The majority of pediatric patients undergoing cardiac surgery are exposed to red blood cell (RBC) transfusions. These transfusions are necessary to improve oxygen-carrying capacity, replace operative blood loss, and improve hemostasis.1 Although RBC transfusions are often necessary, they can be associated with complications including infectious risks, transfusion-related acute lung injury, transfusion-related circulatory overload, hemolytic transfusion reactions, and clerical errors.2 RBC transfusions have been shown to be independently associated with worsening clinical outcomes.3–10 In addition, recent studies in pediatric patients have suggested that conservative transfusion practices do not negatively impact postoperative outcomes.11
Research investigating the effects of blood transfusions in pediatric patients has been primarily focused on transfusion exposure in the postoperative care units. However, in pediatric patients undergoing cardiac surgery, including heart transplantation, a large percentage of transfusions occur in the operating room. The cumulative effects of the total amount of blood transfusions on common postoperative outcomes measures (length of stay, inotrope score, duration of tracheal intubation, and major adverse events [MAEs]) in pediatric heart transplant patients have not been investigated. A majority of pediatric transplant patients will require blood transfusions during their entire perioperative course. The larger volume of transfusions these patients receive may be associated with worsening clinical outcomes as compared with the previous studies evaluating the effects of blood transfusions in other pediatric patients.
Pediatric allografts for heart transplantation are a precious resource, and improvements in clinical outcomes could help conserve this resource. Improved understanding of the effects of blood transfusions on pediatric heart transplant patients may alter patient management and affect outcomes.
This study investigated the relationship between the amount of blood transfused and clinical outcomes in pediatric patients undergoing orthotopic heart transplantation at a single major medical center over the last 6 years. We hypothesized that RBC transfusions both during and after pediatric heart transplantation may be associated with an increase in poor clinical outcomes and MAEs.
A retrospective database review was performed for all pediatric patients (≤18 years of age) undergoing orthotopic heart transplantation from January 1, 2004 to December 31, 2010 at Mattel Children’s Hospital in the University of California Los Angeles Medical Center. A comprehensive database was compiled using both paper and electronic records from outpatient and inpatient databases, the United Network for Organ Sharing (UNOS) pediatric heart transplant database, intraoperative perfusion and anesthesia records, laboratory data, and postoperative pediatric cardiac intensive care records. IRB approval was obtained for patient database review.
Data collected included preoperative (pre-op) demographic variables of age (years), gender, weight (kilograms) of the recipient, UNOS status at time of transplant (1A/B), repeat sternotomy (yes/no), presenting diagnosis (categories of 1 = nonrestrictive cardiomyopathy, 2 = noncardiomyopathy congenital heart disease, 3 = restrictive cardiomyopathy, 4 = repeat cardiac transplantation), pretransplant circulatory assist device, ventilator support, renal failure (defined as creatinine clearance <50 mL/dL or requiring dialysis), and most recent pre-op hematocrit (HCT%) and creatinine (mg/dL). The Index for Mortality Prediction After Cardiac Transplantation (IMPACT) score was calculated pre-op for each patient. IMPACT is a comprehensive pre-op risk stratification score that has been shown to predict short-term mortality after pediatric heart transplantation.12 Intraoperative variables collected included warm and cold ischemia time and the amount of RBC transfusions in milliliters per kilogram (mL/kg) given during cardiopulmonary bypass (CPB) and intraoperatively after CPB. The institution’s clinical blood transfusion protocol was used to guide transfusions in all patients. The clinical protocol is to transfuse patients intra- and postoperatively for (1) a HCT <28%, (2) continued bleeding, or (3) evidence of decreased oxygen-carrying capacity, such as demonstrated by a low mixed venous saturation and/or tachycardia. Transfusions during CPB were also guided by a clinical protocol; the pump is primed with 1 unit of packed RBCs for patients weighing <10 kg, or for an expected dilutional HCT of <23. Blood transfusions were given during CPB if the expected HCT coming off bypass would be <28%. All patients received intraoperative antifibrinolytic therapy (aprotinin from January 2004 to November 2007 and then aminocaproic acid from December 2007 to December 2010). Data were collected in the intensive care unit (ICU) for the first 48 postoperative hours. These data included postoperative HCT (%), highest postoperative creatinine (mg/dL), blood transfused (mL/kg), and inotrope score for the first and second 24 hours.
