Secondary Logo

Journal Logo

Clinical Cardiovascular

Pulsatile Versus Nonpulsatile Flow During Cardiopulmonary Bypass: Extent of Hemolysis and Clinical Significance

Tan, Zihui*; Besser, Martin; Anderson, Simon; Newey, Caroline; Iles, Ray§; Dunning, John; Falter, Florian

Author Information
doi: 10.1097/MAT.0000000000001154
  • Free


For over 50 years, cardiopulmonary bypass (CPB) has been used to provide optimal operating conditions during cardiac surgery. It facilitates cardiac manipulation and allows a bloodless operating field while maintaining hemodynamic stability and excellent tissue oxygenation. One of the recognized limitations of extracorporeal circulation is the potential for erythrocyte damage and hemolysis, regularly seen as ‘pink urine’ after CPB but rarely thought to be clinically relevant.1–3 However, this destructive effect can cause coagulopathy, raised pulmonary and systemic vascular resistance, platelet dysfunction and renal injury, thereby potentially increasing morbidity and mortality.4 Hemolysis is particularly pronounced in young children and in patients on ventricular assist devices or extracorporeal membrane oxygenation but also after prolonged periods of CPB.5–7

The controversy over the optimal perfusion modality during cardiac surgery is not new.8,9 Pulsatile CPB is perceived as being more physiologic, and there has been some evidence that it might confer benefits in postoperative renal function.10 It can be produced by both centrifugal and roller pumps. When using conventional CPB circuits with roller pumps, pulsatile flow (PF) comes at the price of cyclical pressures swings in the circuit.11 We aimed to investigate if pulsatile CPB flow using a conventional circuit with roller pumps would cause a greater extent of cell damage in patients undergoing isolated coronary artery bypass grafting (CABG) when compared to nonpulsatile flow (NPF). We used heme and globin as markers for intravascular hemolysis as well as platelet count as maker for further cell damage.


Inclusion and Exclusion Criteria

After obtaining Institutional Review Board approval (reference number S02319), 62 adult patients scheduled for elective CABG surgery on CPB were included in this prospective observational study. Patients with known or suspected coagulopathy, hemoglobinopathy or anemia likely to require transfusion of packed red blood cells were excluded, as were patients undergoing emergency surgery or those requiring an intra-aortic balloon pump. Patients with a body surface area (BSA) greater than 2.3 m2 were also excluded as it would have been impossible to achieve pulsatile pump flow equivalent to a cardiac index (CI) of 2.4, which is considered ‘full flow’ in our institution, without exceeding maximum safe line pressure in the extracorporeal circuit.

Anesthesia Technique

Anesthesia was induced with midazolam, propofol, fentanyl, or remifentanil followed by tracheal intubation after muscle paralysis with pancuronium, rocuronium, or atracurium. Indwelling arterial and venous lines were placed in the standard way. Before commencement of CPB anticoagulation was established with 300 IU/kg of heparin. An ACT above 400 seconds was considered sufficient for safe bypass and was maintained at this level with further heparin doses until the patient was weaned from CPB. Catecholamine or vasoactive therapy was determined at the discretion of the clinical team. Heparin reversal was achieved with protamine, which was given at a dose matching the pre-CPB heparin 1:1.

Cardiopulmonary Bypass

Both PF and NPF are routinely employed at our institution. The decision whether a patient received PF or NPF was made by the clinical team at the morning team brief before anesthetic induction. This decision is driven by surgical and anesthetic consultant preference. For data analysis, patients were divided in a group with PF and a group with NPF.

We used a LivaNova S5 (London, United Kingdom) heart-lung machine (HLM) with roller pumps only, identical LivaNova disposable tubing packs with a 1/2-inch pump boot and an Inspire 8F oxygenator with integrated arterial line filter for all patients. Arterial cannulation was achieved by cannulating the ascending aorta using a standard 24F Medtronic (Watford, United Kingdom) cannula; the right atrium was cannulated for venous drainage using a Medtronic MC2 two-stage cannula, either 32/40F or 34/46F depending on patient BSA. Vacuum-assisted venous drainage was used routinely in all cases. A standard priming solution, consisting of 1.3 L Hartmann’s solution, 0.3 L Mannitol, and 5000 IU Heparin, was used in all patients. Full flow was considered to be achieved when a cardiac index of 2.4, based upon patient BSA as calculated using the Du Bois formula, was reached.12 High strength Harefield Solution (Terumo, Shibuya, Tokyo, Japan) blood cardioplegia (ratio 4:1) was administered to achieve diastolic arrest after aortic cross-clamping, at which point pulsatile flow was initiated in patients in the PF group.

