INTRODUCTION — Red blood cell (RBC) transfusion provides an immediate increase in oxygen delivery to tissues and is an effective and rapid intervention to treat significant acute anemia. RBC transfusions may reduce the morbidity associated with chronic anemia, such as hemolytic anemia due to Rh or ABO incompatibility and can be lifesaving in neonates with severe blood loss.
The indications for neonatal RBC transfusions and the selection of the proper blood product will be reviewed here. Anemia of prematurity and indications for red cell transfusion in infants and children are discussed separately. (See "Anemia of prematurity (AOP)" and "Red blood cell transfusion in infants and children: Indications".)
INDICATIONS
Overview — Symptomatic anemia occurs when the RBC mass is not adequate to meet the oxygen demands of tissues. Signs that may reflect symptomatic anemia include increased resting heart rate, acidosis, poor growth, decreased energy to nipple feed, apnea, and need for increased respiratory support.
Oxygen supply to tissues is defined as the product of cardiac output and arterial oxygen content. The arterial oxygen content is dependent upon the hemoglobin concentration, arterial oxygen saturation, oxygen carrying capacity of hemoglobin (1.34 mL/gm hemoglobin), and, to a minor extent, the solubility of oxygen. Oxygen supply is increased by increasing cardiac output, arterial oxygen saturation, or hemoglobin concentration.
The indications for neonatal RBC transfusions differ based upon the rate of fall in hemoglobin (acute versus chronic anemia). The need for transfusion in an infant with acute blood loss is generally dependent upon persistent clinical signs of inadequate oxygen delivery following intravascular volume restoration. In chronic anemia, the need for RBC transfusion is, likewise, based upon clinical signs of inadequate oxygen delivery (eg, acidosis) and the degree of respiratory support needed by the infant. Although target hemoglobin and hematocrit (HCT) (values below which a RBC transfusion is administered) have been used as clinical indicators for RBC transfusion, it remains uncertain what target HCT or hemoglobin will optimally balance the risk and benefits of this intervention.
Transfusions are a temporary measure and have the disadvantages of further inhibiting erythropoiesis and being associated with risks of infection [1], graft-versus-host disease, transfusion related acute lung injury (TRALI), transfusion associated circulatory overload (TACO), and toxic effects of anticoagulants or preservatives. RBC transfusions have been associated with an increased risk of death, necrotizing enterocolitis (NEC), extension of intraventricular hemorrhage (IVH), retinopathy of prematurity (ROP), or transient increase in respiratory support; however, high quality evidence for a causal relationship with any of these adverse events is lacking [2-6]. There is also observational data that suggest increased number and volume of transfusions are associated with worse developmental outcomes that need to be confirmed prospectively [7,8]. All of these potential complications underscore the need to carefully evaluate a neonate's ability to deliver oxygen to tissues prior to ordering a nonemergent transfusion, and to document benefit following the transfusion. (See "Transfusion-associated graft-versus-host disease" and "Immunologic transfusion reactions" and "Neonatal necrotizing enterocolitis: Pathology and pathogenesis", section on 'Anemia and RBC transfusion'.)
One of the major challenges in deciding when to transfuse is distinguishing the neonate who is anemic and requires an RBC transfusion from one who has adapted to a low HCT, and is best treated conservatively to avoid the associated risks of transfusion.
Determining optimal HCT levels for neonates and young infants has been an active area of investigation. Various factors influence the threshold, including acuity of blood loss, gestational age (GA), severity of illness, need for respiratory support, or if the infant will undergo major surgery within 72 hours (table 1).
The diagnostic approach to anemia in newborns is discussed separately (algorithm 1). (See "Approach to the child with anemia", section on 'Age of patient'.)
Acute blood loss — Neonates with significant acute blood loss require immediate fluid resuscitation, but may or may not require a RBC transfusion. For example, some term infants may tolerate perinatal blood loss up to one-third of their total blood volume. Infants with a hemoglobin level ≥10 gm/dL following volume expansion may have adequate oxygen delivery and generally only require iron supplementation to replace iron losses due to the hemorrhage.
