Neonates constitute one of the most heavily transfused patient groups in the hospital. Neonatal transfusion practices differ substantially from adult and pediatric transfusion practices because of unique differences in neonatal physiology. Neonates have small blood volumes when compared with older children and adults but high blood volume per body weight. Their immature organ system function predisposes them to metabolic derangements from blood products and additive solutions, and to the infectious and immunomodulatory hazards of transfusion such as transfusion-transmitted cytomegalovirus (TT-CMV) infection and transfusion-associated graft versus host disease (TA-GVHD). Neonates undergo rapid growth but have a limited capacity to expand their blood volume. In addition, passive transfer of maternal antibodies to the immunologically naïve newborn creates unique compatibility scenarios not commonly seen in children or adults. Their responses to stresses, including hypothermia, hypovolemia, hypoxia and acidosis are dependent on gestational age, birth weight, and co-morbidities. These considerations necessitate special approaches to transfusion therapy in the neonate.42,56 Hypoglycemia (see Chapter 95) can result from the combination of decreased glucose infusion rates during transfusion and impaired glycogenolysis and gluconeogenesis within the liver of the preterm neonate. Continuous glucose infusion rates of greater than 3 to 4 mg/kg per minute are often required in preterm infants; if maintenance fluids are suspended during transfusion, glucose infusion rates can decrease to approximately 0.2 mg/kg per minute for citrate-phosphate-dextrose-adenine (CPDA-1) preserved red blood cells (RBCs) and 0.5 mg/kg per minute for Adsol (AS-1) preserved RBCs. In a previous report of 31 fresh (<5 days old) small-volume RBC transfusions in 16 preterm infants (mean birth weight and gestational age: 863 grams and 26 weeks, respectively), 15% of infants receiving AS-1 preserved RBCs and 64% of infants receiving CPDA-1 preserved RBCs required supplemental dextrose infusions during the transfusion owing to hypoglycemia (blood glucose <40 mg/dL or symptoms of hypoglycemia).34 Subsequent analysis has shown similar frequency of transfusion-associated hypoglycemia when older (5 to 21 days old) AS-1 preserved units were used. Furthermore, reported incidences of hypoglycemia in neonates either during or after exchange transfusions range from 1.4% to 3.6% with no difference in incidence when group O whole blood (WB) or reconstituted WB was used. Hypoglycemia occurring after exchange transfusion is believed to be caused by intraprocedural hyperglycemia, which causes rebound hypoglycemia from insulin secretion.31 Preventive measures for transfusion-associated hypoglycemia include recognizing those infants at risk for developing glucose homeostatic imbalances, continuing the infusion of maintenance fluids rich in dextrose (albeit at a slower rate) to maintain an adequate glucose infusion rate during simple transfusions, and close monitoring of blood glucose during both small and large volume transfusions. The risk of transfusion-induced hyperkalemia is directly related to the magnitude of the K+ load delivered with RBC transfusion, which depends on the age, plasma volume, and plasma concentration of K+ of the unit, and the rate at which K+ is delivered (potassium infusion rate). Potassium infusion rates greater than 0.01 mEq/kg per minute may pose a potential risk of transfusion-associated hyperkalemia and have been shown to result in cardiac arrhythmia/arrest in neonates.31 As stored red blood cells age, potassium leaks out into the plasma and raises the extracellular potassium level within the component. Units of RBCs in extended-storage media (AS-1, AS-3, AS-5) have a hematocrit (Hct) of approximately 60%, and RBCs stored in CPDA-1 have an Hct of approximately 70%. Furthermore, some pediatric transfusion centers centrifuge RBC aliquots before transfusions for neonates to further reduce the plasma component to 20% (e.g., RBC unit: Hct >80%). After 42 days of storage, K+ levels in the plasma of an AS-1 preserved RBC unit approximate to 0.05 mEq/mL. Therefore, infusing 15 mL/kg over 3 hours would yield a potassium dose 0.33 mEq/kg and a potassium infusion rate of 0.002 mEq/kg per minute (e.g., 1-kg infant receiving 15 mL RBC, 6 mL plasma). Conversely, after 35 days of storage of CPDA-1 preserved RBCs, potassium levels in the plasma approximate to 0.08 mEq/mL. Transfusing 15 mL/kg of the CPDA-1 stored product over 3 hours would yield a potassium dose of 0.36 mEq/kg (e.g., 1-kg infant receiving 15 mL RBC, 4.5 mL plasma) at a potassium infusion rate of 0.002 mEq/kg per minute.31 Given that daily potassium requirement is approximately 2 to 3 mEq/kg, simple RBC transfusions transfused slowly over 2 to 4 hours (2-5 mL/kg per hour) do not pose a threat for hyperkalemia (Table 89-1). TABLE 89-1 Dose and Infusion Rate for Potassium in Small-Volume and Large-Volume Transfusion *AS-1, AS-3 or AS-5 RBC units. †Potassium infusion rate may pose a potential risk of transfusion-associated hyperkalemia, which may result in cardiac arrhythmia/arrest, especially if given through a central line. Data from Fasano RM, Paul WM, Pisciotto P. Complications of neonatal transfusion. In: Popovsky MA, ed. Transfusion reactions. 