Pulsatile flow was achieved by regular increases and decreases in rotational speed, augmenting the pressure above baseline to mimic normal pulsatile blood flow. For PF, we set the pulse frequency at 50/min, the pulse width (i.e., the time flow stays in simulated systole) at 50% and base flow (i.e., percentage by which a patient’s calculated full flow is reduced in order to generate the diastole) at 40%. In patients in the PF group, PF was maintained from the time of aortic cross-clamping until the cross-clamp was removed; NPF was maintained at a continuous flow matching a CI of 2.4 in the NPF group.

The pressure within the CPB circuit was measured post oxygenator by an analog sphygmomanometer gauge, referred to as a Tycos gauge. The line pressure was displayed as a digital reading from the same transducer, showing the current maximum circuit pressure on the control panel of the HLM.

Sample Collection

Blood samples were collected from patients’ arterial lines at baseline, upon establishing safe anticoagulation before commencing CPB, at 30-minute intervals while on CPB and after protamine administration.

Serum Analysis for Hemolysis

Clotted blood was centrifuged at 1500g for 10 minutes, 0.5 mL of serum was frozen for later analysis off-site. Sample analysis took place at MAP Sciences Laboratory (Bedford, United Kingdom), identification and quantitative analysis for hemolysis were done by matrix-assisted laser desorption ionization (MALDI) time of flight mass spectrometry using the Shimadzu-Kratos MALDI-8020 (Shimadzu-Kratos, Manchester, United Kingdom). This mass spectral method is 1000 times more sensitive than colorimetric assays for hemolysis and identifies free globin of hemoglobin from releases of heme porphyrins.

Samples were diluted at 1:60 in mass spectrometry grade water (Romil, Cambridgeshire, United Kingdom) and prepared for rapid mass spectral analysis. Heme and globins from hemoglobin were determined with cyano-4-hydroxycinnamic acid and 3,5-dimethoxy-4-hydroxycinnamic acid (SA) matrices, respectively. A sandwich method was used to plate the diluted sample onto a stainless-steel plate, where 1 µL of sample and 2 µL of matrix were deposited. Each sample plate run also had seven serum samples spiked with lysed whole blood of a known hemoglobin concentration (2.4, 4.8, 9.7, 19.3, 38.7, 77.3, and 137.8 µg/L) serving as a reference for identification and standard response calibration for quantification. A total of 500 profiles, 17 shots each, were collected per sample, and the average of these 8500 shots was exported as mzML data format. The peak intensities of free heme and free globin were collected onto excel spreadsheets. The concentrations of free heme and free globin were determined after calibration curve fitting.

Demographic and intraoperative lab data as well as preoperative and postoperative renal function were collected retrospectively from patients’ electronic records.

Statistical Analysis

Data were analyzed using t-tests, Mann-Whitney tests, and χ2 tests where appropriate using MedCalc Statistical Software (MedCalc Software, Ostend, Belgium).


Baseline Demographics and Intraoperative Data

A total of 62 patients were recruited into the study (32 in the pulsatile group and 30 in the nonpulsatile group). The baseline demographics are summarized in Table 1, the details of CPB in Table 2.

Table 1. - Baseline Demographics
Nonpulsatile (n = 30) Pulsatile (n = 32) p Value
Age (years) 67.2 ± 10.33 70.5 ± 8.35 0.17
Male 22 (73.3%) 29 (90.6%) 0.07
BSA (m2) 1.92 ± 0.37 1.89 ± 0.32 0.99
Hypertension 17 (56.7%) 19 (59.4%) 0.83
Diabetic 7 (23.3%) 9 (28%) 0.67
Left ventricular function
 Good 22 (73.3%) 25 (78.1%) 0.66
 Moderate 7 (23.3%) 7 (21.9%) 0.89
 Poor 1 (0.03%) 0 (0%) 0.92
Rhythm (sinus) 28 (93.3%) 28 (87.5%) 0.44
EuroScore II 4.0 ± 2.3 3.94 ± 2.01 0.62
logistic EuroSCORE 3.86 ± 2.97 3.53 ± 2.7 0.97
BSA, body surface area.