Primary end points were evaluated for the first 30 postoperative days and included length of ICU hospital stay, duration of tracheal intubation, inotrope score, morbidity in the form of MAEs, and mortality. Length of ICU stay was calculated from the date of postoperative admission until discharge from ICU. Duration of tracheal intubation was from the date of postoperative admission until extubation. Inotrope score was calculated using the vasoactive-inotrope score calculation initially described by Wernovsky et al.13 and then expanded by Gaies et al.14 The vasoactive-inotrope score was calculated using the equation dopamine (µk/kg/min) + dobutamine (µg/kg/min) + 100 × epinephrine (µg/kg/min) + 10 × milrinone dose (µk/kg/min). MAEs included postoperative sepsis, extracorporeal membrane oxygenation (ECMO), open chest, acute kidney injury (AKI) requiring dialysis, and graft failure. Postoperative sepsis was defined as hypotension of an infectious etiology requiring inotropic support. ECMO was used as short-term circulatory support in patients unable to be weaned from CPB due to inadequate perfusion or oxygenation. Open chest was defined as any patient who arrived in the pediatric cardiac ICU without closure of the sternum or required reopening of sternum. Acute kidney injury requiring dialysis was defined as any patient needing postoperative dialysis for the first time. Graft failure was defined as any patient requiring retransplantation or a postoperative circulatory assist device as a bridge to retransplantation due to graft dysfunction. Mortality was defined as death within 30 days after transplantation.
Histograms and quartile plots were examined in a bivariate analysis to determine whether distributions of continuous outcomes were reasonably modeled by a normal (Gaussian) distribution on the original or log scale. P values for comparing continuous variables or ordinal variables between groups were computed using the nonparametric Kruskal-Wallis test as most continuous variables did not follow a normal (Gaussian) distribution. For example, the Kruskal-Wallis test was used to compare ICU stay by gender or to compare the total amount of blood transfused by MAE (yes or no). Spline regression methods were used to assess whether the relation between continuous predictors versus continuous outcomes were linear. The nonparametric Spearman correlation was also computed to assess these relations (please refer to Appendices A1, A2, and B).
To perform the multivariate analysis, the relationship between continuous outcomes (length of ICU stay and inotrope score) on the log scale versus log total blood transfused was modeled using linear regression simultaneously controlling for up to 12 potential confounders/covariates (age, weight, gender, diagnosis, UNOS 1A/B status, reoperation, pre-op circulation device assist, pre-op ventilator support, pre-op HCT, pre-op renal failure, and warm and cold ischemia time). The log scale was used since these outcomes conformed to the normal distribution and were more linearly related to the potential predictors on the log scale. Splines were used to determine whether the impact of age, weight, pre-op HCT, and warm or cold ischemia time was linear. The geometric and arithmetic mean ratios were the same. The arithmetic mean ratio based on the regression coefficients is reported in Results section. For example, in the regression model log10(Y) = a + b1 X1 + b2 X2 + e, where the variance of the errors (e) is assumed constant, 10b1 is both the geometric and arithmetic mean ratio (MR) for a 1 unit increase in X1. For example, if b = 0.15, MR = 100.15 = 1.4 implies that, on average, Y is 1.4 times larger for each 1 unit increase in X. The resampling (bootstrap) method was used with 200 replications to estimate the regression coefficients and their standard errors on the log scale. The 95% confidence bounds for the arithmetic MRs were computed using the generalized pivotal method using 20,000 Monte Carlo simulations.15–17 In the simulations performed by Chen and Zhao,17 this method had between 94% and 96% coverage for a nominal 95% confidence interval (CI) (Appendix C).
The simultaneous relation between numbers of days intubated versus log total blood transfused controlling for the other covariates was assessed using ordinal logistic regression with 6 (J = 6) ordered intubation day categories (0, 1, 2, 3, 4–6, and 7 or more days) and a cumulative odds model. The resampling method was used as mentioned earlier. The corresponding odds ratios (ORs) are reported. For example, an OR = 2.4 for predictor X implies that the odds of intubation lasting up to category J + 1 is 2.4 times the odds of only lasting up to intubation level J for a 1-unit increase in X. The Poisson regression model was also examined and shown not to fit the data. For all regression models, the subset of the 12 potential covariates that were simultaneously significant was found by performing a backward stepdown regression where all covariates were initially used as candidates. The sample size did not allow the investigation of interactions, thus we assumed that the effect of up to 12 covariates was additive.
