Transfusion of erythrocytes in newborn infants, particularly in premature newborns, is a common practice. The indications are those that hold for any other time of life; hypovolemia and anemia. Transfusion for anemia has created much controversy. Anemia is defined as a hemoglobin concentration lower than that which is normal for the given patient. This concept of normality is difficult to apply to the premature infant, who as part of his/her normal course may require respiratory assistance, may have apneic spells, and may have much blood sampled for laboratory tests. There is significant controversy about whether a low hemoglobin concentration is harmful to infants or whether transfusions to maintain an arbitrary hemoglobin concentration improve their clinical condition. Guidelines for administering a RBC transfusion are presented in Table 46-17 and, although lacking definitive proof, are provided as an aid with the recognition that
practice is variable and that criteria should be developed and updated as required by an appropriate process at each local institution.

In newborn infants almost all RBC transfusions are packed RBCs as compared to whole blood or reconstituted whole blood which should be considered a special request and indicated only in specific clinical situations (Table 46-18).

Small volume (10-15 ml/kg) transfusions of packed RBCs are frequently required in extremely low birth weight (<1000 g) or very low birth weight (<1500 g) newborn infants. There are two phases of RBC transfusion need following preterm delivery: (a) early, in infants requiring major surgery or intensive care; and (b) later, during the period of the physiologic anemia of prematurity. Factors that influence the need for RBC transfusion in premature infants include (a) the initial endowment of the infant with blood at birth reflected by the initial red cell mass; (b) the magnitude of iatrogenic blood losses, related to the degree and duration of intensive care; and (c) failure of erythropoiesis (469). As a result of the impact of placental transfusion at birth (a source of autologous blood) (470,471), more restrictive practices for RBC transfusion in LBW infants (472,473), and the elimination of a fixed iatrogenic blood loss as a “trigger” for RBC transfusion, significantly fewer small volume RBC transfusions are given to low birth-weight infants (473,474 and 475). As an example in the study reported by Widness and colleagues (475) the percentage of infants with a birth weight of 1.0 to 1.5 kg who were given RBC transfusions was 83% in 1982, 67% in 1989 and 34% in 1993; the average number of RBC transfusions (15 ml/kg) given per patient was 7.0 in 1982, 5.0 in 1987, and 2.3 in 1993. In a more recent study of extremely low birth weight infants (birth weight <1000 g) reported by Maier and colleagues (473) the median number of RBC transfusions decreased from 7 in the period 1989 to 1991 to 2 in the period 1995 to 1997 with a corresponding reduction in the median number of blood donors per infant from 5 to 1; impressively, the percent of extremely low birth weight (ELBW) infants not transfused with packed RBC in the same periods increased from 3% (1989-1991) to 25% (1995-1997). These changes in transfusion practices need to be taken into consideration when evaluating the results of erythropoietin (EPO) trials in newborn infants, and there is general consensus among experts that, based on available data, the more liberal transfusion practices of the 1980s and early 1990s can be replaced by a more restrictive transfusion policy in small preterm infants without any significant adverse short-term affects; future trials of RBC transfusion and EPO administration in low birth weight infant should focus on long-term outcomes such as neurological development (472). Efforts to reduce iatrogenic blood loss by restricting phlebotomy, miniaturized analyses in the laboratory and transcutaneous blood gas monitoring in ill, small preterm infants need to be stressed (473).

The type of RBC product for use in small volume (15 + 5 ml/kg) transfusions has been reviewed in detail by Strauss and colleagues (476). Based on a critical review of available literature, it is apparent that a dedicated single-donor system, in which RBCs are collected into AS-1 or AS-3 anticoagulant from either unrelated donors or biologic parents, and stored for up to 42 days, is able to supply all small volume RBC transfusions needed by individual patients without adverse effects (476). Their protocol at the University of Iowa Hospital and Clinics is of interest in this regard; depending on the anticipated RBC needs of individual preterm infants up to 50% of a fresh RBC unit is reserved for use by each infant throughout the 42 days of RBC storage. With this dedicated single-donor system, 88%
of transfused preterm infants received RBCs from only one donor, with the remaining 12% being exposed to only two donors. The ideal goal of only one donor per infant was almost achieved without compromising safety (476), and with acceptable cost-effectiveness (477). In addition to recommending use of RBCs stored in extended-storage anticoagulant or preservative solutions to limit donor exposures (as compared to relatively fresh RBCs), Strauss and colleagues (476) recommended the use of CMV safe (CMV seronegative or leukocyte-reduced RBCs) to minimize and potentially eliminate transfusion-transmitted CMV, and γ-irradiation to prevent transfusion-associated graft versus host disease (TA GVHD).

One area of interest is the choice of blood products in newborn infants with activation of the Thomsen-Friedenreich neoantigen (T antigen), a condition often reported in association with NEC. The condition may be associated with hemolysis and is thought to reflect exposure of the Thomsen-Friedenreich cryptantigen (T antigen) by removal of N-acetylneuraminidase acid (also known as sialidase). Neuraminidase is produced by a large number microbial agents including bacteria, viruses and protozoa. In vivo T activation has been reported in association with anaerobic, particularly clostridial sepsis, and other infections. In one recent study 48 of 375 (12.8%) neonates admitted to a tertiary referral center were found to have T or t-variant activated RBCs (478). Thirteen of the 48 infants (27%) developed at least one episode of sepsis and 9 (19%) NEC during their hospital stay; however, T-activation was not always temporally associated with NEC or sepsis. Twenty-three of the 48 (48%) of infants with T-activated RBCs received standard blood products without evidence of transfusion-associated hemolysis. Based on these observations the investigators concluded that, in the neonatal population, routine screening for T-activation of infants with sepsis or NEC is not justified and that the routine provision of low-titer anti-T blood components, washed RBCs or platelet suspended in additive is not warranted and should be considered only in the very small percentage of infants with true T-activation (as compared to T-variant activation) and clinically significant transfusion-associated hemolysis (478). These findings and recommendations are at slight variance to those of Williams and colleagues who reported an incidence of 11% of T-antigen activation in 72 infants with NEC, four of whom experienced hemolysis (one severe) associated with the transfusion of standard blood products (479). The investigators proposed a transfusion protocol that included the use of washed RBCs/platelets or blood components with low anti-T reactivity. Although it is clear that prudence should be exercised in selecting blood components for neonates with RBC T-activation, it is important that such infants, many of whom are small and ill, not be denied the benefits of essential hemostatic components (e.g. plasma, cryoprecipitate, platelets) because of a perceived potential risk of hemolysis from standard donor products that are likely to contain anti-T antibodies (480).