4th ed, Bethesda, MD: AABB Press; 2012:471-518. It has been shown in multiple studies that transfusing dedicated RBC units to their expiration dates does not cause hyperkalemia even in extremely preterm infants.31,52,94,41,55 Therefore the routine practice of washing older RBCs is unnecessary for most small volume RBC transfusions (10-20 mL/kg) in infants, including those with birth weights less than 1.5 kg.31 Irradiation of RBC components damages the erythrocyte membranes, causing K+ leakage and a linear increase in plasma potassium concentration over time. Irradiation substantially increases plasma K+ compared with nonirradiated RBC components both within the first 24 hours after irradiation and for the life of the RBC component.27,99 Large-volume RBC transfusions (>25 mL/kg), particularly if infused rapidly, may pose a significant risk to the neonate. There have been reports of hyperkalemia-induced electrocardiac abnormalities and cardiac arrest when RBC transfusions (fresh and old) have been administered via rapid infusion (10-20 mL/kg over 10-15 minutes) to neonates with concurrent low cardiac output states when the RBCs were irradiated more than 24 hours prior to infusion and/or when they were given via central line directly into the inferior vena cava.36,10,89,99 Whenever possible, fresh RBC units (<7-10 days) should be issued for large-volume RBC transfusions. When fresh RBC units are unavailable, they should be washed and transfused as soon as possible after washing to minimize K+ reaccumulation. Previously irradiated and stored (≥24 hours) units may have plasma K+ unsafe for large-volume transfusion to neonates, especially if administered rapidly. Therefore, they should be issued immediately post-irradiation, or washed or volume-reduced to remove extracellular K+ that accumulates after processing.42,55 In emergent circumstances of unexpected massive bleeding, when neither fresh nor washed nor volume-reduced RBC units are available, infusion rate should not exceed 0.5 mL/kg per minute. Special considerations should be taken when RBCs need to be transfused rapidly, and should be coordinated with the transfusion center so that proper preparation can be performed to avoid hyperkalemia. In-line potassium adsorption filters have been shown to remove extracellular K+ in stored AS-3 RBC units to minimal levels in vitro; however, their use in neonatal RBC transfusions requires more extensive study.104 Infants, especially premature infants, are particularly susceptible to hypocalcemia (see Chapter 96) within the first week of life, owing to multiple factors. Because of the immaturity in neonatal liver and kidney function, and the low amount of skeletal muscle mass, transfusion of citrate-enriched blood can result in hypocalcemia from citrate toxicity. The amount of citrate infused into a neonate during a small-volume transfusion (10-15 mL/kg) is very unlikely to cause hypocalcemia; however, the citrate load during an exchange transfusion can reach very high levels and lead to symptomatic hypocalcemia. In a retrospective review of 106 infants undergoing 140 exchange transfusions, symptomatic hypocalcemia was one of the most common serious side effects. Eighty-one infants were classified as “healthy” if indication for exchange was solely asymptomatic hyperbilirubinemia; 25 infants were classified as “ill” if co-morbid conditions existed. Incidences of asymptomatic and symptomatic hypocalcemia (defined as irritability, jitteriness, or ECG changes) were 34.6% and 5%, respectively, in the “healthy” infant population, and 40% and 8%, respectively, in the “ill” infant population.40 Because clinical manifestations of hypocalcemia are often subtle and/or variable in premature infants, many recommend monitoring ionized calcium levels and/or QT intervals throughout exchange transfusion procedures, in addition to minimizing potentiating factors such as hypomagnesemia, hyperkalemia, alkalosis, and hypothermia in high-risk (ill) patients. Acute hemolytic transfusion reactions occur when RBCs are transfused to a recipient with preformed antibodies to