It has been suggested that pulsatile blood flow is superior to continuous flow (CF) in cardiopulmonary bypass (CPB). However, adoption of pulsatile flow (PF) technology has been limited because of practicality and complexity of creating a consistent physiologic pulse. A pediatric pulsatile rotary ventricular pump (PRVP) was designed to address this problem. We evaluated the PRVP in an animal model and determined its ability to generate PF during CPB. The PRVP (modified peristaltic pump, with tapering of the outlet of the pump chamber) was tested in four piglets (10–12 kg). Cannulation was performed with right atrial and aortic cannulae, and pressure sensors were inserted into the femoral arteries. Pressure curves were obtained at different levels of flow and compared with both the animal’s baseline physiologic function and a CF roller pump. Pressure and flow waveforms demonstrated significant pulsatility in the PRVP setup compared with CF at all tested conditions. Measurement of hemodynamic energy data, including the percentage pulsatile energy and the surplus hydraulic energy, also revealed a significant increase in pulsatility with the PRVP (p < 0.001). The PRVP creates physiologically significant PF, similar to the pulsatility of a native heart, and has the potential to be easily implemented in pediatric CPB.
From the *Department of Surgery, University of Michigan, Ann Arbor, Michigan; †MC3 Inc., Ann Arbor, Michigan; and ‡Department of Cardiac Surgery, University of Michigan, Ann Arbor, Michigan.
Submitted for consideration July 2013; accepted for publication in revised form January 2014.
Disclosure: The authors have no conflicts of interest to report.
The described research was sponsored by the National Institutes of Health: NIH SBIR 1 R43 HL085986-01A1 and NIH SBIR 2 R44 HL085986-02 and was the sole funding source of purchased property and tested technology. The authors had full control of the design of the study, methods used, outcome parameters and results, analysis of data, and production of the written report.
Reprint Requests: Alvaro Rojas-Peña, MD, ECLS Laboratory B560 MSRB II, 1150 W. Medical Center Drive, Ann Arbor, MI 48109. Email: firstname.lastname@example.org.
Cardiopulmonary bypass (CPB) revolutionized cardiovascular surgery over 50 years ago when a heart–lung machine was used successfully in the clinical setting by John Gibbon in 1953 during the repair of atrial septum defects.1 Cardiopulmonary bypass is now frequently used for multiple surgical procedures. The utilization of CPB has evolved over the years, but several questions remain about the optimal characteristics of a CPB system. One controversy relates to the use of pulsatile flow (PF) versus continuous flow (CF). This issue has been studied in many experimental and clinical models. Despite these studies, there is no consensus regarding optimal flow mechanics in CPB.
It has been suggested that PF is superior to CF. Alkan et al.2 demonstrated that pediatric heart surgery patients who had PF as a component of their CPB required less inotropic support and had shorter intensive care unit and hospital length of stay compared with those in the non-PF group. In a prospective, randomized trial of adult cardiac surgery patients, the pulsatile group had lower rates of balloon pump use, myocardial infarction, and death.3 Pulsatile flow is associated with improved microcirculation, less inflammatory response, decrease in inotropic support, shorter hospital stay, and superior organ perfusion including renal preservation. Pulsatile flow has also been shown to increase end-organ and cerebral perfusion.4–10 Several mechanisms have been proposed including reduction of proinflammatory cytokine release with PF, likely because of decreased endothelial injury.11,12 Non-PF causes greater increase in systemic vascular resistance because of vasoconstriction.13 Despite these and other studies demonstrating the clear advantages to PF, this technology is used in only 7% of US centers.14 Possible reasons for this have been identified and refuted elsewhere.15,16 The use of PF has been additionally slowed by lack of a practical, easily adopted means of generating a high-quality physiologic pulse.17 Generation of PF depends on an energy gradient; precise quantification in terms of hemodynamic energy levels is therefore a necessity, not an option.
A novel pulsatile device has been previously described, referred to as the pediatric pulsatile rotary ventricular pump (PRVP). The PRVP is capable of creating normal physiologic pulse in a hydraulic benchtop model of a pediatric sized patient.18 The purpose of this study was to test the PRVP system in an infant/pediatric CPB model in piglets. It was hypothesized that this pump would create adequate PF with minimal priming volume.
