Received date: April 28, 2013; Accepted date: August 24, 2013; Published date: August 27, 2013
Citation: Canty DJ, Kim M, Royse CF, Andrews D, Botrell S, et al. (2013) The Impact of Routine Norepinephrine Infusion on Hemodilution and Blood Transfusion in Cardiac Surgery. J Anesthe Clinic Res 4:342. doi: 10.4172/2155-6148.1000342
Copyright: © 2013 Canty DJ, et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Visit for more related articles at Journal of Anesthesia & Clinical Research
Background: Hemodilution and blood transfusion are associated with poor outcome after cardiac surgery. We hypothesized that routine norepinephrine infusion commenced prior to anesthesia induction during on-bypass cardiac surgery would reduce intraoperative hemodilution and red cell transfusion.
Methods: Two cohorts of consecutive cardiac surgery patients at different time periods were retrospectively reviewed for perioperative hemoglobin, creatinine concentrations and units of red cells transfused. Patients in group NE (n=72, in 2010) all received standardized hemodynamic management by a single anesthesiologist with low dose norepinephrine infusion commenced at 3-5 μg.min-1 (18-30 nmol.min-1, 0.24-0.4 nmol.kg-1.min-1) commencing prior to anesthetic induction and continued into the postoperative period. In the absence of blood loss, hemodynamic stability was achieved using vasopressors and inotropes rather than fluid administration, in an attempt to reduce hemodilutional anemia and trigger for red cell transfusion. Controls (n=94, in 2005) received selective norepinephrine infusion post cardiopulmonary bypass for persistent hypotension and vasodilation. There were no major changes to surgical or perfusion technique in the time period between cohorts, and the transfusion trigger remained the same at Hb<70 g/L.
Results: Intraoperatively, hemoglobin concentrations were higher in group NE compared with controls (p<0.0001) despite lower baseline values (139 ± 19 vs 133 ± 15, P=0.028). Additionally, fewer units of red cells were transfused in the NE group intraoperatively (0.2 ± 0.6 units/patient) compared with controls (0.53 ± 1.47, p=0.041). Maximum postoperative rise in serum creatinine concentration (μmol.L-1) was not significantly different (NE 26 ± 32, controls 30 ± 57, p=0.49 and at discharge 3 ± 53 vs. 5 ± 30, p=0.39). NE group patients were at increased risk of bleeding, having received more extensive surgery (p=0.042), longer clamp-time (p=0.009) and no aprotinin compared to 74% of controls.
Conclusions: This study shows proof of concept that during on-bypass cardiac surgery, routine low dose norepinephrine infusion is associated with reduced hemodilution and intraoperative red cell transfusion without increasing postoperative serum creatinine.
During cardiac surgery, little blood is discarded, with most Red Blood Cells (RBC) returned to the patient. Anemia is primarily caused by hemodilution from intravenous fluid administration as priming volume for the Cardiopulmonary Bypass (CPB) circuit, and as volume expansion to treat hypotension. Vasodilation is a common cause of hypotension and causes include the inflammatory response to surgery, anesthesia and CPB; which commonly persists into postoperatively (9-44%) as a result of systemic inflammation and sedative drugs . Venodilation leads to a relative central hypovolemic state due to expansion of the intravascular capacity. Hypotension is thus usually managed by fluid administration and boluses of short acting vasopressors. Hemodilution, anemiafrequently occurs from fluids, CPB prime, cardioplegia, and platelets and fresh frozen plasma if required. Vasoconstrictors such as Norepinephrine (NE) are generally reserved for hypotension resistant to further fluid challenge or persistent low systemic vascular resistance [1,2]. Vasoconstrictors are often avoided due to concerns of undesirable vasoconstriciton of renal, splanchnic and even arterial coronary bypass conduits including radial arteries, without good supporting evidence.
The rationale for pre-emptive routineNE before anesthetic induction is to maintain normal vascular tone and capacitance rather than cause supranormal vasoconstriction. There is usually little blood loss before CBP and the principal cause of hypotension is vasodilation from anesthesia and commencement of CPB. Since the vascular space is not expanded, there is no need for fluid administration.