Table 2. - Cardiopulmonary Bypass Characteristics
Nonpulsatile (n = 30) Pulsatile (n = 32) p Value
CPB time (minutes) 80.76 ± 23.5 80.74 ± 28.06 0.663
AOX time (minutes) 49.78 ± 17.35 48.66 ± 22.31 0.472
Flow/BSA 2.27 ± 0.44 2.42 ± 0.4 1
Average patient MAP on CPB (mm Hg) 61.32 ± 4.43 59.83 ± 5.70 0.286
Lactate (mmol/L)
 Baseline 1.44 ± 0.55 1.47 ± 0.6 0.744
 CPB 5 min 2.37 ± 0.65 2.53 ± 0.69 0.268
 AOX 30 min 2.01 ± 0.49 2.14 ± 0.61 0.329
 AOX off 1.84 ± 0.43 2.26 ± 0.78 0.027
 %change Baseline/AOX off 28.17 ± 33.1 37.84 ± 36.3 0.267
Heparin dose (U) 28446 ± 7854 27044 ± 6926 0.567
Protamine dose (mg) 256 ± 70.9 275 ± 64.1 0.116
Max tycos pressure (mmHg) 190.45 ± 42.26 257.02 ± 73.07 <0.0001
Min tycos pressure (mmHg) 156.64 ± 35.45 106.16 ± 38.96 <0.0001
Tycos amplitude (mmHg) 34.07 ± 14.55 151.52 ± 56.08 <0.0001
Arterial line amplitude (mmHg) 5.63 ± 2.45 30.15 ± 7.68 <0.0001
AOX, aortic cross clamp; BSA, body surface area; CPB, cardiopulmonary bypass; MAP, mean arterial blood pressure.

There were no differences in baseline characteristics between the groups. The CPB and aortic cross-clamp times, as well as the pump flow achieved, were similar in both groups. The maximum (257.02 ± 73.07 vs. 190.45 ± 42.26, p < 0.0001) and the minimum (156.64 ± 35.45 vs. 106.16 ± 38.96, p < 0.0001) Tycos pressure (mmHg) as well as the amplitude (151.52 ± 56.08 vs. 34.07 ± 14.55, p<0.0001) were statistically significantly higher in group PF.

The average mean arterial blood pressure (MAP) throughout CPB was similar in both groups (61.32 ± 4.43 vs. 59.83 ± 5.70 mmHg, p = 0.286). As expected, the amplitude observed on the patients’ arterial line was significantly higher in group PF (30.15 ± 7.68 vs. 5.63 ± 2.45 mmHg, p < 0.0001). Lactate was similar between the two groups at baseline (p = 0.744) but significantly higher in the pulsatile group at CPB 3 (p = 0.027). However, there was no difference between the groups in the overall percentage change of lactate from baseline to CPB3. The amount of heparin and protamine administered was similar in both groups.

Table 3 gives an overview of hemoglobin, heme and globin levels from baseline to measurements after protamine. Again, there were no differences between NPF and PFat baseline and no differences in hemoglobin levels throughout the operation, although both heme and globin levels were mostly higher in the pulsatile flow group. Heme levels were significantly higher in group PF after aortic unclamping (0.234 ± 0.23 vs. 0.107 ± 0.1 µg/L, p = 0.015), while globin levels were significantly higher in the same group at 30 minutes of aortic cross-clamping (6.66 ± 6.5 vs. 3.357 ± 2.89 µg/L, p = 0.017). The finding that both heme and globin levels not only stay high but also rise after protamine had been administered is somewhat surprising. None of the patients studied received any blood transfusion (Figures 1 and 2).

Table 3. - Hemoglobin, Heme and Globin Levels
Nonpulsatile (n = 30) Pulsatile (n = 32) p Value
Hemoglobin (g/L)
 Baseline 126.7 ± 24.68 122.8 ± 23.84 0.797
 CPB 5 min 95.44 ± 22.14 91.94 ± 17.41 0.603
 AOX 30 min 96.16 ± 21.88 89.75 ± 17.23 0.158
 AOX off 97.25 ± 24.03 89.44 ± 19.64 0.193
 After protamine 98.95 ± 22.67 93.59 ± 17.85 0.281
Heme (µg/l)
 Baseline 0.113 ± 0.14 0.111 ± 0.22 0.49
 CPB 5 min 0.113 ± 0.14 0.126 ± 0.34 0.12
 AOX 30 min 0.261 ± 0.73 0.235 ± 0.5 0.87
 AOX off 0.107 ± 0.10 0.234 ± 0.23 0.015
 After protamine 0.212 ± 0.24 0.275 ± 0.27 0.23
Globin (µg/l)
 Baseline 1.775 ± 1.25 1.655 ± 1.41 0.61
 CPB 5 min 2.127 ± 1.92 2.783 ± 2.84 0.54
 AOX 30 min 3.357 ± 2.89 6.66 ± 6.51 0.017
 AOX off 4.435 ± 3.51 5.59 ± 4.63 0.33
 After protamine 7.739 ± 7.13 11.447 ± 8.36 0.067
AOX, aortic cross clamp; CPB, cardiopulmonary bypass.