antigens on the transfused RBCs. Almost all acute hemolytic transfusion reaction fatalities are the result of transfusion of ABO incompatible blood because of clerical errors and misidentification; however, nonimmune causes of acute hemolysis may also occur. These include hemolysis from shear and/or heat stress imposed on erythrocytes by extracorporeal circuits, infusion devices, filters, blood warmers, or phototherapy light exposure. These reactions are characterized by fever, chills, diaphoresis, abdominal pain, hypotension, and hemoglobinuria with potential progression to disseminated intravascular coagulation (DIC) and acute renal failure. When a hemolytic transfusion reaction is suspected, the transfusion should be immediately stopped, blood cultures (from patient and blood component(s)) should be obtained, and the transfusion service should be notified. A clerical check, blood component inspection, post-transfusion hemolysis check, and direct antiglobulin test (DAT) should be completed by the transfusion service. The patient’s hemoglobin/hematocrit, serum bilirubin, lactate dehydrogenase (LDH), and urobilinogen should be monitored and intravenous fluids should be administered to offset hypotension and ensure adequate urine output. Mannitol may be administered to force diuresis, but osmotic diuresis in neonates is controversial because of concerns about alterations in cerebral microcirculation and risk of intraventricular hemorrhage. Although infants less than 4 months old have an absence of A and B hemagglutinins and other RBC alloantibodies, maternal IgG antibodies can cross the placenta, causing hemolysis of transfused RBCs, and, therefore, should be considered when transfusing infants.25 Delayed hemolytic transfusion reactions (DHTRs) occur 3 to 10 days following RBC transfusion and manifest as unexplained anemia, hyperbilirubinemia, and abdominal pain. As with acute hemolytic reactions, the diagnosis is confirmed by a positive DAT, hyperbilirubinemia, and a reduction in hemoglobin. DHTRs are extremely rare in neonates because of the immaturity of the immune system. Even though there have been case reports of anti-E and anti-Kell formation in infants as young as 18 days of life, the majority of reports have supported the infrequency of RBC alloimmunization and DHTRs in infants less than 4 months of age.31,95 Febrile nonhemolytic transfusion reactions (FNHTRs) are characterized by fever, chills, and diaphoresis. These reactions are believed to result from the release of pyrogenic cytokines by leukocytes within the plasma during storage. The incidence of FNHTRs has been decreased dramatically since the implementation of prestorage leukoreduction of RBCs and platelet products in 1999. Whereas FNHTRs occurred in approximately 10% of transfusions in the past, the incidence for all products since the introduction of leukoreduction, is now 0.1% to 3% (approximately 0.2% for prestorage leukoreduction).48,73 When FNHTR is suspected, the transfusion should be stopped and more serious reactions ruled out. A sample of blood from the patient may be sent for DAT, plasma hemoglobin quantification, serum lactate dehydrogenase, and bilirubin level to ensure that the patient is not experiencing a hemolytic transfusion reaction. Bacterial contamination should be assessed via cultures of the transfused product and the patient’s blood; empiric antibiotic therapy may be warranted. Most FNHTRs respond to antipyretics, and meperidine may be used for rigors.45, 85 Allergic transfusion reactions (ATRs) are marked by urticaria and itching, but can include flushing, bronchospasm, and anaphylaxis in severe cases. For mild or localized cases, transfusion can be continued once symptoms have subsided; however, severe allergic reactions (anaphylactoid or anaphylactic reactions) may require treatment with corticosteroids and/or epinephrine. The same blood unit should never be restarted in severe cases, even after symptoms have abated. Leukoreduction does not decrease the incidence of ATRs as it has for FNHTRs.73 Premedication with antihistamines with or without steroids is recommended for ATRs. Because these reactions are caused by an antibody response in a sensitized recipient to soluble plasma proteins within the blood product, washed RBCs and platelets may be used for severe or recurrent ATRs nonresponsive to medication. Severe ATRs leading to anaphylaxis can often be caused by the development of anti-IgA antibodies in recipients who are IgA-deficient. In these instances, IgA-deficient–plasma products may be obtained, but require the use of rare donor registries.98 Transfusion-associated graft versus host disease (TA-GVHD) occurs when an immunosuppressed or immunodeficient