The key trigger for transfusion is the Hemoglobin (Hb) concentration and not a reduction in the total Red Blood Cell (RBC) mass or the metabolic or physiologic performance of the patient. These triggers are based on guidelines , rather than a clinically related measurement of the patient at the time and may lead to RBC transfusion despite lack of blood loss.
The transfusion trigger has progressively been lowered [3,4] due to evidence of increased morbidity and mortality associated with RBC transfusion  and reduced blood viscosity during CPB . However, the Hb threshold for vital organ tissue oxygenation is unknown, influenced by patient co morbidities and accumulating evidence of morbidity [7-11] and mortality [7,9,10,12] associated with hemodilutional anemia raises concerns about where to strike the balance between blood transfusion versus anemia [6,13,14]. Learned societies have recommended routine prevention ofhemodilutional anemia during cardiac surgery [3,15-17].
We hypothesized that routine use of low-dose NE intraoperatively would reduce hemodilutional anemia the need for red cell transfusion.
This study was approved by the Melbourne Health ethics committee as an audit of hospital practice and requirement for informed consent was waivered.
In this retrospective cohort study, data was retrieved and compared from databases and medical records from consecutive patients who underwent on-CPB cardiac surgery by a single surgeon at the Royal Melbourne and Melbourne Private Hospitals over two separate 12-month periods. Data from a 12 months consecutive cohort in 2010 when intraoperative NE infusion was used routinely (NE group) was compared to a consecutive cohort in 2005 when NE was reserved only for persistent hypotension and vasodilation after CPB. In the NEgroup, NE infusion was commenced at μ min-1 (0.04 μg/kg/min) prior to anesthesia and continued intraoperatively until transferred to ICU. The infusion was adjusted to maintain desired mean arterial pressure (70- 90 mm Hg) with minimal use of intravascular fluids. The aim was to prevent vasodilation and consequent fluid administration in an effort to reduce hemodilution and RBC transfusion.
The operative techniques and transfusion trigger for intraoperative RBC transfusion was the same in both groups (Hb less than 70.0 g.L-1). The intravenous fluids used included hartmanns crystalloid (Baxter, NSW and Australia), gelofusin (Bbraun, Switzerland), plasma-lyte 148 for CPB (Baxter, NSW, Australia) and 4% normal serum albumin (CSL Biotherapies, Broadmeadows, VIC, Australia). Intravenous starches were not used.
Data was reviewed retrospectively by investigators not involved in medical care of the subjects. Baseline characteristics recorded included age, gender, height, weight, history of diabetes mellitus, NYHA classification and type of surgery (CABG, valve, aorta and redosternotomy). Intraoperative data recorded included antifibrinolytic drug administration, intraoperative hemofiltration, postoperative centrifugal hemoconcentration of CPB blood and duration of CPB and aortic cross-clamp. Preoperative hemoglobin and Creatinine (Cr) measurements were taken in the pre-admission clinic in the weeks leading up to elective surgery or on the day of hospital admission for urgent surgery. Hemoglobin was subsequentlyrecorded after induction of anesthesia prior to CPB, after commencement of CPB, the lowest during CPB and on transfer to ICU. Creatinine was recorded on the day of discharge and the highest postoperative measurement. The highest postoperative Cr and at discharge. The numbers of units of RBC’s transfused intraoperatively were recorded.
All patients were operated on by the same surgeon under similar conditions between the two hospitals. Anticoagulant medications were discontinued prior to surgery when possible (intravenous heparin at least 6 hrs, warfarin 5 days, antiplatelet drugs 1 week). Cardiopulmonary bypass and intra-operative cell salvage was used routinely. Acute normovolemichemodilution and retrograde autologous priming of the CPB circuit were not used. General anesthesia was induced with midazolam, fentanyl, and propofol, and maintained with desflurane (1 MAC) and pancuronium for neuromuscular blockade. Analgesia included high thoracic epidural ropivicaine 2% and morphine 20 μ/mL, or intravenous fentanyl (5-10 μ/kg) when epidural contraindicated. Intraoperative monitoring included arterial, central venous and pulmonary artery pressure, intermittent cardiac output by thermodilution, 5 lead ECG, pulse oximetry, capnography, bispectral index EEG, nasopharyngeal temperature and transesophageal echocardiography.