We computed the total amount of blood transfused for each patient and the presence or absence of a MAE to determine the threshold amount of blood transfused that best discriminates between those with MAEs and those without them. The optimal amount of blood was chosen by performing a nonparametric receiver operating characteristic (ROC) curve analysis such that the accuracy was maximized. The accuracy was defined as the average of the sensitivity and specificity (accuracy = 0.5 sensitivity + 0.5 specificity). For the amount of blood transfused, we report the accuracy, sensitivity, and specificity. Statistical analyses were performed with SPSS 16.0 (SPSS Inc., Chicago IL) and SAS 9.2 (SAS Inc, Cary, NC) and STATA 11.2 (STATACorp, College Station, TX) and R 2.15 (Free Software Foundation, Inc., Boston, MA <http://fsf.org/>, http://www.r-project.org). Statistical significance was defined as P ≤ 0.05.
Between 2004 and 2010, 108 pediatric patients underwent orthotopic heart transplantation. Thirteen patients were omitted due to incomplete datasets. One patient was not able to come off CPB and thus did not fit the study inclusion criteria. Data analysis was performed on the final cohort of 94 patients. Demographic data can be seen in Table 1. The median patient age was 11.7 years, weight was 37.4 kg, cold ischemia time was 207.0 minutes, and warm ischemia time was 118.0 minutes. The majority of patients were UNOS status 1A, and there was a slight predominance of male patients. The median IMPACT score, of 50 possible was 5.0 (range 0–26), and the individual variables can be seen in Table 1. The primary diagnosis for heart transplantation can be seen in Table 2. Diagnosis included nonrestrictive cardiomyopathy (n = 50), complex congenital heart disease (n = 28, 8 patients were cyanotic at the time of transplant and 20 were acyanotic), restrictive cardiomyopathy (n = 12), and repeat cardiac transplantation (n = 4).
Red Blood Cell Transfusions
The median pretransplant HCT was 36.1% (19.4%–63.7%) and posttransplant HCT was 28.9% (16.1%–42.7%). The median total blood transfused was 38.7 mL/kg. The majority of patients (88%) were given an RBC transfusion either intraoperative or in the first 48-postoperative hours (Table 3). The distribution of RBC transfusions given during the intraoperative and postoperative course can be seen in Figure 1. The majority of RBC transfusions were given intraoperatively both on and off CPB. The largest amount of blood was transfused in the youngest patients (mL/kg); infants (1–12 months of age): 95.03 ± 53.16, children (≥12 months–144 months): 30.60 ± 27.79, and adolescents (>144 months of age): 9.22 ± 11.43. The largest amount of blood was also transfused in the patients weighing the least (amount transfused in mL/kg); <10 kg: 93.66 ± 48.08, 10–20 kg: 43.84 ± 26.66, 21–30 kg: 16.13 ± 15.03, 31–40 kg: 13.08 ± 5.41, 41–50 kg: 14.56 ± 14.60, 51–60 kg: 8.70 ± 9.97, >60 kg: 3.05 ± 5.32. There was no significant difference (P = 0.22) between total amount of blood transfused in the patients receiving aprotinin (n = 61, median 35.9 ± 91.6) versus aminocaproic acid (n = 33, median 41.0 ± 41.6).
Length of ICU Stay
The median length of ICU stay was 6.0 days. When patients were categorized into transfusion exposure groups (based on prior demonstration of risk stratification in these groups) of none, low (<15 mL/kg), and high (>15 mL/kg), there was an increase in the median length of ICU with increasing transfusion amounts8 (Table 4, Fig. 2).
The bivariate analysis, as seen in Table 5, identified warm ischemia and total RBCs transfused as significant for increasing length of ICU stay. All other variables were not significantly correlated with length of ICU stay. In the multivariate analysis, only the amount of blood transfused was found to be independently associated with increased length of ICU stay (MR = 1.34; CI, 1.03–1.76 per 10-fold increase; P = 0.03). The statistical model demonstrated that, after controlling for the other covariates, a 10-fold increase in total blood transfused was associated with a 1.34-fold increase in the length of ICU stay. This final statistical model accounted for 45% of the variation in ICU stay. The relationship between lengths of stay versus age was quadratic (u-shaped) with the nadir at age 10 and an increase in length of stay at both younger and older extremes of age (Tables 5 and 6). The MR from age 4.1 (1st quartile) to age 11.7 was 1.24 (CI, 1.01–1.62; P = 0.0013).
