Neonates who undergo congenital cardiac surgery requiring cardiopulmonary bypass (CPB) are at high risk for significant postoperative bleeding for multiple reasons. First, CPB circuit priming volumes are often as large as their circulating blood volume, thus diluting components of the coagulation system required to achieve adequate hemostasis. Second, the immature coagulation system at birth results in qualitative and quantitative deficiencies of many coagulation factors that further contribute to post-CPB bleeding. Additionally, neonates often undergo long, complex cardiac repairs at low temperatures with extensive suture lines again contributing to excessive post-CPB bleeding. Of the various groups of pediatric patients presenting for cardiac surgery requiring CPB, neonates are likely to benefit the most from efforts to reduce post-CPB blood loss.
Excessive postoperative bleeding is predominantly treated with blood transfusion products, including packed red blood cells (pRBCs), platelets, cryoprecipitate, and fresh-frozen plasma. Transfusion with any of these products carries significant risks that may affect morbidity and mortality after cardiac surgery. Adult data have demonstrated an increase in mortality in association with pRBC transfusion after coronary artery bypass grafting and a negative effect on both long- and short-term survival after all types of adult cardiac surgery.1,2 Pediatric data too have demonstrated an increased morbidity in association with blood transfusions, including an increased duration of mechanical ventilation and duration of stay in the intensive care unit (ICU).3,4
Most investigations examining the adverse effects of bleeding and transfusion after congenital cardiac surgery combine data from children of various ages and sizes despite the fact that age and size contribute significantly to postoperative bleeding. The primary goal of this study was to examine the relationship between excessive postoperative bleeding/transfusion requirements and major postoperative adverse events in a group of neonates after CPB. Our anticipation is that defining these relationships will aid in predicting outcome and in allocating the resources needed for managing adverse events. To accomplish this goal, we performed a retrospective analysis of neonates who underwent complex congenital heart surgery during 2-year period.
After approval by the IRB, we retrospectively reviewed the medical records of neonates undergoing complex congenital heart surgery requiring CPB between January 1, 2010, and December 31, 2011. During this time period, institutional practices, CPB equipment, and personnel in the cardiac operating rooms were stable. One hundred sixty-nine neonates were included in the study; 2 were excluded because of a substantial amount of missing surgical and/or anesthetic data. For neonates having >1 cardiac surgical procedure during their hospitalization, only the first procedure was included in the database. None of the neonates had a second surgical procedure, other than extracorporeal membrane oxygenation (ECMO) cannulation, within the first 24 postoperative hours. Therefore, 167 neonates undergoing 167 cardiac surgical procedures were included in the final analysis.
Chart reviews were performed by 3 of the authors of this manuscript who standardized among themselves precise definitions of the collected variables. Multiple perioperative data were collected and analyzed. Preoperative data included patient demographics: age, weight, prematurity (defined as <36 weeks’ gestation),5 and presence of a major noncardiac structural anomaly. Cardiac diagnosis, surgical procedure, and the Risk Adjustment for Congenital Heart Surgery (RACHS-1) score5 also were recorded. Neonates with a combination of cardiac surgical procedures were assigned the RACHS-1 score of the greatest risk procedure and, when applicable, the presence of a combination procedure was treated as an independent variable. Preoperative baseline creatinine (Cr) levels also were recorded. Intraoperative data included CPB time, aortic cross-clamp time, the use of regional perfusion and its duration, the use of deep hypothermic circulatory arrest (DHCA) and its duration, and the lowest temperature achieved during CPB. The use of fresh frozen plasma (FFP) and post-CPB modified ultrafiltration (MUF) was also noted. In our postoperative data collection, we recorded data on postoperative bleeding and blood product administration both intraoperatively and postoperatively for the first 24 hours. Bleeding was measured by chest tube output (CTO) in the first 24 hours postoperatively. Postoperative Cr levels were recorded at 24 and 72 hours. We then focused on specific outcomes. In addition to duration of mechanical ventilation and duration of ICU stay, we recorded the presence of major postoperative adverse events, including renal dysfunction, dialysis (peritoneal or hemodialysis), thrombosis, the need for ECMO support, and in-hospital mortality. Definitions were standardized to promote consistency within the dataset. Renal dysfunction was defined as a >50% increase in the preoperative Cr concentration at 72 hours postoperatively.6,7 The presence of thrombosis required documentation by either Doppler ultrasound, echocardiography, or cardiac catheterization. Collection of bleeding and blood product transfusion data stopped with the institution of ECMO.