patient receives cellular blood products, which possess immunologically competent lymphocytes.53 The transfused donor lymphocytes are able to proliferate and engraft within the immunologic incompetent recipient because the recipient is unable to detect and reject foreign cells. The degree of similarity between HLA antigens of donor and recipient also increases the likelihood of developing TA-GVHD. As an example, in the setting of directed donation from a first-degree relative, donor lymphocyte homozygosity for an HLA haplotype for which the recipient is haploidentical predisposes to recipient tolerance, donor lymphocyte engraftment, and alloreaction leading to TA-GVHD.30 The clinical signs and symptoms of TA-GVHD include fever; generalized, erythematous rash that may progress to desquamation; watery diarrhea; mild hepatitis to fulminant liver failure; respiratory distress; and pancytopenia, which is usually severe because hematopoietic progenitor cells are preferentially affected. Onset is traditionally 3 to 30 days following transfusion of lymphocyte-replete cellular blood components in older children and adults; however, there may be a longer latency period before the onset of clinical manifestations of TA-GVHD and a longer course of disease before death in neonates. In a literature review of 27 cases of TA-GVHD in neonates in Japan, the median interval of clinical manifestations was 28 (fever), 30 (rash), and 43 (leukopenia) days, and death occurred in all affected patients at a median interval of 51 days. Prolonged latency of clinical manifestations and death is believed to result from thymic and/or extrathymic semi-tolerance for allogeneic cytotoxic T lymphocytes.71 Neonates at “high risk” for TA-GVHD include those with impaired cellular immunity, such as severe combined immunodeficiency or Wiskott-Aldrich syndrome, those receiving intrauterine transfusions and/or neonatal exchange transfusion, and those receiving cellular blood components from family members or those who are genetically similar to the recipient. Extremely premature neonates are also considered by many to be at significant risk for TA-GVHD.93,40 No effective therapy is available for TA-GVHD, and owing to bone marrow hypoplasia, the mortality rate is 90% in the pediatric population. Fortunately, this complication can be prevented by pretransfusion gamma irradiation of cellular blood components at a dose of 2.5 Gy, which effectively abolishes lymphocyte proliferation.53 The shelf life of irradiated red blood cells is 28 days; however, no data currently exist on the safety in the neonatal population of gamma-irradiated RBCs that are stored for this amount of time. Because potassium and free hemoglobin increase after irradiation and storage of RBCs, it is preferable to irradiate cellular blood products close to administration time for neonates, who may not be able to tolerate high potassium loads.26 There exists no “standard of care” regarding irradiation of blood products for otherwise non–high-risk infants. Many transfusion services irradiate all cellular blood products given to preterm infants born weighing 1.0 to 1.2 kg or less, whereas some irradiate all cellular blood products for infants less than 4 months of age, citing the lack of clinical studies on the incidence of TA-GVHD in the neonatal population and the concern for failure to recognize an infant with an undiagnosed congenital immunodeficiency. When an infant requires irradiated blood components, all cellular blood components for that infant should be irradiated; however, it is not necessary to irradiate acellular blood components, such as fresh frozen plasma (FFP). The known and presumed indications for irradiation of blood components for neonates are listed in Box 89-1. Transfusion-related acute lung injury (TRALI) is an uncommon, potentially fatal acute immune-related transfusion reaction, which typically occurs within 4 hours of transfusion and presents with respiratory distress caused by noncardiogenic pulmonary edema (normal central venous pressure and pulmonary capillary wedge pressure), hypotension, fever, and severe hypoxemia. Differentiation from transfusion-associated circulatory overload (TACO), an acute, nonimmune transfusion reaction that presents with respiratory distress, cardiogenic pulmonary edema, and hypertension caused by volume overload, is important because treatments differ. Furthermore, transient leukopenia, which is commonly seen with TRALI but is absent in TACO, can aid in differentiation of these reactions. Symptoms of TRALI usually improve within 48 to 96 hours; however, three fourths of patients require aggressive respiratory support. Treatment is mainly supportive, including fluid and/or