Figure 1.
Figure 1.:
Effect of CPB on heme (µg/L). CPB, cardiopulmonary bypass.

There was no statistically significant difference in creatinine levels at baseline or 24 hours postoperatively in both groups. The percentage change between preoperative and postoperative creatinine levels is marginally higher in the pulsatile group, although this did not reach statistical significance. There is no statistically significant difference in intraoperative urine output between the two groups, although it is trending higher in group NPF (515.8 ± 327.3 vs. 473.1 ± 308.5 mL, p = 0.63) (Table 4).

Table 4. - Preoperative and Postoperative Creatinine Levels and % Change, Intraoperative Urine Output
Nonpulsatile (n = 30) Pulsatile (n = 32) p Value
Creatinine μmol/L (preoperative) 79.88 ± 18.31 91.74 ± 34.01 0.19
Creatinine μmol/L (postoperative) 79.25 ± 23.78 94.23 ± 40.47 0.09
% change in creatinine 0.57 ± 21.11 2.78 ± 21.72 0.39
Urine output intraoperative (mL) 515.8 ± 327.3 473.1 ± 308.5 0.63

Table 5 shows changes in white blood cell (WBC) and platelet count between levels taken during the preoperative workup and on postoperative day 1. Neither WBC levels preoperatively and postoperatively nor percentage change in the levels between NPF and PF show any significant differences. However, there is a marked percentage change in preoperative and postoperative platelet count between NPFand PF (16.93 ± 20.49% vs. 25.91 ± 17.39, p = 0.07), although this fails to reach statistical difference.

Table 5. - White Blood Cell and Platelet Count
Nonpulsatile (n = 30) Pulsatile (n = 32) p Value
White blood cells (109/L)
 Preoperative 7.22 ± 1.66 7.41 ± 2.10 0.694
 Postoperative 9.68 ± 2.95 9.94 ± 3.29 0.746
 % change 35.5 ± 32.7 39.54 ± 52 11 0.716
Platelets (109/L)
 Preoperative 236.0 ± 72.16 251.71 ± 66.44 0.139
 Postoperative 192.87 ± 64.35 182.19 ± 53.11 0.513
 % change 16.93 ± 20.49 25.91 ± 17.39 0.07

No patient was transfused with pooled platelets.


During the early development of CPB, nonpulsatile flow was the only available mode of perfusion. The availability of simple roller pump systems eventually allowed the adoption of pulsatile flow which is perceived to be more physiologic.13

Roller pumps generate pulsatility by positive displacement of blood when the roller is compressing the tubing. Advantages include low cost and ease of use. However, substantial risks are associated with generating pulsatile flow using roller pumps: tubing fracture, increased spallation, overcompression or undercompression and most importantly cavitation and subsequent air embolism are the main hazards.14

Using centrifugal blood pumps has a lower risk of air embolism and causes less blood trauma.15 Comparing pulsatile and nonpulsatile flow Gu et al.16 found a small increase in energy equivalent pressure and surplus hemodynamic energy associated with centrifugal pump pulsatile flow, although they were not able to demonstrate improved clinical outcomes.16 Roller pumps were used throughout the current study.