For all neonates, nonpulsatile hypothermic CPB was performed via a nonheparin-coated system, a Terumo RX-05 hollow-fiber membrane oxygenator (Terumo Cardiovascular Systems, Ann Arbor, MI) and COBE SMArt neonatal circuits (Sorin Group USA, Inc., Arvada, CO). All circuits contained a 250-mL priming volume with pRBCs added to the circuit as needed to achieve and maintain a hematocrit of 30% throughout the duration of CPB. In accordance with our institutional policy, all neonates having cardiac surgery received pRBCs that were <14 days old. Anticoagulation was achieved by the administration of 400 U/kg of porcine heparin with 1000 units of heparin added to the CPB prime. Kaolin-activated clotting time (ACT) values exceeding 480 seconds were confirmed before the initiation of CPB. Additional heparin was administered as necessary during CPB to maintain an ACT >480 seconds, and all neonates received an additional 100 U/kg of heparin at the time of rewarming. Diuretic therapy was standardized for all neonates. The CPB prime contained 1 mg/kg of mannitol. In addition, all neonates received a single dose of lasix (2 mg/kg) and mannitol (1 mg/kg) during CPB at the time of rewarming. As per institutional protocol, all neonates were treated with tranexamic acid (100 mg/kg as a load to the patient, 100 mg/kg as a load to the pump, and a continuous infusion of 10 mg/kg/h throughout the duration of the operation). MUF was performed immediately after separation from CPB at the surgeon’s discretion.
After MUF, cell saver blood to the extent of volume remaining in the CPB circuit was used for all neonates to decrease donor exposures. Protamine (4 mg/kg) was used to neutralize heparin upon completion of MUF. After confirmation of heparin neutralization by ACT, bleeding was treated with transfusion of nonvolume reduced apheresis platelets followed by cryoprecipitate. Persistent bleeding after this initial therapy was assessed by thromboelastography and treated as deemed necessary by the attending anesthesiologist. In the ICU, the attending intensivist was the primary director of postoperative care.
Variables were tested for normality and summarized as median (interquartile range: 25th to 75th percentile) for continuous variables and N (%) for categorical variables. We used Spearman correlation to determine correlations between perioperative variables and 24-hour postoperative CTO and blood product requirements for the first 24 hours postoperatively. Next, we subdivided our patient population into bleeders and nonbleeders, defining bleeders as those neonates who bled >75th percentile. Nonparametric Wilcoxon rank sum or Kolmogorov-Smirnov tests and χ2 or Fisher exact tests were used to compare variables between the 2 groups. Multivariable logistic regression models were constructed for all binary outcomes. Variables in the models were tested for logistic regression assumptions, and continuous variables not meeting assumptions, such as CPB time, were transformed to achieve linearity with the link function. Full models contained variables with a P value >0.10 in univariate analysis as well as clinically significant covariates. A stepwise method was used to create the final models informed by the Hosmer-Lemeshow test, Akaike information criteria, and adjusted R-squared values. Because of the high incidence of outcomes in the study population, odds ratios and confidence intervals (CIs) were transformed to risk ratios using the methods suggested by Zhang and Yu.8 To assess our definition of bleeders as those neonates who bled >75th percentile, we also performed a sensitivity analysis around this threshold. All statistical analyses were performed using SAS Version 9.2 (SAS Institute Inc., Cary, NC). Statistical significance was defined as a P value ≤0.05.
