Bleeding after cardiopulmonary bypass (CPB) can result in hemodynamic instability, morbidity, and mortality.1 Evidence exists to show that the smaller the patient, the greater the propensity to bleeding complications. The evidence suggests that smaller patients are more likely to have quantitative factor deficiencies of infancy,2–4 whereas postoperative chest tube output has been demonstrated to be inversely related to age, and the incidence of re-exploration for bleeding is significantly greater in infants than older children.5 The same study demonstrated that the smaller the patient, the greater the volume of blood product transfused after CPB. Analysis of pediatric patients undergoing cardiac surgery revealed that younger, smaller children with complex surgery requiring prolonged CPB times and deep hypothermia, were more prone to fibrinolysis than their older counterparts.6 In addition, despite efforts to use smaller prime volumes, the effect of dilutional coagulopathy remains an important issue in the population of neonates and infants who are undergoing open heart surgery. Although much has been published regarding coagulopathic states induced by CPB in the pediatric population, much of our current practice with blood component therapy remains empiric, guided by previously recommended approaches, and visual cues in the operating room rather than by objective, goal-directed data specific to the patient.
Thromboelastography (TEG, Hemascope Corporation, Skokie, IL) measures the viscoelastic properties of clotting blood and assesses all stages of the clotting process, from initiation through attainment of maximum clot strength followed by clot retraction or lysis. It is a functional measurement of dynamic clot formation and not a measurement of individual factor levels. To measure the clot dynamics, TEG has a pin suspended in an oscillating cup containing 0.36 mL of whole blood and kaolin to stimulate the intrinsic cascade. The drag between the cup and pin is measured and is monitored by a computer as a function of time.
Results of TEG are displayed as a tracing (Fig. 1) from which a number of values are measured. The R value is the time necessary for initial clot formation (normal: 4–8 minutes), is prolonged by coagulation factor deficiencies or heparin administration, and is shortened by hypercoagulable states. The angle (α) assesses the rate of clot formation and reflects the adequacy of fibrinogen in the sample (normal: 47°–74°). The angle may also be affected by platelet dysfunction. Maximum clot strength (MA) is a reflection of the quantity and functional quality of platelets, although it also secondarily reflects the interaction of platelets and fibrinogen (normal: 55–73 mm).
When compared with routine transfusion therapy guided by standard laboratory coagulation testing, a transfusion algorithm based on TEG reduced transfusion requirements in adults undergoing cardiac surgery.7 This study was undertaken to examine how our adoption of the transfusion algorithm based on TEG changed our use of blood products and its effect on overall hemostasis in our patient population of infants less than 6 months of age undergoing open heart surgery. Because cyanotic patients tend to bleed more than their acyanotic counterparts, we also looked at these subpopulations.
PATIENTS AND METHODS
This study received approval from the North Texas Institutional Review Board at Medical City with exempt status. The analysis consisted of a retrospective chart review of 70 consecutive patients (48/70 or 68.6% cyanotic cases) operated upon from November 2003 to January 2005 (pre-TEG era) and data collected on 112 consecutive patients (46/112 or 41.1% cyanotic cases) whose surgery took place between April 2005 and January 2007 (TEG era). All 182 patients underwent CPB in their first 6 months of life at Medical City Children’s Hospital. Patient sex, age, weight, preoperative medications, classification of the underlying heart lesion as acyanotic (acyanotic cases include repairs or closures of atrial (ASD) and ventricular septal defects (VSD), foramen ovale, atrioventricular canals, patent ductus, pulmonary artery banding, and aortic arch procedures) or cyanotic (cyanotic procedures include repairs of Tetralogy of Fallot, truncus arteriosus, and Norwood procedures including the shunts used), type of surgery, details of CPB and data on routine hematologic and blood clotting studies, TEG values, blood product usage, and chest tube output were recorded. All anesthesia, surgical procedures, and perioperative care were provided by the one team of physicians. Patients requiring extracorporeal membrane oxygenation support were excluded from the study. Systemic anticoagulation was prescribed, monitored, and treated by a combination of heparin or protamine titration measurement and activated clotting time as measured and performed by the Hepcon HMS whole blood anticoagulation monitor. Heparin levels were monitored and maintained at appropriate levels while on CPB. To achieve anticoagulant reversal after separation from bypass, protamine was administered at 1.3 times the heparin level measured immediately before termination of bypass. Reversal was then confirmed by activated clotting time measurements.
Nonpulsatile CPB was performed with a hollow fiber membrane oxygenator (Terumo Corporation “Baby RX”) primed with 150 mL of plasmalyte-A, 15 meq of sodium bicarbonate, 30 mg/kg of methylprednisone, 5000 units of heparin, 200 mg CaCl2, ½ unit of saline washed packed red blood cells (PRBCs), and ½ unit of fresh frozen plasma (FFP). The remainder of the unit of PRBCs was administered shortly after initiation of CPB.
Prime volume ranged from 400 to 500 mL depending on the sizes of the units of PRBCs and FFP provided by the blood bank. Hematocrit (Hct) was maintained greater than 25% on CPB, and an Hct >30% was reached before termination of bypass for patients with cyanotic heart lesions. Blood conservation techniques for all patients included ultrafiltration while on CPB and modified ultrafiltration after termination of CPB using the HCO5 hemoconcentrator (Terumo). Autotransfusion with cell salvage was not used in this patient population.
Laboratory assessment included prothrombin time, partial thromboplastin time, Hct, and platelet count preoperatively; TEG after induction of anesthesia but before institution of CPB; TEG, Hct, and platelet count 15 minutes before discontinuation of CPB; TEG 2 hours post-CPB; and TEG, Hct, platelet count, PT and PTT at 3 am, the morning, after surgery. Before the start of the use of TEG (pre-TEG), blood component therapy was based on Hct and platelet counts obtained while on CPB. If the platelet count is <100,000, then 10 mL/kg of platelets were empirically given at discontinuation of bypass. In addition, based on visual inspection of the surgical field and perceived ongoing coagulopathy, 10 mL/kg of cryoprecipitate was generally administered before consideration of FFP; if excessive microvascular bleeding persisted, then FFP was also administered in a volume of 10 mL/kg.
However, when TEG was instituted, therapy was based on Hct and platelet count, the R value, the angle, and the MA while on bypass. A prolonged R value (≥8 minutes) was used as an indication for administration of FFP, a narrow angle (≤45°) for administration of cryoprecipitate, and a low MA (≤54 mm) for platelet administration. When the R and the angle were both notably abnormal, cryoprecipitate was administered before deciding on whether or not to give FFP. With the use of TEG, we felt confident in allowing platelet counts to go as low as 50,000 on bypass, before administering 10 mL/kg of platelets at discontinuation. If the platelet count on bypass was between 50,000 and 100,000, the MA was only marginally low, a marked derangement of either the R or the angle resulted in FFP or cryoprecipitate (10 mL/kg) being administered before giving platelets.
Blood administration was recorded for two time periods: intraoperative (from induction of anesthesia until leaving the operating room) and postoperative (from arrival in the intensive care unit [ICU] through the first 24 hours postoperatively). All blood products administered were normalized to body weight and included in the analysis. Postoperative blood loss was recorded as chest tube output beginning from the time of chest closure and continuing for 24 hours postoperatively also normalized to body weight. Postoperative minimum acceptable Hct value was 30% for patients with acyanotic heart lesions and 35% for those with cyanotic lesions.
The collected clinical data were analyzed using SAS 9.2 (SAS Institute, Cary, NC). All continuous variables were tested for normality. Those failing tests for normality (Anderson-Darling) were analyzed using nonparametric statistics (Wilcoxon), whereas the rest were compared using the t test. Categorical variables were subjected to χ2 or Fisher exact tests for statistical significance. Measurements of volumes were normalized by dividing by the patient’s weight. All tests were two-tailed, and a P value of 0.05 or less was considered statistically significant.
All Patients: Pre-TEG Versus TEG
Comparison of the demographics of the two patient groups (Table 1) showed no gender bias or significant difference in age, weight, height, preoperative Hct, or platelet count between pre-TEG and TEG groups. There was also no difference in number of redo sternotomies, use of Aprotinin or aspirin (ASA), CPB time, or degree of hypothermia.
Before the start of surgery, Hct and hemoglobin were not different between groups (Table 2); PT showed small differences that reached statistical significance but are unlikely to be clinically significant. There were a significantly larger proportion of cyanotic patients in the pre-TEG group (68.6%) versus the TEG cohort (41.1%) (P < 0.001).
Comparing the blood product usage in pre-TEG and TEG patients (Table 2), the normalized volume of PRBCs did not differ, but the postoperative use of cryoprecipitate decreased by more than 50% in the TEG era (failing to reach statistical significance because of the large standard deviation). Using the TEG R value as a guide to therapy, TEG era patients received significantly more FFP in the operating room. Conversely, although comparable amounts of platelets were administered intraoperatively between groups, in the TEG era, a significantly smaller volume of platelets needed to be transfused postoperatively and overall. Postoperatively, there was no significant difference between the pre-TEG and TEG groups in the Hct or hemoglobin, although the pre-TEG group ended up with a higher platelet count (Table 2). Although postoperative PTT also did not differ between groups, the postoperative PT was significantly shorter in the TEG era, consistent with more FFP being administered. At all time points, there was significantly less chest tube drainage in TEG era patients.
The need for reoperation for bleeding was examined in both groups. In the pre-TEG era, 4/70 (5.7%) of patients had to be re-explored, whereas in the TEG era, 5/112 (4.5%) were returned to the operating room with bleeding (P = 0.74).
Acyanotic Subset of Patients
Subset analysis of acyanotic patients showed higher TEG era postoperative Hcts and hemoglobin, whereas the postoperative platelet counts were the same in both cohorts (Table 3). The measurements show that TEG era patients received more PRBC during surgery, less after, and slightly more overall; platelet volumes administered were the same at all time points. Cryoprecipitate volumes administered perioperatively were higher, whereas postoperative volumes were lower; total volumes were increased when using TEG directed therapy. FFP usage was higher, although not statistically so. In the TEG era patients, postoperative chest tube output was less at all times, although only the 1 hour difference was statistically significant. In the TEG era, more intraoperative blood products were used overall, which led to a need for less products postoperative.
Cyanotic Subset of Patients
Although not statistically significant, the use of TEG resulted in a trend to more PRBC at all times, with less cryoprecipitate use postoperative and total. There was significantly more FFP being used intraoperatively (with a slight decrease in postoperative volumes used) and decreased usage of platelets especially in the postoperative period, ultimately leading to a slightly lower platelet count after surgery. However, as shown by the chest tube measurements, bleeding was better controlled using TEG data. Using the TEG data to direct blood product usage, led to significantly less postoperative bleeding as measured by chest tube output, with a 40% drop at 1 hours, 25% decreases at 2 hours and at 24 hour, all statistically significant differences (Table 4).
Comparison of TEG Era Blood Product Usage in Acyanotic and Cyanotic Subsets
Cyanotic patients begin with a higher hemoglobin and Hct because of their cyanotic condition. Cyanotic patients received significantly more volume of total FFP (P = 0.048), total cryoprecipitate (P = 0.002), and total PRBCs (P = 0.063). Cyanotic patients required transfusion of more overall blood products also. Acyanotic and cyanotic patients had similar chest tube outputs at all times measured. Postoperatively, hemoglobin (P = 0.09), Hct (P = 0.09), platelet count (P = 0.64), and PT (P = 0.16) also were not different between groups (Table 5).
Comparison of Measured TEG Values in Acyanotic and Cyanotic Subsets
Before initiation of CPB, while the R and MA values were comparable between groups, acyanotic patients had a significantly larger angle than their cyanotic counterparts, although the values for both groups fall within the normal range. Not surprisingly, this persisted with institution of bypass and, additionally, cyanotic patients developed a significantly smaller MA than their acyanotic counterparts while on bypass. This latter finding suggests that the cyanotic patient’s platelets and fibrinogen may be more sensitive to hemodilution. TEG measured 2 hours postbypass revealed increased clot strength (MA) in cyanotic patients, once again consistent with these patients receiving more FFP. As measured at 24 hours postoperatively, the values for R, angle, and MA were again equivalent between acyanotic and cyanotic subsets (Table 6).
Rather than measuring levels of any individual clotting protein, TEG analysis evaluates the overall integrity of the clotting system and provides direction as to how to most effectively correct abnormalities. Studies on adults reveal that while MA values correlate with fibrinogen levels and platelet counts, there is otherwise a poor correlation between TEG and routine clotting studies, such as PT and PTT.8 The explanation for this is that unlike routine clotting tests that terminate with the formation of initial fibrin strands, TEG measures the hemostatic process in whole blood from the initiation of clot formation to the final stages of clot retraction. A study evaluating the usefulness of coagulation tests routinely ordered to assess clotting abnormalities in children undergoing cardiac surgery demonstrated platelet count during CPB to be the variable most significantly associated with intraoperative blood loss and chest tube output in the first 12 hours after surgery.9 Specifically, platelet count during CPB <108,000 had the greatest sensitivity and specificity for prediction of excessive blood loss. That study also demonstrated that MA from TEG obtained while on CPB was the only variable associated with total blood products transfused. Thus, those investigators concluded that TEG data combined with platelet count seemed to predict clinically significant coagulopathies better than conventional clotting tests. There was no difference in platelet count on bypass between the pre-TEG and TEG eras in our study to correlate with the significant difference in chest tube output seen between these two patient populations. Significantly blunted angles and smaller MAs demonstrated in the TEG plots of our cyanotic patients while on CPB resulted in these patients receiving significantly more blood products to achieve chest tube output comparable with their acyanotic counterparts.
We focused our analysis on neonates and infants less than 6 months, as multiple previous studies have demonstrated this group of patients to be at the highest risk for postoperative bleeding. In a study by Miller et al9 evaluating coagulation changes after CPB, children <8 kg had more bleeding and required more coagulation products than children >8 kg. In addition, they found that platelet count after protamine administration and fibrinogen level correlated independently with chest tube drainage in the first 24 hours postoperatively in children weighing <8 kg as did postprotamine platelet count and TEG angle and MA in children weighing >8 kg. Not surprisingly, they found duration of CPB to be predictive of chest tube drainage in the first 24 hours after surgery, and no preoperative coagulation tests were useful in predicting post-CPB bleeding. They concluded that FFP offered no benefits in correcting coagulopathy following CPB in children and followed this with another study providing indirect evidence that fibrinogen of children <12 months of age with congenital heart disease is qualitatively dysfunctional.10 As shown in Table 6, reduced fibrinogen function on CPB is suggested by a larger than 30% decrease of the TEG angle from baseline in both acyanotic and cyanotic patients (Figs. 2 and 3.) Correction of fibrinogen resulted in return toward a normal angle on TEG measurements performed 2 hours after CPB in both acyanotic and cyanotic subsets (Fig. 4). Our practice using TEG as a guide to goal-directed therapy revealed FFP to be a helpful component in correcting post-CPB coagulopathy. Specifically, when indicated by TEG, administration of FFP as front line therapy not only corrects a prolonged R value but also likely enables the existing platelets to function more effectively demonstrating the importance of thrombin’s role in activating platelets. This larger volume of FFP administered also explains the significantly shorter postoperative PT seen in the TEG group. Thrombin is one of the most important and powerful sources of platelet activation, such that the burst of thrombin achieved with administration of the increased volumes of FFP results in maximization of the function of the platelets present during and postbypass, requiring less platelet transfusion. Platelet function is of greater importance than sheer number.
How to most rapidly assess and most effectively treat coagulopathy in young infants undergoing cardiac surgery is of great clinical importance. Before the institution of TEG at our institution, intraoperative and postoperative administration of blood products was guided by previously published recommendations1 tempered by clinical judgment. Since the institution of TEG, we use platelet count and values obtained from TEG mapping while on bypass and in the ICU postoperatively to guide our blood component therapy. This resulted in some increase in PRBCs and significantly more FFP being administered, and approximately a 30% decrease in the average amount of platelets administered. Overall, an increase in use of FFP during and after surgery led to large decreases in postoperative bleeding as seen by chest tube output.
Comparing the pre- and post-TEG–cohorts broken into acyanotic and cyanotic subgroups, we saw that using TEG resulted in more blood products being administered to the cyanotic group than the acyanotics, expected because the cyanotic infants tend to have more serious coagulopathies.14 As a result of this therapy, the postoperative bleeding (chest tube output) in the cyanotics was equivalent to the acyanotics. In both groups, this resulted in the chest tube output being significantly less after surgery compared with those that received blood component therapy in the pre-TEG era.
Currently, TEG technology allows real time monitoring of the tests on a computer screen in the OR, with all testing performed within 30 minutes. New reagents will reduce test times to 10 minutes. Using TEG to assess the integrity of the patient’s hemostatic profile allows correction of a coagulopathy before departing to the ICU from the operating room. Goal-directed blood product replacement as determined by TEG resulted in optimization of blood component therapy with effective correction of clotting abnormalities, resulting in a significant decrease in chest tube output at 1, 2, and 24 hours after open heart surgery and need for administration of less overall blood products in the postoperative interval.
TEG performed in severely premature but clinically stable infants (gestational age 27–31 weeks) revealed no defects in global coagulation, documenting the functional integrity of coagulation when compared with adults. That study and others demonstrated that healthy neonates and infants in fact demonstrate significantly shorter R times (initiation of clot formation) when compared with adults. This may be explained by reduced levels of plasma inhibitors and increased levels of procoagulants seen in these babies. Specifically, plasma levels of inhibitors, such as antithrombin and proteins C and S, are reduced in neonates, whereas major procoagulant factors, such as fibrinogen, factor V, factor VIII, and vWF, are at or above adult values. Increased levels of vWF may contribute to an MA measured by TEG equal to that of adults despite the known hyporeactivity of platelets during the first 2 weeks of life.3 Thus, there is functional maturity of the coagulation system in young, otherwise healthy children despite quantitative deficiencies in coagulation factors.
Using TEG during the patient’s hospitalization adds a small increment to the costs of the procedure. This hospital charges approximately $50 to $75 TEG.
LIMITATIONS OF THIS STUDY
The major limitation of the study is that it is not a prospective randomized study. There may be some change in the patient population seen for surgery and improvements in treatment and equipment over time since comparisons were made to a historical cohort of patients.