The median inotrope score for the first 24 hours postoperatively was 17 and for the second 24 hours was 11. When patients were categorized into transfusion exposure groups (based on prior demonstration of risk stratification in these groups) of none, low (<15 mL/kg), and high (>15 mL/kg), there was an increase in the median inotrope score in the first 25 hours, with increasing transfusion amounts8 (Table 4, Fig. 3).
When looking at all covariates one at a time (bivariate), only the total amount of blood transfused and cold ischemia time were found to be significantly associated with inotrope score in the first 24 hours, none of the other factors were significant at P <0.05. Controlling for RBCs, none of the other factors was simultaneously significantly associated with inotrope score in the first 24 hours, including cold ischemia time. Larger RBC transfusions were associated with higher inotrope scores in the first 24 hours, for every 1 log unit (10-fold) increase in total blood transfused; the inotrope score increased an average of 0.1 log units, corresponding to an MR = 100.01 = 1.26 (CI, 1.04–1.52; P = 0.035). Higher weight (MR = 0.99; CI, 0.98–1.0 per kilogram; P = 0.003), age (MR = 0.96; CI, 0.94–0.99 per year; P = 0.01), and increasing RBC transfusions (MR = 1.35; CI, 1.03–1.77; P = 0.027) were bivariately associated with increasing inotrope score in the second 24 hours. The multivariate analysis found that, controlling for weight and pre-op HCT, blood transfusion was still associated with an increased inotrope score in the second 24 hours, but the relationship was not statistically significant (MR = 1.26; CI, 1.05–1.50; P = 0.19) (Tables 5 and 6).
Duration of Intubation
The median number of postoperative days that patients were tracheally intubated and their lungs mechanically ventilated was 1.0. Patients were intubated for a median of 1 day (0–5) in the no transfusion group, 1 day (0–6) low transfusion group, and 1 day (0–77) high transfusion group (Table 4).
Total blood transfused (OR = 2.42; CI, 1.34–4.39 per 10-fold increase; P = 0.005), warm ischemia time, IMPACT score, age, and pre-op ventilator support were found to be bivariately significant with increasing duration of postoperative intubation. All other variables were not significantly associated. The multivariate analysis demonstrated that after controlling for UNOS status (OR = 0.27; CI, 0.08–0.92; P = 0.04), warm ischemia time (OR = 1.67; CI, 0.87–3.21 per hour; P = 0.13), age (OR = 1.02; CI, 1.01–1.03; P = 0.003), and IMPACT score (OR = 1.12; CI, 1.02–1.21; P = 0.02), increasing total RBC transfusions were found to be associated with longer duration of intubation (OR = 1.46; CI, 0.63–3.39 per 10-fold increase; P = 0.37; Tables 4–6).
Major Adverse Events
Postoperative MAEs were seen in 20 patients. MAEs included posttransplant ECMO (8), sepsis (5), open chest (10), AKI requiring dialysis (6), and graft failure (3). The following results can be seen in Figures 4 and 5. Patients suffering from MAEs received significantly larger median amounts of blood transfusions, 81.9 mL/kg (0–580.2) as compared with patients with no MAE, 28.3 mL/kg (0–191.5) (P = 0.002). The ROC analysis showed that the threshold for increased likelihood of MAEs was 60 mL/kg in total RBC transfusions. Using this optimal threshold, the sensitivity is 80%, the specificity is 72%, and the unweighted accuracy is 76%. The area under the ROC curve is 0.772.
The 30-day postoperative mortality was 2% (2 cases). These 2 patients suffered from severe intraoperative hemorrhage and graft failure after CPB. No additional data analysis was run due to the small sample size. A power calculation determined a sample size of 200 would be needed to determine a statistically significant mortality result. Results can be seen in Table 4.
In a cohort of 94 pediatric patients undergoing orthotopic heart transplantation at a major university medical center, 88% of the patients received RBC transfusions either during or after surgery. A major finding of this study is that RBC transfusions were independently associated, in a dose-dependent manner, with an increase in length of ICU stay, postoperative inotrope score, and MAEs. Importantly, we find that patients receiving ≥60 mL/kg in total RBC transfusions were more likely to have a postoperative MAE such as ECMO, sepsis, open chest, AKI requiring dialysis, and graft failure. Other pre-op factors such as age, UNOS status, and ischemia time were not found to be independently associated with the studied outcomes.
The association of RBC transfusions with increased length of ICU stay is indicative of prolonged recovery and leads to increased use of hospital resources.18 Previous studies in both adult and pediatric patients suggest that blood transfusions may be independently associated with worsened outcomes.18–21 Although the effects of blood transfusions have not been studied in a large population of pediatric heart transplant patients, our results support those published in general pediatric patients.
One noteworthy difference in our study, however, is that we investigated the total amount of blood transfused in each pediatric patient, both during surgery and in the first 48 hours after heart transplantation. Two previous studies have investigated the relationship between the total amount of blood transfusion exposure and MAEs in general pediatric cardiac patients. Although Szekely et al.10 found an association between blood transfusions and risk of infection, the study by Cholette et al.22 was inadequately powered to determine clinical outcomes. The majority of the studies investigating effects of blood transfusions have been performed in the postoperative period observing transfusions in the ICU.9,18 We found that more than half of RBC transfusions in our patients were given in the operating room. Salvin et al.8 found that the median blood transfused postoperatively in pediatric cardiac patients was 14.7 mL/kg and those transfusions of >15 mL/kg were associated with increased length of hospital stay. We found that compared with the amount transfused postoperatively, a much larger amount, 38.7 mL/kg (median), was transfused in the operating room. While looking at the association between the total perioperative RBC transfusions and outcomes, we observed not only an increase in the length of ICU stay but also an increase in inotrope score and MAEs. Our study, consistent with previous studies, shows a dose-dependent relationship between worsening outcomes and increasing amounts of blood transfusions.23–25 Given that a significant amount of blood is typically transfused in operating rooms, it is essential to account for its contribution in the total blood transfusion exposure and subsequent outcomes in each patient.
Our study also demonstrated an independent association between escalating amounts of blood transfused and increasing inotrope scores in the first 24 postoperative hours. The inotrope score is a measure of pharmacologic cardiovascular support. Increasing amounts of cardiovascular support after cardiac surgery have been associated with higher likelihood of postoperative complications.14 When analyzing the data by the amount of transfusion received, a significant increase in inotrope score was not seen between the low and high transfusion groups. This may have been due to the higher postoperative pharmacologic cardiovascular support needed in patients receiving fewer blood transfusions. The lack of association seen between transfused blood and inotrope score in the second postoperative 24 hours may have been due to the smaller number of patients requiring blood transfusions and inotropes during that time.
RBC transfusions have been associated with immunomodulation leading to increased morbidity and mortality.1 Studies have proposed that recipient immune response to donor leukocytes in blood transfusions may be associated with worsening outcomes.22,26 However, increased rates of morbidity have been seen even in the setting of leukocyte-depleted blood transfusions suggesting that the immune response to blood transfusions is not completely understood.10,27 Pediatric heart transplant patients are subject to complex immunologic reactions. The immunomodulation due to blood transfusions could have had varying effects on patient outcomes, and therefore, the previous studies in pediatric patients may not be applicable to this patient population. The association found between RBC transfusions and the MAEs of sepsis may be due to immunosuppression from RBC transfusions, though further studies need to confirm the strength of this relationship.
Our study has several limitations. The primary limitations are secondary to the retrospective study design and the limited sample size of our patient population. Although we used a multivariate regression analysis and demonstrated an association between RBC transfusions and increased adverse outcomes, we cannot determine causality. All cofounding risk factors shown in previous studies to be associated with worsening outcomes were included in our multivariate analysis; however, there may be other variables affecting the outcome that were not included. With a retrospective study of this nature, there is always the possibility for confounding variables that are not accounted for, to be related to the described outcomes. As stated by Glance et al.28, “surgical patients with significant intraoperative bleeding are more likely to receive blood transfusions, and patients who bleed are also more likely to have worse outcome.” Furthermore, it is also conceivable that physiologic perturbations that may be seen during blood loss can also serve as potential risk factors in affecting outcomes. Therefore without a prospective randomized study, it cannot be known for certain whether the adverse effects seen were due to increased transfusion amounts or the results of increasing blood loss in the patients needing to receive blood transfusions.
Our sample size of 94 patients is large for a single center pediatric heart transplant study; however, it was not sufficiently powered to show a difference in the incidence of postoperative mortality. There were 13 patients excluded from the study due to incomplete datasets. The data that were missing were largely 30-day follow-up in patients transferred out of our medical center. There is no reason to believe their short-term blood transfusion data would be significantly different than our studied cohort. Although this study investigated the effect of blood transfusions in a single institution, a recent query of the Society of Thoracic Surgeons database published in 2012 showed that the percentage of patients receiving intraoperative blood transfusions in our institution is comparable with the national average (88% vs 83%).a Furthermore, Szeleky et al.10 stated that median amounts of total blood transfused are comparable to the amount transfused in this study (33.2 mL/kg vs 38.7 mL/kg). Even though all transfusions were guided by a standard clinical transfusion protocol for administering RBCs on CPB, intraoperatively, and postoperatively, given the retrospective design, we cannot completely exclude any bias that may have led to transfusions outside the clinical protocol. Recent literature has shown that the age of blood transfused has been associated with increased morbidity and mortality.29,30 Given the retrospective nature of this study, we did not have access to the age of transfused RBCs. Future prospective studies will need to examine the causality between blood transfusions and postoperative outcomes as well as look at the effects of the age of blood transfused in this patient population.
This study demonstrates that larger amounts of RBC transfusions are associated with increases in length of ICU stay, inotrope scores, and postoperative MAEs. Pediatric heart allografts are a limited resource with the need far exceeding the demand. Advancements in both the intra and postoperative care of pediatric heart transplant patients can help improve outcomes and conserve this scarce resource. A significant association was seen between blood transfusions and worsening postoperative outcomes; however, it is not known whether the observed outcomes were due to increasing transfusion amounts or increased surgical bleeding. Future studies should prospectively investigate the role of both blood transfusion and conservation strategies to determine whether restrictive transfusion practices may improve clinical outcomes.
Appendix A1. Histograms and Tables for Outcomes and Red Blood Cell (RBC) Transfused
Length of Stay in ICU
Log10 Length of Stay in ICU
Inotrope at 1st 24 Hours
Log10 Inotrope at 1st 24 Hours
Log10 RBC Transfused
Appendix A2. Normal Quantile Plots for Outcomes and Red Blood Cell (RBC) Transfused
Length of Stay in ICU
Log10 Length of Stay in ICU
Inotrope Score in 1st 24 Hours
Log10 Inotrope Score in 1st 24 Hours
Total RBC Transfused
Log10 Total RBC Transfused
Appendix B. Regression Diagnostics and Residuals
Log10 Length of Stay Model—Linear Regression
Ordered Intubation Duration Categories—Ordinal Logistic Model: Leverage Plot. Standardized residuals not well defined for ordinal outcomes.
Log Inotrope at 24 hours—linear Regression Model.
Name: Kimberly Howard-Quijano, MD.
Contribution: This author helped design and conduct the study, analyze the data, and prepare the manuscript.
Name: Johanna C. Schwarzenberger, MD.
Contribution: This author helped conduct the study, analyze the data, and prepare the manuscript.
Name: Jennifer C. Scovotti, MA.
Contribution: This author helped conduct the study, analyze the data, and prepare the manuscript.
Name: Alexandra Alejos, BA.
Contribution: This author helped conduct the study.
Name: Jason Ngo, BS.
Contribution: This author helped conduct the study.
Name: Jeffrey Gornbein, PhD.
Contribution: This author helped analyze the data and prepare the manuscript.
Name: Aman Mahajan, MD, PhD.
Contribution: This author helped design and conduct the study, analyze the data, and prepare the manuscript.
This manuscript was handled by: Peter J. Davis, MD.
We thank Franklin Dexter, MD, PhD, for his insightful suggestions on the statistical methods used in this study.
a Society of Thoracic Surgeons Database Query. STS Congenital Cardiac Surgery Database of Heart Transplant Procedures, January 2010–June 2011 via University of California at Los Angeles database coordinator, Nancy Satou. Cited Here...
1. Guzzetta NA. Benefits and risks of red blood cell transfusion in pediatric patients undergoing cardiac surgery. Pediatr Anesth. 2010;21:504–11
2. Lavoie J. Blood Transfusion risks and alternative strategies in pediatric patients. Pediatr Anesth. 2011;21:14–24
3. Corwin HL, Gettinger A, Pearl RG, Fink MP, Levy MM, Abraham E, MacIntyre NR, Shabot MM, Duh MS, Shapiro MJ. The CRIT study: anemia and blood transfusion in the critically ill—current clinical practice in the United States. Crit Care Med. 2004;32:39–52
4. Hebert PC, Blajchman MA, Cook DJ, Yetisir E, Wells G, Marshall J, Schweitzer I.. Transfusion Requirements in Critical Care Investigators for the Canadian Critical Care Trials Group. Do blood transfusions improve outcomes related to mechanical ventilation? Chest. 2001;119:1850–7
5. Vincent JL, Baron JF, Reinhart K, Gattinoni L, Thijs L, Webb A, Meier-Hellmann A, Nollet G, Peres-Bota D.. ABC (Anemia and Blood Transfusion in Critical Care) Investigators. Anemia and blood transfusion in critically ill patients. JAMA. 2002;288:1499–507
6. Vincent JL, Sakr Y, Sprung C, Harboe S, Damas PSepsis Occurrence in Acutely Ill Patients (SOAP) Investigators. . Are blood transfusions associated with greater mortality rates? Results of the Sepsis Occurrence in Acutely Ill Patients study. Anesthesiology. 2008;108:31–9
7. 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. 2002;3:269–74
8. Salvin JW, Scheurer MA, Laussen PC, Wypij D, Polito A, Bacha EA, McGowan FX, Costello JM, Thiagarajan RR. Blood transfusion after Pediatric Cardiac Surgery is associated with prolonged hospital stay. Ann Thorac Surg. 2011;91:204–11
9. Kneyber MC, Hersi MI, Twisk JW, Markhorst DG, Plotz FB. Red blood cell transfusion in critically ill children is associated with increased mortality. Intensive Care Med. 2007;33:1414–22
10. Szekely A, Cserep Z, Sapi E, Breuer T, Nagy CA, Vargha P, Hartyanszky I, Szatmari A, Treszel A. Risks and predictors of blood transfusion in pediatric patients undergoing open heart operations. Ann Thorac Surg. 2009;87:187–97
11. Willems A, Harrington K, Lacroix J, Biarent D, Joffe AR, Wensley D, Ducruet T, Hébert PC, Tucci M.TRIPICU investigators; Canadian Critical Care Trials Group; Pediatric Acute Lung Injury and Sepsis Investigators (PALISI) Network. . Comparison of two red-cell transfusion strategies after pediatric cardiac surgery: a subgroup analysis. Crit Care Med. 2010;38:649–56
12. Kilic A, Allen J G, Arnaoutakis, George TJ, Cameron DE, Vricella LA, Weiss ES. Adult-derived Index for Mortality Prediction After Cardiac Transplantation (IMPACT) risk score predicts short-term mortality after pediatric heart transplantation. Ann Thorac Surg. 2012;93:1228–34
13. Wernovsky G, Wypij D, Jonas RA, Mayer JE Jr, Hanley FL, Hickey PR, Walsh AZ, Chang AC, Castañeda AR, Newburger JW, Wessel DL. Postoperative course and hemodynamic profile after the arterial switch operation in neonates and infants: a comparison of low profile cardiopulmonary bypass and circulatory arrest. Circulation. 1998;92:2226–35
14. Gaies MG, Gurney JG, Yen AH, Napoli ML, Gajarski RJ, Ohye RG, Charpie JR, Hirsch JC. Vasoactive-inotropic score as a predictor of morbidity and mortality in infants after cardiopulmonary bypass. Pediatr Crit Care Med. 2010;11:234–38
15. Ledolter J, Dexter F, Epstein RH. Analysis of variance of communication latencies in anesthesia: comparing means of multiple log-normal distributions. Anesth Analg. 2011;113:888–96
16. Iyer H, Patterson D A Recipe for Constructing Generalized Pivotal Quantities and Generalized Confidence Intervals. Technical Report 2–10. 2002 Fort Collins, CO Colorado State University
17. Chen Y-H, Zhou X-H. Interval estimates for the ratio and the difference of two log-normal means. Stat Med. 2006;25:4099–113
18. Murphy GJ, Reeves BC, Rogers CA, Rizvi SI, Culliford L, Angelini GD. Increased mortality, postoperative morbidity, and cost after red blood cell transfusion in patients having cardiac surgery. Circulation. 2007;116:2544–52
19. Bateman ST, Lacroix J, Boven K, Forbes P, Barton R, Thomas NJ, Jacobs B, Markovitz B, Goldstein B, Hanson JH, Li HA, Randolph AG.. Pediatric Acute Lung Injury and Sepsis Investigators Network. Pediatric Acute Lung Injury and Sepsis Investigators Network. Anemia, blood loss and blood transfusions in North American Children in the intensive care unit. Am J Respir Crit Care Med. 2008;178:26–33
20. Hebert PC, Wells G, Blajchman MA, Marshall J, Martin C, Pagliarello G, Tweeddale M, Schweitzer I, Yetisir E. A multicenter randomized, controlled trial of transfusion requirements in critical care. Transfusions Requirements in Critical Care Investigators, Canadian Critical Care Trials Group. Crit Care Med. 1998;26:482–7
21. Newburger JW, Jonas RA, Soul J, Kussman BD, Bellinger DC, Laussen PC, Robertson R, Mayer JE Jr, del Nido PJ, Bacha EA, Forbess JM, Pigula F, Roth SJ, Visconti KJ, du Plessis AJ, Farrell DM, McGrath E, Rappaport LA, Wypij D. Randomized trial of hematocrit 25% vs 35% during hypothermic cardiopulmonary bypass in infant heart surgery. J Thorac Cardiovasc Surg. 2008;135:347–54
22. Cholette JM, Henrichs KF, Alfieris GM, Powers KS, Phipps R, Spinelli SL, Swartz M, Gensini F, Daugherty LE, Nazarian E, Rubenstein JS, Sweeney D, Eaton M, Lerner NB, Blumberg N. Washing red blood cells and platelets transfused in cardiac surgery reduces postoperative inflammation and number of transfusions: Results of a prospective, randomized, controlled clinical trial. Pediatr Crit Care Med. 2012;13:290–9
23. Karkouti K, Wijeysundera DN, Yau TM, Beattie WS, Abdelnaem E, McCluskey SA, Ghannam M, Yeo E, Djaiani G, Karski J. The independent association of massive blood loss with mortality in cardiac surgery. Transfusion. 2004;44:1453–62
24. Hill GE, Frawley WH, Griffith KE, Forestner JE, Minel JP. Allogenic blood transfusion increases the risk of postoperative bacterial infection: a meta-analysis. J Trauma. 2003;54:908–14
25. Costello JM, Graham DA, Morrow DF, Morrow J, Potter-Bynoe G, Sandora TJ, Pigula FA, Laussen PC. Risk factors for surgical site infection after cardiac surgery in children. Ann Thorac Surg. 2010;89:1833–42
26. van de Watering LM, Hermans J, Houbiers JG, van den Broek PJ, Bouter H, Boer F, Harvey MS, Huysmans HA, Brand A. Beneficial effects of leukocyte depletion of transfused blood on postoperative complications in patients undergoing cardiac surgery: a randomized clinical trial. Circulation. 1998;97:562–8
27. Fergusson D, Hebert PC, Lee SK, Walker CR, Barrington KJ, Joseph L, Blajchman MA, Shapiro S. Clinical outcomes following institution of universal leukoreduction of blood transfusions for premature infants. JAMA. 2003;298:1950–6
28. Glance LG, Dick AW, Mukamel DB, Fleming FJ, Zollo RA, Wissler R, Salloum R, Meredith UW, Osler TM. Association between intraoperative blood transfusion and mortality and morbidity in patients undergoing noncardiac surgery. Anesthesiology. 2011;114:283–92
29. Koch CG, Li L, Sessler DI, Figueroa P, Hoeltge GA, Mihaljevic T, Blackstone EH. Duration of red blood cell storage and complications after cardiac surgery. N Engl J Med. 2008;358:1229–39
© 2013 International Anesthesia Research Society
30. Karem O, Tucci M, Bateman ST, Ducruet T, Spinella PC, Randolph AG, Lacroix J. Association between length of storage of red blood cell unit and outcome of critically ill children: a prospective observational study. Crit Care. 2010;14:R57