ceased and nitroglycerin infusion if persistent. For low cardiac index (<2.0 L.min-1.m-2), the initial inotrope added was dobutamine, with epinephrineormilrinone infusions administered infrequently. Anticoagulation was achieved with heparin 300-400 IU/kg of body weight and ACT >480 seconds was required before initiating CPB. Pump flow was 2.3-2.5 L.min-2 and mean arterial pressure maintained between 70-90 mm Hg. The CPB circuit was primed with 2 L of Plasma-lyte 148 crystalloid solution and maintained at a temperature of 34-35°C. Median sternotomy and standard arterial and venous cannulation were used as appropriate to the surgical procedure. Temperature was allowed to drift progressively to 34°C until rewarming, with the heat exchanger temperature not exceeding 37°C. The ascending aorta was cross-clamped and cardiac arrest induced by administration of tepid blood cardioplegia (20–25°C, blood: cardioplegia 4:1) with both antegrade and retrograde delivery. Further doses of maintenance cardioplegia were given following completion of each graft anastomosis or within 15 minutes. After weaning from CPB, protamine 4 mg/kg was given to neutralize heparin. Intraoperative RBC scavenging was routine in both groups. Post CPB residual circuit blood was centrifugally hemoconcentrated selectively in control and routinely inNE group patients.All patients were transferred to ICU mechanically ventilated and sedated with a propofol infusion.
Coronary arterybypasses grafting included total arterial revascularization with mammary and radial arteries where possible. There was no change to the use ofNE consequent on the use of arterial grafts; and this institution has used vasopressors with arterial grafts since 1995.
The primary endpoint was the Hb concentration over the intraoperative period. The ICU admission Hb was considered the final intraoperative measurement. Secondary endpoints included the number of units of RBC transfused, and the postoperative Cr rise from baseline.
Statistical analysis was performed using one-way mixed factorial repeated measured analysis of variance allowing for multisampling asphericity by applying either the Huynh-Feldt or Greenhouse-Geisser corrections for ε < 0.75 and ε >0.75 respectively . This allowed for a comparison of continuous variables over time. Continuous data and categorical data were compared with Student’s two-tailed t test and Fisher’s exact test respectively. Data were coded and stored in Microsoft Excel 2010 and analyzed using SPSS V20 (SPSS Inc, Chicago, IL, USA). Statistical significance was defined as a value of p<0.05.
Of the 171 patients retrieved from the hospital databases who received on-bypass cardiac surgery from the investigators, there were 2 patients from each group excluded due to missing hospital records and one patient excluded from theNE group due to incorrect surgical coding leaving 94 patients in the control group and 72 patients in theNE group. Preoperative data is presented in Table 1, intraoperative data in Table 2 and transfusion data in Table 3. NE group patients had more complex surgery (p=0.042), longer duration of CPB (114.5 ± 46.6 vs. 99.6 ± 47.9 min, p=0.046) and cross clamp (90.9 ± 39.7 vs. control 73.9 ± 42.1 min, p=0.009). The control group had a higher baseline Hb (p=0.028) and Cr (p=0.006). Aprotinin was used more frequently in the control group (74% vs. 0%); as it was withdrawn from hospital use in the time between cohorts. No other antifibrinolytics or parenteral synthetic hemostatic agents were used. After surgery, centrifugal hemoconcentration of residual CPB blood occurred more frequently in theNE group 97% than the control group 12% (p<0.0001). Other characteristics were similar between groups. The patients receiving CABG surgery was similar in both groups as were the proportions receiving total arterial revascularization and radial artery graftingIntraoperative.
|Variable||Control n = 94||NE n = 72||p|
|Age; years||67 (11)||66 (12)||0.456|
|Height; m||1.68 (0.1)||1.68 (1.1)||0.711|
|Weight; kg||81.0 (16.3)||78.4 (16.6)||0.306|
|Diabetes mellitus (%)||33||40||0.415|
|Baseline preopHb; g.L-1
Baseline preop Cr; mmol.L-1
Values are mean (SD) or number (proportion).*p<0.05
NE: Norepinephrine group; NYHA: New York Heart Association; Hb: Hemoglobin concentration; Cr: Creatinine concentration
Table 1: Pre-operative characteristics of control and norepinephrine group patients.
n = 94
n = 72
|Type of surgery|
|Isolated CABG (%)||61||44||0.042*|
|Valve/complex (%) #||39||56||0.042*|
|Aortic aneurysm (%)||2||1||1|
|Aortic dissection (%)||0||3||1|
|All CABG (%)
Total arterial grafts (%)
Radial artery used (%)
|CPB time, min||99.6 (47.9)||114.5 (46.6)||0.046*|
|Aortic cross-clamp, min||73.9 (42.1)||90.9 (39.7)||0.009*|
|Hemofiltration on-CPB (%)||10||4|
|Hemoconcentration post-CPB (%)||12||97||0.234|
|Postoperative hemofiltration for renal failure in ICU (%)||14||4||<0.0001*|
|Mortality n (%)||1 (1.1)||1 (1.4)||0.06
Values are mean (SD) or number (proportion). *p<0.05, #Valve ± CABG/redo,
multiple valves or aortic surgery.
NE: Norepinephrine group; CABG: Coronary Artery Bypass Grafting; CPB: Cardiopulmonary Bypass; ICU: Intensive Care Unit.
Table 2: Intraoperative data of control and norepinephrine group patients.
n = 94
n = 72
|Units RBC transfused||50||11||0.041*|
|Units per patient||0.53 (1.47)||0.15 (0.60)||0.041*|
|Patients transfused, n (%)||16 (17)||6 (8)||0.11|
|≥4 RBC units, n (%)||5 (5)||1 (1)||0.37|
Values are mean (SD) or number (proportion).
NE: Norepinephrine group; RBC: Red Blood Cells
Table 3: Intraoperative red blood cell transfusion requirements of control and norepinephrine group patients.
The Hb was better preserved in theNE group than controls (p<0.0001, Figure 1), with the greatest difference at the completion of surgery and transfer to ICU,NE group 111.9 ± 18.7, control group 100.7 ± 13.7, p<0.0001. There were fewer units of RBCs per patient transfused in theNE group (p=0.041, Table 3 and Figure 2). Despite a higher baseline Cr in the control group, the maximum increases in postoperative Cr from baseline were not different (NE group 26 ± 32, vs. control 30 ± 57 mmol.L-1, p=0.49, Figure 3). However there was a non-significant trend towards more postoperative hemofiltration in the control group 14% vs. 4%, p=0.06.
This study shows proof of concept that routine use of norepinephrine infusion to counteract the anaesthetic, CPB and surgically induced inflammatory vasodilation may assist in restrictive fluid management during on-bypass cardiac surgery. We found less intraoperative hemodilution and red cell transfusion. These findings occurred despite a higher baseline Hb, less complex surgery and shorter CPB times in the control group. Routine infusion of norepinephrine was not associated with increased Cr.
The use of norepinephrine or of a restrictive fluid strategy was not controlled in the ICU and therefore adherence to the technique was variable. Our data does not therefore allow us to comment on the efficacy on the strategy throughout the hospital stay, and we have restricted out primary endpoint to the intraoperative period up to the time of ICU admission. Further research is required to identify if maintenance of the vasopressor infusion to normalize vascular resistance prior to fluid administration will result in less anemia and fewer transfusions.
A degree of hemodilution is inevitable with use if CPB, due to the pump prime volume of 2 L of fluid. However, the anaesthetic, CPB and surgery commonly results in a systemic inflammatory response which causes vasodilation, hypotension and a normal or increased cardiac output [19-21]. The traditional treatment is to initially administer intravenous fluids and to reserve vasopressor infusion (such as NA) when intravenous fluids are no-longer effective. This results in greater hemodilution and may trigger RBC transfusion, both of which are independent predictors of morbidity and mortality in cardiac surgery [5,11,22-24]. In this study the aim ofNE infusion was to maintain venous and arterial vascular tone to prevent hypotension and maintain the normal vascular space volume, thereby facilitating a fluid restrictive strategy to minimize hemodilutional anemia.