Regardless of the mode of perfusion, red blood cells are subjected to significant mechanical forces as they travel through the various components of the extracorporeal circuit and extracorporeal circulation is always associated with a degree of hemolysis.3 But there is speculation that the aggressive acceleration and deceleration during pulsatile CPB may increase red cell damage.14 This cell damage is potentially aggravated due to the large swings in pressure within the circuit using roller pumps during pulsatile flow.11,17 Two recent studies have produced conflicting results.18 Our findings support Voss et al. who conclude that the extent of hemolysis is higher in pulsatile flow compared to nonpulsatile perfusion. In our study, we report large pressure swings in the circuit during pulsatile flow. The Tycos amplitude during pulsatile flow was at least 4.5 times that during nonpulsatile flow. Also, the maximum Tycos pressure was at least 60 mmHg higher than during nonpulsatile flow. This correlated with higher levels of heme and globin seen in the pulsatile flow group. These results reached statistical significance after aortic unclamping for heme and after approximately 30 minutes of pulsatile flow for globin. The globin levels in the pulsatile group were consistently higher than those in the nonpulsatile group, whereas the heme levels increased more in the pulsatile group 30 minutes into bypass. The bigger difference may indicate that haptoglobin is less able to compensate for an increase in free hemoglobin than hemopexin in the context of pulsatile CPB flow.

The observation that both heme and globin kept rising after termination of CPB and administration of protamine was unexpected, and we are currently unable to explain this. Protamine has previously been shown to lyse mouse red cells in a temperature-dependent fashion, but this is not currently one of side effects listed in the specific product characteristics.19

Red blood cells, being enucleate, are easily deformed, an essential functional feature as they pass through the microcirculation. Upon destruction, the heme and globin components are cleared by the body’s natural scavenging system. Free globins of hemoglobin are only seen in serum or plasma when vascular hemolysis has occurred whilst free heme is seen in both vascular hemolysis and where hepato-splenic destruction of damaged red cells is increased. The globin component is broken down into amino acids by circulating macrophages. Iron is removed from the heme component, binds to transferrin, and gets delivered to the bone marrow where it is used to synthesize new heme molecules. Free heme is converted to biliverdin by heme oxygenase and subsequently to bilirubin.20 When this clearing mechanism is overwhelmed in excessive hemolysis, hemoglobin is released, and levels of heme and globin in blood and urine will start to increase. It is estimated that the hemopexin of a healthy adult can bind 6 µg/mL of heme in 1:1 molar ratio21; haptoglobin, in turn, can bind an estimated 1 mg/1 mL of hemoglobin before it is depleted.22 Heme is lipophilic and initially binds to hemopexin until this mechanism is saturated. When hemopexin is depleted, heme will bind to albumin and form methemalbumin. Haptoglobin will be depleted before hemopexin is consumed. After depletion of haptoglobin, hemoglobin is excreted through the renal glomeruli. Hemosiderin indicates toxic levels of iron and subsequently hemoglobinuria occurs.22 Hemopexin levels increase in the context of persistent hemolysis. Iron, when released at an excessive amount, can cause iron overload, which can lead to acute kidney injury.

At a microvascular level, upon saturation of the scavenging system, plasma free hemoglobin binds irreversibly to nitric oxide (NO) derived from the endothelium.20 NO depletion results in increases in pulmonary and systemic vascular resistance as well as platelet activation and aggregation with a subsequently increased risk of bleeding.23

Although the difference in drop-in preoperative to postoperative platelet count did not reach statistical significance, the nearly 10% difference between NP and P can be seen as further evidence of cell damage caused by pulsatile flow. This can be said with reasonable confidence as none of the included patients received a platelet transfusion, there was no difference in hemoglobin levels between the two groups and CPB prime is standardized in our institution, that is, the level of hemodilution will have been the same in both groups.

Modern CPB tubing and cannulas have been designed to reduce the amount of turbulence during bypass to a minimum. If used according to recommendations, levels of hemolysis as a result of circuit components are within the capacity of the endogenous clearing mechanisms.3 Increased levels of hemolysis in a pulsatile flow group should be seen as evidence for increased cell damage when compared to a nonpulsatile flow group using identical extracorporeal circuits and bypass characteristics. Pulsatile and nonpulsatile circuits used in this study were built identically and used identical equipment. There were no differences in patient MAP and flow rate on CPB, bypass, or cross-clamp time. Any differences in rates of hemolysis are unlikely to be related to different bypass characteristics.

Our results do not support some of the findings of previous studies, which demonstrated improved postoperative kidney function in patients undergoing pulsatile perfusion.24,25 Intriguingly, one of the studies most strongly in favor of PF used intra-aortic balloon counterpulsation rather than roller pumps on the CPB circuit to generate pulsatility.26 It is fair to say that there is an equally large body of evidence that does not support the claim that pulsatility generated by roller pumps leads to better renal outcomes.27,28 A very recent retrospective analysis of nearly 2500 patients came to the conclusion that there is either no association between pulsatile perfusion and reduced kidney injury or that the difference is extremely small.29 Although not statistically significant, and not powered for this outcome, we saw an increase in postoperative creatinine levels and less intraoperative urine output in the pulsatile flow group. Increased hemolysis and raised heme levels, in particular, might be an explanation why any potential advantages in renal function are negated.