Indications for a RBC transfusion in a term or preterm neonate after volume resuscitation with restoration of the effective circulating volume following an acute blood loss include one or more of the following:
●>20 percent blood loss
●10 to 20 percent blood loss with evidence of inadequate oxygen delivery, such as persistent acidosis
●Ongoing hemorrhage
Percent of external blood loss is usually determined by dividing the volume of blood loss (measuring the difference in weight of bandages, gauze, and blankets before and after surgery [blood soaked]) by the volume of the infant's total blood volume (85 mL/kg). However, internal blood loss is more difficult to estimate; some calculations have been determined based on clinical context. For example, the volume loss due to a subgaleal hemorrhage can be estimated at 38 mL for every 1 cm increase in head circumference.
Prior to transfusion, it is critical to determine whether a significantly low HCT of an infant at birth is due to an acute fall near the time of delivery or to a chronic in utero process, because this may alter the type of transfusion administered (ie, simple versus exchange). For example, in twin-twin transfusion syndrome or feto-maternal hemorrhage, the degree and timing of bleeding is variable. Bleeding may occur just before delivery or might have begun in the second trimester and be long-standing at the time of delivery, both of which may result in an HCT below 20 percent at birth. In the latter case, a partial exchange transfusion should be considered in an infant in whom an increase in tissue oxygen delivery is necessary. A simple transfusion may cause a significant increase in blood volume, resulting in heart failure, because these infants often have elevated circulating blood volumes. Exchange transfusions should be considered in infants with severe hemolytic anemia. (See "Alloimmune hemolytic disease of the newborn: Postnatal diagnosis and management", section on 'Postnatal management'.)
Chronic anemia — Newborn infants have a gradual fall in their HCT after birth due to decreased production of erythropoietin (EPO), referred to as physiologic anemia. In preterm infants, this decline occurs earlier and is more pronounced in its severity than in term infants. This hyporegenerative anemia associated with low EPO concentrations is defined as anemia of prematurity. (See "Anemia of prematurity (AOP)", section on 'Pathogenesis'.)
In addition, infants cared for in the neonatal intensive care unit (NICU) are at risk for developing chronic anemia because of blood loss from phlebotomy. This is especially true for extremely low birth weight (ELBW) infants (BW <1000 g) who are often more severely ill and require frequent blood tests, and who have a lower total volume of blood than more mature, larger neonates [9]. In addition, the RBCs circulating in preterm infants have a reduced life span compared with those in full term infants, which contributes to the anemia of prematurity.
Target hemoglobin or hematocrit — Results of clinical trials in preterm infants, including ELBW infants, have consistently reported that a restrictive (low) transfusion threshold compared with a liberal (high) threshold does not increase mortality or serious morbidity, and reduces exposures to transfusion [10-15]. We use a restrictive approach for neonatal transfusions (table 1) that is based upon HCT triggers and the respiratory support/illness severity of the infant, supported by data from a metanalysis from 2011 and 2012 [10,12] and the subsequent following largest to date clinical trials:
●In a large multicenter study of 1013 ELBW infants (ETTNO study), the primary outcome of death or neurodevelopmental impairment (CP, cognitive deficit defined as a Bayley II MDI score <85, or severe visual or hearing impairment) at 24 months corrected age (CA) was similar between the restrictive versus liberal threshold groups (42.9 versus 44.4 percent, odds ratio 0.97, 95% CI 0.78-1.22) [14]. The risk of secondary outcomes was also similar between the restrictive and liberal threshold groups including death alone (9 versus 8.3 percent), cognitive deficit (34 versus 38 percent) and CP (6 versus 4 percent) at 24 months corrected age, necrotizing enterocolitis (NEC) requiring surgery (5 versus 4 percent), BPD (26 versus 28 percent), treatment for ROP (8 versus 9 percent), and growth.
The restrictive threshold group had a lower incidence of any transfusion (60 versus 79 percent), lower cumulative median volume of transfused blood (19 versus 40 mL), and lower weekly mean hematocrit values that were 3 percentage points lower than the liberal threshold group. In this clinical trial, threshold for both restrictive and liberal targets were based on gestational age and whether the infant was critically ill (eg, requiring respiratory support, sepsis, NEC requiring inotropic/vasopressor support).
●In another large, randomized multicenter trial of 1824 ELBW infants performed by the NICHD Neonatal Research Network (TOP trial), the primary outcome of death or neurodevelopmental impairment (moderate or severe CP, severe vision or hearing loss, cognitive delay (defined as a composite cognitive score <85 on the Bayley III Scales of Infant and Toddler Development at 22 to 26 months corrected age) was similar between the low and high target threshold groups (49.8 versus 50.1 percent , adjusted relative risk 1.00, 95% CI 0.90-1.08) [15]. At two years of age, death rates were similar for low and high target groups (15 versus 16.2 percent), as were the rates for neurodevelopmental impairment (40.3 versus 39.6 percent) and serious adverse events (21.7 versus 22.7 percent). During the treatment period (until infants reached 36 weeks postmenstrual age or were discharged), the mean hemoglobin level was 1.9 g/dL lower in the low versus high target threshold groups. The low threshold group also received fewer transfusions (mean transfusions 4.4 versus 6.2).
In this study, transfusion criteria were set by protocols based on surveyed clinical practice at each site. Blinding was not feasible at the bedside and more transfusions not based on protocol were administered in the low target threshold group; however, a significant difference in mean hemoglobin level was still achieved between the two groups.
In developed countries, the number of transfusions given to neonates has decreased from an average of seven transfusions in the late 1980s to two transfusions per infant in 2009 during their initial birth hospitalization due to a more restrictive approach to transfusions [16,17]. Despite this decrease, most ELBW infants still receive at least one blood transfusion [14].
Preoperative target — It remains uncertain what the optimal preoperative target hemoglobin should be for neonates who undergo major surgery. In a study of neonates undergoing non-cardiac surgery based on data from the American College of Surgeons National Surgical Quality Improvement Program, multiple regression analysis demonstrated that preoperative HCT was an independent risk factor for mortality [18].
Timing — Early transfusion may be associated with increased risk of severe ROP due to early introduction of adult hemoglobin versus fetal hemoglobin, resulting in a shift in the oxygen dissociation curve towards increased oxygen delivery during a time when extremely preterm tissues are still developing adaptive mechanisms to tolerate increased oxygen concentrations [19,20]. In a single center retrospective study of very preterm infants (GA <32 weeks), binary logistic regression analysis that controlled for GA and illness severity (eg, bronchopulmonary dysplasia, postnatal corticosteroid use, inotropic therapy, and persistent patent ductus arteriosus) showed early transfusion within the first 10 days of life was associated with an increase in severe ROP defined as requiring retinal laser ablation or bevacizumab injection [21]. However, there are substantial limitations of this study due to significant differences in birth weight and GA between groups, the lack of data regarding supplemental oxygen use and hemoglobin oxygen saturation levels, the use of ventilation, and detailed retinal findings. This report should be considered hypothesis-generating; further study is needed to confirm these findings prior to making changes on timing of transfusions.
Erythropoietin — Infants who receive appropriate and timely doses of EPO or darbepoetin (a long-acting EPO) therapy require fewer transfusions and are exposed to fewer donors [22-26]. Ongoing studies are evaluating dosing and dosing schedules to maximize increases in red cell mass. However, guidelines published from some organizations, for example, the 2016 guidelines of the British Committee for Standards in Haematology do not recommend the use of erythropoiesis-stimulating agents in neonates [13].
Review and potential revisions to these guidelines are needed based on subsequently published data that include:
●Systematic reviews from 2017 and 2019 that reported early and late administration of EPO reduced the number of RBC transfusions [23,24]. Early administration (before eight days of life) but not late administration also reduced the total volume of transfused blood, reduced donor exposure in preterm infants with a birth weight <2500 g, and was associated with decreased rates of IVH, periventricular leukomalacia, and NEC.
●In the PENUT (Preterm EPO for Neuroprotection) trial, posthoc analysis demonstrated that ELBW infants randomized to high-dose EPO (1000 units/kg every other day for six doses) followed by hematopoietic doses of EPO (400 units/kg three times per week) compared with those randomized to placebo required fewer transfusions (mean 3.2 versus 4.9 transfusions), and had lower cumulative volumes of transfused pRBC (mean 43.3 versus 70.6 mL), and were exposed to fewer donors (mean 1.6 versus 2.4 donors) [27].
In addition, another analysis of 628 infants enrolled in the PENUT trial reported an increase in each red cell transfusion was associated with a decrease in cognitive (0.96 points per transfusion), motor (1.51 points per transfusion), and language (1.10 points per transfusion) score on the Bayley Scales of Infant Development, III [7]. As a result, EPO and darbepoetin have been used in the management of preterm infants, who are at-risk for anemia (see "Anemia of prematurity (AOP)", section on 'Erythropoiesis stimulating agents (ESAs)').
Search for other transfusion markers — Although HCT and hemoglobin have primarily been the triggers to indicate when transfusion is indicated in neonates, they are not adequate measures. This is illustrated by the observation that some infants remain asymptomatic at a given low HCT (hemoglobin), while others exhibit signs that might reflect anemia (tachycardia, poor weight gain, increased requirement of supplemental oxygen, or increased frequency of apnea or bradycardia) at a similar or higher concentration.
Research efforts have focused on identifying a more accurate indicator for transfusions; however, a reliable and sensitive marker has yet to be identified. Investigations to test a more accurate marker for transfusion have included studies of direct or indirect oxygen delivery (eg, cerebral oxygen saturation, peripheral fractional oxygen extraction and oxygen consumption) [28-31], echocardiographic changes in cardiovascular circulation [32-34], and the use of biochemical markers, such as serum lactate or vascular endothelial growth factor [35,36]. Research continues on near-infrared spectroscopy (NIRS) to monitor cerebral and splanchnic saturations, and to identify ratios or cutoffs where regional oxygenation might be improved with transfusion [37]. However despite these ongoing efforts, a marker specifically identifying the need for a transfusion remains elusive.
Our approach — Based upon the available data, the following transfusion guidelines have been adopted in several NICUs, including our own (table 1):
●Acute blood loss that is greater than 20 percent of blood volume.
●Acute blood loss that is greater than 10 percent of blood volume with symptoms of decreased oxygen delivery (such as persistent acidosis) after volume resuscitation.
●The following guidelines for infants with chronic anemia suggest that transfusions should be considered based upon the respiratory support needed by the infant. They are dependent upon an HCT or hemoglobin value that is preferably measured from either a central venous or arterial sample. HCT/hemoglobin values are based on a restrictive transfusion threshold supported by strong evidence from clinical trials. (See 'Target hemoglobin or hematocrit' above.)
If a heelstick specimen is used, it should be obtained after adequate warming of the heel. The final decision for a transfusion is at the discretion of the neonatology team.
•For infants requiring moderate or significant mechanical ventilation, defined as fraction of inspired oxygen (FiO2) ≥0.4, and mean airway pressure (MAP) >8 cm H2O on a conventional ventilator or MAP >14 on a high frequency ventilator, the HCT trigger is <30 percent (hemoglobin ≤10 g/dL).
•For infants requiring minimal mechanical ventilation, defined as fraction of inspired oxygen (FiO2) <0.4, and MAP ≤8 cm H2O on a conventional ventilator or MAP ≤14 on a high frequency ventilator, the HCT trigger is <25 percent (hemoglobin ≤8 g/dL).
•The HCT trigger is <25 percent (hemoglobin ≤8 g/dL) for infants requiring supplemental low- or high-flow oxygen but not mechanical ventilation, and one or more of the following: tachycardia (heart rate ≥180 beats per minute) for ≥24 hours, tachypnea (respiratory rate ≥60 breaths per minute) for ≥24 hours, doubling of oxygen requirement from the previous 48 hours, metabolic acidosis as indicated by a pH 7.2 or serum lactate ≥2.5 mEq/L, weight gain <10 g/kg per day over the previous four days while receiving ≥120 kcal/kg per day, or if the infant undergoes major surgery within 72 hours. For infants requiring oxygen without any signs, a transfusion is not considered until signs occur.
•In asymptomatic infants, the HCT trigger is 21 percent (hemoglobin ≤7 g/dL) with an absolute reticulocyte <100,000/microL (<2 percent). Infants without signs or oxygen requirements who are actively producing new red cells and have an elevated reticulocyte count likely do not require a red cell transfusion. Other centers and societies have transfusion guidelines with higher HCT (hemoglobin) triggers, which are based upon similar requirements for respiratory support [13,38-40].
SELECTION OF RED BLOOD CELL PRODUCTS — Once the decision to transfuse RBCs has been made, the appropriate RBC product is chosen based upon the clinical setting. Donated whole blood used for transfusion may be modified in several ways that remove varying proportions of non-red cell components. These modifications are particularly important in the neonate because of their increased vulnerability to certain infections, such as cytomegalovirus, potential increased risk of graft-versus-host disease due to transfusion, and the possibility of alloimmune hemolytic disease of the newborn.
Products for acute life-threatening blood loss — In the setting of acute blood loss, especially in life-threatening circumstances, any available RBC product that is compatible with the infant's blood type can be administered. If O-negative RBCs are available in the delivery room as an emergency transfusion for the mother, this blood also can be used in the neonate. O-negative whole blood should only be used as a last resort in critically ill neonates as it may contain antibodies directed against A and B blood groups, the Rhesus D antigen, and against leukocytes.
Hemolytic disease of the newborn — Alloimmune hemolytic disease of the fetus and newborn (HDFN) is a condition in which the red cells of the fetus or newborn are destroyed by maternally derived alloantibodies. These antibodies arise in the mother as the direct result of a blood group incompatibility between the mother and fetus. Infants often require exchange transfusion to remove the maternal alloantibodies, reduce the rate of hemolysis, and prevent significant hyperbilirubinemia. In some cases, a simple transfusion is administered because an exchange transfusion cannot be done in a timely manner. (See "Alloimmune hemolytic disease of the newborn: Postnatal diagnosis and management", section on 'Antenatal management'.)
The different types of HDFN include (table 2):
●Rhesus D (RhD) – The most significant setup for HDFN occurs in the Rh-positive infant (ie, D antigen positive) of an Rh-negative mother who has become sensitized and has produced RhD antibodies. An infant who needs either simple transfusion or exchange transfusion because of HDFN should receive RhD-negative red cells of either the appropriate ABO type or group O. In some institutions in which group O, RhD-negative RBCs are used, the RBCs are washed to remove any alloantibodies to group A or B antigens in the plasma.
●ABO incompatibility – Humans have four major blood groups in the ABO system (A, B, AB, and O) named for the antigen(s) on the red cell. The ABO system may cause HDFN, although most cases are less severe than that caused by D antigen isoimmunization. In infants with HDFN due to ABO incompatibility, donor O cells are washed to remove any plasma that contains alloantibodies. These cells may be suspended in plasma that is compatible with both the infant's red cells and the transfused (donor) cells. AB plasma is compatible in all cases, since it does not contain anti-A or anti-B alloantibodies.
●Other Rh antibodies – Other antibodies including Rh antibodies in the C and E systems, and Kell antibodies can cause HDFN or acute or delayed hemolytic transfusion reaction. In these cases, red cells that are negative for the antigen, against which the antibody is directed, are used.
Red blood cells — RBCs are used for replacement during surgery, RBC loss, and sporadic transfusion therapy. The hematocrit (HCT) of the unit varies depending upon the preservative solution. In the United States, commonly used solutions include CPDA-1 (citrate, phosphate, dextrose, adenine) with an HCT of 65 to 70 percent, and nutrient preparations, such as Adsol (AS-1; adenine, glucose, mannitol, and sodium chloride) and Nutricel (AS-2; citrate, phosphate, glucose, adenine, and sodium chloride), with an HCT of 50 to 60 percent. RBCs in the nutrient solutions (AS-1 and AS-2) can be used for up to 42 days after collection, and those preserved in CPDA-1 for up to 28 days [41-43]. (See "Practical aspects of red blood cell transfusion in adults: Storage, processing, modifications, and infusion".)
It has been suggested that the use of fresh RBC transfusions in preterm infants would result in lower rates of organ dysfunction, nosocomial infection, and length of stay. However, a Canadian multicenter trial of preterm infants (birth weight [BW] <1250 g) demonstrated no benefit in the use of fresh RBC (stored for seven days or less) transfusion compared with transfusions using RBCs stored under standard blood bank practice [44]. Specifically, there was no difference in the primary composite outcome of major neonatal morbidities that included necrotizing enterocolitis (NEC), retinopathy of prematurity (ROP), bronchopulmonary dysplasia (BPD), and intraventricular hemorrhage (IVH) between groups of infants who received fresh RBCs (mean age of RBCs, 5.1 ± 2.1 days) and those administered standard transfusions (mean age of RBCs, 14.6 ± 8.3 days). Infants in the fresh RBC group were exposed to a greater number of donors compared with the standard RBC group (3.7 ± 2.7 donors versus 2.1 ± 1.6 donors). Although the relationship between timing of red cell transfusion and the incidence of NEC continues to be evaluated [45], the Canadian study findings show that no alteration for the standard RBCs storage time practices is needed for preterm infants of neonatal RBC transfused products. (See "Neonatal necrotizing enterocolitis: Pathology and pathogenesis", section on 'Anemia and RBC transfusion'.)
Leukoreduced and irradiated red cells — Leukoreduction filters remove approximately 99.9 percent of white blood cells from packed red blood cells (PRBCs), which reduces febrile non-hemolytic transfusion reactions, prevents alloimmunization, and reduces (but does not completely prevent) the transmission of certain infections (notably cytomegalovirus [CMV]). However, leukoreduction does not eliminate all lymphocytes and cannot prevent transfusion-associated graft-versus-host disease (TA-GVHD).
Irradiation prevents TA-GVHD in susceptible recipients. The dose of radiation is not sufficient to kill viruses and irradiation does not provide a CMV-safe product.
Leukoreduced PRBCs should be used in all neonates. In addition, most blood banks will also irradiate leukoreduced PRBCs prior to neonatal transfusion.
CMV-safe red cells — Cytomegalovirus (CMV)-safe products (including CMV-seronegative and leukoreduced red cells) reduce the transmission of CMV in seronegative recipients. In our practice, we use CMV-seronegative PRBCs for all neonates. In centers in which this practice is not available, CMV-seronegative PRBCs should be used for the following populations:
●Infants awaiting or undergoing transplantation
●Immunocompromised infants
●Preterm infants of CMV-seronegative mothers
Cord blood — Cord blood has been suggested as a form of autologous blood donation; however, most birthing institutions are not prepared for the collection and storage of neonatal cord blood. Even with a program that collects and processes umbilical cord blood, autologous blood was only available in one-third of neonates that required transfusion [46]. Although the collection of cord blood for autologous transfusion in neonates requires further study, the use of cord blood for initial blood tests can be easily instituted. In one study, the use of cord blood for initial laboratory evaluation resulted in an increased initial HCT at 12 to 24 hours after birth, and decreased the rate of early transfusions and vasopressors in very low birth weight (VLBW) infants (BW <1500 g) [47].
Another method that may reduce erythrocyte transfusions is delayed clamping of the umbilical cord, especially in preterm infants. (See "Labor and delivery: Management of the normal first stage".)
ADMINISTRATION — The volume of transfusion is dependent upon the desired rise in hematocrit (Hct).
The volume of transfusion in mL is equal to the following calculation:
Wt (kg) X blood volume per kg X (Desired HCT- Observed HCT)/HCT PRBCs
In the newborn, the blood volume is approximately 85 mL/kg (or 90 mL/kg in very low birth weight [VLBW] infants). Transfusions generally are given as packed red blood cells (PRBCs), in aliquots of 10 to 20 mL/kg, over two to four hours. In some circumstances, such as hemodynamic instability or hypovolemia due to blood loss, a smaller volume (10 mL/kg) is given more rapidly (over one to two hours). Data are inconclusive regarding the optimal volume for neonatal PRBC transfusion. In a systematic review of the literature, there were only four trials with 146 infants that compared transfusion volumes of 10 versus 20 mL/kg [12]. Although there were no reports of difference in neonatal outcomes including mortality, the number of patients is too small to reach any definitive conclusion regarding choice of aliquot volume for PRBC transfusion.
In extremely low birth weight (ELBW) and VLBW infants, the small volumes, which may be as low as 7 mL of blood for a single transfusion, require the use of special equipment to maximize the use of a single unit from a donor. In addition, these systems allow for serial transfusions to an individual neonate from the same donor.