Data from 167 neonates undergoing the same number of cardiac surgical procedures requiring CPB were included in our final analysis. Patient demographics and CPB data are summarized in Table 1. Intraoperative and postoperative transfusion requirements, 24-hour postoperative CTO, and postoperative outcomes are shown in Table 2. Because our primary focus was bleeding, as measured by CTO and transfusion requirements for the first 24 hours postoperatively, we initially determined that these 2 measures were congruent and correlated with each other (r = 0.470, P < 0.0001), thus indicating that both variables were acceptable to monitor postoperative bleeding. We then assessed significant correlations between all preoperative and CPB data with postoperative CTO and transfusion amounts. Moderately strong correlations were found between 24-hour CTO and postoperative blood product transfusion with weight, RACHS-1 score, CPB time, and lowest temperature. Spearman correlation coefficients for the statistically significant correlations are shown in Table 3.
Next, we stratified neonates into those who experienced excessive bleeding versus those who did not. We defined excessive bleeding as those neonates who experienced more than the 75th percentile for 24-hour postoperative CTO. Univariate analysis between the 2 groups is shown in Table 4. We identified multiple demographic and CPB-related differences, including weight, CPB time, aortic cross-clamp time, lowest temperature on CPB, prematurity, and RACHS-1 score. The occurrence of regional perfusion and DHCA and their durations also were greater in neonates who bled excessively (Table 4). Postoperative outcomes were worse in neonates who bled excessively. They received more total donor exposures, had a longer duration of mechanical ventilation, and longer ICU stays (Table 5). Their baseline Cr and Cr at 72 hours after surgery were greater. After adjusting for multiple variables, we found in a logistic regression analysis that neonates in the top 75th percentile had a statistically significant increased incidence of postoperative dialysis and ECMO support (Table 6). Neonates who bled more also tended to have a greater risk of in-hospital mortality (Table 6). RACHS-1 score was a significant predictor of in-hospital mortality (P = 0.03).
To assess our definition of excessive bleeding (more than the 75th percentile for 24-hour postoperative CTO), we performed the same analysis for the 70th and 80th percentiles. When bleeding was defined as more than the 70th percentile, logistic regression analysis found that excessive bleeding was an independent predictor of ECMO (relative risk [RR] 16.0; CI, 4.34–44.98; P = 0.0007) and in-hospital mortality (RR 4.97; CI, 1.66–13.17; P = 0.007). We could not estimate a RR for dialysis because none of the neonates who bled less than the 70th percentile required dialysis. When bleeding was defined as more than the 80th percentile, logistic regression analysis found that excessive bleeding was an independent predictor of dialysis (RR 17.1; CI, 2.18–73.4; P = 0.008), ECMO (RR 12.13; CI, 4.26–35.46; P = 0.0002), and in-hospital mortality (RR 4.2; CI, 1.65–11.56; P = 0.01).
In this study, we examined the relationship between excessive postoperative bleeding and major adverse events in neonates after CPB. We demonstrated that those neonates who bled in the top quartile had a statistically significant increased risk of postoperative dialysis and ECMO support. We also found that neonates who bled the most tended to be at a greater risk for in-hospital mortality. RACHS-1 score however was a statistically significant predictor of in-hospital mortality. In addition, we identified several perioperative variables within the neonatal population that significantly correlated with increased blood loss and transfusion requirements in the immediate post-CPB period. These variables included weight, RACHS-1 score, CPB time, and lowest temperature. Previous studies performed in pediatric patients of all ages undergoing CPB have found similar risk factors (younger age, lower weight, and longer CPB time) to be associated with an increased risk of postoperative bleeding.9–12
Several studies in pediatric cardiac patients undergoing CPB have shown an increased risk of morbidity and mortality in association with excessive blood component transfusion. In a secondary analysis of infants younger than 9 months undergoing 2 ventricular repair without aortic arch obstruction, Kipps et al.3 showed that administration of larger amounts of intraoperative and postoperative blood products was an important risk factor for longer duration of mechanical ventilation. In a retrospective review of 804 pediatric cardiac surgery patients, Salvin et al.4 found that pRBC transfusion was associated with a longer hospital stay and that the strongest association was in the high transfusion group. Another investigation analyzing pRBC transfusions after pediatric heart transplantation found that transfusions were independently associated with an increased inotrope score in the first 24 hours postoperatively and an increased duration of ICU stay.13 The authors also showed that patients with major postoperative adverse events, including dialysis and ECMO support, were administered significantly larger amounts of pRBCs. Finally, in a recent investigation of infants who underwent CPB, the authors reported that early postoperative hemorrhage in the ICU was associated independently with an increased mortality.14 Once again, the risk was greatest in the highest quartile for bleeding. Our study adds to the proliferation of literature supportive of a significant relationship between excessive postoperative bleeding and perioperative morbidity in pediatric cardiac patients after CPB. Our study differs from previous investigations because we strictly focused on the neonatal population. To our knowledge, this is the first study to report on the major adverse consequences associated with excessive postoperative bleeding in a group of neonates undergoing CPB.
Our finding that excessive bleeding and transfusion in the first 24 hours postoperatively are associated with an increased risk of postoperative dialysis and ECMO is important because both of these adverse events lead to a known substantial increase in mortality. Neonates requiring dialysis after cardiac surgery are 6.4 times more likely to die before hospital discharge than those not requiring dialysis.15 Likewise, survival rates for children supported by ECMO for cardiac reasons are lowest in neonatal patients.16 Our data also showed that neonates who bled in the upper quartile had a numerically greater risk of mortality before hospital discharge than those who bled less, although our study was not powered to assess mortality. RACHS-1 score was the only variable significantly associated with an increased risk of in-hospital mortality.
Previous investigations have demonstrated that age and weight are important predictors of postbypass bleeding.9–12 However, no study has evaluated predictors of blood loss exclusively within the neonatal population. Perhaps one reason that most studies include children of all ages is because it is difficult to enroll enough neonatal patients to formulate robust conclusions. Our results indicate that within the neonatal population, age is not an important variable. Weight is the significant variable associated with greater postoperative bleeding, thus suggesting that neonates who are small for gestational age are at an equally increased risk. A combination procedure did not influence postoperative bleeding aside from its effect on CPB time. Neonates who bled excessively, as defined by the top quartile for 24-hour postoperative CTO, displayed multiple factors associated with increased risk of post-CPB bleeding, including higher RACHS-1 score, longer CPB times with lower temperatures, and more regional perfusion and/or DHCA. Clearly, the complexity of the surgery also plays an important role in postoperative bleeding and transfusion. Neonates who bled in the top quartile experienced poorer postoperative outcomes: longer duration of mechanical ventilation, longer ICU stays, and a higher Cr at 72 hours after surgery. As previously discussed, these neonates were also at greatest risk to require postoperative dialysis and ECMO support.