The intraoperative RBC transfusion trigger of 70 g.L-1 was used in both groups, and so the reduced intraoperative transfusion in theNE group is likely to be due to better preservation of Hb concentration from less hemodilution. In a small randomized trial by Vretsakis et al. demonstrated reduced red cell transfusion in 100 patients with a fluid restrictive strategy compared with 92 controls with similar risk factors for bleeding . Techniques to prevent hemodilutional anemia include minibypass circuits, retrograde autologous blood priming, ultra filtration on CPB and microplegia , however to date we are unaware of any reports of using routine vasoconstrictors to minimize hemodilution. The optimal conditions for blood function include normal concentration and temperature. A low Hb is a marker for dilution of all of the blood components and it may be expected that all functions of blood including clotting may be affected.
The mean Hb was higher in the NE group than controls at all four intraoperative periods (anesthesia, commencement of CPB, lowest on CPB and arrival in ICU) but the most marked difference was on arrival in ICU. The Hb increased in both groups from CPB to transfer to ICU due to the transfusion of residual CPB blood. The greater rise in the NE group may have been due to more frequent centrifugal hemoconcentration of residual CPB blood prior to re-transfusion (97% vs. 12%). However, this does not account for the higher Hb concentrations at the other 3 time periods.
There are concerns thatNE may exacerbate CPB associated renal dysfunction by potential regional vasoconstriction and it is usually reserved for persistent postoperative hypotension and vasodilation [26,27]. However this was not evident in this study. In fact there was a non-significant trend of increased postoperative hemofiltration in the control group.This may have been contributed to from their higher baseline Cr and frequent use of aprotinin (associated with renal injury ). Similar findings were found by Morimatsu et al. who reported no difference in postoperative Cr between 100 on-bypass cardiac surgery patients with postoperative vasodilation treated withNE compared with 100 without .
One key concern amongst some surgeons is the belief that use of vasoconstrictors may cause arterial coronary bypass grafts to spasm and fail. There is little if any evidence for this, and the concept arose from the original paper by Acar  who postulated that a key reason for the early failure of radial artery grafts in the 1970s  was spasm and recommended the use of calcium channel antagonists. This became standard practice for radial artery grafts. In our institution however, we have used vasopressors after CPB and in the ICU since 1995. Our practice is routine total arterial revascularization  where the majority of multivessel coronary grafting rely on the radial artery [33,34]. We reject that the use of vasoconstrictors that aim to restore a vasodilated state toward normal vascular tone will result in spasm of these arterial grafts. This centre has one of the largest experiences in the world of radial artery grafting [33,35].
Due to the retrospective study design the two groups are not identical. However, the NE group started with lower Hb and had more complex surgery. If the groups had been equal there may have been a greater difference observed. An effort to minimize bias was made by performance of the study by research staff not involved in patient care, to capture whole cohorts, and to use a single surgeon. The Hb and RBC transfusion rates may have been influenced by differing anesthesiologists, however it would be expected that a more restrictive transfusion strategy would result in lower Hb concentration. Total intraoperative blood loss and fluid administration were not recorded and hence hemodilution is therefore assumed rather than proven. It is possible that there was less intraoperative blood loss in theNE group; yet all intraoperative blood loss was scavenged into the CPB reservoir or the cell saver and thus not discarded in either group. Additionally, theNE group patients were expected to be at greater risk of bleeding as they had more extensive surgery, longer CPB and aortic cross clamp times and less antifibrinolytics than the control group .
The routine use of norepinephrine to treat vasodilation was associated with less intraoperative anemia and reduced intraoperative red cell transfusion, without detrimental effect on renal function.
We are grateful for the assistance of a number of staff at the Royal Melbourne Hospital, including Mrs. Zelda Williams and Deborah Deguigand (research nurses). There was no funding supplied.