Koning et al.10 showed that pulsatile flow preserves microcirculatory perfusion in the early postoperative period. Neil et al.30 demonstrated a lower peak lactate level postoperatively in patients who received pulsatile perfusion. We did not see any differences in metabolic effects. Lactate on bypass was broadly similar between the groups at all measuring points apart from aortic cross clamp off, when it was significantly higher in the PF group. Although the study was not powered for this outcome, it still raises questions about the often-stated advantages PF is believed to have in maintaining physiologic perfusion to the microvascular bed.


The observational study design comes with a number of limitations. Neither surgical nor anesthetic technique was standardized, but extracorporeal circuit design and equipment were identical in all cases. By choosing to only include isolated CABG patients, we were able to exclude other sources of hemolysis associated with open chamber heart surgery. Unfortunately, this means that the data cannot be extrapolated to valve or aortic surgery. The study is based on institutional practices and specific bypass circuits. We, therefore, urge caution when attempting to generalize our results to other institutions.

Another factor associated with increasing hemolysis is the duration of CPB and aortic cross-clamping. The duration of pulsatility was relatively short in our study and was similar in Group NPF and PF (49.78 ± 17.35 vs. 48.66 ± 22.31 minutes, p = 0.472). This short duration may explain why we were not able to demonstrate statistical significance in heme and globin levels at more measuring points.

We did not have any longer-term follow-up data available for analysis.

We used lactate as a surrogate for the adequacy of microcirculation. Although lactate is easy to measure and regularly used during CPB, it is not the most specific parameter. Other modalities, such as tissue near infra-red reflectance spectroscopy or side stream dark-field imaging, are potentially more targeted but far more cumbersome to use.10,31


Pulsatile CPB flow using roller pumps subjects red blood cells to significantly higher pressures in the extracorporeal circuit and to higher cyclical pressure drops when compared with NPF. These shear forces lead to higher maker levels of heme and of globin and a larger drop-in platelet count. This increased level of hemolysis goes some way to explain why we did not see a beneficial impact of pulsatile CPB flow on renal function or intraoperative lactate levels.

Figure 2.
Figure 2.:
Effect of CPB on globin (µg/L). CPB, cardiopulmonary bypass.