Materials and Methods
The PRVP is a scaled embodiment of previously described M-pump technology (MC3 Inc., Ann Arbor, MI) designed for CF CPB support. It consists of a collapsible conduit called the pump chamber, wrapped under tension around freely rotating rollers mounted on a simple, three-roller pumphead. The design benefits include starling principle operation, pressure limited output, suction limited input, afterload insensitivity, an inability to pump air or drain a venous reservoir, and avoidance of retrograde flow. These features have been described in detail.19–21 In addition to reduced scale, the PRVP includes modifications to the pump chamber designed to create the desired PF, with a priming volume of 8 ml. The M-Pump was originally marketed and used successfully for adult CPB procedures where its intuitive operation and improved safety above that offered by centrifugal pumps won ready acceptance by perfusionists.
As shown in Figure 1A, the M-pump consists of a collapsible conduit with unique cross-sectional design called the pump chamber, wrapped under tension around three freely rotating rollers mounted on a rotor. In its “free” condition, the pump’s cross-sectional shape is collapsed and occluded (Figure 1A). The pump chamber will prime only when fluid is supplied at a pressure above the atmospheric pressure and will assume the cross-sectional shape shown in Figure 1B. Once the inlet region of the pump chamber is primed and the rotor rotates, fluid is driven by peristaltic motion toward the outlet region of the pump chamber.
The PRVP incorporates innovations resulting in improved PF generation without the use of active electronic controls, while maintaining constant revolutions per minute. Fluid remains isolated between the rollers as it advances toward the outlet. The outlet region is characterized by a tapering cross section that results in pressurization of the advancing fluid. As the lead roller disengages the tubing, an initial pressurized discharge into the extracorporeal circuit results, followed by a period of steady reduced flow as the roller passes over the smaller cross section portion of the outlet region. The resulting flow and pressure are pulsatile and periodic with each roller pass.
This study was approved by the University of Michigan University Committee on Use and Care of Animals. All piglets received humane care in compliance with the Guide for the Care and Use of Laboratory Animals. Four infant female piglets (10–12 kg) were anesthetized, ventilated, and instrumented. General anesthesia was administered and maintained with isoflurane (0.8–3.0%) using a Narkomed 6000 ventilator (Draeger Medical Inc., Telford, PA) while the animal was surgically instrumented. A cutdown was performed to the right femoral vein and artery for intravenous access and measurement of mean arterial pressure (MAP). Arterial blood gases were obtained every 30 min after intubation to verify appropriate metabolic status. The left femoral artery was dissected and a 5 Fr pressure probe (Millar Instruments, Houston, TX) was inserted into the descending aorta within approximately 1–2 cm of the arch to measure MAP. A median sternotomy was made to access the thoracic cavity. A 10 mm vascular flow probe (Transonic Systems, Ithaca, NY) was inserted around the aorta between the two cephalic branches to monitor aortic blood flow. A norepinephrine infusion (0.5–2.0 μg/min) was used as needed to maintain MAP >40 mm Hg during sternotomy and manipulation of the heart and great vessels; after surgical instrumentation, the infusion was discontinued. After heparinization (450 U/kg), the aorta was cannulated with a 12 Fr aortic cannula (Medtronic DLP Model 77012, Minneapolis, MN) inserted approximately 1–1.5 cm into the aorta and secured by dual purse string suture. The right atrium was cannulated with a 24 Fr right angle metal tip drainage cannula (Medtronic DLP model 69324). The cannulae were connected to the CPB circuit (described below). The average time on bypass was 2 hours. Activated coagulation times were >600 sec throughout the experiment.
Cardiopulmonary Bypass Circuit
The CPB circuit used was a basic clinical pediatric CPB circuit, except that no vents or cardioplegia were used. The PRVP pulsatile pump was tested and compared in parallel with a Stöckert 6″ nonpulsatile roller pump that was used to collect CF data. The CPB circuit consisted of 1/4″ ID × 1/16″ wall tubing for venous drainage and 1/4″ ID × 3/32″ wall tubing for arterial reinfusion. Venous drainage was via gravity to a Terumo Baby RX open venous reservoir and Terumo CAPIOX RX-05 (Terumo Medical Corp, Ann Arbor, MI). Blood exiting the reservoir was diverted in parallel to either the CF roller pump (1/4″ ID tubing) or the PRVP via a “Y” connection (Figure 2). The pump not in use was isolated by clamps from the circuit to minimize flow artifact. Arterial blood exiting the gas exchange device was infused into the ascending aorta in three of four animals. In one animal, the blood was pumped through a pediatric arterial filter (Terumo Capiox AF02, Terumo Corp) before reinfusion into the ascending aorta. An extracorporeal flow probe (Transonic Systems) was attached just proximal to the reinfusion cannula to measure systemic blood flow. Priming volume of the system was 500 ml.
Data Acquisition and Calculation
Pressure and flow rate data were sampled and recorded using customized LabVIEW software (National Instruments, Austin, TX) and a compact DAQ (NI cDAQ-9172), including general-purpose bridge amplifier (NI9237) and analogue input (NI9219) modules. All pressure sensors received before and after two point calibrations using standing water columns of known height. At each measurement point, pressure and flow waveforms were sampled at a rate of 100 Hz for 30 sec using our LabVIEW-based data acquisition system. For each measurement, the full 30 sec data sample was used to calculate MAP, energy equivalent pressure (EEP), surplus hydraulic energy (SHE), and percentage pulsatile energy (PPE) using the equations shown in Table 1. In addition, the maximum rate of change of pressure (dP/dtmax), maximum rate of change of flow rate (dQ/dtmax), and pulse pressure (systolic–diastolic) were derived and recorded within each sample. To evaluate the effects of PF versus CF on aortic hemodynamics, a group average and standard deviation for EEP, SHE, and PPE were generated at each standard flow rate across all animals.
Baseline data for each study were gathered from the piglet native circulation after instrumentation but before cannulation for CPB. This established a basis of comparison between native hemodynamics and those of CF and PF perfusion that would be used during CPB. Bypass was initiated using the roller pump in CF mode. The blood flow rate was gradually increased until full bypass flows were attained. The animal was stabilized on bypass for approximately 15–30 min. Arterial blood gases were drawn to confirm adequacy of perfusion. Baseline data at full CF were recorded. The pump flow rate was then increased in 250 ml increments to achieve flows of 0.5, 0.750, and 1.0 L/min. Pressure and flow rate waveforms were recorded at each flow increment. A second set of data was obtained by increasing the flow rate in similar 250 ml increments until full bypass flows were again established. At that point, the arterial pump was weaned off and the PRVP engaged so that the same data sequence could be collected in PF mode. The data cycles were repeated two to three times in PF mode and a second time in CF mode.
Statistical analysis was performed in Microsoft Office Excel 2007 (Redmond, WA) using a one-way analysis of variance approach, where values of p < 0.05 were considered statistically significant. The independent variables were pump type (roller, PRVP, no pump). The dependent variables were PPE, EEP, SHE, perfusion flow rate, aortic pulse, femoral pulse, aortic pressure, femoral pressure, peak cannula flow rate, and maximum change in cannula flow rate. All results are expressed as mean values and standard error.
All piglets were supported successfully on CPB after instrumentation. There were no immediate complications from surgical manipulation. No clot formation was observed at the end of the study in any circuit loop or component in all the experiments. The maximum CPB flow obtained was 122 ± 12 ml/min/kg. Representative flow and pressure waveforms for the CF (roller pump) and PF (PRVP) are shown in Figures 3 and 4. The PRVP waveforms reveal significantly more pulsatility when compared with those of the roller pump. The PRVP is characterized by a more rapid rise of pressure during the systolic period followed by a gradual fall in pressure during the diastolic phase. This characteristic is preserved over the range of flows tested and remains intact in the femoral artery (Figure 5).
Hemodynamic energy data comparison between PRVP and roller pump is summarized in Table 2 and was derived using the aortic cannula flow and the aortic pressure waveforms. The percentage increase in hemodynamic energy content during PF over an equivalent steady flow at the measured MAP is presented as the PPE. At each flow rate, the PRVP’s PPE was significantly higher (p < 0.001) than the roller pump and under all conditions remained above 10%, reaching as high as 21%. The increase in energy content per unit volume of PF compared with equivalent steady flow at the measured MAP is given as SHE. Surplus hydraulic energy showed a similar trend as the PPE, ranging from 2.4 to 7.0 times greater for the PRVP than the roller pump. The absolute value of the hemodynamic energy was calculated as EEP. Energy equivalent pressure is highly dependent on the magnitude of the MAP at the time of measurement, rendering direct comparisons between flows difficult in light of varying MAPs throughout our animal study. We were unable to distinguish a difference in EEP between pumps. Analysis of PRVP waveforms at each flow rate is summarized in Table 3. The maximum rate of change of pressure (dP/dtmax), pulse pressure (systolic–diastolic), peak cannula flow rate (dQ/dtmax), and beat rate (BPM) were statistically similar between baseline animal and PRVP at 1.0 L/min flow. Values of dP/dtmax and pulse pressure were reduced with decreasing perfusion flow rates, most notably in the femoral artery where the dP/dtmax at 0.5 L/min fell to 32% of the value recorded at 1.0 L/min, and the pulse pressure was 50% of that at full perfusion flow rate. This is in contrast to the hydraulic energy values of the pulse flow that trended higher at lower perfusion rates (Table 2). Pulse rates for the PRVP were in the range of 50–120 pulses per minute over the range of flows tested. Extracorporeal circuit pressures measured just proximal to the aortic cannula increased with increasing flow rates and remained below 350 mm Hg peak pressure in all cases (Figure 6).
Pulsatile perfusion has been thought of as the best physiologic mode for CPB to address the morbidities associated with pediatric bypass; yet, this remains controversial. This controversy is rooted, in part, in a historical lack of consistent reporting of quantifiable measures of PF. Classic measures of pulsatility have included pulse pressure (systolic pressure–diastolic pressure) and maximum rate of change of pressure (dP/dtmax). Significant efforts have been made more recently to standardize a set of metrics for quantifying PF. Undar22 and Shepard et al.23 have advocated characterizing pulsatility in terms of EEP, SHE, and PPE. Energy equivalent pressure refers to the ratio of the area under the hemodynamic power curve and the area under the pump flow curve. In non-PF, the EEP will be equivalent to the MAP. Surplus hydraulic energy is the difference between the EEP and the MAP multiplied by a constant and refers to the surplus energy created by a pulsation. This method seems to be the most accurate and clinically relevant because the generation of a pulse alone does not necessarily result in transmission of surplus energy relative to nonpulsatile pumps.24 It is hypothesized that it is this excess energy that is the mechanism for improved tissue perfusion with PF.25 Percentage pulsatile energy is related to the ratio of the EEP and the MAP.
The results of this study confirm that PRVP design is capable of safely and simply producing a physiologically significant pulse in an in vivo piglet model. The pulse generated by the PRVP system is similar to the animal’s native heart and more closely resembles its normal physiology than the roller pump. In addition, this model has previously been studied by other groups because of the piglets having similar hemodynamics, cardiovascular physiology, and vascular resistance characteristics to those of human.26,27
Modern centrifugal pumps have been shown capable of generating PF28 but have been largely excluded from pediatric bypass for other reasons. The nonocclusive nature of centrifugal pumps poses risk of gravity-induced retrograde flow with possible risk of air entrainment at the aortic cannulation site.29 In addition, the high-speed impeller rotation necessary to produce pressure-driven flows poses a risk of cavitation and blood damage under conditions of low blood flow rate and high operating pressures that can be characteristic during pediatric CPB.30–33 Furthermore, centrifugal pumps are afterload sensitive, making precise control of blood flow rate challenging in the pediatric population.34 Conventional roller pumps can produce PF but are rarely able to generate a physiologic waveform.35 Significant variation in hemodynamic energy created by various pulsatile roller pumps has been demonstrated.21,27,28 The PRVP avoids these drawbacks and provides consistent pulsatile energy that can be used in CPB.
Undar et al.8,36 developed in the late 1990s a physiologic pulsatile pump (PPP) using a hydraulic driven pump with dual pumping chamber mechanism, one located immediately after the venous reservoir and the second immediately after the oxygenator. The PPP was tested successfully in piglets, but the authors concluded that all components of the circuit (oxygenator, arterial filter, and aortic cannulae) need to be carefully chosen. In addition, this system required the use of unidirectional valves and large priming volumes. Wang et al.37 published an in vitro model of a nonocclusive pediatric pulsatile roller pump for acute (CPB) and chronic support (extracorporeal life support and biventricular assist systems), in which roller pump could demonstrate effective PF in the three simulated circulatory support circuits, especially at higher flow rates. However, this system has limitation including the pump chamber inability to be completely flat, which increases the risk of generating negative pressure (upstream of pump tubing blockage) or air embolism when the chamber is empty at high flow rates.37
There are several limitations to this system: 1) the pump rate and rpm, determine the pulse frequency and it cannot be adjusted manually; 2) this study only evaluated the mechanics of producing PF and did not evaluate the clinical impact of the pulsatility, with regard to hemolysis, end-organ perfusion, or myocardial preservation; and 3) small sample size. Future studies with the piglet model must address these concerns.
