplicated in the independent circulation, which may be characterized by greater variability. Additionally, the CO, SV, and HR values measured in this study were those used therapeutically for circulatory optimization in TAH HF and, although similar to those found in high normal circulation, were greater than those of untreated HF. Importantly this validation at high normal outputs has implications for management of conditions associated with high CO, such as sepsis. Further study in subjects with low outputs will determine the usefulness of the device in the broader setting of HF; however, finding a reliable beat-to-beat gold standard reference measure may be difficult as PAC has established limitations for measurement of low CO.14,15,25–27
Another limitation is that this study was of artificially generated circulation and so the Doppler waveforms produced by the CardioWest may not be duplicated in other artificial or native hearts. However, the CardioWest driver has been engineered to generate optimized circulation simulating high normal native CO, thus creating flow profiles that are ostensibly similar to native signals. Additionally, CW Doppler has been well validated in industrial applications, circulatory models, and in clinical subjects across a range of COs and rhythms over 35 yr of clinical application suggesting that, although the profiles may differ in different clinical instances, the principle of Doppler measurement of hemodynamics allows extrapolation to other circulations.
This study found excellent agreement between the hemodynamic measurements made by the USCOM and the CardioWest across a range of COs from 5.2 to 9.3 L/min in TAH-treated HF subjects. USCOM is a feasible and accurate method for noninvasive measurement and monitoring of CO in TAH HF patients and may have additional diagnostic and therapeutic applications in circulatory management.
Uscom Ltd. did provide the USCOM device for the purpose of the study.
1. Stevenson LW, the ESCAPE investigators, ESCAPE coordinators. Evaluation study of congestive heart failure and pulmonary artery catheterization effectiveness. JAMA 2005;294:1625–33
2. Shah MR, Hasselblad V, Stevenson LW, Binanay C, O’Connor CM, Sopko G, Califf RM. Impacts of the pulmonary artery catheter in critically ill patients: meta-analysis of randomized trials. JAMA 2005;294:1664–70
3. Hall JB. Searching for evidence to support pulmonary artery catheter use in critically ill patients. JAMA 2005;294:1693–4
4. Iberti TJ, Fischer EP, Leibowitz AB, Panacek EA, Silverstein JH, Albertson TE. Pulmonary Artery Catheter Study Group. A multicenter study of physicians’ knowledge of the pulmonary artery catheter. JAMA 1990;264:2928–32
5. Johnston IG, Jane R, Fraser JF, Kruger P, Hickling K. Survey of intensive care nurses’ knowledge relating to the pulmonary artery catheter. Anaesth Intensive Care 2004;32:564–8
6. Lester AC, Zhi YP, Benny SF, Jules F, Simon CW, Anna L, Robert AP. Testing the reliability of a new ultrasonic cardiac output monitor, the USCOM, using aortic flow probes in anaesthetized dogs. Anesth Analg 2005;100:748–53
7. Knobloch K, Lichtenberg A, Winterhalter M, Rossner D, Pichlmaier M, Phillips R. Non-invasive cardiac output determination by two-dimensional independent doppler during and after cardiac surgery. Ann Thorac Surg 2005;80:1479–84
8. Tan HL, Pinder M, Parsons R, Roberts B, van Heerden PV. Clinical evaluation of the USCOM ultrasonic cardiac output monitor in cardiac surgical patients in the intensive care unit. Br J Anaesth 2005;94:287–91
9. Chand R, Mehta Y, Trehan N. Cardiac output estimation with a new Doppler device after off-pump coronary artery bypass surgery. J Cardiothorac Vasc Anesth 2006;20:315–9
10. Arora D, Chand R, Mehta Y, Trehan N. Cardiac output estimation after off-pump coronary artery bypass: a comparison of two different techniques. Ann Card Anaesth 2007;10:132–6
11. Copeland JG, Smith RG, Arabia FA, Nolan PE, Sethi GK, Tsau PH, McClellan, Slepian MJ. Cardiac replacement with a total artificial heart as a bridge to transplantation. N Engl J Med 2004;351:859–67
12. Von Grondelle A, Ditchey RV, Groves BM, Wagner WW Jr, Reeves JT. Thermodilution method overestimates low cardiac output in humans. Am J Physiol 1983;245:H690–H692
13. Cigarroa RG, Lange RA, Williams RH, Bedotto JB, Hillis LD. Underestimation of cardiac output by thermodilution in patients with tricuspid regurgitation. Am J Med 1989;86:417–20
14. Renner LE, Morton MJ, Sakuma GY. Indicator amount, temperature, and intrinsic cardiac output affect thermodilution cardiac output accuracy and reproducibility. Crit Care Med 1993;21:586–97
15. Nishikawa T, Dohi S. Errors in the measurement of cardiac output by thermodilution. Can J Anaesth 1993;40:142–53
16. Otto CM, Stoddard M, Waggoner A, Zoghbi WA. Recommendations for quantification of Doppler echocardiography: a report from the Doppler Quantification Task Force of the Nomenclature and Standards Committee of the American Society of Echocardiography. J Am Soc Echocardiogr 2002;15:167–84
17. Critchley LAH, Critchley JAJH. A meta-analysis of studies using bias and precision statistics to compare cardiac output measurement techniques. J Clin Monit 1999;15:85–91
18. Phillips RA, Hood SG, Jacobson BM, Lichtenthal PR, Burstow DJ, West MJ, May CN. Measurement of CO by flow probe, USCOM and PAC in conscious sheep at rest and after dobutamine. 26th ISICEM 2006. Crit Care 2006;10(suppl 1):138
19. Gunn SR, Pinsky MR. Implications of arterial pressure variation in patients in the intensive care. Curr Opin Crit Care 2001;7:212–7
20. Reuter DA, Kirchner A, Felbinger TW, Weiss FC, Kilger E, Lamm P, Goetz AE. Usefulness of left ventricular stroke volume variation to assess fluid responsiveness in patients with reduced cardiac function. Crit Care Med 2003;31:1399–1403
21. Feissel M, Michard F, Mangin I, Ruyer O, Faller JP, Teboul JL. Respiratory changes in aortic blood velocity as an indicator of fluid responsiveness in ventilated patients with septic shock. Chest 2001;119:867–73
22. Hofer CK, Muller SM, Furrer L, Klaghofer R, Genoni M, Zollinger A. Stroke volume and pulse pressure variation for prediction of fluid responsiveness in patients undergoing off-pump coronary artery bypass grafting. Chest 2005;128:848–54
23. Michard F, Teboul JL. Predicting fluid responsiveness in ICU patients: a critical analysis of the evidence. Chest 2002;121:2000–2008
24. Monnet X, Teboul JL. Passive leg raising. Int Care Med 2008;34:659–63
25. Von Grondelle A, Ditchey RV, Groves BM, Wagner WW Jr, Reeves JT. Thermodilution method overestimates low cardiac output in humans. Am J Physiol 1983;245:H690–H692
26. Latson TW, Whitthen CW, O’Flaherty D. Ventilation, thermal noise, and errors in cardiac output measurment after cardiopulmonary bypass. Anesthesiology 1993;79:1233–43
27. Mackenzie JD, Haites NE, Rawles JM. Method of assessing the reproducibility of blood flow measurement: factors influencing the performance of thermodilution cardiac output computers. Br Heart J 1986;55:14–24
© 2009 International Anesthesia Research Society