vasopressor support in the face of hypotension. Whereas aggressive diuresis is often required in TACO, this should be avoided in TRALI.51 Although the exact mechanism of TRALI remains uncertain, it is generally believed to be caused by the passive transmission of HLA or neutrophil antibodies directed against recipient leukocyte antigens. These antibodies activate and sequester recipient neutrophils within the endothelium of the lungs, ultimately leading to the production of vasoactive mediators and capillary leak. Plasma products (FFP, apheresis platelets) account for the majority of severe TRALI cases, and multiparous women are the most commonly implicated donors.51,83 Because of these findings, various preventative measures have been adopted in the United States and elsewhere. These include the use of male-only, high-volume plasma products (FFP, platelets), or the selection of donor products from donors who have a low likelihood of being alloimmunized via pregnancy or prior transfusions. Although there have been no definitive cases of TRALI documented in the neonatal population, TRALI has been well documented in the pediatric population. A case has been reported of a 4-month-old girl who experienced respiratory distress, hypoxemia, hypotension, and fever within 2 hours of completion of an RBC transfusion from her mother. HLA antibodies were identified in the mother’s serum, demonstrating the possible role of HLA antibodies in the pathogenesis of TRALI in the setting of a designated blood transfusion between mother and infant.83,105 “T-activation” is a phenomenon that can cause immune-mediated hemolysis in neonates, ranging from minor to fulminant and fatal. Removal of N-acetyl neuraminic (sialic) acid residues from the O-linked oligosaccharides on glycophorins (A, B, and C) on RBC membranes by neuraminidase-producing bacteria, particularly Clostridiam bacteroides and Streptococcus pneumoniae results in exposure of the normally masked Thompsen-Friedenreich (T) cryptantigen on the RBC surface of the neonate. Transfusion of adult blood products containing plasma with naturally occurring anti-T antibodies into neonates with T-activation can present with intravascular hemolysis following transfusion, or unexplained failure to achieve the expected post-transfusion hemoglobin increment. Alternatively, T-activation may be detected in the laboratory without any evidence of clinical hemolysis, making broad-based screening impractical. T-activation has been reported mainly in neonates with necrotizing enterocolitis, especially in those with severe disease requiring surgical intervention but also in septic infants with other surgical problems.79 The incidence of T-activation has been reported in 9% to 27% of infants with necrotizing enterocolitis (NEC).79 Williams et al.102 reported a frequency of T-antigen activation as a laboratory phenomenon of 0.6% of all infants (N = 1672) admitted for intensive care over a 3-year period, but the frequency increased to 11% in infants with NEC and 28% in those with NEC who required surgical intervention. In another study of 62 infants with suspected NEC, 17 (27%) had evidence of T-antigen activation. Those infants with T-activation had Clostridia cultured from blood, peritoneal fluid, or stool in 14 of 16 (88%) of cases and were more likely to have intestinal perforation at laparoscopy.50 In a recent retrospective report, laboratory evidence of T-antigen activation was detected by lectin agglutination testing in 9% of 43 Taiwanese infants with stage II/III NEC, which did not result in hemolysis regardless of whether washed or unwashed components were administered, or contribute to an increase in NEC-related mortality.100 Others have reported that transfusion-associated hemolysis is often not seen in T-activated infants when transfused with donor anti–T-containing serum. However, others have shown that infants with confirmed NEC with laboratory evidence of T-activation had more frequent and severe hemolysis, hyperkalemia, renal impairment, and hypotension than those without T-activation.72 Routine cross-matching techniques may not detect the polyagglutination owing to T-activation when monoclonal ABO antiserum is used. Minor cross-matching of neonatal T-activated red blood cells with donor anti–T-containing serum may show agglutination, but this is not performed routinely. Infants with discrepancies in forward and reverse blood typing and evidence of hemolysis on smear should be suspected of T-activation. The diagnosis is confirmed by specific agglutination tests using peanut lectin Arachis hypogea and Glycine soja. Further hemolysis may be prevented by using washed RBCs and platelets, and low-titer anti-T plasma if available.31 