1. Wright G. Haemolysis during cardiopulmonary bypass: Update. Perfusion 2001.16: 345–351.
2. Rajbhandhari N, Rahman Z, Islam N, et al. Cardiopulmonary bypass induced hematological changes in patients undergoing cardiac surgery. Bangladesh Hear J 2015.30:53–57.
3. Vercaemst L. Hemolysis in cardiac surgery patients undergoing cardiopulmonary bypass : A review in search of a treatment algorithm. J Extracorpor Technol2008.40: 257–267.
4. Van Bijnen AA, Ganushchak YM, Heijmans JH, et al. Hemolysis during cardiac surgery is associated with increased intravascular nitric oxide consumption and perioperative kidney and intestinal tissue damage. Front Physiol 2014.5:1–9.
5. Omar HR, Mirsaeidi M, Socias S, et al. Plasma free hemoglobin is an independent predictor of mortality among patients on extracorporeal membrane oxygenation support. PLoS One 2015.10: e0124034.
6. Meyer A, Wiles A, Rivera O, et al. Hemolytic and thrombocytopathic characteristics of extracorporeal membrane oxygenation systems at simulated flow rate for neonates. Pediatr Crit Care Med. 2012.13: e255–e261.
7. Lodge AJ, Turi JL. Cardiopulmonary bypass is associated with hemolysis and acute kidney injury in neonates, infants and children. Pediatr Crit Care Med 201515:1–18.
8. Alghamdi AA, Latter DA. Cardiopulmonary bypass flow : An evidence-based approach. J Card Surg 2005.21: 347–354.
9. Haines N, Wang S, Undar A, Alkan T, Akcevin A. Clinical outcomes of pulsatile and non-pulsatile mode of perfusion. J Extra Corpor Technol 2009.41: P26–P29.
10. Koning NJ, Vonk AB, van Barneveld LJ, et al. Pulsatile flow during cardiopulmonary bypass preserves postoperative microcirculatory perfusion irrespective of systemic hemodynamics. J Appl Physiol (1985) 2012.112: 1727–1734.
11. Jameel S, Colah S, Klein A. Recent advances in cardiopulmonary bypass techniques. Contin Educ Anaesthesia, Crit Care Pain 2010.10: 20–23.
12. Burton RF. Estimating body surface area from mass and height: Theory and the formula of du bois and du bois. Ann Hum Biol 2008.35: 170–184.
13. Wilkens H, Regelson W, Hoffmeister FS. The physiolgic importance of pulsatile blood flow. N Engl J Med 1962.267: 443–446.
14. Gourlay T, Taylor KM. Pulsatile Cardiopulmonary Bypass. in Cardiopulmonary Bypass: Principles and Practice, Philadelphia, PA, Lippincott Williams & Wilkins, 2000.2nd ed.
15. Sarkar M, Vishal P. Basics of cardiopulmonary bypass. Indian J Anaesth 2017.61: 760–767.
16. Gu YJ, van Oeveren W, Mungroop HE, et al. Clinical effectiveness of centrifugal pump to produce pulsatile flow during cardiopulmonary bypass in patients undergoing cardiac surgery. Artif Organs 2011.35: E18–E26.
17. Mulholland JW, Shelton JC, Luo XY. Blood flow and damage by the roller pumps during cardiopulmonary bypass. J Fluids Struct 2005.20: 129–140.
18. Simons AP, Wortel P, van Kan RA, van der Veen FH, Weerwind PW, Maessen JG. Pulse conductance and flow-induced hemolysis during pulsatile cardiopulmonary bypass. Artif Organs 2010.34: 289–294.
19. FF B. Studies on the hemolytic properties of protamine. J Gen Physiol 1961.44: 433–442.
20. Thomas C, Chb MB, Lumb AB, et al. Physiology of haemoglobin. Contin Educ Anaesthesia, Crit Care Pain 2012.12: 251–256.
21. Michael M, Jaeger U. Autoimmune Hemolytic Anaemia, Hematology. Elsevier, Amsterdam, The Netherlands, 2018, pp. 648–662.
22. Delanghe JR, Langlois MR. Hemopexin: A review of biological aspects and the role in laboratory medicine. Clin Chim Acta 2001.312: 13–23.
23. Rapido F. The potential adverse effects of haemolysis. Blood Transfus 2017.15: 218–221.
24. Domenico A, Dodonov M, Van Oeveren W, et al. Pulsatile cardiopulmonary bypass and renal function in elderly patients undergoing aortic valve surgery †. Eur J Cardio-Thoracic Surg 2015.47: 291–298.
25. Poswal P, Mehta Y, Juneja R. Comparative Study of pulsatile and nonpulsatile flow. Ann Card Anaesth 2004.7: 44–50.
26. Serraino GF, Marsico R, Musolino G, et al. Pulsatile cardiopulmonary bypass with intra-aortic balloon pump improves organ function and reduces endothelial activation. Circ J 2012.76: 1121–1129.
27. Farid S, Povey H, Anderson S, Nashef SA, Abu-Omar Y. The effect of pulsatile cardiopulmonary bypass on the need for haemofiltration in patients with renal dysfunction undergoing cardiac surgery. Perfusion 2016.31: 477–481.
28. Baraki H, Gohrbandt B, Del Bagno B, et al. Does pulsatile perfusion improve outcome after cardiac surgery? a propensity-matched analysis of 1959 patients. Perfusion 2012.27: 166–174.
29. Coulson TG, McPhilimey E, Falter F, Abu-Omar Y, Klein AA. The Association between pulsatile cardiopulmonary bypass and acute kidney injury after cardiac surgery: A Before-and-After Study. J Cardiothorac Vasc Anesth 2020.34: 108–113.
30. Neil MPO, Fleming JC, Badhwar A, et al. Cardiopulmonary bypass : Microcirculatory and systemic effects. ATS 2012.94: 2046–2053.
31. Murkin JM. Monitoring and optimization of the microcirculation during CPB. J Thorac Dis 2019.11(suppl 10): S1489–S1491.

cardiopulmonary bypass; hemolysis; pulsatile flow; non-pulsatile flow

Copyright © 2020 by the ASAIO