These include the following two systems:
●Small bags may come as part of a transfusion set (referred to as satellite packs) in which four or six aliquots can be made from a single red cell unit. If the blood bank has a sterile connecting device, the small bags can be connected to the large blood unit, and an appropriate amount withdrawn at any time.
●Another convenient device for small-volume transfusions is a syringe set in which the syringe is sterilely connected to the original unit. Blood is drawn through a filter into a syringe, and can be used within four hours without further filtration. Such systems have been shown to be safe [48] and can increase the number of transfusions from a single unit.
With both systems, the original expiration date of the unit is maintained if the sampling device remains connected to the original unit in a sterile manner. This period may be up to six weeks for use of a single unit in a nutrient solution such as Adsol (AS-1) and Nutricel (AS-2). If a single unit is designated for a premature infant and is used until its expiration date, as many as 13 individual transfusions can be made from a single donor unit, which markedly decreases donor exposure [49].
SUMMARY AND RECOMMENDATIONS — Anemia occurs when the red blood cell (RBC) mass does not adequately meet the oxygen demands of tissues. Neonates may require RBC transfusions because of significant anemia due to acute blood loss or chronic anemia due to physiologic anemia and blood loss from phlebotomy. In particular, anemia in preterm infants (referred to as anemia of prematurity) is more severe and presents earlier in life. (See "Anemia of prematurity (AOP)".)
●Neonates with significant acute blood loss require immediate fluid resuscitation. Indications for RBC transfusion in a neonate with acute blood loss following volume resuscitation include persistent acidosis indicating inadequate oxygen delivery, ongoing bleeding, and an estimated blood loss greater than 20 percent of the infant's blood volume. (See 'Acute blood loss' above.)
●In the acute setting of blood loss, especially in life-threatening circumstances, any available RBC product that is compatible with the infant's blood type can be administered. O-negative whole blood should not be used as it contains antibodies directed against A or B blood group, and leukocytes. O-negative whole blood should only be used as a last resort in critically ill neonates. (See 'Products for acute life-threatening blood loss' above.)
●In neonates with chronic anemia, target hemoglobin (Hb) and hematocrit (Hct) are used as clinical indicators for RBC transfusion based on gestational age and the clinical status of the infant. Evidence based on clinical trials has shown that a restrictive transfusion strategy (low Hb/Hct threshold) versus a liberal approach (high Hb/Hct threshold) results in fewer transfusions and does not increase the risk of death or serious morbidity. We recommend a restrictive approach for neonatal transfusions that is based upon HCT triggers and the respiratory support required by the infant (table 1) (Grade 1B). (See 'Target hemoglobin or hematocrit' above.)
●Blood products for neonates with hemolytic disease of the fetus and newborn (HDFN) who require either exchange or simple transfusion need to be compatible with the infant's blood type. (See 'Hemolytic disease of the newborn' above.)
●We suggest that leukoreduced packed red blood cells (PRBCs) should be used in all neonatal transfusions because it reduces febrile non-hemolytic transfusion reactions, prevents alloimmunization, and reduces the transmission of certain infections (notably cytomegalovirus [CMV]) (Grade 2C). Most neonatal intensive care units (NICUs) also use irradiated RBCs, which do not provide a CMV-safe product, but prevent transfusion-associated-graft-versus-host disease. (See 'Leukoreduced and irradiated red cells' above.)
●CMV-seronegative PRBCs reduce the transmission of CMV in seronegative recipients. We suggest that CMV-seronegative PRBCs be used for all neonates (Grade 2C). We recommend the use of CMV-seronegative PRBCs for immunodeficient neonates, those who are awaiting transplantation, and preterm infants of seronegative mothers (Grade 1C). (See 'CMV-safe red cells' above.)
●In the neonate, transfusions generally are given as packed red blood cells (PRBCs) in aliquots of 10 to 20 mL/kg, over two to four hours. In some circumstances, such as hemodynamic instability or hypovolemia due to blood loss, a smaller volume (10 mL/kg) is given more rapidly (over one to two hours). In extremely low birth weight (ELBW) infants (BW <1000 g) small volumes may require the use of special equipment that allows several aliquots to be administered from a single unit. (See 'Administration' above.)