Exchange transfusion with plasma-reduced components may be necessary for infants with severe ongoing hemolysis. Alternatively, hemolysis associated with T-activation may occur independently of transfusion with anti–T-containing serum, and may be the result of sepsis or disseminated intravascular coagulation and cannot be prevented. Necrotizing enterocolitis is a devastating gastrointestinal emergency that primarily afflicts premature infants but can occur in at-risk term gestation infants as well. Severity and treatment can range from full recovery with conservative medical management to intestinal perforation and bowel necrosis requiring surgical intervention. NEC-related morbidity and mortality remains quite high, with up to 30% of affected infants succumbing to the disease, many with an acute presentation followed by rapid deterioration and death. For full description of clinical presentations, pathophysiology, diagnosis, and treatment of NEC, refer to Chapter 94. Although the exact mechanism of NEC is not completely elucidated, ischemic insult to the gastrointestinal tract has been proposed as one major contributor to NEC. Some hypothesize that even subtle reductions in blood flow and subsequent reperfusion occurring in response to hypoxia may contribute to bowel injury. It has been demonstrated that high vascular resistance of the superior mesenteric artery determined by Doppler flow velocity on the first day of life in preterm infants is associated with an increased risk of NEC.63 Furthermore, the stressed newborn has been shown to exhibit abnormal regulation of intestinal vasomotor mediators (i.e., nitric oxide, endothelium, substance P, norepinephrine, and angiotensin), resulting in compromised intestinal flow.18,68 It has been proposed that anemia may impair blood flow to the intestine, with subsequent transfusion of RBCs leading to reperfusion injury. This phenomenon may be exacerbated by abnormalities associated with stored RBCs, including reduced erythrocyte deformability, increased erythrocyte adhesion/aggregation, and decreased nitric oxide stores.80 This last effect may predispose to vasoconstriction and further ischemic insult owing to transfused RBCs acting as a nitric oxide sink within the intestinal microvasculature.21 The use of near-infrared spectroscopy has gained popularity in recent years with its ability to measure regional tissue saturations. A prospective study to examine potential alterations in mesenteric tissue oxygenation demonstrated wide fluctuations and decreases in mesenteric oxygenation patterns in infants who developed transfusion-related NEC.59 Proposed, but unproven, pathologic mechanisms for transfusion-associated NEC include: a transfusion immunologic injury to the intestine similar to what is seen in transfusion-related acute lung injury (TRALI), and/or transfusion-related ischemia/reperfusion injury as described previously. Several retrospective studies have demonstrated a temporal association between RBC transfusion and neonatal NEC. They report that 25% to 38% of NEC cases occur within 48 hours of RBC transfusion and that the risk of transfusion-associated NEC increases with decreasing gestational age of the infant.57,44,14,75 Neonates with transfusion-associated NEC were generally older than those with NEC unrelated to transfusions (3-5 weeks versus 1-3 weeks of age). In one study, all of the neonates in the transfusion-associated NEC group were stable, growing, not ventilated, receiving full enteral feedings, and had no other active medical problems except anemia (hematocrit 24 ± 3%), in contrast to the non-transfusion NEC group, who were more often ventilated, receiving less than 50% of fluids orally, and were transfused for an average hematocrit of 37 ± 7%.57 Blau et al.14 found a convergence of transfusion-associated NEC at 31 weeks’ gestation, the age of presentation of O2 toxicity and other neovascularization syndromes. In another retrospective report, Singh et al.88 displayed a strong association of transfusion within 24 hours and NEC (OR = 7.60, p = .001), a significant albeit decreased association for transfusion within 48 hours (OR = 5.55, p = .001), and a statistically insignificant (absent) association for transfusion within 96 hours (OR = 2.13, p = .07) in their multivariate analysis. Although attempts were made to minimize the confounding effects of multiple variables, the effect of infants’ nadir hematocrit level on the risk of developing NEC remained statistically significant, making it impossible to separate the influence of hematocrit and RBC transfusion. However, Bednarek et al.11 reported that there was no significant difference in the incidence of NEC between high versus low hematocrit