The novel PRVP is capable of providing full flow support of a piglet during CPB. The PRVP creates greater pulsatility than a CF roller pump over the range of flow rates used during pediatric CPB. The mechanically induced pulsatility is similar to that observed with the native heart. The PRVP offers a safe, simple, drop-in replacement for conventional roller pumps for pediatric open heart surgery that may provide significant physiologic PF resulting in improved end-organ perfusion and potentially better outcomes.
1. Miller BJ, Gibbon JH Jr, Greco VF, Smith BA, Cohn CH, Allbritten FF Jr. The production and repair of interatrial septal defects under direct vision with the assistance of an extracorporeal pump-oxygenator circuit. J Thorac Surg. 1953;26:598–616
2. Alkan T, Akçevin A, Undar A, Türkoğlu H, Paker T, Aytaç A. Benefits of pulsatile perfusion on vital organ recovery during and after pediatric open heart surgery. ASAIO J. 2007;53:651–654
3. Murkin JM, Martzke JS, Buchan AM, Bentley C, Wong CJ. A randomized study of the influence of perfusion technique and pH management strategy in 316 patients undergoing coronary artery bypass surgery. I. Mortality and cardiovascular morbidity. J Thorac Cardiovasc Surg. 1995;110:340–348
4. Presta P, Onorati F, Fuiano L, et al. Can pulsatile cardiopulmonary bypass prevent perioperative renal dysfunction during myocardial revascularization in elderly patients? Nephron Clin Pract. 2009;111:c229–c235
5. Onorati F, Presta P, Fuiano G, et al. A randomized trial of pulsatile perfusion using an intra-aortic balloon pump versus nonpulsatile perfusion on short-term changes in kidney function during cardiopulmonary bypass during myocardial reperfusion. Am J Kidney Dis. 2007;50:229–238
6. Orime Y, Shiono M, Nakata K, et al. The role of pulsatility in end-organ microcirculation after cardiogenic shock. ASAIO J. 1996;42:M724–M729
7. Su XW, Guan Y, Barnes M, Clark JB, Myers JL, Undar A. Improved cerebral oxygen saturation and blood flow pulsatility with pulsatile perfusion during pediatric cardiopulmonary bypass. Pediatr Res. 2011;70:181–185
8. Undar A, Masai T, Beyer EA, Goddard-Finegold J, McGarry MC, Fraser CD Jr. Pediatric physiologic pulsatile pump enhances cerebral and renal blood flow during and after cardiopulmonary bypass. Artif Organs. 2002;26:919–923
9. Undar A, Masai T, Yang SQ, Goddard-Finegold J, Frazier OH, Fraser CD Jr. Effects of perfusion mode on regional and global organ blood flow in a neonatal piglet model. Ann Thorac Surg. 1999;68:1336–1342
10. Wang W, Bai SY, Zhang HB, Bai J, Zhang SJ, Zhu DM. Pulsatile flow improves cerebral blood flow in pediatric cardiopulmonary bypass. Artif Organs. 2010;34:874–878
11. Orime Y, Shiono M, Hata H, et al. Cytokine and endothelial damage in pulsatile and nonpulsatile cardiopulmonary bypass. Artif Organs. 1999;23:508–512
12. Sezai A, Shiono M, Nakata K, et al. Effects of pulsatile CPB on interleukin-8 and endothelin-1 levels. Artif Organs. 2005;29:708–713
13. Heiss KF, Clark RH, Cornish JD, et al. Preferential use of venovenous extracorporeal membrane oxygenation for congenital diaphragmatic hernia. J Pediatr Surg. 1995;30:416–419
14. Cecere G, Groom R, Forest R, Quinn R, Morton J. A 10-year review of pediatric perfusion practice in North America. Perfusion. 2002;17:83–89
15. Undar A, Frazier OH, Fraser CD Jr. Defining pulsatile perfusion: Quantification in terms of energy equivalent pressure. Artif Organs. 1999;23:712–716
16. Undar A, Rosenberg G, Myers JL. Major factors in the controversy of pulsatile versus nonpulsatile flow during acute and chronic cardiac support. ASAIO J. 2005;51:173–175
17. Allen GS, Murray KD, Olsen DB. The importance of pulsatile and nonpulsatile flow in the design of blood pumps. Artif Organs. 1997;21:922–928