This retrospective study has its limitations. First, although associations were demonstrated, a cause-and-effect relationship cannot be assumed because of potential unmeasured confounding variables. Second, we did not measure specific biomarkers to evaluate the incidence of post-CPB acute kidney injury in these neonates. Instead, we defined renal dysfunction based on changes in Cr values. Neonates undergoing CPB experience significant CPB-induced hemolysis. Given the important role that free hemoglobin plays in the pathogenesis of acute kidney injury, it is likely that, had we measured specific biomarkers, we would have found a greater incidence of post-CPB renal dysfunction. On the other hand, our routine use of pRBCs that are <14 days old was beneficial since the cell membranes of RBCs become increasingly stiff and predisposed to hemolysis with increasing storage times. Third, after the initial transfusion of platelets and cryoprecipitate, transfusion practices varied among practitioners in both the operating room and ICU. We did however demonstrate a positive correlation between CTO and blood product transfusion in the first 24 hours postoperatively, indicating that these 2 variables were congruent. Additionally, because CTO and transfusion are co-related, it is difficult to postulate whether bleeding or transfusion is the true underlying cause of the adverse events. Fourth, in our logistic regression model, we used the actual duration of CPB and not a forecasted estimate or scheduled duration. Because only the latter could be known preoperatively, our results cannot be accurately applied to a prospective intent-to-treat study.17 Finally, it should be noted that our study was not specifically powered to assess mortality despite its strength in representing a relatively large collection of neonates from a single institution using fairly consistent clinical practices. Going forward, we will need larger study populations to accurately assess the risk of mortality related to postoperative bleeding and transfusion in neonates after CPB. This will likely require multi-institutional studies.
In summary, we demonstrated in this retrospective analysis that excessive bleeding in neonates within the first 24 hours after CPB is independently associated with an increased risk of certain major postoperative adverse events, namely dialysis and ECMO support. We also found that neonates who bled within the top quartile had an increased incidence of death before hospital discharge. We hesitate to draw conclusions regarding this association since our study was not powered to address mortality. In addition, we identified several perioperative variables that correlated with postoperative bleeding and blood transfusion requirements including weight, RACHS-1 score, CPB time, and lowest temperature. We anticipate that our results will aid clinicians in predicting major postoperative adverse events after neonatal CPB and in allocating resources necessary to manage these events.
Name: Nina A. Guzzetta, MD, FAAP.
Contribution: This author helped design the study, conduct the study, analyze the data, and write the manuscript.
Attestation: Nina A. Guzzetta has seen the original study data, reviewed the analysis of the data, approved the final manuscript, and is the author responsible for archiving the study files.
Name: Nadine N. Allen, MD.
Contribution: This author helped analyze the data and write the manuscript.
Attestation: Nadine N. Allen has seen the original study data, reviewed the analysis of the data, and approved the final manuscript.
Name: Elizabeth C. Wilson, MD.
Contribution: This author helped design the study, conduct the study, and analyze the data.
Attestation: Elizabeth C. Wilson has seen the original study data, reviewed the analysis of the data, and approved the final manuscript.
Name: Gregory S. Foster, BS.
Contribution: This author helped conduct the study and analyze the data.
Attestation: Gregory S. Foster has seen the original study data, reviewed the analysis of the data, and approved the final manuscript.
Name: Alexandra C. Ehrlich, MPH.
Contribution: This author helped design the study, analyze the data, and write the manuscript.
Attestation: Alexandra C. Ehrlich has seen the original study data, reviewed the analysis of the data, and approved the final manuscript.
Name: Bruce E. Miller, MD.
Contribution: This author helped design the study, analyze the data, and write the manuscript.
Attestation: Bruce E. Miller has seen the original study data, reviewed the analysis of the data, and approved the final manuscript.
This manuscript was handled by: James A. DiNardo, MD.