threshold transfusion practices among six NICUs, and the Premature Infants in Need of Transfusion (PINT) trial did not show a difference in the incidence of NEC between the low versus high hematocrit transfusion threshold groups.49 Despite the temporal relationship observed between RBC transfusion and neonatal NEC, it remains unclear whether RBC transfusion contributes to the pathogenesis of the disease or whether transfusion is an epiphenomenon. Further study is warranted to determine if RBC transfusions directly contribute to the pathogenesis of NEC. A newly postulated risk of transfusions among VLBW neonates has emerged based on an association between RBC transfusion and the development of, and progression to, severe intraventricular hemorrhage (IVH), which has been observed in retrospective case control reports. In a six-site prospective study, Bednarek et al. found a trend toward a higher incidence of severe IVH (adjusted for birth weight and illness severity) in NICUs administering RBC transfusions “liberally,” compared with other NICUs using transfusions in a more restricted fashion.11 Subsequently, a retrospective analysis of 155 VLBW neonates with normal head ultrasounds during the first week of life indicated an association between RBC transfusion and the risk of developing a severe IVH within the first month, independent of hemoglobin level, initial pH, sepsis, ventilation, coagulation studies, or severe thrombocytopenia. During the first 72 hours of life and when the head ultrasound was normal, 67% of those neonates who later developed severe IVH versus 31% who did not develop IVH, received one or more RBC transfusions (p < .001).8 Additionally, each subsequent RBC transfusion during the first week was determined to double the risk of a severe IVH (each transfusion increased relative risk 2.02; 95% CI 1.54-3.33). In another report by the same investigators, Baer et al. retrospectively compared 55 VLBW neonates in whom a grade I IVH evolved into a grade II or higher IVH, to 362 VLBW neonates matched for demographic, level of illness, and coagulation parameters, who had a grade I IVH that resolved completely with no extension to a higher grade. On logistic (and Lasso-fit) regression analysis, an association was found between RBC transfusion up to and on the day of the grade I IVH and extension of the IVH (OR 2.92; 95% CI 2.19-3.90).9 However, these associations do not necessarily indicate a cause-and-effect relationship because of inherent limitations in retrospective studies such as failure to recognize confounding variables and susceptibility to bias. Furthermore, no difference in risk of IVH (all grades, or ≥grade III) was seen in those infants randomized to liberal versus restrictive transfusion practices in either the multi-institutional Canadian PINT study49 or the Bell study.12 Additional studies are needed to determine if early RBC transfusion plays any pathogenic role in IVH development or extension. Many infectious agents can be transmitted by blood or blood component transfusion (Table 89-2). These include viruses, bacteria, protozoa, and other pathogens. Current transfusion-transmitted disease testing for allogeneic blood donation includes hepatitis B virus surface antigen (HBsAg), hepatitis B core antibody (anti-HBc), anti-hepatitis C antibody (anti-HCV), antibody to HIV-1 and HIV-2 (anti-HIV-1/2), antibody to human T-lymphotropic virus (HTLV-I), HTLV-II, serology for syphilis and Trypanosoma cruzi, and nucleic acid testing (NAT) for HIV-1, HIV-2, HCV, HBV, and West Nile virus (WNV).33 TABLE 89-2 Transfusion-Transmitted Infections5,19,35 *Infection risk cited prior to initiation of donor hepatitis B virus NAT testing. Infection risk post initiation of donor hepatitis B virus NAT is not currently known. †May be as high as 1 in 1800 in highly endemic areas (northeast USA). The prevalence of human cytomegalovirus (CMV) is 30% to 70% in blood donors and varies based on demographic differences within areas of the United States. This DNA virus remains latent within the leukocytes of immune persons and can be transmitted by transfusion of cellular blood components into seronegative recipients. Primary CMV infection occurs in a seronegative recipient from a blood component from a donor who has either active or latent infection. There is wide variation in clinical sequelae from transfusion-transmitted CMV (TT-CMV), ranging from asymptomatic serological conversion, to significant morbidity and mortality from CMV-related pneumonia, cytopenias, and hepatic dysfunction. Premature, seronegative neonates less than 1250 grams, fetuses receiving intrauterine transfusions, severely immunocompromised individuals, and recipients of hematopoietic stem cell