18. Mazur DE, Osterholzer KR, Toomasian JM, Merz SI. A novel, low cost, disposable, pediatric pulsatile rotary ventricular pump for cardiac surgery that provides a physiological flow pattern. ASAIO J. 2008;54:523–528
19. Montoya JP, Merz SI, Bartlett RH. Laboratory experience with a novel, non-occlusive, pressure-regulated peristaltic blood pump. ASAIO J. 1992;38:M406–M411
20. Montoya JP, Merz SI, Bartlett RH. Significant safety advantages gained with an improved pressure-regulated blood pump. J Extra Corpor Technol. 1996;28:71–78
21. Montoya JP, Merz SI, Bartlett RH. A versatile positive displacement pump with independently controllable suction and discharge pressures. 19971997 ASME Fluids Engineering Division Summer Meeting 3355;
22. Undar A. Pulsatile versus nonpulsatile cardiopulmonary bypass procedures in neonates and infants: From bench to clinical practice. ASAIO J. 2005;51:vi–x
23. Shepard RB, Simpson DC, Sharp JF. Energy equivalent pressure. Arch Surg. 1966;93:730–740
24. Undar A, Eichstaedt HC, Masai T, Bigley JE, Kunselman AR. Precise quantification of pulsatility is a necessity for direct comparisons of six different pediatric heart-lung machines in a neonatal CPB model. ASAIO J. 2005;51:600–603
25. Haines NM, Wang S, Kunselman A, Myers JL, Undar A. Comparison of pumps and oxygenators with pulsatile and nonpulsatile modes in an infant cardiopulmonary bypass model. Artif Organs. 2009;33:993–1001
26. Pantalos GM, Giridharan G, Colyer J, et al. Effect of continuous and pulsatile flow left ventricular assist on pulsatility in a pediatric animal model of left ventricular dysfunction: Pilot observations. ASAIO J. 2007;53:385–391
27. Undar A, Eichstaedt HC, Masai T, et al. Comparison of six pediatric cardiopulmonary bypass pumps during pulsatile and nonpulsatile perfusion. J Thorac Cardiovasc Surg. 2001;122:827–829
28. Ressler N, Rider AR, Kunselman AR, et al. A hemodynamic evaluation of the Levitronix Pedivas centrifugal pump and Jostra Hl-20 roller pump under pulsatile and nonpulsatile perfusion in an infant CPB model. ASAIO J. 2009;55:106–110
29. Clark JB, Guan Y, McCoach R, Kunselman AR, Myers JL, Undar A. An investigational study of minimum rotational pump speed to avoid retrograde flow in three centrifugal blood pumps in a pediatric extracorporeal life support model. Perfusion. 2011;26:185–190
30. Kolff J, McClurken JB, Alpern JB. Beware centrifugal pumps: Not a one-way street, but a dangerous siphon! [Letter] Ann Thorac Surg. 1990;50:512
31. Arcari M, Phillips SD, Gibbs P, Rela SM, Heaton ND. An investigation into the risk of air embolus during veno-venous bypass in orthotopic liver transplantation. Transplantation. 1999;68:150–152
32. Augustin S, Horton A, Butt W, Bennett M, Horton S. Centrifugal pump inlet pressure site affects measurement. Perfusion. 2010;25:313–320
33. Bachus L, Custodio A Know and Understand Centrifugal Pumps.
. 20031st ed Oxford, UK Elsevier
34. Nishida H, Akazawa T, Nishinaka T, Aomi S, Endo M, Koyanagi H. Afterload-dependent flow fluctuation of centrifugal pump: Should it be actively fixed? Artif Organs. 1998;22:362–365
35. Kay PH, Munsch CM Techniques in Extracorporeal Circulation
. 20044th ed. New York, NY Oxford University Press Inc:p. 135
36. Undar A, Masai T, Inman R, et al. Evaluation of a physiologic pulsatile pump system for neonate-infant cardiopulmonary bypass support. ASAIO J. 1999;45:53–58
37. Wang S, Durandy Y, Kunselman AR, Ündar A. A nonocclusive, inexpensive pediatric pulsatile roller pump for cardiopulmonary bypass, extracorporeal life support, and left/right ventricular assist systems. Artif Organs. 2013;37:48–56
pediatric; cardiopulmonary bypass; CABG new technology