1. DeFoe GR, Ross CS, Olmstead EM, Surgenor SD, Fillinger MP, Groom RC, Forest RJ, Pieroni JW, Warren CS, Bogosian ME, Krumholz CF, Clark C, Clough RA, Weldner PW, Lahey SJ, Leavitt BJ, Marrin CA, Charlesworth DC, Marshall P, O’Connor GT. Lowest hematocrit on bypass and adverse outcomes associated with coronary artery bypass grafting. Northern New England Cardiovascular Disease Study Group. Ann Thorac Surg. 2001;71:769–76
2. Bhaskar B, Dulhunty J, Mullany DV, Fraser JF. Impact of blood product transfusion on short and long-term survival after cardiac surgery: more evidence. Ann Thorac Surg. 2012;94:460–7
3. Kipps AK, Wypij D, Thiagarajan RR, Bacha EA, Newburger JW. Blood transfusion is associated with prolonged duration of mechanical ventilation in infants undergoing reparative cardiac surgery. Pediatr Crit Care Med. 2011;12:52–6
4. Salvin JW, Scheurer MA, Laussen PC, Wypij D, Polito A, Bacha EA, Pigula FA, McGowan FX, Costello JM, Thiagarajan RR. Blood transfusion after pediatric cardiac surgery is associated with prolonged hospital stay. Ann Thorac Surg. 2011;91:204–10
5. Jenkins KJ, Gauvreau K, Newburger JW, Spray TL, Moller JH, Iezzoni LI. Consensus-based method for risk adjustment for surgery for congenital heart disease. J Thorac Cardiovasc Surg. 2002;123:110–8
6. Karkouti K, Beattie WS, Dattilo KM, McCluskey SA, Ghannam M, Hamdy A, Wijeysundera DN, Fedorko L, Yau TM. A propensity score case-control comparison of aprotinin and tranexamic acid in high-transfusion-risk cardiac surgery. Transfusion. 2006;46:327–38
7. Bellomo R, Ronco C, Kellum JA, Mehta RL, Palevsky PADQI workgroup. . Acute renal failure—definition, outcome measures, animal models, fluid therapy and information technology needs: the second international consensus conference of the Acute Dialysis Quality Initiative (ADQI) group. Critical Care. 2004;8:R204–12
8. Zhang J, Yu KF. What’s the relative risk? A method of correcting the odds ratio in cohort studies of common outcomes. JAMA. 1998;280:1690–1
9. Miller BE, Mochizuki T, Levy JH, Bailey JM, Tosone SR, Tam VK, Kanter KR. Predicting and treating coagulopathies after cardiopulmonary bypass in children. Anesth Analg. 1997;85:1196–202
10. Williams GD, Bratton SL, Riley EC, Ramamoorthy C. Association between age and blood loss in children undergoing open heart operations. Ann Thorac Surg. 1998;66:870–5
11. Williams GD, Bratton SL, Riley EC, Ramamoorthy C. Coagulation tests during cardiopulmonary bypass correlate with blood loss in children undergoing cardiac surgery. J Cardiothorac Vasc Anesth. 1999;13:398–404
12. Williams GD, Bratton SL, Ramamoorthy C. Factors associated with blood loss and blood product transfusions: a multivariate analysis in children after open-heart surgery. Anesth Analg. 1999;89:57–64
13. Howard-Quijano K, Schwarzenberger JC, Scovotti JC, Alejos A, Ngo J, Gornbein J, Mahajan A. Increased red blood cell transfusions are associated with worsening outcomes in pediatric heart transplant patients. Anesth Analg. 2013;116:1295–308
14. Wolf MJ, Maher KO, Kanter KR, Kogon BE, Guzzetta NA, Mahle WT. Early postoperative bleeding is independently associated with increased surgical mortality in infants after cardiopulmonary bypass. J Thorac Cardiovasc Surg. 2014;148:631–6.e1
15. Morgan CJ, Zappitelli M, Robertson CM, Alton GY, Sauve RS, Joffe AR, Ross DB, Rebeyka IMWestern Canadian Complex Pediatric Therapies Follow-Up Group. . Risk factors for and outcomes of acute kidney injury in neonates undergoing complex cardiac surgery. J Pediatr. 2013;162:120–7.e1
16. Brown KL, Ichord R, Marino BS, Thiagarajan RR. Outcomes following extracorporeal membrane oxygenation in children with cardiac disease. Pediatr Crit Care Med. 2013;14:S73–83
17. Dexter F, Dexter EU, Ledolter J. Importance of appropriately modeling procedure and duration in logistic regression studies of perioperative morbidity and mortality. Anesth Analg. 2011;113:1197–201