and solid-organ transplants are recipient groups at increased risk for post-transfusion CMV-related morbidity and mortality.56 Infants born to seropositive mothers apparently have decreased risk for acquiring TT-CMV, but perinatal infection with a different strain of CMV has been reported infrequently. Because the prevalence of CMV seropositivity among blood donors limits the availability of seronegative components, supplementary strategies can be used to minimize the risk of TT-CMV infection in high-risk infants. Use of third-generation leukoreduction filters at the time of collection of cellular products has been recommended for recipients at risk for TT-CMV. The American Association of Blood Banks states that “leukoreduced” blood products must contain fewer than 5 × 106 total WBCs per unit. Current third-generation leukocyte reduction filters consistently provide WBC reduction in accordance with these standards, with some filters yielding less than 1 × 106 per product. European standards maintain a more stringent definition of leukoreduced as less than 1 × 106 total WBCs per unit.56 Leukocyte reduction has been shown to be effective in preventing CMV infection in neonates, among other groups; however, whether leukocyte reduction is as efficacious as the use of CMV-seronegative blood has been disputed widely. In one study, equivalent rates of post-transfusion CMV infection in allogeneic hematopoietic stem cell transplant patients were found for CMV-seronegative units and leukoreduced units (1.4% vs. 2.4%, respectively).16 A subsequent study demonstrated similar rates of TT-CMV for leukoreduced and CMV-seronegative platelet products, but not for RBC products.66 More recently, a prospective observational study of 23 allogeneic HSCT in CMV-negative donor-patient paired individuals transfused a total of 1847 leukoreduced cellular blood products from CMV “untested” donors and demonstrated no CMV DNA within the blood or CMV-associated clinical complications in the patients greater than 100 days post-transplant despite anti-CMV IgG seroconversion in 17 of 23 patients.97 Although these reports support the notion that leukoreduced blood products are “CMV safe,” no formal consensus on the debate of equivalency has been developed, leading many to advise against the elimination of “dual inventories” of blood products for CMV-seronegative and seropositive units. Nonetheless, variable strategies for preventing TT-CMV currently exist depending on the number of high-risk patients treated at a given center, the regional donor demographics, and product availability. Furthermore, an ongoing prospective birth cohort study is under way to estimate the efficacy of combined CMV-negative, leukoreduced blood product support in preventing TT-CMV in preterm infants in the NICU.43
Blood Component Therapy for the Neonate
Special Considerations for Transfusion Therapy in the Neonate
Risks of Transfusion Therapy
Metabolic Complications
Hypoglycemia
Hyperkalemia
CPDA-1
(35-Day)
Additive Solution*
(42-Day)
Additive Solution*
(20-Day)
Unit hematocrit
70%-75%
57%-60%
57%-60%
Plasma (K+)
.080 mEq/mL
.050 mEq/mL
.030 mEq/mL
15 mL/kg over 3 hours (0.08 mL/kg/min)
K+ dose
K+ infusion rate
0.36 mEq/kg
0.002 mEq/kg/min
0.33 mEq/kg
0.002 mEq/kg/min
0.20 mEq/kg
0.001 mEq/kg/min
25 mL/kg over 3 hours (0.14 mL/kg/min)
K+ dose
K+ infusion rate
0.60 mEq/kg
0.003 mEq/kg/min
0.54 mEq/kg
0.003 mEq/kg/min
0.32 mEq/kg
0.002 mEq/kg/min
25 mL/kg over 1 hour (0.42 mL/kg/min)
K+ dose
K+ infusion rate
0.60 mEq/kg
0.01 mEq/kg/min†
0.54 mEq/kg
0.009 mEq/kg/min
0.32 mEq/kg
0.005 mEq/kg/min
25 mL/kg over 30 min (0.84 mL/kg/min)
K+ dose
K+ infusion rate†
0.60 mEq/kg
0.02 mEq/kg/min†
0.54 mEq/kg
0.018 mEq/kg/min†
0.32 mEq/kg
0.011 mEq/kg/min†
Hypocalcemia
Immunologic Complications
Transfusion Reactions
Hemolytic Transfusion Reactions.
Febrile Nonhemolytic Transfusion Reactions.
Allergic Transfusion Reactions.
Transfusion-Associated Graft Versus Host Disease.
Transfusion-Related Acute Lung Injury.
T-Antigen Activation.
Transfusion-Related Necrotizing Enterocolitis
Transfusion-Related Intraventricular Hemorrhage
Infectious Complications
Infectious Agent
Infectious Risk
Bacterial contamination
Platelets
1 in 5,000
RBCs
1 in 5 million
Hepatitis A
1 in 10 million
Hepatitis B
1 in 250,000 to 350,000*
Hepatitis C
1 in 1.8 million
HIV
1 in 2.3 million
Cytomegalovirus
Unknown
Human T-lymphotropic virus
1 in 641,000 to 921,000
West Nile virus
Extremely low
Chagas disease (T. cruzi)
Extremely low
Syphilis
Virtually nonexistent
Malaria
1 in 4 million
Babesiosis (Babesia microti)
Extremely low†
Creutzfeldt-Jakob disease
No cases reported in US
Cytomegalovirus Infection
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree